{
  "dataset": "benchmark-leaderboard",
  "title": "OpenRouter benchmark leaderboard",
  "description": "One row per benchmark, model version, and provider: the aggregate values displayed on the OpenRouter leaderboard pages (GPQA Diamond, tau2-bench airline). Derived from the benchmark-runs dataset.",
  "schema_version": "v1",
  "snapshot_date": "2026-10-25",
  "retrieved_at": "2026-10-25T01:45:32.918Z",
  "update_cadence": "daily",
  "license": {
    "name": "CC BY 4.0",
    "url": "https://creativecommons.org/licenses/by/4.0/"
  },
  "attribution": "Source: OpenRouter (openrouter.ai), licensed under CC BY 4.0: reuse and republish with attribution to OpenRouter.",
  "harness": {
    "name": "OpenRouter benchmark harness",
    "url": "https://github.com/OpenRouterTeam/benchmark-harness"
  },
  "source_pages": [
    "https://openrouter.ai/benchmarks/gpqa-diamond",
    "https://openrouter.ai/benchmarks/tau2-bench-airline"
  ],
  "methodology": "Rows aggregate the benchmark-runs dataset exactly as the leaderboard pages do, over each benchmark's non-zero-score runs from the newest 90 days: accuracy is the task-weighted mean of per-run accuracy, accuracy_std_dev is the population standard deviation of per-run accuracy (null with a single run), and the per-task averages divide summed cost, generation time, and output tokens by summed tasks. Rows are ordered by accuracy descending, then model name.",
  "readme_url": "https://openrouter.ai/api/v1/datasets/exports/README.md",
  "columns": [
    {
      "name": "benchmark",
      "type": "string",
      "nullable": false,
      "description": "Benchmark identifier."
    },
    {
      "name": "benchmark_page_url",
      "type": "string",
      "nullable": false,
      "description": "openrouter.ai leaderboard page that displays this row."
    },
    {
      "name": "model_permaslug",
      "type": "string",
      "nullable": false,
      "description": "Immutable OpenRouter model version identifier."
    },
    {
      "name": "model_slug",
      "type": "string",
      "nullable": true,
      "description": "Current OpenRouter model slug, null when the version is no longer listed."
    },
    {
      "name": "model_name",
      "type": "string",
      "nullable": false,
      "description": "Display name as shown on the leaderboard page."
    },
    {
      "name": "provider_name",
      "type": "string",
      "nullable": false,
      "description": "Inference provider that served the aggregated runs."
    },
    {
      "name": "accuracy",
      "type": "number",
      "nullable": false,
      "description": "Task-weighted mean accuracy across runs, 0 to 1."
    },
    {
      "name": "accuracy_std_dev",
      "type": "number",
      "nullable": true,
      "description": "Population standard deviation of per-run accuracy; null when only one run exists."
    },
    {
      "name": "run_count",
      "type": "integer",
      "nullable": false,
      "description": "Number of runs aggregated."
    },
    {
      "name": "total_questions",
      "type": "integer",
      "nullable": false,
      "description": "Tasks attempted across all aggregated runs (sample size)."
    },
    {
      "name": "avg_cost_per_task_usd",
      "type": "number",
      "nullable": true,
      "description": "Summed run cost divided by summed tasks, in USD."
    },
    {
      "name": "avg_duration_per_task_ms",
      "type": "number",
      "nullable": true,
      "description": "Summed generation time divided by summed tasks, in milliseconds."
    },
    {
      "name": "avg_output_tokens_per_task",
      "type": "number",
      "nullable": true,
      "description": "Summed completion tokens divided by summed tasks."
    },
    {
      "name": "last_run_timestamp",
      "type": "string",
      "nullable": false,
      "description": "ISO 8601 UTC timestamp of the most recent aggregated run."
    }
  ],
  "row_count": 2274,
  "rows": [
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-4-argon-20261013",
      "model_slug": "google/gemini-4-argon",
      "model_name": "Google: Gemini 4 Argon",
      "provider_name": "Google",
      "accuracy": 0.964646,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.09468125757575757,
      "avg_duration_per_task_ms": 101980.28282828283,
      "avg_output_tokens_per_task": 8439.414141414141,
      "last_run_timestamp": "2026-10-14T15:00:03.588Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-4-argon",
      "model_slug": "google/gemini-4-argon",
      "model_name": "Gemini 4 Argon",
      "provider_name": "auto-routing",
      "accuracy": 0.959596,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.029260161616161615,
      "avg_duration_per_task_ms": 40252.82323232323,
      "avg_output_tokens_per_task": 2870.0252525252527,
      "last_run_timestamp": "2026-10-23T07:38:37.725Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.8-flash-20260902",
      "model_slug": "google/gemini-3.8-flash",
      "model_name": "Google: Gemini 3.8 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.95637,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 191,
      "avg_cost_per_task_usd": 0.06822796858638744,
      "avg_duration_per_task_ms": 132197.8062827225,
      "avg_output_tokens_per_task": 18049.068062827224,
      "last_run_timestamp": "2026-10-01T08:31:33.481Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-astra-pro-20260903",
      "model_slug": "openai/gpt-6-astra-pro",
      "model_name": "OpenAI: GPT-6 Astra Pro",
      "provider_name": "auto-routing",
      "accuracy": 0.954545,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.332555398989899,
      "avg_duration_per_task_ms": 75446.5505050505,
      "avg_output_tokens_per_task": 4530.388888888889,
      "last_run_timestamp": "2026-10-01T08:27:35.414Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.8-flash-20260902",
      "model_slug": "google/gemini-3.8-flash",
      "model_name": "Google: Gemini 3.8 Flash",
      "provider_name": "Google AI Studio",
      "accuracy": 0.952862,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.07423223737373738,
      "avg_duration_per_task_ms": 86748.28787878787,
      "avg_output_tokens_per_task": 19646.409090909092,
      "last_run_timestamp": "2026-10-01T08:29:11.347Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.1-pro-preview-20260219",
      "model_slug": "google/gemini-3.1-pro-preview",
      "model_name": "Google: Gemini 3.1 Pro Preview",
      "provider_name": "Google",
      "accuracy": 0.95202,
      "accuracy_std_dev": 0.0008420000000000094,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.20364760606060606,
      "avg_duration_per_task_ms": 135706.18686868687,
      "avg_output_tokens_per_task": 16847.588383838385,
      "last_run_timestamp": "2026-10-01T19:14:26.157Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-astra-pro-20260903",
      "model_slug": "openai/gpt-6-astra-pro",
      "model_name": "OpenAI: GPT-6 Astra Pro",
      "provider_name": "OpenAI",
      "accuracy": 0.9511785,
      "accuracy_std_dev": 0.0016835000000000044,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.33620659090909094,
      "avg_duration_per_task_ms": 70544.52272727272,
      "avg_output_tokens_per_task": 4588.830808080808,
      "last_run_timestamp": "2026-10-02T23:23:01.262Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-astra-pro-20260903",
      "model_slug": "openai/gpt-6-astra-pro",
      "model_name": "OpenAI: GPT-6 Astra Pro",
      "provider_name": "Azure",
      "accuracy": 0.949495,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.35495546717171716,
      "avg_duration_per_task_ms": 72102.17676767676,
      "avg_output_tokens_per_task": 4622.60101010101,
      "last_run_timestamp": "2026-10-01T08:29:03.480Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.1-pro-preview-20260219",
      "model_slug": "google/gemini-3.1-pro-preview",
      "model_name": "Google: Gemini 3.1 Pro Preview",
      "provider_name": "Google AI Studio",
      "accuracy": 0.9483723333333332,
      "accuracy_std_dev": 0.004827144451490511,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.20306945454545455,
      "avg_duration_per_task_ms": 134835.03535353535,
      "avg_output_tokens_per_task": 16798.409090909092,
      "last_run_timestamp": "2026-10-01T19:12:42.155Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.8-flash-20260902",
      "model_slug": "google/gemini-3.8-flash",
      "model_name": "Google: Gemini 3.8 Flash",
      "provider_name": "Google",
      "accuracy": 0.9482614885496183,
      "accuracy_std_dev": 0.0012429999999999941,
      "run_count": 2,
      "total_questions": 393,
      "avg_cost_per_task_usd": 0.07536056743002545,
      "avg_duration_per_task_ms": 132814.94147582698,
      "avg_output_tokens_per_task": 18975.51145038168,
      "last_run_timestamp": "2026-10-15T21:00:04.244Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "inclusionai/ling-3.1-flash-20261002",
      "model_slug": "inclusionai/ling-3.1-flash",
      "model_name": "inclusionAI: Ling 3.1 Flash",
      "provider_name": "DeepInfra",
      "accuracy": 0.947917,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 96,
      "avg_cost_per_task_usd": 0.009331666666666667,
      "avg_duration_per_task_ms": 232878.71875,
      "avg_output_tokens_per_task": 10278.75,
      "last_run_timestamp": "2026-10-14T18:00:04.673Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6.1-sol-20260929",
      "model_slug": "openai/gpt-6.1-sol",
      "model_name": "OpenAI: GPT-6.1 Sol",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.947811,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.021931015151515153,
      "avg_duration_per_task_ms": 68600.9191919192,
      "avg_output_tokens_per_task": 1835.4242424242425,
      "last_run_timestamp": "2026-10-03T01:13:05.002Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-astra-20260903",
      "model_slug": "openai/gpt-6-astra",
      "model_name": "OpenAI: GPT-6 Astra",
      "provider_name": "OpenAI",
      "accuracy": 0.9472503333333333,
      "accuracy_std_dev": 0.0015872190215034077,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.11162585016835018,
      "avg_duration_per_task_ms": 48234.978114478115,
      "avg_output_tokens_per_task": 2076.191919191919,
      "last_run_timestamp": "2026-10-02T23:16:05.967Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "sakana/fugu-ultra",
      "model_slug": "sakana/fugu-ultra",
      "model_name": "Fugu Ultra",
      "provider_name": "auto-routing",
      "accuracy": 0.9461279999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 1.148739303030303,
      "avg_duration_per_task_ms": 328779.10101010103,
      "avg_output_tokens_per_task": 33271.606060606064,
      "last_run_timestamp": "2026-09-15T01:07:47.987Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.1-pro-preview-20260219",
      "model_slug": "google/gemini-3.1-pro-preview",
      "model_name": "Google: Gemini 3.1 Pro Preview",
      "provider_name": "auto-routing",
      "accuracy": 0.945286,
      "accuracy_std_dev": 0.0008420000000000094,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.20203182828282829,
      "avg_duration_per_task_ms": 133595.55303030304,
      "avg_output_tokens_per_task": 16712.858585858587,
      "last_run_timestamp": "2026-10-01T19:16:14.342Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-astra-20260903",
      "model_slug": "openai/gpt-6-astra",
      "model_name": "OpenAI: GPT-6 Astra",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.945286,
      "accuracy_std_dev": 0.0008420000000000094,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.11987038888888889,
      "avg_duration_per_task_ms": 35572.169191919194,
      "avg_output_tokens_per_task": 2020.9419191919192,
      "last_run_timestamp": "2026-10-01T08:24:02.411Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-astra-20260903",
      "model_slug": "openai/gpt-6-astra",
      "model_name": "OpenAI: GPT-6 Astra",
      "provider_name": "auto-routing",
      "accuracy": 0.944444,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.11333201010101009,
      "avg_duration_per_task_ms": 57056.792929292926,
      "avg_output_tokens_per_task": 2108.121212121212,
      "last_run_timestamp": "2026-10-01T08:25:16.514Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6.1-sol-20260929",
      "model_slug": "openai/gpt-6.1-sol",
      "model_name": "OpenAI: GPT-6.1 Sol",
      "provider_name": "OpenAI",
      "accuracy": 0.944444,
      "accuracy_std_dev": 0.0027491439879836154,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.020607563973063972,
      "avg_duration_per_task_ms": 47519.37542087542,
      "avg_output_tokens_per_task": 1902.2373737373737,
      "last_run_timestamp": "2026-10-02T23:20:22.253Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6.1-sol-20260929",
      "model_slug": "openai/gpt-6.1-sol",
      "model_name": "OpenAI: GPT-6.1 Sol",
      "provider_name": "auto-routing",
      "accuracy": 0.944444,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.02136650505050505,
      "avg_duration_per_task_ms": 52411.32323232323,
      "avg_output_tokens_per_task": 1978.1313131313132,
      "last_run_timestamp": "2026-10-01T08:24:51.870Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-astra-20260903",
      "model_slug": "openai/gpt-6-astra",
      "model_name": "OpenAI: GPT-6 Astra",
      "provider_name": "Azure",
      "accuracy": 0.9431713315363881,
      "accuracy_std_dev": 0.006147978549897494,
      "run_count": 4,
      "total_questions": 742,
      "avg_cost_per_task_usd": 0.10982969407008086,
      "avg_duration_per_task_ms": 117613.10242587602,
      "avg_output_tokens_per_task": 1949.2371967654988,
      "last_run_timestamp": "2026-10-01T08:26:36.876Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.7-flash-20260813",
      "model_slug": "google/gemini-3.7-flash",
      "model_name": "Google: Gemini 3.7 Flash",
      "provider_name": "Google AI Studio",
      "accuracy": 0.9427606666666666,
      "accuracy_std_dev": 0.0036369299629708685,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.029872336700336698,
      "avg_duration_per_task_ms": 28492.575757575756,
      "avg_output_tokens_per_task": 7819.474747474747,
      "last_run_timestamp": "2026-10-01T16:16:03.631Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-sol-pro-20260709",
      "model_slug": "openai/gpt-5.6-sol-pro",
      "model_name": "OpenAI: GPT-5.6 Sol Pro",
      "provider_name": "auto-routing",
      "accuracy": 0.9393935,
      "accuracy_std_dev": 0.0016835000000000044,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.29793838636363634,
      "avg_duration_per_task_ms": 106861.7196969697,
      "avg_output_tokens_per_task": 11692.656565656565,
      "last_run_timestamp": "2026-10-01T16:10:33.208Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.7-flash-20260813",
      "model_slug": "google/gemini-3.7-flash",
      "model_name": "Google: Gemini 3.7 Flash",
      "provider_name": "Google",
      "accuracy": 0.9391951653543308,
      "accuracy_std_dev": 0.009884880879403648,
      "run_count": 4,
      "total_questions": 762,
      "avg_cost_per_task_usd": 0.026313035433070862,
      "avg_duration_per_task_ms": 56131.107611548556,
      "avg_output_tokens_per_task": 7729.320209973754,
      "last_run_timestamp": "2026-10-15T21:01:11.190Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.7-flash-20260813",
      "model_slug": "google/gemini-3.7-flash",
      "model_name": "Google: Gemini 3.7 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.938528,
      "accuracy_std_dev": 0.004357499999999959,
      "run_count": 2,
      "total_questions": 385,
      "avg_cost_per_task_usd": 0.0293307974025974,
      "avg_duration_per_task_ms": 59248.22337662338,
      "avg_output_tokens_per_task": 7683.532467532467,
      "last_run_timestamp": "2026-10-01T16:17:32.289Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6.1-sol-20260929",
      "model_slug": "openai/gpt-6.1-sol",
      "model_name": "OpenAI: GPT-6.1 Sol",
      "provider_name": "Azure",
      "accuracy": 0.9382713333333333,
      "accuracy_std_dev": 0.004627357405210382,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.022108563973063967,
      "avg_duration_per_task_ms": 32928.50084175084,
      "avg_output_tokens_per_task": 1914.9974747474748,
      "last_run_timestamp": "2026-10-01T08:40:50.203Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.5-20260423",
      "model_slug": "openai/gpt-5.5",
      "model_name": "OpenAI: GPT-5.5",
      "provider_name": "auto-routing",
      "accuracy": 0.9371493333333333,
      "accuracy_std_dev": 0.004199386807089248,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.30925494444444446,
      "avg_duration_per_task_ms": 152328.37037037036,
      "avg_output_tokens_per_task": 10179.092592592593,
      "last_run_timestamp": "2026-10-01T19:12:11.505Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.5-20260423",
      "model_slug": "openai/gpt-5.5",
      "model_name": "OpenAI: GPT-5.5",
      "provider_name": "OpenAI",
      "accuracy": 0.9360266666666668,
      "accuracy_std_dev": 0.0036369299629708854,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.3141816919191919,
      "avg_duration_per_task_ms": 142474.1195286195,
      "avg_output_tokens_per_task": 10342.412457912458,
      "last_run_timestamp": "2026-10-02T23:25:41.614Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "Io Net",
      "accuracy": 0.9353121872146118,
      "accuracy_std_dev": 0.0026674999999999893,
      "run_count": 2,
      "total_questions": 219,
      "avg_cost_per_task_usd": 0.014858730593607304,
      "avg_duration_per_task_ms": 395134.2511415525,
      "avg_output_tokens_per_task": 29583.762557077625,
      "last_run_timestamp": "2026-10-24T08:19:41.879Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.8-flash",
      "model_slug": "google/gemini-3.8-flash",
      "model_name": "Gemini 3.8 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.9343430000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.023175040404040403,
      "avg_duration_per_task_ms": 44381.02525252525,
      "avg_output_tokens_per_task": 6131.126262626262,
      "last_run_timestamp": "2026-10-23T07:37:43.257Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-sol-pro-20260709",
      "model_slug": "openai/gpt-5.6-sol-pro",
      "model_name": "OpenAI: GPT-5.6 Sol Pro",
      "provider_name": "Azure",
      "accuracy": 0.9337823333333333,
      "accuracy_std_dev": 0.008398773614178557,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.3842955404040404,
      "avg_duration_per_task_ms": 82918.1936026936,
      "avg_output_tokens_per_task": 11996.59090909091,
      "last_run_timestamp": "2026-10-01T16:36:19.657Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-sol-pro-20260709",
      "model_slug": "openai/gpt-5.6-sol-pro",
      "model_name": "OpenAI: GPT-5.6 Sol Pro",
      "provider_name": "OpenAI",
      "accuracy": 0.9332206666666667,
      "accuracy_std_dev": 0.00965467639137749,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.24476150168350166,
      "avg_duration_per_task_ms": 101605.01851851853,
      "avg_output_tokens_per_task": 11508.424242424242,
      "last_run_timestamp": "2026-10-02T23:21:48.863Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "sakana/fugu-ultra-v2",
      "model_slug": "sakana/fugu-ultra-v2",
      "model_name": "Fugu Ultra V2",
      "provider_name": "auto-routing",
      "accuracy": 0.93266,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.5025571414141414,
      "avg_duration_per_task_ms": 194915.41919191918,
      "avg_output_tokens_per_task": 13806.227272727272,
      "last_run_timestamp": "2026-09-15T01:07:48.052Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "typesafe/jev-router",
      "model_slug": "typesafe/jev-router",
      "model_name": "TypeSafe: Jev Router",
      "provider_name": "auto-routing",
      "accuracy": 0.93266,
      "accuracy_std_dev": 0.004761657064510171,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.03490100505050505,
      "avg_duration_per_task_ms": 57907.71885521885,
      "avg_output_tokens_per_task": 6070.841750841751,
      "last_run_timestamp": "2026-10-09T23:23:36.213Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.5-20260423",
      "model_slug": "openai/gpt-5.5",
      "model_name": "OpenAI: GPT-5.5",
      "provider_name": "Azure",
      "accuracy": 0.9326596250000001,
      "accuracy_std_dev": 0.00862529761714779,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.32610833964646463,
      "avg_duration_per_task_ms": 161836.85416666666,
      "avg_output_tokens_per_task": 10112.390782828283,
      "last_run_timestamp": "2026-10-01T19:14:07.472Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.6-flash-20260721",
      "model_slug": "google/gemini-3.6-flash",
      "model_name": "Google: Gemini 3.6 Flash",
      "provider_name": "Google AI Studio",
      "accuracy": 0.9325395714285714,
      "accuracy_std_dev": 0.004401689845956867,
      "run_count": 4,
      "total_questions": 756,
      "avg_cost_per_task_usd": 0.06429072222222224,
      "avg_duration_per_task_ms": 63071.10317460317,
      "avg_output_tokens_per_task": 11529.871693121693,
      "last_run_timestamp": "2026-10-01T19:12:42.162Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-sonnet-5.5-20260928",
      "model_slug": "anthropic/claude-sonnet-5.5",
      "model_name": "Anthropic: Claude Sonnet 5.5",
      "provider_name": "Google",
      "accuracy": 0.9308640666666669,
      "accuracy_std_dev": 0.007021551143602255,
      "run_count": 15,
      "total_questions": 2970,
      "avg_cost_per_task_usd": 0.026242806060606066,
      "avg_duration_per_task_ms": 18771.915488215487,
      "avg_output_tokens_per_task": 2231.500336700337,
      "last_run_timestamp": "2026-10-24T08:30:19.120Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-sonnet-5.5-20260928",
      "model_slug": "anthropic/claude-sonnet-5.5",
      "model_name": "Anthropic: Claude Sonnet 5.5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.9306396,
      "accuracy_std_dev": 0.006683344480123684,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.025279440404040406,
      "avg_duration_per_task_ms": 20334.968686868688,
      "avg_output_tokens_per_task": 2296.2020202020203,
      "last_run_timestamp": "2026-10-24T08:11:43.558Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.6-flash-20260721",
      "model_slug": "google/gemini-3.6-flash",
      "model_name": "Google: Gemini 3.6 Flash",
      "provider_name": "Google",
      "accuracy": 0.930415,
      "accuracy_std_dev": 0.004827415699246697,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.06216489141414142,
      "avg_duration_per_task_ms": 64761.851851851854,
      "avg_output_tokens_per_task": 11730.429292929293,
      "last_run_timestamp": "2026-10-01T19:12:42.153Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-sonnet-5.5-20260928",
      "model_slug": "anthropic/claude-sonnet-5.5",
      "model_name": "Anthropic: Claude Sonnet 5.5",
      "provider_name": "Anthropic",
      "accuracy": 0.9303028,
      "accuracy_std_dev": 0.00377921650081072,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.02488049898989899,
      "avg_duration_per_task_ms": 19525.145454545454,
      "avg_output_tokens_per_task": 2247.960606060606,
      "last_run_timestamp": "2026-10-24T08:10:23.014Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "x-ai/grok-4.6-20260810",
      "model_slug": "x-ai/grok-4.6",
      "model_name": "SpaceXAI: Grok 4.6",
      "provider_name": "auto-routing",
      "accuracy": 0.928451,
      "accuracy_std_dev": 0.004209000000000018,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.12326839393939394,
      "avg_duration_per_task_ms": 361758.7878787879,
      "avg_output_tokens_per_task": 20322.222222222223,
      "last_run_timestamp": "2026-10-01T16:09:08.142Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Parasail",
      "accuracy": 0.9280388242811503,
      "accuracy_std_dev": 0.010739874350082303,
      "run_count": 12,
      "total_questions": 2191,
      "avg_cost_per_task_usd": 0.1293467288909174,
      "avg_duration_per_task_ms": 132537.31081697854,
      "avg_output_tokens_per_task": 8542.681424007302,
      "last_run_timestamp": "2026-10-24T08:42:17.989Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-sonnet-5.5-20260928",
      "model_slug": "anthropic/claude-sonnet-5.5",
      "model_name": "Anthropic: Claude Sonnet 5.5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.9276092,
      "accuracy_std_dev": 0.002380828544855767,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.025585643434343434,
      "avg_duration_per_task_ms": 19359.4303030303,
      "avg_output_tokens_per_task": 2318.8838383838383,
      "last_run_timestamp": "2026-10-24T08:01:41.235Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-sonnet-5.5-20260928",
      "model_slug": "anthropic/claude-sonnet-5.5",
      "model_name": "Anthropic: Claude Sonnet 5.5",
      "provider_name": "auto-routing",
      "accuracy": 0.9271882499999999,
      "accuracy_std_dev": 0.006560703749408305,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.025096623737373737,
      "avg_duration_per_task_ms": 19787.91414141414,
      "avg_output_tokens_per_task": 2272.02904040404,
      "last_run_timestamp": "2026-10-24T08:11:40.974Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Fireworks",
      "accuracy": 0.9266735746971735,
      "accuracy_std_dev": 0.007620373920907321,
      "run_count": 36,
      "total_questions": 6687,
      "avg_cost_per_task_usd": 0.16851088829071328,
      "avg_duration_per_task_ms": 159036.4706146254,
      "avg_output_tokens_per_task": 9187.432929564828,
      "last_run_timestamp": "2026-10-24T08:50:28.834Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.6-flash-20260721",
      "model_slug": "google/gemini-3.6-flash",
      "model_name": "Google: Gemini 3.6 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.9264870000000001,
      "accuracy_std_dev": 0.004418715424192887,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.05958464141414141,
      "avg_duration_per_task_ms": 58983.43771043771,
      "avg_output_tokens_per_task": 11740.584175084176,
      "last_run_timestamp": "2026-10-01T19:11:04.522Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Decart",
      "accuracy": 0.9259260000000001,
      "accuracy_std_dev": 0.004610436378478734,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.10788488484848485,
      "avg_duration_per_task_ms": 187765.80707070706,
      "avg_output_tokens_per_task": 9499.632323232323,
      "last_run_timestamp": "2026-10-24T08:37:09.424Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.5-flash-20260519",
      "model_slug": "google/gemini-3.5-flash",
      "model_name": "Google: Gemini 3.5 Flash",
      "provider_name": "Google AI Studio",
      "accuracy": 0.925926,
      "accuracy_std_dev": 0.005498287975967322,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.1441699393939394,
      "avg_duration_per_task_ms": 61658.986531986535,
      "avg_output_tokens_per_task": 15894.8367003367,
      "last_run_timestamp": "2026-10-01T19:12:44.398Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.5-flash-20260519",
      "model_slug": "google/gemini-3.5-flash",
      "model_name": "Google: Gemini 3.5 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.9259256666666668,
      "accuracy_std_dev": 0.0036366213562713575,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.1413088872053872,
      "avg_duration_per_task_ms": 82468.60606060606,
      "avg_output_tokens_per_task": 15584.752525252525,
      "last_run_timestamp": "2026-10-01T19:18:40.706Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Relace",
      "accuracy": 0.925795996178344,
      "accuracy_std_dev": 0.007145577022492671,
      "run_count": 8,
      "total_questions": 1570,
      "avg_cost_per_task_usd": 0.1095707592356688,
      "avg_duration_per_task_ms": 200872.2719745223,
      "avg_output_tokens_per_task": 9447.89617834395,
      "last_run_timestamp": "2026-10-24T08:14:36.241Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.5-20260423",
      "model_slug": "openai/gpt-5.5",
      "model_name": "OpenAI: GPT-5.5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.9255440309278352,
      "accuracy_std_dev": 0.0047577793839843644,
      "run_count": 3,
      "total_questions": 582,
      "avg_cost_per_task_usd": 0.3407643092783505,
      "avg_duration_per_task_ms": 132309.46391752578,
      "avg_output_tokens_per_task": 10197.671821305841,
      "last_run_timestamp": "2026-10-01T19:13:06.559Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-sol-20260709",
      "model_slug": "openai/gpt-5.6-sol",
      "model_name": "OpenAI: GPT-5.6 Sol",
      "provider_name": "Azure",
      "accuracy": 0.9248735,
      "accuracy_std_dev": 0.00862277742377712,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.10364464204545455,
      "avg_duration_per_task_ms": 55123.33648989899,
      "avg_output_tokens_per_task": 3517.8642676767677,
      "last_run_timestamp": "2026-10-01T19:42:48.407Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "DeepInfra",
      "accuracy": 0.9247310806451613,
      "accuracy_std_dev": 0.013119500000000006,
      "run_count": 2,
      "total_questions": 372,
      "avg_cost_per_task_usd": 0.005701161290322581,
      "avg_duration_per_task_ms": 222245.45967741936,
      "avg_output_tokens_per_task": 19159.182795698925,
      "last_run_timestamp": "2026-10-24T08:13:25.170Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "x-ai/grok-4.6-20260810",
      "model_slug": "x-ai/grok-4.6",
      "model_name": "SpaceXAI: Grok 4.6",
      "provider_name": "xAI",
      "accuracy": 0.9246375695652174,
      "accuracy_std_dev": 0.006863146440316182,
      "run_count": 7,
      "total_questions": 1380,
      "avg_cost_per_task_usd": 0.12292555942028985,
      "avg_duration_per_task_ms": 340480.1579710145,
      "avg_output_tokens_per_task": 19662.684057971015,
      "last_run_timestamp": "2026-10-01T16:09:31.184Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-sol-20260922",
      "model_slug": "openai/gpt-6-sol",
      "model_name": "OpenAI: GPT-6 Sol",
      "provider_name": "Azure",
      "accuracy": 0.924599700084246,
      "accuracy_std_dev": 0.006819818440072705,
      "run_count": 6,
      "total_questions": 1187,
      "avg_cost_per_task_usd": 0.02751449789385004,
      "avg_duration_per_task_ms": 24957.139848357205,
      "avg_output_tokens_per_task": 2420.7127211457455,
      "last_run_timestamp": "2026-10-01T08:24:52.927Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.5-flash-20260519",
      "model_slug": "google/gemini-3.5-flash",
      "model_name": "Google: Gemini 3.5 Flash",
      "provider_name": "Google",
      "accuracy": 0.9242421666666667,
      "accuracy_std_dev": 0.0048599613653014136,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.14878865656565657,
      "avg_duration_per_task_ms": 85040.50420875421,
      "avg_output_tokens_per_task": 15641.649831649831,
      "last_run_timestamp": "2026-10-01T19:13:26.156Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "unbiased/pareto-26.10-preview",
      "model_slug": "unbiased/pareto-26.10-preview",
      "model_name": "Pareto 26.10 Preview",
      "provider_name": "auto-routing",
      "accuracy": 0.9242420000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0030723434343434344,
      "avg_duration_per_task_ms": 37807.181818181816,
      "avg_output_tokens_per_task": 903.9090909090909,
      "last_run_timestamp": "2026-10-15T14:24:31.863Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-sol-20260709",
      "model_slug": "openai/gpt-5.6-sol",
      "model_name": "OpenAI: GPT-5.6 Sol",
      "provider_name": "OpenAI",
      "accuracy": 0.9234005000000001,
      "accuracy_std_dev": 0.008625346413333228,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.05533751388888889,
      "avg_duration_per_task_ms": 61705.84722222222,
      "avg_output_tokens_per_task": 3564.286616161616,
      "last_run_timestamp": "2026-10-02T23:20:36.509Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "InferenceNet",
      "accuracy": 0.9230042575935954,
      "accuracy_std_dev": 0.028834365810508387,
      "run_count": 22,
      "total_questions": 4247,
      "avg_cost_per_task_usd": 0.12647735083588418,
      "avg_duration_per_task_ms": 173368.8584883447,
      "avg_output_tokens_per_task": 9831.030138921591,
      "last_run_timestamp": "2026-10-24T08:45:15.682Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-sol-20260709",
      "model_slug": "openai/gpt-5.6-sol",
      "model_name": "OpenAI: GPT-5.6 Sol",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.9227054347826086,
      "accuracy_std_dev": 0.0037342422507786015,
      "run_count": 4,
      "total_questions": 759,
      "avg_cost_per_task_usd": 0.09927257048748353,
      "avg_duration_per_task_ms": 52555.73781291173,
      "avg_output_tokens_per_task": 3495.6416337285905,
      "last_run_timestamp": "2026-10-01T19:12:57.119Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Makora",
      "accuracy": 0.9220161548643283,
      "accuracy_std_dev": 0.015489389982329555,
      "run_count": 9,
      "total_questions": 1511,
      "avg_cost_per_task_usd": 0.1115360423560556,
      "avg_duration_per_task_ms": 205913.44606221045,
      "avg_output_tokens_per_task": 8295.92322964924,
      "last_run_timestamp": "2026-10-24T08:11:11.433Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Sail Research",
      "accuracy": 0.9214363999999999,
      "accuracy_std_dev": 0.007595474019001951,
      "run_count": 15,
      "total_questions": 2970,
      "avg_cost_per_task_usd": 0.13534853400673405,
      "avg_duration_per_task_ms": 180051.0377104377,
      "avg_output_tokens_per_task": 11618.102693602694,
      "last_run_timestamp": "2026-10-24T09:13:18.915Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Modal",
      "accuracy": 0.9214090472319009,
      "accuracy_std_dev": 0.0068651349363197224,
      "run_count": 14,
      "total_questions": 2583,
      "avg_cost_per_task_usd": 0.14661487727448702,
      "avg_duration_per_task_ms": 125761.86372435153,
      "avg_output_tokens_per_task": 9628.765001935733,
      "last_run_timestamp": "2026-10-24T08:48:19.650Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.92087525,
      "accuracy_std_dev": 0.004610545920766882,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.14864516792929292,
      "avg_duration_per_task_ms": 124980.99747474748,
      "avg_output_tokens_per_task": 9671.253787878788,
      "last_run_timestamp": "2026-10-24T08:58:50.031Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-sol-20260709",
      "model_slug": "openai/gpt-5.6-sol",
      "model_name": "OpenAI: GPT-5.6 Sol",
      "provider_name": "auto-routing",
      "accuracy": 0.9208750000000001,
      "accuracy_std_dev": 0.004761657064510223,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.06043574579124578,
      "avg_duration_per_task_ms": 62971.38047138047,
      "avg_output_tokens_per_task": 3586.8703703703704,
      "last_run_timestamp": "2026-10-01T19:12:28.573Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-sol-20260922",
      "model_slug": "openai/gpt-6-sol",
      "model_name": "OpenAI: GPT-6 Sol",
      "provider_name": "OpenAI",
      "accuracy": 0.9208749999999999,
      "accuracy_std_dev": 0.002749143987983661,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.025590255892255893,
      "avg_duration_per_task_ms": 33502.22222222222,
      "avg_output_tokens_per_task": 2400.506734006734,
      "last_run_timestamp": "2026-10-02T23:17:42.236Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Alibaba",
      "accuracy": 0.920127,
      "accuracy_std_dev": 0.007058817385999527,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.16989762008978673,
      "avg_duration_per_task_ms": 298475.25589225587,
      "avg_output_tokens_per_task": 9659.37598204265,
      "last_run_timestamp": "2026-10-24T08:46:01.110Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "DigitalOcean",
      "accuracy": 0.9200335,
      "accuracy_std_dev": 0.004208500000000004,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.021749656565656566,
      "avg_duration_per_task_ms": 75214.94191919192,
      "avg_output_tokens_per_task": 20570.64898989899,
      "last_run_timestamp": "2026-10-24T08:29:16.810Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "ParetoAlpha",
      "accuracy": 0.9196215425531915,
      "accuracy_std_dev": 0.0047267721426877685,
      "run_count": 3,
      "total_questions": 564,
      "avg_cost_per_task_usd": 0.14787351063829787,
      "avg_duration_per_task_ms": 388234.25177304965,
      "avg_output_tokens_per_task": 9645.446808510638,
      "last_run_timestamp": "2026-10-24T08:25:42.796Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "BaseTen",
      "accuracy": 0.9194288167597765,
      "accuracy_std_dev": 0.012289026624910344,
      "run_count": 14,
      "total_questions": 2685,
      "avg_cost_per_task_usd": 0.12730921601489759,
      "avg_duration_per_task_ms": 123644.88379888269,
      "avg_output_tokens_per_task": 8486.700558659219,
      "last_run_timestamp": "2026-10-24T08:44:47.181Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "auto-routing",
      "accuracy": 0.9194090050632909,
      "accuracy_std_dev": 0.01100095968888121,
      "run_count": 12,
      "total_questions": 2370,
      "avg_cost_per_task_usd": 0.1314646818565401,
      "avg_duration_per_task_ms": 272126.83670886076,
      "avg_output_tokens_per_task": 10234.781012658228,
      "last_run_timestamp": "2026-10-24T08:16:43.073Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-sol-20260922",
      "model_slug": "openai/gpt-6-sol",
      "model_name": "OpenAI: GPT-6 Sol",
      "provider_name": "auto-routing",
      "accuracy": 0.9191919999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.02528019191919192,
      "avg_duration_per_task_ms": 32477.28787878788,
      "avg_output_tokens_per_task": 2369.5,
      "last_run_timestamp": "2026-10-01T08:24:24.034Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-sol-20260922",
      "model_slug": "openai/gpt-6-sol",
      "model_name": "OpenAI: GPT-6 Sol",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.9191915,
      "accuracy_std_dev": 0.005050500000000013,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.026391550505050505,
      "avg_duration_per_task_ms": 26392.646464646463,
      "avg_output_tokens_per_task": 2240.712121212121,
      "last_run_timestamp": "2026-10-01T08:37:17.903Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "NextBit",
      "accuracy": 0.918981375,
      "accuracy_std_dev": 0.00679620392089402,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.12805398863636364,
      "avg_duration_per_task_ms": 541652.8712121212,
      "avg_output_tokens_per_task": 37116.541666666664,
      "last_run_timestamp": "2026-10-03T09:05:48.324Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "Venice",
      "accuracy": 0.9180774433042584,
      "accuracy_std_dev": 0.01435885589070384,
      "run_count": 10,
      "total_questions": 1949,
      "avg_cost_per_task_usd": 0.24670780708055412,
      "avg_duration_per_task_ms": 414365.66136480245,
      "avg_output_tokens_per_task": 36895.286813750645,
      "last_run_timestamp": "2026-10-24T08:18:32.510Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Wafer",
      "accuracy": 0.9172978750000002,
      "accuracy_std_dev": 0.007642562159339972,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.09133790782828283,
      "avg_duration_per_task_ms": 290431.10353535356,
      "avg_output_tokens_per_task": 24957.965909090908,
      "last_run_timestamp": "2026-10-24T09:01:06.737Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-sonnet-5.5-20260928",
      "model_slug": "anthropic/claude-sonnet-5.5",
      "model_name": "Anthropic: Claude Sonnet 5.5",
      "provider_name": "Azure",
      "accuracy": 0.9171716,
      "accuracy_std_dev": 0.0037495390703391955,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.02583652727272727,
      "avg_duration_per_task_ms": 19190.16717171717,
      "avg_output_tokens_per_task": 2225.3469696969696,
      "last_run_timestamp": "2026-10-24T08:28:10.076Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Moonshot AI",
      "accuracy": 0.9165345411718132,
      "accuracy_std_dev": 0.00728827618680496,
      "run_count": 13,
      "total_questions": 2526,
      "avg_cost_per_task_usd": 0.14153307442596988,
      "avg_duration_per_task_ms": 283509.3226444972,
      "avg_output_tokens_per_task": 9267.701108471892,
      "last_run_timestamp": "2026-10-24T08:44:42.687Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "SiliconFlow",
      "accuracy": 0.9158245,
      "accuracy_std_dev": 0.011784500000000031,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.026372479797979794,
      "avg_duration_per_task_ms": 193752.02525252526,
      "avg_output_tokens_per_task": 21866.449494949495,
      "last_run_timestamp": "2026-10-24T08:46:05.254Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Baidu",
      "accuracy": 0.9152636666666667,
      "accuracy_std_dev": 0.006963814615568124,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.09072644781144781,
      "avg_duration_per_task_ms": 657647.413580247,
      "avg_output_tokens_per_task": 40463.242985409655,
      "last_run_timestamp": "2026-10-24T08:13:28.067Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Baidu",
      "accuracy": 0.9147832881355932,
      "accuracy_std_dev": 0.01218950000000002,
      "run_count": 2,
      "total_questions": 354,
      "avg_cost_per_task_usd": 0.02172540395480226,
      "avg_duration_per_task_ms": 124806.83333333333,
      "avg_output_tokens_per_task": 19165.833333333332,
      "last_run_timestamp": "2026-10-24T08:44:43.907Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Mancer 2",
      "accuracy": 0.9142155073529412,
      "accuracy_std_dev": 0.012753000000000014,
      "run_count": 2,
      "total_questions": 272,
      "avg_cost_per_task_usd": 0.014779014705882354,
      "avg_duration_per_task_ms": 301053.6323529412,
      "avg_output_tokens_per_task": 14660.216911764706,
      "last_run_timestamp": "2026-10-24T08:32:55.335Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Minimax",
      "accuracy": 0.9138823076923077,
      "accuracy_std_dev": 0.013004232597265677,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.03398491258741259,
      "avg_duration_per_task_ms": 227434.108003108,
      "avg_output_tokens_per_task": 28148.493395493395,
      "last_run_timestamp": "2026-10-24T16:21:24.997Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Morph",
      "accuracy": 0.9138354912891987,
      "accuracy_std_dev": 0.029291397755522402,
      "run_count": 24,
      "total_questions": 4592,
      "avg_cost_per_task_usd": 0.14595076415505223,
      "avg_duration_per_task_ms": 388256.5635888502,
      "avg_output_tokens_per_task": 9056.975174216028,
      "last_run_timestamp": "2026-10-24T08:17:56.389Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "GMICloud",
      "accuracy": 0.9132995,
      "accuracy_std_dev": 0.002525500000000014,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.022403944444444445,
      "avg_duration_per_task_ms": 147814.9898989899,
      "avg_output_tokens_per_task": 23595.60606060606,
      "last_run_timestamp": "2026-10-24T08:18:32.509Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "AtlasCloud",
      "accuracy": 0.9131792857142857,
      "accuracy_std_dev": 0.008577350568532835,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.15955438744588746,
      "avg_duration_per_task_ms": 565778.9992784993,
      "avg_output_tokens_per_task": 39975.08946608946,
      "last_run_timestamp": "2026-10-24T08:28:29.525Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "AkashML",
      "accuracy": 0.9131455263157894,
      "accuracy_std_dev": 0.01187053215487832,
      "run_count": 10,
      "total_questions": 1349,
      "avg_cost_per_task_usd": 0.11814906300963676,
      "avg_duration_per_task_ms": 186376.30022238696,
      "avg_output_tokens_per_task": 8107.576723498888,
      "last_run_timestamp": "2026-10-24T08:02:27.411Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Chutes",
      "accuracy": 0.9118396713596139,
      "accuracy_std_dev": 0.015391401769750214,
      "run_count": 13,
      "total_questions": 2486,
      "avg_cost_per_task_usd": 0.1705156729686243,
      "avg_duration_per_task_ms": 504972.3576025744,
      "avg_output_tokens_per_task": 11645.380128720837,
      "last_run_timestamp": "2026-10-24T08:25:10.474Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "GMICloud",
      "accuracy": 0.9117844,
      "accuracy_std_dev": 0.006079420781620585,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.03790721616161616,
      "avg_duration_per_task_ms": 129456.97070707071,
      "avg_output_tokens_per_task": 28484.87575757576,
      "last_run_timestamp": "2026-10-03T09:28:48.247Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Decart",
      "accuracy": 0.911616,
      "accuracy_std_dev": 0.0025249999999999995,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.007887881313131314,
      "avg_duration_per_task_ms": 185118.85353535353,
      "avg_output_tokens_per_task": 21770.37626262626,
      "last_run_timestamp": "2026-10-24T08:29:45.567Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "DekaLLM",
      "accuracy": 0.911111096,
      "accuracy_std_dev": 0.009498497417355022,
      "run_count": 4,
      "total_questions": 750,
      "avg_cost_per_task_usd": 0.136844476,
      "avg_duration_per_task_ms": 321154.29333333333,
      "avg_output_tokens_per_task": 35843.412,
      "last_run_timestamp": "2026-10-24T08:11:11.441Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Alibaba",
      "accuracy": 0.9108864666666666,
      "accuracy_std_dev": 0.015101135925780192,
      "run_count": 15,
      "total_questions": 2970,
      "avg_cost_per_task_usd": 0.13889531952861953,
      "avg_duration_per_task_ms": 583838.6407407407,
      "avg_output_tokens_per_task": 39957.99865319865,
      "last_run_timestamp": "2026-10-24T08:16:23.307Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Alibaba",
      "accuracy": 0.9099324999999999,
      "accuracy_std_dev": 0.005892500000000023,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.02721213888888889,
      "avg_duration_per_task_ms": 211865.40151515152,
      "avg_output_tokens_per_task": 21376.03787878788,
      "last_run_timestamp": "2026-10-24T08:25:08.258Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "AtlasCloud",
      "accuracy": 0.9099324999999999,
      "accuracy_std_dev": 0.004208500000000004,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.02111154797979798,
      "avg_duration_per_task_ms": 145444.96212121213,
      "avg_output_tokens_per_task": 21297.967171717173,
      "last_run_timestamp": "2026-10-24T08:04:44.510Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Nebius",
      "accuracy": 0.909652,
      "accuracy_std_dev": 0.0028613344206273234,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.025303424242424242,
      "avg_duration_per_task_ms": 91659.5925925926,
      "avg_output_tokens_per_task": 20911.968013468013,
      "last_run_timestamp": "2026-10-24T08:24:39.071Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-luna-pro-20260709",
      "model_slug": "openai/gpt-5.6-luna-pro",
      "model_name": "OpenAI: GPT-5.6 Luna Pro",
      "provider_name": "Azure",
      "accuracy": 0.9094274,
      "accuracy_std_dev": 0.002474202384608023,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.07117444141414142,
      "avg_duration_per_task_ms": 135218.17474747475,
      "avg_output_tokens_per_task": 27695.128282828282,
      "last_run_timestamp": "2026-10-01T16:11:23.394Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "NextBit",
      "accuracy": 0.9094272,
      "accuracy_std_dev": 0.00547072746168187,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.025930728282828285,
      "avg_duration_per_task_ms": 177886.11212121212,
      "avg_output_tokens_per_task": 28121.050505050505,
      "last_run_timestamp": "2026-10-03T09:34:35.533Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Reka",
      "accuracy": 0.9090909999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.1305062777777778,
      "avg_duration_per_task_ms": 540279.6767676767,
      "avg_output_tokens_per_task": 38448.40404040404,
      "last_run_timestamp": "2026-10-01T09:14:23.001Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "GLM 5.3 Flash",
      "provider_name": "digitalocean",
      "accuracy": 0.9090909999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0054125,
      "avg_duration_per_task_ms": 235839.91414141413,
      "avg_output_tokens_per_task": 10749.717171717171,
      "last_run_timestamp": "2026-10-01T23:50:45.024Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.7-max-20260520",
      "model_slug": "qwen/qwen3.7-max",
      "model_name": "Qwen: Qwen3.7 Max",
      "provider_name": "AtlasCloud",
      "accuracy": 0.9090909999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.2221867676767677,
      "avg_duration_per_task_ms": 298883.5202020202,
      "avg_output_tokens_per_task": 29352.99494949495,
      "last_run_timestamp": "2026-09-27T00:06:19.757Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "CoreWeave",
      "accuracy": 0.908733465037911,
      "accuracy_std_dev": 0.006850022329140406,
      "run_count": 12,
      "total_questions": 2374,
      "avg_cost_per_task_usd": 0.030451491996630154,
      "avg_duration_per_task_ms": 408677.97219882056,
      "avg_output_tokens_per_task": 31497.24094355518,
      "last_run_timestamp": "2026-10-24T16:30:24.394Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Phala",
      "accuracy": 0.9084910475789473,
      "accuracy_std_dev": 0.04945617807940908,
      "run_count": 12,
      "total_questions": 2375,
      "avg_cost_per_task_usd": 0.14244542484210523,
      "avg_duration_per_task_ms": 331138.315368421,
      "avg_output_tokens_per_task": 10400.368421052632,
      "last_run_timestamp": "2026-10-24T08:48:54.139Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Fireworks",
      "accuracy": 0.9082801312101909,
      "accuracy_std_dev": 0.007583269228208922,
      "run_count": 4,
      "total_questions": 785,
      "avg_cost_per_task_usd": 0.021053796178343947,
      "avg_duration_per_task_ms": 83570.12993630573,
      "avg_output_tokens_per_task": 20735.173248407642,
      "last_run_timestamp": "2026-10-24T08:31:24.564Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Together",
      "accuracy": 0.9080194545454545,
      "accuracy_std_dev": 0.015874287061234574,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.035897662534435264,
      "avg_duration_per_task_ms": 311230.74242424243,
      "avg_output_tokens_per_task": 29833.370982552802,
      "last_run_timestamp": "2026-10-24T16:24:24.501Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "DeepInfra",
      "accuracy": 0.9077097088435375,
      "accuracy_std_dev": 0.01602150486845725,
      "run_count": 4,
      "total_questions": 735,
      "avg_cost_per_task_usd": 0.12614392789115647,
      "avg_duration_per_task_ms": 1486736.511564626,
      "avg_output_tokens_per_task": 50215.49115646259,
      "last_run_timestamp": "2026-10-24T08:43:13.447Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "InferenceNet",
      "accuracy": 0.9074072500000001,
      "accuracy_std_dev": 0.008828586845441347,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.009289530303030303,
      "avg_duration_per_task_ms": 172567.80050505052,
      "avg_output_tokens_per_task": 18481.776515151516,
      "last_run_timestamp": "2026-10-24T08:44:46.341Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Ionstream",
      "accuracy": 0.907407,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.018446676767676768,
      "avg_duration_per_task_ms": 140631.38888888888,
      "avg_output_tokens_per_task": 20189.31818181818,
      "last_run_timestamp": "2026-10-24T08:54:33.388Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Relace",
      "accuracy": 0.907407,
      "accuracy_std_dev": 0.003367000000000009,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.012070199494949497,
      "avg_duration_per_task_ms": 98657.74747474748,
      "avg_output_tokens_per_task": 21841.54797979798,
      "last_run_timestamp": "2026-10-24T08:21:56.264Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "auto-routing",
      "accuracy": 0.907407,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.010421979797979798,
      "avg_duration_per_task_ms": 107076.06565656565,
      "avg_output_tokens_per_task": 18400.97474747475,
      "last_run_timestamp": "2026-10-24T08:14:51.110Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/switchyard",
      "model_slug": "nvidia/switchyard",
      "model_name": "NVIDIA: Switchyard",
      "provider_name": "auto-routing",
      "accuracy": 0.9071381724573594,
      "accuracy_std_dev": 0.03371213456235009,
      "run_count": 8,
      "total_questions": 1583,
      "avg_cost_per_task_usd": 0.015826173089071385,
      "avg_duration_per_task_ms": 64047.12571067593,
      "avg_output_tokens_per_task": 3604.423246999368,
      "last_run_timestamp": "2026-09-30T03:11:21.812Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Together",
      "accuracy": 0.906910242707118,
      "accuracy_std_dev": 0.038342984458019556,
      "run_count": 13,
      "total_questions": 2571,
      "avg_cost_per_task_usd": 0.14211418241929213,
      "avg_duration_per_task_ms": 216494.0859587709,
      "avg_output_tokens_per_task": 9532.819914430183,
      "last_run_timestamp": "2026-10-24T08:46:33.301Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "SiliconFlow",
      "accuracy": 0.9067057499999999,
      "accuracy_std_dev": 0.02181467158101399,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.15476828703703704,
      "avg_duration_per_task_ms": 601020.0946969697,
      "avg_output_tokens_per_task": 39038.01388888889,
      "last_run_timestamp": "2026-10-24T08:50:20.872Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Phala",
      "accuracy": 0.9066696814159291,
      "accuracy_std_dev": 0.02221528527057198,
      "run_count": 11,
      "total_questions": 2034,
      "avg_cost_per_task_usd": 0.150266023107178,
      "avg_duration_per_task_ms": 558942.4459193707,
      "avg_output_tokens_per_task": 37986.727138643066,
      "last_run_timestamp": "2026-10-24T09:27:08.802Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Sail Research",
      "accuracy": 0.9065654999999999,
      "accuracy_std_dev": 0.008117672649842422,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.0029748573232323233,
      "avg_duration_per_task_ms": 218585.40025252526,
      "avg_output_tokens_per_task": 19647.71212121212,
      "last_run_timestamp": "2026-10-24T08:58:07.435Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "GMICloud",
      "accuracy": 0.9065625122807017,
      "accuracy_std_dev": 0.025017422490793098,
      "run_count": 13,
      "total_questions": 2565,
      "avg_cost_per_task_usd": 0.035798040545808964,
      "avg_duration_per_task_ms": 238869.84015594542,
      "avg_output_tokens_per_task": 30114.881091617935,
      "last_run_timestamp": "2026-10-24T16:31:25.833Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Novita",
      "accuracy": 0.9064251666666665,
      "accuracy_std_dev": 0.020964442642377977,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.03777529545454546,
      "avg_duration_per_task_ms": 260085.33417508416,
      "avg_output_tokens_per_task": 31644.78872053872,
      "last_run_timestamp": "2026-10-24T16:42:35.935Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "BaseTen",
      "accuracy": 0.9063117514792899,
      "accuracy_std_dev": 0.0007095000000000296,
      "run_count": 2,
      "total_questions": 338,
      "avg_cost_per_task_usd": 0.0210026775147929,
      "avg_duration_per_task_ms": 63992.53550295858,
      "avg_output_tokens_per_task": 17351.381656804733,
      "last_run_timestamp": "2026-10-24T08:13:28.071Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Baidu",
      "accuracy": 0.9060604000000001,
      "accuracy_std_dev": 0.0032644069966840917,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.02883772424242424,
      "avg_duration_per_task_ms": 166818.5292929293,
      "avg_output_tokens_per_task": 28846.466666666667,
      "last_run_timestamp": "2026-10-24T08:33:28.371Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "StreamLake",
      "accuracy": 0.9059828461538462,
      "accuracy_std_dev": 0.02626261808627545,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.04094654972804972,
      "avg_duration_per_task_ms": 283710.56487956486,
      "avg_output_tokens_per_task": 31884.02874902875,
      "last_run_timestamp": "2026-10-24T16:17:39.031Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "StreamLake",
      "accuracy": 0.9058946753684209,
      "accuracy_std_dev": 0.029872349099497063,
      "run_count": 12,
      "total_questions": 2375,
      "avg_cost_per_task_usd": 0.10017258273684211,
      "avg_duration_per_task_ms": 566875.4896842105,
      "avg_output_tokens_per_task": 39376.88463157895,
      "last_run_timestamp": "2026-10-24T08:54:07.383Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta/muse-spark-open-20261015",
      "model_slug": "meta/muse-spark-open",
      "model_name": "Meta: Muse Spark Open",
      "provider_name": "auto-routing",
      "accuracy": 0.905724,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.06900820707070707,
      "avg_duration_per_task_ms": 101485.72222222222,
      "avg_output_tokens_per_task": 16020.974747474747,
      "last_run_timestamp": "2026-10-24T08:27:04.527Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Relace",
      "accuracy": 0.9057143142857144,
      "accuracy_std_dev": 0.001963571554591265,
      "run_count": 4,
      "total_questions": 700,
      "avg_cost_per_task_usd": 0.18660921714285714,
      "avg_duration_per_task_ms": 435582.91714285716,
      "avg_output_tokens_per_task": 41735.86142857143,
      "last_run_timestamp": "2026-10-24T08:46:32.813Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta/muse-spark-open-20261015",
      "model_slug": "meta/muse-spark-open",
      "model_name": "Meta: Muse Spark Open",
      "provider_name": "Modal",
      "accuracy": 0.9056120000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 196,
      "avg_cost_per_task_usd": 0.06647105102040816,
      "avg_duration_per_task_ms": 64791.01530612245,
      "avg_output_tokens_per_task": 15327.612244897959,
      "last_run_timestamp": "2026-10-24T08:14:36.238Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "InferenceNet",
      "accuracy": 0.905376,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 155,
      "avg_cost_per_task_usd": 0.09219292258064517,
      "avg_duration_per_task_ms": 892432.9419354838,
      "avg_output_tokens_per_task": 36790.02580645162,
      "last_run_timestamp": "2026-10-02T06:44:44.722Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Makora",
      "accuracy": 0.9053030000000001,
      "accuracy_std_dev": 0.004187691070267732,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.026182512626262627,
      "avg_duration_per_task_ms": 138194.00252525252,
      "avg_output_tokens_per_task": 21714.06186868687,
      "last_run_timestamp": "2026-10-24T08:02:23.005Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Alibaba",
      "accuracy": 0.9052427142857143,
      "accuracy_std_dev": 0.007830218017659632,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.03489360533910534,
      "avg_duration_per_task_ms": 228842.85353535353,
      "avg_output_tokens_per_task": 28483.24531024531,
      "last_run_timestamp": "2026-10-24T08:15:15.300Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Morph",
      "accuracy": 0.9051626666666666,
      "accuracy_std_dev": 0.0055553002519115975,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.009504934343434343,
      "avg_duration_per_task_ms": 118639.06902356903,
      "avg_output_tokens_per_task": 16467.956228956227,
      "last_run_timestamp": "2026-10-24T08:45:01.104Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Modal",
      "accuracy": 0.904882,
      "accuracy_std_dev": 0.005892000000000008,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.022279229797979798,
      "avg_duration_per_task_ms": 122775.77777777778,
      "avg_output_tokens_per_task": 22121.87626262626,
      "last_run_timestamp": "2026-10-24T08:33:42.016Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "auto-routing",
      "accuracy": 0.9047062578763128,
      "accuracy_std_dev": 0.0178723281096552,
      "run_count": 13,
      "total_questions": 2571,
      "avg_cost_per_task_usd": 0.03395674834694671,
      "avg_duration_per_task_ms": 313164.32672112016,
      "avg_output_tokens_per_task": 29780.11707506807,
      "last_run_timestamp": "2026-10-24T16:12:49.069Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "ModelRun",
      "accuracy": 0.9044775343283582,
      "accuracy_std_dev": 0.043758500000000033,
      "run_count": 2,
      "total_questions": 335,
      "avg_cost_per_task_usd": 0.01697948358208955,
      "avg_duration_per_task_ms": 145286.75820895523,
      "avg_output_tokens_per_task": 24074.814925373135,
      "last_run_timestamp": "2026-10-24T08:44:22.197Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "SiliconFlow",
      "accuracy": 0.9044612500000001,
      "accuracy_std_dev": 0.009172855468037184,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.03501683964646465,
      "avg_duration_per_task_ms": 199099.8143939394,
      "avg_output_tokens_per_task": 29016.732954545456,
      "last_run_timestamp": "2026-10-24T08:11:11.436Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "StreamLake",
      "accuracy": 0.904461125,
      "accuracy_std_dev": 0.008737580020198659,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.02118971275252525,
      "avg_duration_per_task_ms": 207950.51199494948,
      "avg_output_tokens_per_task": 28934.199494949495,
      "last_run_timestamp": "2026-10-24T08:11:11.332Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "Novita",
      "accuracy": 0.9040400000000001,
      "accuracy_std_dev": 0.004761657064510192,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.03280091414141414,
      "avg_duration_per_task_ms": 814621.6686868687,
      "avg_output_tokens_per_task": 37346.743434343436,
      "last_run_timestamp": "2026-10-24T08:23:31.992Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta/muse-spark-open-20261015",
      "model_slug": "meta/muse-spark-open",
      "model_name": "Meta: Muse Spark Open",
      "provider_name": "Makora",
      "accuracy": 0.9039929892328398,
      "accuracy_std_dev": 0.013615822071674532,
      "run_count": 4,
      "total_questions": 743,
      "avg_cost_per_task_usd": 0.061841156123822336,
      "avg_duration_per_task_ms": 74303.45625841184,
      "avg_output_tokens_per_task": 14413.555854643339,
      "last_run_timestamp": "2026-10-24T08:25:42.800Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Venice",
      "accuracy": 0.9036160645842126,
      "accuracy_std_dev": 0.011349068793175144,
      "run_count": 12,
      "total_questions": 2369,
      "avg_cost_per_task_usd": 0.028033349514563106,
      "avg_duration_per_task_ms": 292905.91599831154,
      "avg_output_tokens_per_task": 23207.665681722247,
      "last_run_timestamp": "2026-10-24T16:03:25.397Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Alibaba",
      "accuracy": 0.9036070050669563,
      "accuracy_std_dev": 0.012155177188745377,
      "run_count": 14,
      "total_questions": 2763,
      "avg_cost_per_task_usd": 0.027082988418385807,
      "avg_duration_per_task_ms": 186339.8273615635,
      "avg_output_tokens_per_task": 25900.92254795512,
      "last_run_timestamp": "2026-10-24T08:48:00.664Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "x-ai/grok-4.6-20260810",
      "model_slug": "x-ai/grok-4.6",
      "model_name": "SpaceXAI: Grok 4.6",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.903508802631579,
      "accuracy_std_dev": 0.0984941779678147,
      "run_count": 3,
      "total_questions": 456,
      "avg_cost_per_task_usd": 0.11285829166666667,
      "avg_duration_per_task_ms": 423954.7916666667,
      "avg_output_tokens_per_task": 16846.184210526317,
      "last_run_timestamp": "2026-10-01T16:04:59.818Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "DeepInfra",
      "accuracy": 0.9033163440134907,
      "accuracy_std_dev": 0.008355089002931883,
      "run_count": 9,
      "total_questions": 1779,
      "avg_cost_per_task_usd": 0.015168129286115795,
      "avg_duration_per_task_ms": 392884.92523889826,
      "avg_output_tokens_per_task": 28918.40921866217,
      "last_run_timestamp": "2026-10-24T08:18:49.230Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Wafer",
      "accuracy": 0.9031985,
      "accuracy_std_dev": 0.007575500000000013,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.008858196969696969,
      "avg_duration_per_task_ms": 108844.08838383839,
      "avg_output_tokens_per_task": 17385.492424242424,
      "last_run_timestamp": "2026-10-24T09:00:40.099Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "DeepInfra",
      "accuracy": 0.9031626720031235,
      "accuracy_std_dev": 0.019975453712141735,
      "run_count": 13,
      "total_questions": 2561,
      "avg_cost_per_task_usd": 0.11988065286997265,
      "avg_duration_per_task_ms": 343221.44474814524,
      "avg_output_tokens_per_task": 8205.648574775478,
      "last_run_timestamp": "2026-10-24T08:49:03.637Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.4-20260305",
      "model_slug": "openai/gpt-5.4",
      "model_name": "OpenAI: GPT-5.4",
      "provider_name": "auto-routing",
      "accuracy": 0.902918,
      "accuracy_std_dev": 0.0007933738084913016,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.14672305387205387,
      "avg_duration_per_task_ms": 133620.78956228957,
      "avg_output_tokens_per_task": 9652.131313131313,
      "last_run_timestamp": "2026-10-01T19:14:23.097Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Novita",
      "accuracy": 0.9026110979153505,
      "accuracy_std_dev": 0.006660054560587301,
      "run_count": 8,
      "total_questions": 1583,
      "avg_cost_per_task_usd": 0.028997156664560955,
      "avg_duration_per_task_ms": 168876.1749842072,
      "avg_output_tokens_per_task": 28895.867340492736,
      "last_run_timestamp": "2026-10-24T08:51:57.001Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta/muse-spark-open-20261015",
      "model_slug": "meta/muse-spark-open",
      "model_name": "Meta: Muse Spark Open",
      "provider_name": "Fireworks",
      "accuracy": 0.9023569999999999,
      "accuracy_std_dev": 0.003367000000000009,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.06824257323232323,
      "avg_duration_per_task_ms": 275518.9696969697,
      "avg_output_tokens_per_task": 15830.017676767677,
      "last_run_timestamp": "2026-10-24T08:04:43.716Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-opus-5.5-20260921",
      "model_slug": "anthropic/claude-opus-5.5",
      "model_name": "Anthropic: Claude Opus 5.5",
      "provider_name": "auto-routing",
      "accuracy": 0.9023568,
      "accuracy_std_dev": 0.005733794394639539,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.05410575151515151,
      "avg_duration_per_task_ms": 31371.09595959596,
      "avg_output_tokens_per_task": 2496.787878787879,
      "last_run_timestamp": "2026-10-24T08:04:44.407Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-opus-5.5-20260921",
      "model_slug": "anthropic/claude-opus-5.5",
      "model_name": "Anthropic: Claude Opus 5.5",
      "provider_name": "Anthropic",
      "accuracy": 0.9023566666666667,
      "accuracy_std_dev": 0.003636621356271357,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.053918992424242425,
      "avg_duration_per_task_ms": 26868.823232323233,
      "avg_output_tokens_per_task": 2480.0361952861954,
      "last_run_timestamp": "2026-10-24T08:14:51.096Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "reflection/beam-1.0-501b-a23b-20261016",
      "model_slug": "reflection/beam-1.0-501b-a23b",
      "model_name": "Reflection: Beam 1.0 501B A23B",
      "provider_name": "Nebius",
      "accuracy": 0.9017956666666668,
      "accuracy_std_dev": 0.003459479954495376,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.014097868686868686,
      "avg_duration_per_task_ms": 101233.50336700337,
      "avg_output_tokens_per_task": 13920.223905723906,
      "last_run_timestamp": "2026-10-24T08:29:16.809Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.8-opus-20260528",
      "model_slug": "anthropic/claude-opus-4.8",
      "model_name": "Anthropic: Claude Opus 4.8",
      "provider_name": "Azure",
      "accuracy": 0.9017254583333333,
      "accuracy_std_dev": 0.0058683477301193655,
      "run_count": 24,
      "total_questions": 4752,
      "avg_cost_per_task_usd": 0.2056731323653199,
      "avg_duration_per_task_ms": 97891.09090909091,
      "avg_output_tokens_per_task": 7837.397516835017,
      "last_run_timestamp": "2026-10-24T08:44:41.805Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "auto-routing",
      "accuracy": 0.9016089931034483,
      "accuracy_std_dev": 0.017035826057784066,
      "run_count": 11,
      "total_questions": 2175,
      "avg_cost_per_task_usd": 0.10899811586206896,
      "avg_duration_per_task_ms": 561470.3995402299,
      "avg_output_tokens_per_task": 37112.96183908046,
      "last_run_timestamp": "2026-10-24T08:01:41.138Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-1017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "Deepseek V4.1 Flash 1017",
      "provider_name": "auto-routing",
      "accuracy": 0.901515,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.006598151515151515,
      "avg_duration_per_task_ms": 72990.44444444444,
      "avg_output_tokens_per_task": 11623.267676767677,
      "last_run_timestamp": "2026-10-23T06:09:54.243Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Together",
      "accuracy": 0.901515,
      "accuracy_std_dev": 0.0025249999999999995,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.022678441919191918,
      "avg_duration_per_task_ms": 49322.41161616162,
      "avg_output_tokens_per_task": 18730.81818181818,
      "last_run_timestamp": "2026-10-24T08:41:23.349Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Cloudflare",
      "accuracy": 0.9009352959183673,
      "accuracy_std_dev": 0.0002650000000000152,
      "run_count": 2,
      "total_questions": 392,
      "avg_cost_per_task_usd": 0.013316757653061225,
      "avg_duration_per_task_ms": 199818.51020408163,
      "avg_output_tokens_per_task": 22103.426020408162,
      "last_run_timestamp": "2026-10-24T08:16:43.075Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.8-opus-20260528",
      "model_slug": "anthropic/claude-opus-4.8",
      "model_name": "Anthropic: Claude Opus 4.8",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.9008097837837836,
      "accuracy_std_dev": 0.0057041149259447295,
      "run_count": 37,
      "total_questions": 7326,
      "avg_cost_per_task_usd": 0.21664884998635,
      "avg_duration_per_task_ms": 93919.68495768496,
      "avg_output_tokens_per_task": 7898.381790881791,
      "last_run_timestamp": "2026-10-24T08:14:03.138Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Google",
      "accuracy": 0.9006734999999999,
      "accuracy_std_dev": 0.005050500000000013,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.020602151515151518,
      "avg_duration_per_task_ms": 398490.0909090909,
      "avg_output_tokens_per_task": 16966.28787878788,
      "last_run_timestamp": "2026-10-24T08:44:49.435Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.9006730000000001,
      "accuracy_std_dev": 0.006733999999999962,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.026410239898989902,
      "avg_duration_per_task_ms": 92596.52777777778,
      "avg_output_tokens_per_task": 21830.75,
      "last_run_timestamp": "2026-10-24T08:32:55.338Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta/muse-spark-open-20261015",
      "model_slug": "meta/muse-spark-open",
      "model_name": "Meta: Muse Spark Open",
      "provider_name": "Together",
      "accuracy": 0.900673,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.07405119191919192,
      "avg_duration_per_task_ms": 94171.00505050505,
      "avg_output_tokens_per_task": 17264.035353535353,
      "last_run_timestamp": "2026-10-24T08:14:51.112Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Parasail",
      "accuracy": 0.9002912013585639,
      "accuracy_std_dev": 0.008613175564538058,
      "run_count": 12,
      "total_questions": 2061,
      "avg_cost_per_task_usd": 0.024704503153808833,
      "avg_duration_per_task_ms": 166852.90538573507,
      "avg_output_tokens_per_task": 20435.77389616691,
      "last_run_timestamp": "2026-10-24T16:26:43.118Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "Xiaomi",
      "accuracy": 0.8999997999999999,
      "accuracy_std_dev": 0.0027351667883330276,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.03378601313131313,
      "avg_duration_per_task_ms": 836485.5171717171,
      "avg_output_tokens_per_task": 38467.88484848485,
      "last_run_timestamp": "2026-10-24T08:37:44.264Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Baidu",
      "accuracy": 0.8999869065420562,
      "accuracy_std_dev": 0.0055303889812583175,
      "run_count": 13,
      "total_questions": 2568,
      "avg_cost_per_task_usd": 0.012645620327102805,
      "avg_duration_per_task_ms": 176614.65965732088,
      "avg_output_tokens_per_task": 26316.880841121496,
      "last_run_timestamp": "2026-10-24T08:45:10.619Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.4-20260305",
      "model_slug": "openai/gpt-5.4",
      "model_name": "OpenAI: GPT-5.4",
      "provider_name": "Azure",
      "accuracy": 0.8999248888888889,
      "accuracy_std_dev": 0.006354339085213833,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.15479229124579125,
      "avg_duration_per_task_ms": 130090.1857463524,
      "avg_output_tokens_per_task": 9547.675645342311,
      "last_run_timestamp": "2026-10-01T19:18:24.536Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-opus-5.5-20260921",
      "model_slug": "anthropic/claude-opus-5.5",
      "model_name": "Anthropic: Claude Opus 5.5",
      "provider_name": "Google",
      "accuracy": 0.8998315555555554,
      "accuracy_std_dev": 0.004532988273904737,
      "run_count": 18,
      "total_questions": 3564,
      "avg_cost_per_task_usd": 0.05740673849607183,
      "avg_duration_per_task_ms": 25170.2278338945,
      "avg_output_tokens_per_task": 2482.2177328843995,
      "last_run_timestamp": "2026-10-24T08:41:23.935Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Parasail",
      "accuracy": 0.8998315,
      "accuracy_std_dev": 0.007575499999999957,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.02710091414141414,
      "avg_duration_per_task_ms": 156920.601010101,
      "avg_output_tokens_per_task": 22433.31818181818,
      "last_run_timestamp": "2026-10-24T08:34:01.458Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "Cohere",
      "accuracy": 0.8998315,
      "accuracy_std_dev": 0.0025254999999999583,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.11557131818181818,
      "avg_duration_per_task_ms": 1030735.8762626263,
      "avg_output_tokens_per_task": 45994.38383838384,
      "last_run_timestamp": "2026-10-24T08:14:03.133Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "DeepSeek",
      "accuracy": 0.8996633999999999,
      "accuracy_std_dev": 0.005792805834826492,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.007343192929292929,
      "avg_duration_per_task_ms": 219337.0404040404,
      "avg_output_tokens_per_task": 25809.762626262625,
      "last_run_timestamp": "2026-08-08T09:30:52.456Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.4-20260305",
      "model_slug": "openai/gpt-5.4",
      "model_name": "OpenAI: GPT-5.4",
      "provider_name": "OpenAI",
      "accuracy": 0.899551,
      "accuracy_std_dev": 0.006198077174952426,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.14109027272727273,
      "avg_duration_per_task_ms": 131303,
      "avg_output_tokens_per_task": 9278.422558922559,
      "last_run_timestamp": "2026-10-01T19:14:55.028Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-luna-pro-20260709",
      "model_slug": "openai/gpt-5.6-luna-pro",
      "model_name": "OpenAI: GPT-5.6 Luna Pro",
      "provider_name": "auto-routing",
      "accuracy": 0.899551,
      "accuracy_std_dev": 0.006198077174952426,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0412280202020202,
      "avg_duration_per_task_ms": 141835.40572390571,
      "avg_output_tokens_per_task": 27211.210437710437,
      "last_run_timestamp": "2026-10-01T16:12:34.647Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "auto-routing",
      "accuracy": 0.8995506666666668,
      "accuracy_std_dev": 0.005722799567422267,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.11900113804713805,
      "avg_duration_per_task_ms": 846691.8215488215,
      "avg_output_tokens_per_task": 47399.26936026936,
      "last_run_timestamp": "2026-10-24T08:02:41.727Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "Novita",
      "accuracy": 0.8994372877637131,
      "accuracy_std_dev": 0.01977995405637288,
      "run_count": 6,
      "total_questions": 1185,
      "avg_cost_per_task_usd": 0.03544181097046414,
      "avg_duration_per_task_ms": 152645.21687763714,
      "avg_output_tokens_per_task": 26597.288607594935,
      "last_run_timestamp": "2026-10-03T09:28:39.969Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-opus-5.5-20260921",
      "model_slug": "anthropic/claude-opus-5.5",
      "model_name": "Anthropic: Claude Opus 5.5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.8993638888888889,
      "accuracy_std_dev": 0.004257353623619148,
      "run_count": 18,
      "total_questions": 3564,
      "avg_cost_per_task_usd": 0.058306676767676764,
      "avg_duration_per_task_ms": 37477.73063973064,
      "avg_output_tokens_per_task": 2520.0830527497196,
      "last_run_timestamp": "2026-10-24T08:44:59.329Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "NextBit",
      "accuracy": 0.8993638888888887,
      "accuracy_std_dev": 0.004029367042778795,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.030411617283950615,
      "avg_duration_per_task_ms": 237553.6509539843,
      "avg_output_tokens_per_task": 26633.1290684624,
      "last_run_timestamp": "2026-10-03T10:19:34.320Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-fable-5.1-20260831",
      "model_slug": "anthropic/claude-fable-5.1",
      "model_name": "Anthropic: Claude Fable 5.1",
      "provider_name": "auto-routing",
      "accuracy": 0.8993264,
      "accuracy_std_dev": 0.006511664598242117,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.13052546969696968,
      "avg_duration_per_task_ms": 30874.333333333332,
      "avg_output_tokens_per_task": 2403.130303030303,
      "last_run_timestamp": "2026-10-01T08:24:30.121Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Novita",
      "accuracy": 0.8992834083438684,
      "accuracy_std_dev": 0.01959300954191881,
      "run_count": 12,
      "total_questions": 2373,
      "avg_cost_per_task_usd": 0.12703601896333752,
      "avg_duration_per_task_ms": 516266.79730299197,
      "avg_output_tokens_per_task": 38334.854193004634,
      "last_run_timestamp": "2026-10-24T08:37:37.787Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-opus-5.5-20260921",
      "model_slug": "anthropic/claude-opus-5.5",
      "model_name": "Anthropic: Claude Opus 5.5",
      "provider_name": "Azure",
      "accuracy": 0.89913,
      "accuracy_std_dev": 0.005890503218458193,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.05769051388888888,
      "avg_duration_per_task_ms": 26119.287037037036,
      "avg_output_tokens_per_task": 2529.9574915824915,
      "last_run_timestamp": "2026-10-24T08:24:00.716Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.8-opus-20260528",
      "model_slug": "anthropic/claude-opus-4.8",
      "model_name": "Anthropic: Claude Opus 4.8",
      "provider_name": "Google",
      "accuracy": 0.8991226842105264,
      "accuracy_std_dev": 0.006159935919460174,
      "run_count": 38,
      "total_questions": 7524,
      "avg_cost_per_task_usd": 0.2224983919457735,
      "avg_duration_per_task_ms": 94191.69205209994,
      "avg_output_tokens_per_task": 8105.584795321637,
      "last_run_timestamp": "2026-10-24T08:20:25.133Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "DeepSeek",
      "accuracy": 0.8989900000000001,
      "accuracy_std_dev": 0.006733999999999962,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.007148643939393939,
      "avg_duration_per_task_ms": 224904.53787878787,
      "avg_output_tokens_per_task": 25295.88888888889,
      "last_run_timestamp": "2026-08-08T10:06:09.471Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta/muse-spark-open-20261015",
      "model_slug": "meta/muse-spark-open",
      "model_name": "Meta: Muse Spark Open",
      "provider_name": "BaseTen",
      "accuracy": 0.8989899999999998,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.07316585858585858,
      "avg_duration_per_task_ms": 97804.9696969697,
      "avg_output_tokens_per_task": 16993.328282828283,
      "last_run_timestamp": "2026-10-24T08:12:37.563Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-luna-pro-20260709",
      "model_slug": "openai/gpt-5.6-luna-pro",
      "model_name": "OpenAI: GPT-5.6 Luna Pro",
      "provider_name": "OpenAI",
      "accuracy": 0.8989897499999999,
      "accuracy_std_dev": 0.005188652155184398,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.03249949747474747,
      "avg_duration_per_task_ms": 137621.66666666666,
      "avg_output_tokens_per_task": 26156.715909090908,
      "last_run_timestamp": "2026-10-02T23:26:10.109Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "ModelRun",
      "accuracy": 0.8987651999999998,
      "accuracy_std_dev": 0.01944778431321504,
      "run_count": 15,
      "total_questions": 2970,
      "avg_cost_per_task_usd": 0.08230536969696971,
      "avg_duration_per_task_ms": 213609.22895622897,
      "avg_output_tokens_per_task": 30143.43771043771,
      "last_run_timestamp": "2026-10-24T16:37:39.077Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Wafer",
      "accuracy": 0.8987310178571428,
      "accuracy_std_dev": 0.016411985541715763,
      "run_count": 11,
      "total_questions": 2128,
      "avg_cost_per_task_usd": 0.015117266917293233,
      "avg_duration_per_task_ms": 247345.46381578947,
      "avg_output_tokens_per_task": 19483.954417293233,
      "last_run_timestamp": "2026-10-24T08:55:07.297Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-luna",
      "model_slug": "openai/gpt-6-luna",
      "model_name": "GPT-6-Luna",
      "provider_name": "auto-routing",
      "accuracy": 0.898477,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 197,
      "avg_cost_per_task_usd": 0.003442197969543147,
      "avg_duration_per_task_ms": 73312.30456852792,
      "avg_output_tokens_per_task": 6831.9390862944165,
      "last_run_timestamp": "2026-10-08T13:29:57.437Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.7-opus-20260416",
      "model_slug": "anthropic/claude-opus-4.7",
      "model_name": "Anthropic: Claude Opus 4.7",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.8983774545454545,
      "accuracy_std_dev": 0.004829955304962313,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.2299064136822773,
      "avg_duration_per_task_ms": 110399.28466483012,
      "avg_output_tokens_per_task": 8983.414600550965,
      "last_run_timestamp": "2026-10-24T08:27:26.668Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8983162,
      "accuracy_std_dev": 0.008080975000580047,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.1253350707070707,
      "avg_duration_per_task_ms": 994044.4858585858,
      "avg_output_tokens_per_task": 49876.97171717172,
      "last_run_timestamp": "2026-10-24T08:12:02.438Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "DigitalOcean",
      "accuracy": 0.8981654060721064,
      "accuracy_std_dev": 0.006690047440041041,
      "run_count": 8,
      "total_questions": 1581,
      "avg_cost_per_task_usd": 0.01891549778621126,
      "avg_duration_per_task_ms": 414943.30676786846,
      "avg_output_tokens_per_task": 22047.87349778621,
      "last_run_timestamp": "2026-10-24T09:25:31.426Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-opus-5-20260723",
      "model_slug": "anthropic/claude-opus-5",
      "model_name": "Anthropic: Claude Opus 5",
      "provider_name": "Azure",
      "accuracy": 0.898164278659612,
      "accuracy_std_dev": 0.007879477322215058,
      "run_count": 26,
      "total_questions": 5103,
      "avg_cost_per_task_usd": 0.12592846874387617,
      "avg_duration_per_task_ms": 58143.77189888301,
      "avg_output_tokens_per_task": 4709.22437781697,
      "last_run_timestamp": "2026-10-24T08:33:28.370Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "SiliconFlow",
      "accuracy": 0.898148,
      "accuracy_std_dev": 0.012012972737836369,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.03667170117845118,
      "avg_duration_per_task_ms": 157797.89814814815,
      "avg_output_tokens_per_task": 27519.53787878788,
      "last_run_timestamp": "2026-10-03T09:28:24.400Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.8-opus-20260528",
      "model_slug": "anthropic/claude-opus-4.8",
      "model_name": "Anthropic: Claude Opus 4.8",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.8977654545454545,
      "accuracy_std_dev": 0.00783374636212555,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.19846787741046834,
      "avg_duration_per_task_ms": 95325.20982552801,
      "avg_output_tokens_per_task": 7728.792011019284,
      "last_run_timestamp": "2026-10-24T08:17:56.390Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Cloudflare",
      "accuracy": 0.8972558314475225,
      "accuracy_std_dev": 0.033281996803894424,
      "run_count": 13,
      "total_questions": 2563,
      "avg_cost_per_task_usd": 0.1455034553257901,
      "avg_duration_per_task_ms": 811067.1357783847,
      "avg_output_tokens_per_task": 36442.67655091689,
      "last_run_timestamp": "2026-10-24T08:25:27.617Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Novita",
      "accuracy": 0.8972168045540795,
      "accuracy_std_dev": 0.016285736546419217,
      "run_count": 16,
      "total_questions": 3162,
      "avg_cost_per_task_usd": 0.024440066413662233,
      "avg_duration_per_task_ms": 243796.87760910817,
      "avg_output_tokens_per_task": 26883.111321948134,
      "last_run_timestamp": "2026-10-24T08:56:13.509Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.8-opus-20260528",
      "model_slug": "anthropic/claude-opus-4.8",
      "model_name": "Anthropic: Claude Opus 4.8",
      "provider_name": "auto-routing",
      "accuracy": 0.8971659166666668,
      "accuracy_std_dev": 0.011261007344359652,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.19972315109427613,
      "avg_duration_per_task_ms": 95894.5180976431,
      "avg_output_tokens_per_task": 7770.697390572391,
      "last_run_timestamp": "2026-10-24T08:28:56.790Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.4-20260305",
      "model_slug": "openai/gpt-5.4",
      "model_name": "OpenAI: GPT-5.4",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.8968568964143425,
      "accuracy_std_dev": 0.005278733293840445,
      "run_count": 4,
      "total_questions": 753,
      "avg_cost_per_task_usd": 0.15754132403718457,
      "avg_duration_per_task_ms": 84099.57768924303,
      "avg_output_tokens_per_task": 9420.674634794157,
      "last_run_timestamp": "2026-10-01T19:12:25.243Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "reflection/beam-1.0-501b-a23b-20261016",
      "model_slug": "reflection/beam-1.0-501b-a23b",
      "model_name": "Reflection: Beam 1.0 501B A23B",
      "provider_name": "CoreWeave",
      "accuracy": 0.8964644999999999,
      "accuracy_std_dev": 0.0025254999999999583,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.027913444444444446,
      "avg_duration_per_task_ms": 54299.99747474748,
      "avg_output_tokens_per_task": 13871.598484848484,
      "last_run_timestamp": "2026-10-24T08:12:49.568Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Venice",
      "accuracy": 0.8962877525832377,
      "accuracy_std_dev": 0.04356588502231736,
      "run_count": 9,
      "total_questions": 1742,
      "avg_cost_per_task_usd": 0.04917616016073478,
      "avg_duration_per_task_ms": 485084.33180252585,
      "avg_output_tokens_per_task": 15776.056831228472,
      "last_run_timestamp": "2026-10-24T16:26:47.562Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-flash-20260826",
      "model_slug": "qwen/qwen3.8-flash",
      "model_name": "Qwen: Qwen3.8 Flash",
      "provider_name": "Alibaba",
      "accuracy": 0.8962022556962025,
      "accuracy_std_dev": 0.0022689999999999655,
      "run_count": 2,
      "total_questions": 395,
      "avg_cost_per_task_usd": 0.020123096202531644,
      "avg_duration_per_task_ms": 653583.2962025317,
      "avg_output_tokens_per_task": 42659.516455696205,
      "last_run_timestamp": "2026-10-10T00:00:04.637Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.8-opus-20260528",
      "model_slug": "anthropic/claude-opus-4.8",
      "model_name": "Anthropic: Claude Opus 4.8",
      "provider_name": "Anthropic",
      "accuracy": 0.8961839333333333,
      "accuracy_std_dev": 0.006639909331375588,
      "run_count": 15,
      "total_questions": 2970,
      "avg_cost_per_task_usd": 0.2064141511784512,
      "avg_duration_per_task_ms": 100220.49057239057,
      "avg_output_tokens_per_task": 8041.884511784512,
      "last_run_timestamp": "2026-10-24T08:30:35.509Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta/muse-spark-open-20261015",
      "model_slug": "meta/muse-spark-open",
      "model_name": "Meta: Muse Spark Open",
      "provider_name": "Nebius",
      "accuracy": 0.8960434999999999,
      "accuracy_std_dev": 0.007996499999999962,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.06795408585858587,
      "avg_duration_per_task_ms": 278542.00757575757,
      "avg_output_tokens_per_task": 15808.588383838383,
      "last_run_timestamp": "2026-10-24T08:15:15.302Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "DeepInfra",
      "accuracy": 0.8959899189640768,
      "accuracy_std_dev": 0.02896400676270464,
      "run_count": 8,
      "total_questions": 1197,
      "avg_cost_per_task_usd": 0.024112938178780283,
      "avg_duration_per_task_ms": 783093.3592314118,
      "avg_output_tokens_per_task": 22333.22807017544,
      "last_run_timestamp": "2026-10-17T16:03:42.308Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "OpenInference",
      "accuracy": 0.895623,
      "accuracy_std_dev": 0.010100999999999971,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.014240813131313132,
      "avg_duration_per_task_ms": 199874.40656565657,
      "avg_output_tokens_per_task": 21243.522727272728,
      "last_run_timestamp": "2026-10-24T08:54:21.561Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Crusoe",
      "accuracy": 0.8956225,
      "accuracy_std_dev": 0.005050499999999958,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.026159166666666667,
      "avg_duration_per_task_ms": 177167.38888888888,
      "avg_output_tokens_per_task": 21649.300505050505,
      "last_run_timestamp": "2026-10-24T08:22:00.564Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.7-opus-20260416",
      "model_slug": "anthropic/claude-opus-4.7",
      "model_name": "Anthropic: Claude Opus 4.7",
      "provider_name": "Google",
      "accuracy": 0.8954932820512822,
      "accuracy_std_dev": 0.00772304896264559,
      "run_count": 39,
      "total_questions": 7722,
      "avg_cost_per_task_usd": 0.2042425423465424,
      "avg_duration_per_task_ms": 70332.41427091428,
      "avg_output_tokens_per_task": 7437.828541828542,
      "last_run_timestamp": "2026-10-24T08:43:08.078Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-opus-5.5-20260921",
      "model_slug": "anthropic/claude-opus-5.5",
      "model_name": "Anthropic: Claude Opus 5.5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.8953421666666667,
      "accuracy_std_dev": 0.0047035083153133565,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.05463887542087542,
      "avg_duration_per_task_ms": 31241.656565656565,
      "avg_output_tokens_per_task": 2527.686026936027,
      "last_run_timestamp": "2026-10-24T08:08:54.167Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8952379124999998,
      "accuracy_std_dev": 0.0099951687737931,
      "run_count": 17,
      "total_questions": 3360,
      "avg_cost_per_task_usd": 0.013780392857142856,
      "avg_duration_per_task_ms": 312284.5139880952,
      "avg_output_tokens_per_task": 27615.769642857143,
      "last_run_timestamp": "2026-10-24T08:49:27.630Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-opus-5-20260723",
      "model_slug": "anthropic/claude-opus-5",
      "model_name": "Anthropic: Claude Opus 5",
      "provider_name": "Anthropic",
      "accuracy": 0.8951046153846154,
      "accuracy_std_dev": 0.007423472104267645,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.12089025990675989,
      "avg_duration_per_task_ms": 58355.058275058276,
      "avg_output_tokens_per_task": 4615.157342657342,
      "last_run_timestamp": "2026-10-24T08:16:38.969Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Cohere",
      "accuracy": 0.8950615,
      "accuracy_std_dev": 0.01045325149973282,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.008697955387205388,
      "avg_duration_per_task_ms": 170461.9739057239,
      "avg_output_tokens_per_task": 30584.97053872054,
      "last_run_timestamp": "2026-10-24T08:08:54.161Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-terra-20260709",
      "model_slug": "openai/gpt-5.6-terra",
      "model_name": "OpenAI: GPT-5.6 Terra",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.8950130708661417,
      "accuracy_std_dev": 0.00891334037763057,
      "run_count": 4,
      "total_questions": 762,
      "avg_cost_per_task_usd": 0.048151682414698166,
      "avg_duration_per_task_ms": 43476.24934383202,
      "avg_output_tokens_per_task": 3503.036745406824,
      "last_run_timestamp": "2026-10-01T19:11:43.444Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "reflection/beam-1.0-501b-a23b-20261016",
      "model_slug": "reflection/beam-1.0-501b-a23b",
      "model_name": "Reflection: Beam 1.0 501B A23B",
      "provider_name": "BaseTen",
      "accuracy": 0.8947810000000002,
      "accuracy_std_dev": 0.0025249999999999995,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.03846228535353535,
      "avg_duration_per_task_ms": 31568.290404040403,
      "avg_output_tokens_per_task": 13846.992424242424,
      "last_run_timestamp": "2026-10-24T08:14:03.134Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "StreamLake",
      "accuracy": 0.894781,
      "accuracy_std_dev": 0.007575999999999972,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.036287184343434344,
      "avg_duration_per_task_ms": 737979.1590909091,
      "avg_output_tokens_per_task": 37423.542929292926,
      "last_run_timestamp": "2026-10-24T08:21:38.159Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8946608571428573,
      "accuracy_std_dev": 0.021330086138923062,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.025403829725829725,
      "avg_duration_per_task_ms": 208889.86363636365,
      "avg_output_tokens_per_task": 27754.472582972583,
      "last_run_timestamp": "2026-10-24T08:01:41.237Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "DeepInfra",
      "accuracy": 0.8945826942242354,
      "accuracy_std_dev": 0.013340195542461545,
      "run_count": 18,
      "total_questions": 3532,
      "avg_cost_per_task_usd": 0.003962120045300114,
      "avg_duration_per_task_ms": 321568.4875424689,
      "avg_output_tokens_per_task": 21723.382785956965,
      "last_run_timestamp": "2026-10-24T08:41:47.029Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-terra-20260709",
      "model_slug": "openai/gpt-5.6-terra",
      "model_name": "OpenAI: GPT-5.6 Terra",
      "provider_name": "Azure",
      "accuracy": 0.8941264444444444,
      "accuracy_std_dev": 0.009636502052221474,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.05178651795735129,
      "avg_duration_per_task_ms": 45445.343434343435,
      "avg_output_tokens_per_task": 3671.5897867564536,
      "last_run_timestamp": "2026-10-01T19:13:06.514Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.7-opus-20260416",
      "model_slug": "anthropic/claude-opus-4.7",
      "model_name": "Anthropic: Claude Opus 4.7",
      "provider_name": "auto-routing",
      "accuracy": 0.8939451085180864,
      "accuracy_std_dev": 0.009458026084140597,
      "run_count": 13,
      "total_questions": 2571,
      "avg_cost_per_task_usd": 0.22913243290548424,
      "avg_duration_per_task_ms": 109689.0805134189,
      "avg_output_tokens_per_task": 8941.389342668223,
      "last_run_timestamp": "2026-10-24T08:44:41.392Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "GMICloud",
      "accuracy": 0.89372875,
      "accuracy_std_dev": 0.019477743681892395,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.02587082891414141,
      "avg_duration_per_task_ms": 157351.7840909091,
      "avg_output_tokens_per_task": 28618.30744949495,
      "last_run_timestamp": "2026-10-24T08:44:42.937Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "Makora",
      "accuracy": 0.8934709999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 194,
      "avg_cost_per_task_usd": 0.11686864948453608,
      "avg_duration_per_task_ms": 605687.7474226804,
      "avg_output_tokens_per_task": 46496.701030927834,
      "last_run_timestamp": "2026-10-13T03:00:07.016Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Phala",
      "accuracy": 0.8932483811659193,
      "accuracy_std_dev": 0.017336749323389345,
      "run_count": 7,
      "total_questions": 1338,
      "avg_cost_per_task_usd": 0.03355175710014948,
      "avg_duration_per_task_ms": 240320.3542600897,
      "avg_output_tokens_per_task": 32642.40284005979,
      "last_run_timestamp": "2026-10-24T08:41:23.835Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "GMICloud",
      "accuracy": 0.893246,
      "accuracy_std_dev": 0.014262311738201734,
      "run_count": 17,
      "total_questions": 3366,
      "avg_cost_per_task_usd": 0.015338635769459302,
      "avg_duration_per_task_ms": 208853.27035056447,
      "avg_output_tokens_per_task": 25149.005347593582,
      "last_run_timestamp": "2026-10-24T08:42:04.982Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "Tencent",
      "accuracy": 0.8930975,
      "accuracy_std_dev": 0.009560420192125432,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.11553578787878788,
      "avg_duration_per_task_ms": 625203.1957070708,
      "avg_output_tokens_per_task": 46042.48863636364,
      "last_run_timestamp": "2026-10-24T08:01:41.143Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Fireworks",
      "accuracy": 0.8928570861678005,
      "accuracy_std_dev": 0.015340394005370288,
      "run_count": 5,
      "total_questions": 882,
      "avg_cost_per_task_usd": 0.12683532766439912,
      "avg_duration_per_task_ms": 285532.1031746032,
      "avg_output_tokens_per_task": 28366.77097505669,
      "last_run_timestamp": "2026-09-19T09:14:09.704Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-terra-20260709",
      "model_slug": "openai/gpt-5.6-terra",
      "model_name": "OpenAI: GPT-5.6 Terra",
      "provider_name": "OpenAI",
      "accuracy": 0.89267675,
      "accuracy_std_dev": 0.006979218773437292,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.037834488636363635,
      "avg_duration_per_task_ms": 50710.07575757576,
      "avg_output_tokens_per_task": 3480.1502525252527,
      "last_run_timestamp": "2026-10-02T23:21:06.556Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "BaseTen",
      "accuracy": 0.892567903763697,
      "accuracy_std_dev": 0.02022831814050264,
      "run_count": 11,
      "total_questions": 2099,
      "avg_cost_per_task_usd": 0.15797696188661267,
      "avg_duration_per_task_ms": 338559.5883754169,
      "avg_output_tokens_per_task": 39638.616484040016,
      "last_run_timestamp": "2026-09-19T08:51:08.108Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.7-opus-20260416",
      "model_slug": "anthropic/claude-opus-4.7",
      "model_name": "Anthropic: Claude Opus 4.7",
      "provider_name": "Anthropic",
      "accuracy": 0.892262208728653,
      "accuracy_std_dev": 0.011301549411889275,
      "run_count": 16,
      "total_questions": 3162,
      "avg_cost_per_task_usd": 0.20793211669829223,
      "avg_duration_per_task_ms": 100118.31277672359,
      "avg_output_tokens_per_task": 8089.089500316256,
      "last_run_timestamp": "2026-10-24T08:05:46.408Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "Novita",
      "accuracy": 0.8922557999999999,
      "accuracy_std_dev": 0.007299435934371887,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.11377392424242423,
      "avg_duration_per_task_ms": 848578.0848484848,
      "avg_output_tokens_per_task": 45335.65555555555,
      "last_run_timestamp": "2026-10-24T08:01:41.133Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Wafer",
      "accuracy": 0.8922132346791637,
      "accuracy_std_dev": 0.039246562470616454,
      "run_count": 15,
      "total_questions": 2774,
      "avg_cost_per_task_usd": 0.35474507606344624,
      "avg_duration_per_task_ms": 458796.4985580389,
      "avg_output_tokens_per_task": 57478.27541456381,
      "last_run_timestamp": "2026-10-24T16:31:04.415Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.8919751666666667,
      "accuracy_std_dev": 0.008661714291768222,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.024034642255892257,
      "avg_duration_per_task_ms": 245030.94696969696,
      "avg_output_tokens_per_task": 30463.32239057239,
      "last_run_timestamp": "2026-10-24T08:01:41.135Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Sail Research",
      "accuracy": 0.8916709261744967,
      "accuracy_std_dev": 0.019973452544815567,
      "run_count": 10,
      "total_questions": 1937,
      "avg_cost_per_task_usd": 0.13092793236964376,
      "avg_duration_per_task_ms": 553822.8373773877,
      "avg_output_tokens_per_task": 42781.386164171396,
      "last_run_timestamp": "2026-10-24T09:11:22.395Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.7-opus-20260416",
      "model_slug": "anthropic/claude-opus-4.7",
      "model_name": "Anthropic: Claude Opus 4.7",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.8916082307692307,
      "accuracy_std_dev": 0.008673677213747568,
      "run_count": 26,
      "total_questions": 5148,
      "avg_cost_per_task_usd": 0.24150008061383055,
      "avg_duration_per_task_ms": 84772.77195027196,
      "avg_output_tokens_per_task": 8978.732711732711,
      "last_run_timestamp": "2026-10-24T08:14:36.267Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "DeepInfra",
      "accuracy": 0.8915753430656934,
      "accuracy_std_dev": 0.004887565515332395,
      "run_count": 6,
      "total_questions": 1096,
      "avg_cost_per_task_usd": 0.03103391879562044,
      "avg_duration_per_task_ms": 795798.2390510949,
      "avg_output_tokens_per_task": 36089.42244525548,
      "last_run_timestamp": "2026-10-24T08:07:59.343Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Morph",
      "accuracy": 0.8915342857142857,
      "accuracy_std_dev": 0.010905197538189691,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.011785490620490622,
      "avg_duration_per_task_ms": 284433.1883116883,
      "avg_output_tokens_per_task": 21375.483405483406,
      "last_run_timestamp": "2026-10-24T08:45:15.768Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Parasail",
      "accuracy": 0.8914141,
      "accuracy_std_dev": 0.016939939187907378,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.1456811595959596,
      "avg_duration_per_task_ms": 619546.0101010101,
      "avg_output_tokens_per_task": 36709.93131313131,
      "last_run_timestamp": "2026-10-17T08:18:00.876Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Novita",
      "accuracy": 0.891414,
      "accuracy_std_dev": 0.01094299999999998,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.007142083333333334,
      "avg_duration_per_task_ms": 125847.88383838384,
      "avg_output_tokens_per_task": 11714.222222222223,
      "last_run_timestamp": "2026-10-24T08:16:38.970Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "reflection/beam-1.0-501b-a23b-20261016",
      "model_slug": "reflection/beam-1.0-501b-a23b",
      "model_name": "Reflection: Beam 1.0 501B A23B",
      "provider_name": "DigitalOcean",
      "accuracy": 0.891414,
      "accuracy_std_dev": 0.0025249999999999995,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.016427454545454544,
      "avg_duration_per_task_ms": 96439.17424242424,
      "avg_output_tokens_per_task": 13618.247474747475,
      "last_run_timestamp": "2026-10-24T08:14:51.086Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "GMICloud",
      "accuracy": 0.8914139999999999,
      "accuracy_std_dev": 0.010479783311214021,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.005115402777777778,
      "avg_duration_per_task_ms": 224009.38825757575,
      "avg_output_tokens_per_task": 28058.227904040403,
      "last_run_timestamp": "2026-10-24T16:05:36.256Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8907934361904761,
      "accuracy_std_dev": 0.020177999596946557,
      "run_count": 8,
      "total_questions": 1575,
      "avg_cost_per_task_usd": 0.18863606476190478,
      "avg_duration_per_task_ms": 1541820.2831746032,
      "avg_output_tokens_per_task": 55365.57968253968,
      "last_run_timestamp": "2026-10-24T16:20:08.208Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2-thinking-20251106",
      "model_slug": "moonshotai/kimi-k2-thinking",
      "model_name": "MoonshotAI: Kimi K2 Thinking",
      "provider_name": "Novita",
      "accuracy": 0.8907719999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 177,
      "avg_cost_per_task_usd": 0.10733528813559323,
      "avg_duration_per_task_ms": 969537.395480226,
      "avg_output_tokens_per_task": 42767.57627118644,
      "last_run_timestamp": "2026-08-01T17:46:48.312Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3-flash-preview-20251217",
      "model_slug": "google/gemini-3-flash-preview",
      "model_name": "Google: Gemini 3 Flash Preview",
      "provider_name": "Google",
      "accuracy": 0.8905723333333333,
      "accuracy_std_dev": 0.021339306148253502,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.13216584343434343,
      "avg_duration_per_task_ms": 251067.54713804714,
      "avg_output_tokens_per_task": 43934.003367003366,
      "last_run_timestamp": "2026-10-01T19:18:35.588Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "amazon/nova-micro-v1",
      "model_slug": "amazon/nova-micro-v1",
      "model_name": "Amazon: Nova Micro 1.0",
      "provider_name": "Anthropic",
      "accuracy": 0.8900113333333334,
      "accuracy_std_dev": 0.005722799567422281,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.19800758417508418,
      "avg_duration_per_task_ms": 96141.14141414141,
      "avg_output_tokens_per_task": 7695.558922558923,
      "last_run_timestamp": "2026-10-01T19:11:18.427Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-luna-pro-20260922",
      "model_slug": "openai/gpt-6-luna-pro",
      "model_name": "OpenAI: GPT-6 Luna Pro",
      "provider_name": "OpenAI",
      "accuracy": 0.8900113333333334,
      "accuracy_std_dev": 0.0015872190215034077,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.008230473063973063,
      "avg_duration_per_task_ms": 80128.80303030302,
      "avg_output_tokens_per_task": 12935.998316498317,
      "last_run_timestamp": "2026-10-02T23:23:43.005Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "amazon/nova-micro-v1",
      "model_slug": "amazon/nova-micro-v1",
      "model_name": "Amazon: Nova Micro 1.0",
      "provider_name": "auto-routing",
      "accuracy": 0.8900110000000001,
      "accuracy_std_dev": 0.0055552329084086844,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.2164912121212121,
      "avg_duration_per_task_ms": 104660.01851851853,
      "avg_output_tokens_per_task": 8434.90404040404,
      "last_run_timestamp": "2026-10-01T19:10:42.213Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-opus-5-20260723",
      "model_slug": "anthropic/claude-opus-5",
      "model_name": "Anthropic: Claude Opus 5",
      "provider_name": "Google",
      "accuracy": 0.8898858525165823,
      "accuracy_std_dev": 0.009105600456959892,
      "run_count": 39,
      "total_questions": 7689,
      "avg_cost_per_task_usd": 0.133101505137209,
      "avg_duration_per_task_ms": 59235.32214852387,
      "avg_output_tokens_per_task": 4759.289244375082,
      "last_run_timestamp": "2026-10-24T08:16:43.073Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.7-opus-20260416",
      "model_slug": "anthropic/claude-opus-4.7",
      "model_name": "Anthropic: Claude Opus 4.7",
      "provider_name": "Azure",
      "accuracy": 0.8897974953271028,
      "accuracy_std_dev": 0.009445547545688906,
      "run_count": 13,
      "total_questions": 2568,
      "avg_cost_per_task_usd": 0.20473812188473522,
      "avg_duration_per_task_ms": 99561.84657320872,
      "avg_output_tokens_per_task": 7962.808411214954,
      "last_run_timestamp": "2026-10-24T08:44:29.191Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.889547806779661,
      "accuracy_std_dev": 0.017603851112753716,
      "run_count": 3,
      "total_questions": 590,
      "avg_cost_per_task_usd": 0.03740084406779662,
      "avg_duration_per_task_ms": 166568.35593220338,
      "avg_output_tokens_per_task": 31004.76779661017,
      "last_run_timestamp": "2026-10-24T08:41:52.401Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-fable-5.1-20260831",
      "model_slug": "anthropic/claude-fable-5.1",
      "model_name": "Anthropic: Claude Fable 5.1",
      "provider_name": "Anthropic",
      "accuracy": 0.8894498333333334,
      "accuracy_std_dev": 0.02853141346198763,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.136280962962963,
      "avg_duration_per_task_ms": 40331.19781144781,
      "avg_output_tokens_per_task": 2516.2390572390573,
      "last_run_timestamp": "2026-10-01T08:24:17.840Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-luna-pro-20260922",
      "model_slug": "openai/gpt-6-luna-pro",
      "model_name": "OpenAI: GPT-6 Luna Pro",
      "provider_name": "Azure",
      "accuracy": 0.8894498333333334,
      "accuracy_std_dev": 0.009896244159893304,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.008765856902356902,
      "avg_duration_per_task_ms": 59664.589225589225,
      "avg_output_tokens_per_task": 12901.691919191919,
      "last_run_timestamp": "2026-10-01T08:29:53.989Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Reka",
      "accuracy": 0.8887654163757536,
      "accuracy_std_dev": 0.008192069777191544,
      "run_count": 16,
      "total_questions": 3151,
      "avg_cost_per_task_usd": 0.01559116534433513,
      "avg_duration_per_task_ms": 197619.02634084417,
      "avg_output_tokens_per_task": 26956.708981275784,
      "last_run_timestamp": "2026-10-24T08:24:10.667Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Ionstream",
      "accuracy": 0.8886482857142856,
      "accuracy_std_dev": 0.013807459233682208,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.1021348318903319,
      "avg_duration_per_task_ms": 444900.012987013,
      "avg_output_tokens_per_task": 35591.61255411255,
      "last_run_timestamp": "2026-10-24T08:42:40.868Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Relace",
      "accuracy": 0.8885464319465846,
      "accuracy_std_dev": 0.029017071552622252,
      "run_count": 10,
      "total_questions": 1947,
      "avg_cost_per_task_usd": 0.009370765279917823,
      "avg_duration_per_task_ms": 312934.7123780175,
      "avg_output_tokens_per_task": 26808.334874165383,
      "last_run_timestamp": "2026-10-24T08:13:41.086Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Novita",
      "accuracy": 0.888467875,
      "accuracy_std_dev": 0.02891189332021296,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.10564124999999999,
      "avg_duration_per_task_ms": 446348.9172979798,
      "avg_output_tokens_per_task": 36610.40214646464,
      "last_run_timestamp": "2026-10-24T16:12:09.455Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Phala",
      "accuracy": 0.8880469999999999,
      "accuracy_std_dev": 0.01094299999999998,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.02377390909090909,
      "avg_duration_per_task_ms": 131245.66919191918,
      "avg_output_tokens_per_task": 19627.237373737375,
      "last_run_timestamp": "2026-10-24T08:41:24.320Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-terra-20260709",
      "model_slug": "openai/gpt-5.6-terra",
      "model_name": "OpenAI: GPT-5.6 Terra",
      "provider_name": "auto-routing",
      "accuracy": 0.8877663333333334,
      "accuracy_std_dev": 0.005555502277522277,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.03440870875420875,
      "avg_duration_per_task_ms": 50195.98148148148,
      "avg_output_tokens_per_task": 3309.2491582491584,
      "last_run_timestamp": "2026-10-01T19:13:07.806Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "NextBit",
      "accuracy": 0.887537924012158,
      "accuracy_std_dev": 0.020250680116973837,
      "run_count": 5,
      "total_questions": 987,
      "avg_cost_per_task_usd": 0.12033997061803446,
      "avg_duration_per_task_ms": 580755.4113475177,
      "avg_output_tokens_per_task": 34357.5663627153,
      "last_run_timestamp": "2026-10-03T16:44:56.199Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-fable-5.1-20260831",
      "model_slug": "anthropic/claude-fable-5.1",
      "model_name": "Anthropic: Claude Fable 5.1",
      "provider_name": "Azure",
      "accuracy": 0.8872051666666666,
      "accuracy_std_dev": 0.024761493125527695,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.1434989882154882,
      "avg_duration_per_task_ms": 42121.643939393936,
      "avg_output_tokens_per_task": 2660.465488215488,
      "last_run_timestamp": "2026-10-01T08:24:34.535Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Phala",
      "accuracy": 0.8869462307692308,
      "accuracy_std_dev": 0.01086717272801863,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.025130035353535356,
      "avg_duration_per_task_ms": 337121.5742035742,
      "avg_output_tokens_per_task": 26419.156565656565,
      "last_run_timestamp": "2026-10-24T08:11:11.359Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "auto-routing",
      "accuracy": 0.8867541405460061,
      "accuracy_std_dev": 0.003543823280018365,
      "run_count": 5,
      "total_questions": 989,
      "avg_cost_per_task_usd": 0.03347630637007078,
      "avg_duration_per_task_ms": 767457.7462082912,
      "avg_output_tokens_per_task": 38389.49544994944,
      "last_run_timestamp": "2026-10-24T08:26:00.565Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Inceptron",
      "accuracy": 0.8867506138022562,
      "accuracy_std_dev": 0.020605395430793746,
      "run_count": 8,
      "total_questions": 1507,
      "avg_cost_per_task_usd": 0.1588602813536828,
      "avg_duration_per_task_ms": 705608.4233576643,
      "avg_output_tokens_per_task": 51029.565361645655,
      "last_run_timestamp": "2026-10-24T16:27:11.137Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Inceptron",
      "accuracy": 0.8867197848605577,
      "accuracy_std_dev": 0.007952474956944445,
      "run_count": 13,
      "total_questions": 2510,
      "avg_cost_per_task_usd": 0.09700779960159363,
      "avg_duration_per_task_ms": 461798.00398406375,
      "avg_output_tokens_per_task": 28989.141832669324,
      "last_run_timestamp": "2026-10-24T16:06:30.904Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Fireworks",
      "accuracy": 0.8867051248554914,
      "accuracy_std_dev": 0.021208571037200977,
      "run_count": 5,
      "total_questions": 865,
      "avg_cost_per_task_usd": 0.16137210289017342,
      "avg_duration_per_task_ms": 430828.78150289017,
      "avg_output_tokens_per_task": 40500.787283237,
      "last_run_timestamp": "2026-09-19T08:53:57.162Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8866872307692307,
      "accuracy_std_dev": 0.01902814900716969,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.027383982128982128,
      "avg_duration_per_task_ms": 210380.49145299144,
      "avg_output_tokens_per_task": 25687.588966588966,
      "last_run_timestamp": "2026-10-24T08:19:41.879Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Wafer",
      "accuracy": 0.8865963923784493,
      "accuracy_std_dev": 0.13919925741876826,
      "run_count": 21,
      "total_questions": 3805,
      "avg_cost_per_task_usd": 0.15561201261498026,
      "avg_duration_per_task_ms": 231786.43758212877,
      "avg_output_tokens_per_task": 10794.42838370565,
      "last_run_timestamp": "2026-10-24T08:53:46.121Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "GMICloud",
      "accuracy": 0.8863635000000001,
      "accuracy_std_dev": 0.005892500000000023,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.006986138888888889,
      "avg_duration_per_task_ms": 141699.11616161617,
      "avg_output_tokens_per_task": 11730.31313131313,
      "last_run_timestamp": "2026-10-24T08:29:14.214Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Parasail",
      "accuracy": 0.8863540053865651,
      "accuracy_std_dev": 0.012359654596078922,
      "run_count": 16,
      "total_questions": 3156,
      "avg_cost_per_task_usd": 0.007512178707224334,
      "avg_duration_per_task_ms": 424002.4524714829,
      "avg_output_tokens_per_task": 26773.751901140684,
      "last_run_timestamp": "2026-10-24T08:50:11.572Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "InferenceNet",
      "accuracy": 0.8859760038834952,
      "accuracy_std_dev": 0.023393307207604786,
      "run_count": 8,
      "total_questions": 1545,
      "avg_cost_per_task_usd": 0.02020688673139159,
      "avg_duration_per_task_ms": 264412.80194174754,
      "avg_output_tokens_per_task": 30725.21488673139,
      "last_run_timestamp": "2026-10-24T09:10:08.243Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "GMICloud",
      "accuracy": 0.8858024166666666,
      "accuracy_std_dev": 0.035974938034364815,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.1124567276936027,
      "avg_duration_per_task_ms": 508472.5484006734,
      "avg_output_tokens_per_task": 36070.853114478115,
      "last_run_timestamp": "2026-10-24T08:17:53.304Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Novita",
      "accuracy": 0.885662,
      "accuracy_std_dev": 0.014374902133000174,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.11714179755892257,
      "avg_duration_per_task_ms": 572173.5332491583,
      "avg_output_tokens_per_task": 30582.406986531987,
      "last_run_timestamp": "2026-10-24T16:30:32.666Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Sail Research",
      "accuracy": 0.8856471629778674,
      "accuracy_std_dev": 0.009013068317726206,
      "run_count": 8,
      "total_questions": 1491,
      "avg_cost_per_task_usd": 0.010023977867203217,
      "avg_duration_per_task_ms": 270739.96177062375,
      "avg_output_tokens_per_task": 28220.264922870556,
      "last_run_timestamp": "2026-10-24T09:01:43.485Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3-flash-preview-20251217",
      "model_slug": "google/gemini-3-flash-preview",
      "model_name": "Google: Gemini 3 Flash Preview",
      "provider_name": "Google AI Studio",
      "accuracy": 0.8855220000000001,
      "accuracy_std_dev": 0.012598160421267844,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.1415003787878788,
      "avg_duration_per_task_ms": 184472.2676767677,
      "avg_output_tokens_per_task": 47043.42424242424,
      "last_run_timestamp": "2026-10-01T19:12:30.168Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "sakana/fugu-max",
      "model_slug": "sakana/fugu-max",
      "model_name": "Fugu Max",
      "provider_name": "auto-routing",
      "accuracy": 0.885522,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.07555469191919191,
      "avg_duration_per_task_ms": 225091.94949494948,
      "avg_output_tokens_per_task": 12279.550505050505,
      "last_run_timestamp": "2026-09-15T01:07:47.957Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Minimax",
      "accuracy": 0.885522,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.007081343434343435,
      "avg_duration_per_task_ms": 121564.28282828283,
      "avg_output_tokens_per_task": 11603.333333333334,
      "last_run_timestamp": "2026-10-24T08:10:23.026Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-flash-20260826",
      "model_slug": "qwen/qwen3.8-flash",
      "model_name": "Qwen: Qwen3.8 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.885522,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.02146120202020202,
      "avg_duration_per_task_ms": 674351.9393939395,
      "avg_output_tokens_per_task": 45464.17676767677,
      "last_run_timestamp": "2026-10-01T08:31:57.634Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-luna-20260709",
      "model_slug": "openai/gpt-5.6-luna",
      "model_name": "OpenAI: GPT-5.6 Luna",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.8853190481927712,
      "accuracy_std_dev": 0.005795847495189989,
      "run_count": 4,
      "total_questions": 747,
      "avg_cost_per_task_usd": 0.0097919718875502,
      "avg_duration_per_task_ms": 50542.31994645247,
      "avg_output_tokens_per_task": 7272.5394912985275,
      "last_run_timestamp": "2026-10-01T19:09:12.892Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "DekaLLM",
      "accuracy": 0.8850689656218401,
      "accuracy_std_dev": 0.014411932672615443,
      "run_count": 5,
      "total_questions": 989,
      "avg_cost_per_task_usd": 0.026149528816986855,
      "avg_duration_per_task_ms": 355797.47320525785,
      "avg_output_tokens_per_task": 26995.41152679474,
      "last_run_timestamp": "2026-10-10T08:22:45.418Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Moonshot AI",
      "accuracy": 0.884890375,
      "accuracy_std_dev": 0.021787089743340537,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.22976492803030302,
      "avg_duration_per_task_ms": 949653.6900252525,
      "avg_output_tokens_per_task": 57582.06123737374,
      "last_run_timestamp": "2026-10-24T16:07:10.471Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "reflection/beam-1.0-501b-a23b-20261016",
      "model_slug": "reflection/beam-1.0-501b-a23b",
      "model_name": "Reflection: Beam 1.0 501B A23B",
      "provider_name": "Together",
      "accuracy": 0.8846800000000001,
      "accuracy_std_dev": 0.0025249999999999995,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.01685389393939394,
      "avg_duration_per_task_ms": 60063.59848484849,
      "avg_output_tokens_per_task": 13866.29292929293,
      "last_run_timestamp": "2026-10-24T08:11:43.555Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "GMICloud",
      "accuracy": 0.884566476280835,
      "accuracy_std_dev": 0.018280900654354407,
      "run_count": 8,
      "total_questions": 1581,
      "avg_cost_per_task_usd": 0.06262646173308033,
      "avg_duration_per_task_ms": 529742.5496521189,
      "avg_output_tokens_per_task": 35635.60657811512,
      "last_run_timestamp": "2026-10-24T16:07:10.471Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Moonshot AI",
      "accuracy": 0.8842948996342949,
      "accuracy_std_dev": 0.013500101124006438,
      "run_count": 25,
      "total_questions": 4922,
      "avg_cost_per_task_usd": 0.19286014851686306,
      "avg_duration_per_task_ms": 362955.4989841528,
      "avg_output_tokens_per_task": 31469.974603819584,
      "last_run_timestamp": "2026-10-24T16:13:46.475Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-luna-20260709",
      "model_slug": "openai/gpt-5.6-luna",
      "model_name": "OpenAI: GPT-5.6 Luna",
      "provider_name": "Azure",
      "accuracy": 0.884269595505618,
      "accuracy_std_dev": 0.010583520080008851,
      "run_count": 9,
      "total_questions": 1780,
      "avg_cost_per_task_usd": 0.018860816853932583,
      "avg_duration_per_task_ms": 74171.62584269662,
      "avg_output_tokens_per_task": 7862.741573033708,
      "last_run_timestamp": "2026-10-01T19:12:42.994Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3-flash-preview-20251217",
      "model_slug": "google/gemini-3-flash-preview",
      "model_name": "Google: Gemini 3 Flash Preview",
      "provider_name": "auto-routing",
      "accuracy": 0.8838383333333335,
      "accuracy_std_dev": 0.001374572014200141,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.13180945791245793,
      "avg_duration_per_task_ms": 248970.3754208754,
      "avg_output_tokens_per_task": 43817.097643097644,
      "last_run_timestamp": "2026-10-01T19:15:31.426Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-luna-pro-20260922",
      "model_slug": "openai/gpt-6-luna-pro",
      "model_name": "OpenAI: GPT-6 Luna Pro",
      "provider_name": "auto-routing",
      "accuracy": 0.883838,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.008301368686868687,
      "avg_duration_per_task_ms": 72985.27777777778,
      "avg_output_tokens_per_task": 13098.969696969696,
      "last_run_timestamp": "2026-10-01T08:24:38.949Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "reflection/beam-1.0-501b-a23b-20261016",
      "model_slug": "reflection/beam-1.0-501b-a23b",
      "model_name": "Reflection: Beam 1.0 501B A23B",
      "provider_name": "auto-routing",
      "accuracy": 0.883838,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.014282979797979798,
      "avg_duration_per_task_ms": 55075.11111111111,
      "avg_output_tokens_per_task": 13355.585858585859,
      "last_run_timestamp": "2026-10-24T08:02:32.162Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Relace",
      "accuracy": 0.8838218675352877,
      "accuracy_std_dev": 0.0032754605783004037,
      "run_count": 5,
      "total_questions": 921,
      "avg_cost_per_task_usd": 0.10965064820846905,
      "avg_duration_per_task_ms": 399019.9750271444,
      "avg_output_tokens_per_task": 36828.9565689468,
      "last_run_timestamp": "2026-10-24T16:03:25.471Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Wafer",
      "accuracy": 0.8836519487704917,
      "accuracy_std_dev": 0.06806651408353726,
      "run_count": 16,
      "total_questions": 2928,
      "avg_cost_per_task_usd": 0.009323294740437157,
      "avg_duration_per_task_ms": 150365.9463797814,
      "avg_output_tokens_per_task": 24280.998292349726,
      "last_run_timestamp": "2026-10-24T08:48:19.065Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Nebius",
      "accuracy": 0.88341725,
      "accuracy_std_dev": 0.003834202809907171,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.03130394823232323,
      "avg_duration_per_task_ms": 201934.6616161616,
      "avg_output_tokens_per_task": 25861.328282828283,
      "last_run_timestamp": "2026-10-24T08:53:53.931Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "DeepInfra",
      "accuracy": 0.8833333472727273,
      "accuracy_std_dev": 0.012158692807021237,
      "run_count": 7,
      "total_questions": 1100,
      "avg_cost_per_task_usd": 0.007285390000000001,
      "avg_duration_per_task_ms": 942880.7490909091,
      "avg_output_tokens_per_task": 28917.175454545453,
      "last_run_timestamp": "2026-10-17T08:19:36.266Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Together",
      "accuracy": 0.8828761428571429,
      "accuracy_std_dev": 0.004395219722398795,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.029912125541125538,
      "avg_duration_per_task_ms": 82923.05122655122,
      "avg_output_tokens_per_task": 24775.42784992785,
      "last_run_timestamp": "2026-10-24T08:49:27.543Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Decart",
      "accuracy": 0.8828537731629394,
      "accuracy_std_dev": 0.008622317150279293,
      "run_count": 5,
      "total_questions": 939,
      "avg_cost_per_task_usd": 0.006383692225772098,
      "avg_duration_per_task_ms": 468696.3525026624,
      "avg_output_tokens_per_task": 35072.061767838124,
      "last_run_timestamp": "2026-10-24T08:44:41.546Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-opus-5-20260723",
      "model_slug": "anthropic/claude-opus-5",
      "model_name": "Anthropic: Claude Opus 5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.8821585088105727,
      "accuracy_std_dev": 0.01851343261661651,
      "run_count": 14,
      "total_questions": 2724,
      "avg_cost_per_task_usd": 0.1276248480176211,
      "avg_duration_per_task_ms": 62648.52643171806,
      "avg_output_tokens_per_task": 4878.82268722467,
      "last_run_timestamp": "2026-10-24T08:27:56.033Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Phala",
      "accuracy": 0.8821550000000001,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.129083601010101,
      "avg_duration_per_task_ms": 388751.6590909091,
      "avg_output_tokens_per_task": 32112.176767676767,
      "last_run_timestamp": "2026-10-17T16:07:32.047Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "SiliconFlow",
      "accuracy": 0.88215475,
      "accuracy_std_dev": 0.018821912135527027,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.008138106691919191,
      "avg_duration_per_task_ms": 380823.72411616164,
      "avg_output_tokens_per_task": 28681.33712121212,
      "last_run_timestamp": "2026-10-24T16:29:33.430Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-opus-5-20260723",
      "model_slug": "anthropic/claude-opus-5",
      "model_name": "Anthropic: Claude Opus 5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.882043637444875,
      "accuracy_std_dev": 0.009663107099304217,
      "run_count": 39,
      "total_questions": 7483,
      "avg_cost_per_task_usd": 0.13849482640652144,
      "avg_duration_per_task_ms": 62839.13149806228,
      "avg_output_tokens_per_task": 5015.037551784044,
      "last_run_timestamp": "2026-10-24T08:22:55.673Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "StreamLake",
      "accuracy": 0.882034357142857,
      "accuracy_std_dev": 0.011301592736090304,
      "run_count": 14,
      "total_questions": 2772,
      "avg_cost_per_task_usd": 0.006003468614718615,
      "avg_duration_per_task_ms": 401285.20815295813,
      "avg_output_tokens_per_task": 29705.93326118326,
      "last_run_timestamp": "2026-10-24T08:46:33.521Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "CoreWeave",
      "accuracy": 0.8818485454545456,
      "accuracy_std_dev": 0.01327175411002725,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.18638605096418728,
      "avg_duration_per_task_ms": 321116.0505050505,
      "avg_output_tokens_per_task": 46783.020661157025,
      "last_run_timestamp": "2026-10-24T08:02:27.414Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "CoreWeave",
      "accuracy": 0.8816240000000001,
      "accuracy_std_dev": 0.004701000000000011,
      "run_count": 2,
      "total_questions": 390,
      "avg_cost_per_task_usd": 0.020361825641025642,
      "avg_duration_per_task_ms": 90976.65641025642,
      "avg_output_tokens_per_task": 40519.597435897434,
      "last_run_timestamp": "2026-10-24T08:02:32.160Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2-thinking-20251106",
      "model_slug": "moonshotai/kimi-k2-thinking",
      "model_name": "MoonshotAI: Kimi K2 Thinking",
      "provider_name": "auto-routing",
      "accuracy": 0.881603,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 183,
      "avg_cost_per_task_usd": 0.10363212021857923,
      "avg_duration_per_task_ms": 876152.9726775957,
      "avg_output_tokens_per_task": 41457.83606557377,
      "last_run_timestamp": "2026-08-01T17:53:10.275Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-luna-20260922",
      "model_slug": "openai/gpt-6-luna",
      "model_name": "OpenAI: GPT-6 Luna",
      "provider_name": "OpenAI",
      "accuracy": 0.8815936666666665,
      "accuracy_std_dev": 0.00975199857579064,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.002981575757575758,
      "avg_duration_per_task_ms": 52993.44276094276,
      "avg_output_tokens_per_task": 5804.632996632997,
      "last_run_timestamp": "2026-10-02T23:20:31.396Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Novita",
      "accuracy": 0.8815751316455697,
      "accuracy_std_dev": 0.010764200383163109,
      "run_count": 6,
      "total_questions": 1185,
      "avg_cost_per_task_usd": 0.007927140084388185,
      "avg_duration_per_task_ms": 240915.00168776372,
      "avg_output_tokens_per_task": 27983.259071729957,
      "last_run_timestamp": "2026-10-03T16:57:09.267Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Cloudflare",
      "accuracy": 0.8815374099783081,
      "accuracy_std_dev": 0.05700823639454502,
      "run_count": 7,
      "total_questions": 1383,
      "avg_cost_per_task_usd": 0.08337883875632683,
      "avg_duration_per_task_ms": 715683.0332610267,
      "avg_output_tokens_per_task": 31906.826464208243,
      "last_run_timestamp": "2026-10-24T16:21:36.007Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Inceptron",
      "accuracy": 0.8814552816901408,
      "accuracy_std_dev": 0.00965776820360461,
      "run_count": 13,
      "total_questions": 2556,
      "avg_cost_per_task_usd": 0.00679829303599374,
      "avg_duration_per_task_ms": 375447.8978873239,
      "avg_output_tokens_per_task": 19618.11854460094,
      "last_run_timestamp": "2026-10-24T08:46:28.764Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "Parasail",
      "accuracy": 0.8814329055819478,
      "accuracy_std_dev": 0.011397141503221075,
      "run_count": 9,
      "total_questions": 1684,
      "avg_cost_per_task_usd": 0.06679124287410926,
      "avg_duration_per_task_ms": 309462.4631828979,
      "avg_output_tokens_per_task": 18456.302850356293,
      "last_run_timestamp": "2026-10-24T16:08:29.567Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "BaseTen",
      "accuracy": 0.881313,
      "accuracy_std_dev": 0.012626000000000026,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.022825318181818183,
      "avg_duration_per_task_ms": 147723.58080808082,
      "avg_output_tokens_per_task": 45335.5101010101,
      "last_run_timestamp": "2026-10-24T08:12:49.570Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Novita",
      "accuracy": 0.8813128749999999,
      "accuracy_std_dev": 0.02861962867700026,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.19908026893939393,
      "avg_duration_per_task_ms": 958354.9299242424,
      "avg_output_tokens_per_task": 58904.72348484849,
      "last_run_timestamp": "2026-10-24T16:17:57.472Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "DeepInfra",
      "accuracy": 0.8811927142857143,
      "accuracy_std_dev": 0.007888829783113527,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.015891973304473304,
      "avg_duration_per_task_ms": 279407.10173160175,
      "avg_output_tokens_per_task": 24276.554834054834,
      "last_run_timestamp": "2026-10-08T22:18:51.642Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "DigitalOcean",
      "accuracy": 0.8809269425684658,
      "accuracy_std_dev": 0.03229722684698395,
      "run_count": 19,
      "total_questions": 3761,
      "avg_cost_per_task_usd": 0.1444545269875033,
      "avg_duration_per_task_ms": 371121.18771603296,
      "avg_output_tokens_per_task": 10937.842329167774,
      "last_run_timestamp": "2026-10-24T08:22:00.565Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "Together",
      "accuracy": 0.8808292124352332,
      "accuracy_std_dev": 0.0020954999999999724,
      "run_count": 2,
      "total_questions": 386,
      "avg_cost_per_task_usd": 0.03105037564766839,
      "avg_duration_per_task_ms": 282019.7538860104,
      "avg_output_tokens_per_task": 61860.47668393782,
      "last_run_timestamp": "2026-10-24T08:07:42.017Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "DeepInfra",
      "accuracy": 0.8807385965665235,
      "accuracy_std_dev": 0.04941426514861386,
      "run_count": 13,
      "total_questions": 2563,
      "avg_cost_per_task_usd": 0.0748979442060086,
      "avg_duration_per_task_ms": 557063.2532188841,
      "avg_output_tokens_per_task": 39438.98985563793,
      "last_run_timestamp": "2026-10-24T16:40:43.534Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-luna-20260922",
      "model_slug": "openai/gpt-6-luna",
      "model_name": "OpenAI: GPT-6 Luna",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.880471,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0032475000000000004,
      "avg_duration_per_task_ms": 58949.292929292926,
      "avg_output_tokens_per_task": 5746.025252525253,
      "last_run_timestamp": "2026-09-23T06:09:34.299Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "auto-routing",
      "accuracy": 0.8801906666666667,
      "accuracy_std_dev": 0.011226983784119793,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.0227881835016835,
      "avg_duration_per_task_ms": 234700.68434343435,
      "avg_output_tokens_per_task": 25264.190235690236,
      "last_run_timestamp": "2026-10-03T09:30:18.263Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Modal",
      "accuracy": 0.8800047366136035,
      "accuracy_std_dev": 0.007207181987717666,
      "run_count": 7,
      "total_questions": 1382,
      "avg_cost_per_task_usd": 0.03244589001447178,
      "avg_duration_per_task_ms": 148073.17149059335,
      "avg_output_tokens_per_task": 26851.936324167873,
      "last_run_timestamp": "2026-10-24T08:26:33.257Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Decart",
      "accuracy": 0.8798644568527919,
      "accuracy_std_dev": 0.004527248931623837,
      "run_count": 3,
      "total_questions": 591,
      "avg_cost_per_task_usd": 0.0029938477157360403,
      "avg_duration_per_task_ms": 316818.90016920475,
      "avg_output_tokens_per_task": 18321.88663282572,
      "last_run_timestamp": "2026-08-22T10:20:26.941Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openrouter/auto",
      "model_slug": "openrouter/auto",
      "model_name": "Auto Router",
      "provider_name": "auto-routing",
      "accuracy": 0.8795957999999999,
      "accuracy_std_dev": 0.026351833738091142,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.08917041111111111,
      "avg_duration_per_task_ms": 196788.44343434344,
      "avg_output_tokens_per_task": 37445.42121212121,
      "last_run_timestamp": "2026-09-29T14:28:17.204Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "GMICloud",
      "accuracy": 0.8794324650455927,
      "accuracy_std_dev": 0.01587967361881218,
      "run_count": 5,
      "total_questions": 987,
      "avg_cost_per_task_usd": 0.03493195542046606,
      "avg_duration_per_task_ms": 723219.4984802431,
      "avg_output_tokens_per_task": 40176.24822695035,
      "last_run_timestamp": "2026-10-24T08:41:21.450Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "OpenInference",
      "accuracy": 0.879386,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 76,
      "avg_cost_per_task_usd": 0.03120878947368421,
      "avg_duration_per_task_ms": 274583.63157894736,
      "avg_output_tokens_per_task": 9741.526315789473,
      "last_run_timestamp": "2026-10-09T00:00:06.206Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "DigitalOcean",
      "accuracy": 0.879370525058275,
      "accuracy_std_dev": 0.02189493848211963,
      "run_count": 9,
      "total_questions": 1716,
      "avg_cost_per_task_usd": 0.09564452447552448,
      "avg_duration_per_task_ms": 551582.8269230769,
      "avg_output_tokens_per_task": 28335.815268065267,
      "last_run_timestamp": "2026-10-24T08:13:46.723Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Relace",
      "accuracy": 0.8792376607706888,
      "accuracy_std_dev": 0.006337813294770935,
      "run_count": 8,
      "total_questions": 1583,
      "avg_cost_per_task_usd": 0.016976666456096023,
      "avg_duration_per_task_ms": 306778.0650663297,
      "avg_output_tokens_per_task": 30476.99052432091,
      "last_run_timestamp": "2026-10-24T08:43:17.831Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Venice",
      "accuracy": 0.8790282857142857,
      "accuracy_std_dev": 0.019161523885971404,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.03542920851370852,
      "avg_duration_per_task_ms": 140156.7987012987,
      "avg_output_tokens_per_task": 27224.489898989897,
      "last_run_timestamp": "2026-10-24T08:28:56.672Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "Deepseek V4.1 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.878788,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.010643959595959596,
      "avg_duration_per_task_ms": 127627.76767676767,
      "avg_output_tokens_per_task": 14277.949494949495,
      "last_run_timestamp": "2026-10-08T13:29:57.446Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-pro",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Mimo V2.6 Pro",
      "provider_name": "auto-routing",
      "accuracy": 0.878788,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.01249009595959596,
      "avg_duration_per_task_ms": 315799.5808080808,
      "avg_output_tokens_per_task": 14226.873737373737,
      "last_run_timestamp": "2026-10-08T01:55:44.175Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.2-20251211",
      "model_slug": "openai/gpt-5.2",
      "model_name": "OpenAI: GPT-5.2",
      "provider_name": "OpenAI",
      "accuracy": 0.878788,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.13332395959595958,
      "avg_duration_per_task_ms": 138478.24242424243,
      "avg_output_tokens_per_task": 9426.934343434343,
      "last_run_timestamp": "2026-08-01T08:50:28.136Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.7-plus-20260602",
      "model_slug": "qwen/qwen3.7-plus",
      "model_name": "Qwen: Qwen3.7 Plus",
      "provider_name": "auto-routing",
      "accuracy": 0.878788,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.043995565656565656,
      "avg_duration_per_task_ms": 588519.7272727273,
      "avg_output_tokens_per_task": 34166.82828282828,
      "last_run_timestamp": "2026-10-01T08:47:14.818Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-luna-20260709",
      "model_slug": "openai/gpt-5.6-luna",
      "model_name": "OpenAI: GPT-5.6 Luna",
      "provider_name": "auto-routing",
      "accuracy": 0.8787876666666666,
      "accuracy_std_dev": 0.005991714074916775,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.007907055555555554,
      "avg_duration_per_task_ms": 79561.398989899,
      "avg_output_tokens_per_task": 7834.653198653199,
      "last_run_timestamp": "2026-10-01T19:14:22.294Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Fireworks",
      "accuracy": 0.8787801424227066,
      "accuracy_std_dev": 0.016401176693323544,
      "run_count": 11,
      "total_questions": 1973,
      "avg_cost_per_task_usd": 0.011341015205271159,
      "avg_duration_per_task_ms": 261740.6791687785,
      "avg_output_tokens_per_task": 24571.48302078054,
      "last_run_timestamp": "2026-09-19T08:54:58.652Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Cloudflare",
      "accuracy": 0.8787668916494136,
      "accuracy_std_dev": 0.03281417517628624,
      "run_count": 8,
      "total_questions": 1449,
      "avg_cost_per_task_usd": 0.2005868881987578,
      "avg_duration_per_task_ms": 1332066.3747412008,
      "avg_output_tokens_per_task": 50003.33195307109,
      "last_run_timestamp": "2026-10-24T16:26:57.307Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Alibaba",
      "accuracy": 0.8786913361593894,
      "accuracy_std_dev": 0.025462712419203403,
      "run_count": 12,
      "total_questions": 2359,
      "avg_cost_per_task_usd": 0.08448664264518864,
      "avg_duration_per_task_ms": 403631.4383213226,
      "avg_output_tokens_per_task": 31865.207291225095,
      "last_run_timestamp": "2026-10-24T16:13:03.135Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Parasail",
      "accuracy": 0.8786261545293074,
      "accuracy_std_dev": 0.0068412882173039926,
      "run_count": 6,
      "total_questions": 1126,
      "avg_cost_per_task_usd": 0.11316016607460035,
      "avg_duration_per_task_ms": 476858.84191829484,
      "avg_output_tokens_per_task": 32318.380994671403,
      "last_run_timestamp": "2026-10-17T16:07:16.258Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-opus-5-20260723",
      "model_slug": "anthropic/claude-opus-5",
      "model_name": "Anthropic: Claude Opus 5",
      "provider_name": "auto-routing",
      "accuracy": 0.87780575,
      "accuracy_std_dev": 0.016897481104812654,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.1286623606902357,
      "avg_duration_per_task_ms": 62852.628367003366,
      "avg_output_tokens_per_task": 4923.856481481482,
      "last_run_timestamp": "2026-10-24T08:15:51.136Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8776513109987357,
      "accuracy_std_dev": 0.007333672175410435,
      "run_count": 12,
      "total_questions": 2373,
      "avg_cost_per_task_usd": 0.11767066118836916,
      "avg_duration_per_task_ms": 752365.9431099873,
      "avg_output_tokens_per_task": 30843.225874420565,
      "last_run_timestamp": "2026-10-24T16:07:46.989Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Ionstream",
      "accuracy": 0.8774407999999999,
      "accuracy_std_dev": 0.00954717737134911,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.03921924343434343,
      "avg_duration_per_task_ms": 242470.04242424242,
      "avg_output_tokens_per_task": 37354.009090909094,
      "last_run_timestamp": "2026-10-21T11:23:15.101Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Parasail",
      "accuracy": 0.8773447142857143,
      "accuracy_std_dev": 0.012707857195050472,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.03401366883116883,
      "avg_duration_per_task_ms": 154330.79004329004,
      "avg_output_tokens_per_task": 28188.37950937951,
      "last_run_timestamp": "2026-10-24T09:09:44.617Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Reka",
      "accuracy": 0.8772070529695024,
      "accuracy_std_dev": 0.03893659390197737,
      "run_count": 4,
      "total_questions": 623,
      "avg_cost_per_task_usd": 0.09425666452648475,
      "avg_duration_per_task_ms": 461375.2568218299,
      "avg_output_tokens_per_task": 23922.329052969504,
      "last_run_timestamp": "2026-09-30T11:37:23.219Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-luna-20260922",
      "model_slug": "openai/gpt-6-luna",
      "model_name": "OpenAI: GPT-6 Luna",
      "provider_name": "Azure",
      "accuracy": 0.8771043333333335,
      "accuracy_std_dev": 0.007896528133018712,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.003143374579124579,
      "avg_duration_per_task_ms": 35450.33585858586,
      "avg_output_tokens_per_task": 5733.669191919192,
      "last_run_timestamp": "2026-10-01T08:25:15.234Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "GMICloud",
      "accuracy": 0.877104,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.07509321212121212,
      "avg_duration_per_task_ms": 263175.32828282827,
      "avg_output_tokens_per_task": 20738.459595959597,
      "last_run_timestamp": "2026-08-29T19:12:43.653Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Parasail",
      "accuracy": 0.8770437482014388,
      "accuracy_std_dev": 0.016843957465790968,
      "run_count": 8,
      "total_questions": 1529,
      "avg_cost_per_task_usd": 0.1855886553302812,
      "avg_duration_per_task_ms": 777107.6200130804,
      "avg_output_tokens_per_task": 53473.10595160235,
      "last_run_timestamp": "2026-10-24T16:28:49.711Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "SiliconFlow",
      "accuracy": 0.876999,
      "accuracy_std_dev": 0.06406231714705611,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.1307801066919192,
      "avg_duration_per_task_ms": 1601351.4103535353,
      "avg_output_tokens_per_task": 55537.77272727273,
      "last_run_timestamp": "2026-10-24T16:02:47.376Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "StreamLake",
      "accuracy": 0.8769172222222222,
      "accuracy_std_dev": 0.01657946057812884,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.02028404994388328,
      "avg_duration_per_task_ms": 725618.4191919192,
      "avg_output_tokens_per_task": 35595.71436588103,
      "last_run_timestamp": "2026-10-24T16:27:17.186Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Crusoe",
      "accuracy": 0.8767675999999999,
      "accuracy_std_dev": 0.014101374310328764,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.042260786868686874,
      "avg_duration_per_task_ms": 210147.5181818182,
      "avg_output_tokens_per_task": 35072.08888888889,
      "last_run_timestamp": "2026-10-24T08:50:45.433Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "AkashML",
      "accuracy": 0.8765966933797908,
      "accuracy_std_dev": 0.020189124617118887,
      "run_count": 6,
      "total_questions": 1148,
      "avg_cost_per_task_usd": 0.004125730836236933,
      "avg_duration_per_task_ms": 453615.28397212544,
      "avg_output_tokens_per_task": 16054.564459930314,
      "last_run_timestamp": "2026-09-05T08:24:32.148Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-fable-5.1-20260831",
      "model_slug": "anthropic/claude-fable-5.1",
      "model_name": "Anthropic: Claude Fable 5.1",
      "provider_name": "Google",
      "accuracy": 0.8765431666666668,
      "accuracy_std_dev": 0.03217293038884639,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.11818556060606061,
      "avg_duration_per_task_ms": 28260.9031986532,
      "avg_output_tokens_per_task": 2161.995791245791,
      "last_run_timestamp": "2026-10-01T08:25:56.890Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Parasail",
      "accuracy": 0.8765007667238421,
      "accuracy_std_dev": 0.009087011145341243,
      "run_count": 3,
      "total_questions": 583,
      "avg_cost_per_task_usd": 0.09344886277873071,
      "avg_duration_per_task_ms": 568169.3550600343,
      "avg_output_tokens_per_task": 27396.61578044597,
      "last_run_timestamp": "2026-08-22T09:51:24.721Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Reka",
      "accuracy": 0.8763533692307693,
      "accuracy_std_dev": 0.006821595577445394,
      "run_count": 3,
      "total_questions": 585,
      "avg_cost_per_task_usd": 0.027450420512820508,
      "avg_duration_per_task_ms": 188376.252991453,
      "avg_output_tokens_per_task": 30391.492307692308,
      "last_run_timestamp": "2026-10-23T06:17:23.826Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "Novita",
      "accuracy": 0.8763108617977529,
      "accuracy_std_dev": 0.006823199452268141,
      "run_count": 9,
      "total_questions": 1780,
      "avg_cost_per_task_usd": 0.12380380168539326,
      "avg_duration_per_task_ms": 917825.4601123596,
      "avg_output_tokens_per_task": 43434.7488764045,
      "last_run_timestamp": "2026-10-24T16:19:29.598Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Z.AI",
      "accuracy": 0.8760746752655538,
      "accuracy_std_dev": 0.011470076619186123,
      "run_count": 10,
      "total_questions": 1977,
      "avg_cost_per_task_usd": 0.009704542235710673,
      "avg_duration_per_task_ms": 429066.8811330298,
      "avg_output_tokens_per_task": 22619.601416287303,
      "last_run_timestamp": "2026-10-24T08:41:28.565Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "CoreWeave",
      "accuracy": 0.8760220000000001,
      "accuracy_std_dev": 0.014151103773204405,
      "run_count": 14,
      "total_questions": 2772,
      "avg_cost_per_task_usd": 0.11097613275613274,
      "avg_duration_per_task_ms": 457617.5707070707,
      "avg_output_tokens_per_task": 31565.57683982684,
      "last_run_timestamp": "2026-10-24T16:02:47.387Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Cloudflare",
      "accuracy": 0.8754205,
      "accuracy_std_dev": 0.0016835000000000044,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.04156198232323233,
      "avg_duration_per_task_ms": 309570.8888888889,
      "avg_output_tokens_per_task": 39159.9494949495,
      "last_run_timestamp": "2026-10-17T08:26:18.265Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "DeepInfra",
      "accuracy": 0.8754205,
      "accuracy_std_dev": 0.0016835000000000044,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.01960839393939394,
      "avg_duration_per_task_ms": 110709.8308080808,
      "avg_output_tokens_per_task": 43264.59848484849,
      "last_run_timestamp": "2026-10-24T08:12:37.561Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "DigitalOcean",
      "accuracy": 0.8752913076923076,
      "accuracy_std_dev": 0.04445532892411749,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.09291413209013209,
      "avg_duration_per_task_ms": 618152.5388500389,
      "avg_output_tokens_per_task": 41235.80536130536,
      "last_run_timestamp": "2026-10-24T16:04:29.298Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Venice",
      "accuracy": 0.8749675384615386,
      "accuracy_std_dev": 0.013017813792677657,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.006246133255633255,
      "avg_duration_per_task_ms": 265528.69425019424,
      "avg_output_tokens_per_task": 18684.931235431235,
      "last_run_timestamp": "2026-10-24T08:20:25.161Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Decart",
      "accuracy": 0.8749014196217493,
      "accuracy_std_dev": 0.02259480542425625,
      "run_count": 5,
      "total_questions": 846,
      "avg_cost_per_task_usd": 0.12702199408983453,
      "avg_duration_per_task_ms": 548158.9716312056,
      "avg_output_tokens_per_task": 48997.07092198582,
      "last_run_timestamp": "2026-10-03T17:01:24.374Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Baidu",
      "accuracy": 0.8747894999999999,
      "accuracy_std_dev": 0.02495521020748971,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.1433024273989899,
      "avg_duration_per_task_ms": 926726.0309343435,
      "avg_output_tokens_per_task": 61552.77146464647,
      "last_run_timestamp": "2026-10-24T16:07:10.480Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8747014285714287,
      "accuracy_std_dev": 0.09649629293840344,
      "run_count": 12,
      "total_questions": 2373,
      "avg_cost_per_task_usd": 0.02719648798988622,
      "avg_duration_per_task_ms": 204176.03792667508,
      "avg_output_tokens_per_task": 23353.09228824273,
      "last_run_timestamp": "2026-10-24T16:28:20.445Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Reka",
      "accuracy": 0.8746087667870037,
      "accuracy_std_dev": 0.007592482877484693,
      "run_count": 7,
      "total_questions": 1385,
      "avg_cost_per_task_usd": 0.2179415407942238,
      "avg_duration_per_task_ms": 1239989.4288808664,
      "avg_output_tokens_per_task": 48485.83321299639,
      "last_run_timestamp": "2026-10-24T16:18:35.554Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Reka",
      "accuracy": 0.8745790000000001,
      "accuracy_std_dev": 0.04079651270022963,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.010708270202020202,
      "avg_duration_per_task_ms": 125365.15404040404,
      "avg_output_tokens_per_task": 15791.75,
      "last_run_timestamp": "2026-10-24T08:20:32.397Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Together",
      "accuracy": 0.874579,
      "accuracy_std_dev": 0.005892000000000008,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.018065838383838382,
      "avg_duration_per_task_ms": 168823.5202020202,
      "avg_output_tokens_per_task": 29810.62373737374,
      "last_run_timestamp": "2026-10-24T08:41:24.133Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Makora",
      "accuracy": 0.8745317312013828,
      "accuracy_std_dev": 0.011186512454241005,
      "run_count": 6,
      "total_questions": 1157,
      "avg_cost_per_task_usd": 0.0043274416594641305,
      "avg_duration_per_task_ms": 185028.52031114954,
      "avg_output_tokens_per_task": 21870.51080380294,
      "last_run_timestamp": "2026-09-26T08:53:59.525Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "Phala",
      "accuracy": 0.8744854444444443,
      "accuracy_std_dev": 0.005088649135545852,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.08018900897867563,
      "avg_duration_per_task_ms": 386083.189674523,
      "avg_output_tokens_per_task": 22800.944444444445,
      "last_run_timestamp": "2026-10-24T16:39:19.451Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Mancer 2",
      "accuracy": 0.8743608677867055,
      "accuracy_std_dev": 0.030009909369776877,
      "run_count": 15,
      "total_questions": 2738,
      "avg_cost_per_task_usd": 0.011979280861943022,
      "avg_duration_per_task_ms": 340619.6665449233,
      "avg_output_tokens_per_task": 21211.479547114683,
      "last_run_timestamp": "2026-10-24T08:45:03.893Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Wafer",
      "accuracy": 0.8742983333333334,
      "accuracy_std_dev": 0.009152249171044747,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0078061279461279455,
      "avg_duration_per_task_ms": 622665.6515151515,
      "avg_output_tokens_per_task": 44426.93771043771,
      "last_run_timestamp": "2026-10-24T16:20:24.813Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Wafer",
      "accuracy": 0.8742830219449225,
      "accuracy_std_dev": 0.06426649201755963,
      "run_count": 12,
      "total_questions": 2324,
      "avg_cost_per_task_usd": 0.01125052065404475,
      "avg_duration_per_task_ms": 484759.71815834765,
      "avg_output_tokens_per_task": 23203.48364888124,
      "last_run_timestamp": "2026-10-24T08:11:11.333Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Makora",
      "accuracy": 0.8740843414084506,
      "accuracy_std_dev": 0.004005737471597214,
      "run_count": 9,
      "total_questions": 1775,
      "avg_cost_per_task_usd": 0.044288470985915485,
      "avg_duration_per_task_ms": 203522.5785915493,
      "avg_output_tokens_per_task": 38212.04394366197,
      "last_run_timestamp": "2026-10-24T08:44:12.949Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Together",
      "accuracy": 0.8740044607508533,
      "accuracy_std_dev": 0.007605951280996289,
      "run_count": 3,
      "total_questions": 586,
      "avg_cost_per_task_usd": 0.10401818771331058,
      "avg_duration_per_task_ms": 459122.319112628,
      "avg_output_tokens_per_task": 30022.856655290103,
      "last_run_timestamp": "2026-08-22T16:49:15.579Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Together",
      "accuracy": 0.8739276971507354,
      "accuracy_std_dev": 0.014730569376258348,
      "run_count": 11,
      "total_questions": 2176,
      "avg_cost_per_task_usd": 0.09131274172794118,
      "avg_duration_per_task_ms": 205017.0390625,
      "avg_output_tokens_per_task": 23097.8125,
      "last_run_timestamp": "2026-10-24T08:48:36.491Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "CoreWeave",
      "accuracy": 0.8737371428571428,
      "accuracy_std_dev": 0.01647031431245381,
      "run_count": 14,
      "total_questions": 2772,
      "avg_cost_per_task_usd": 0.008272517676767677,
      "avg_duration_per_task_ms": 219776.6417748918,
      "avg_output_tokens_per_task": 29274.76370851371,
      "last_run_timestamp": "2026-10-24T09:11:12.594Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "SambaNova",
      "accuracy": 0.8737370000000001,
      "accuracy_std_dev": 0.006734000000000018,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.03531422474747475,
      "avg_duration_per_task_ms": 145858.30808080808,
      "avg_output_tokens_per_task": 29116.540404040403,
      "last_run_timestamp": "2026-10-24T08:59:28.837Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "auto-routing",
      "accuracy": 0.8737370000000001,
      "accuracy_std_dev": 0.006734000000000018,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.023037462121212122,
      "avg_duration_per_task_ms": 506854.46717171714,
      "avg_output_tokens_per_task": 48027.71464646464,
      "last_run_timestamp": "2026-10-24T08:01:41.049Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-luna-20260922",
      "model_slug": "openai/gpt-6-luna",
      "model_name": "OpenAI: GPT-6 Luna",
      "provider_name": "auto-routing",
      "accuracy": 0.873737,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0029526111111111116,
      "avg_duration_per_task_ms": 49722.71717171717,
      "avg_output_tokens_per_task": 5746.707070707071,
      "last_run_timestamp": "2026-10-01T08:23:37.659Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.7-plus-20260602",
      "model_slug": "qwen/qwen3.7-plus",
      "model_name": "Qwen: Qwen3.7 Plus",
      "provider_name": "Alibaba",
      "accuracy": 0.873737,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.045554454545454544,
      "avg_duration_per_task_ms": 605480.6616161616,
      "avg_output_tokens_per_task": 35385.22727272727,
      "last_run_timestamp": "2026-10-01T08:40:06.225Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Venice",
      "accuracy": 0.873716393364929,
      "accuracy_std_dev": 0.03410588644533702,
      "run_count": 9,
      "total_questions": 1688,
      "avg_cost_per_task_usd": 0.012903949052132702,
      "avg_duration_per_task_ms": 360645.17417061614,
      "avg_output_tokens_per_task": 25970.889218009477,
      "last_run_timestamp": "2026-10-24T08:43:07.915Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Cloudflare",
      "accuracy": 0.8736468057110863,
      "accuracy_std_dev": 0.03036521292219109,
      "run_count": 10,
      "total_questions": 1786,
      "avg_cost_per_task_usd": 0.025540696528555434,
      "avg_duration_per_task_ms": 360273.59238521836,
      "avg_output_tokens_per_task": 24510.389137737962,
      "last_run_timestamp": "2026-10-24T16:24:04.274Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Venice",
      "accuracy": 0.8733163333333334,
      "accuracy_std_dev": 0.030039827993663497,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.17326605218855218,
      "avg_duration_per_task_ms": 357944.43055555556,
      "avg_output_tokens_per_task": 39522.67929292929,
      "last_run_timestamp": "2026-10-24T16:27:20.826Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Cloudflare",
      "accuracy": 0.8732174942820837,
      "accuracy_std_dev": 0.03602354128311638,
      "run_count": 12,
      "total_questions": 2361,
      "avg_cost_per_task_usd": 0.10224943286742906,
      "avg_duration_per_task_ms": 640310.1185938162,
      "avg_output_tokens_per_task": 25410.69970351546,
      "last_run_timestamp": "2026-10-24T16:26:55.135Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "ModelRun",
      "accuracy": 0.8727272475524476,
      "accuracy_std_dev": 0.029233686851601524,
      "run_count": 8,
      "total_questions": 1430,
      "avg_cost_per_task_usd": 0.09700304615384615,
      "avg_duration_per_task_ms": 194874.37272727274,
      "avg_output_tokens_per_task": 26397.99300699301,
      "last_run_timestamp": "2026-09-26T09:19:13.012Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-luna-20260709",
      "model_slug": "openai/gpt-5.6-luna",
      "model_name": "OpenAI: GPT-5.6 Luna",
      "provider_name": "OpenAI",
      "accuracy": 0.8724745,
      "accuracy_std_dev": 0.008858340547190563,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.008325782828282828,
      "avg_duration_per_task_ms": 77307.88005050505,
      "avg_output_tokens_per_task": 7842.704545454545,
      "last_run_timestamp": "2026-10-02T23:19:10.742Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "Chutes",
      "accuracy": 0.8723343333333334,
      "accuracy_std_dev": 0.005338803632109208,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.06334109932659931,
      "avg_duration_per_task_ms": 262990.18855218857,
      "avg_output_tokens_per_task": 21026.934343434343,
      "last_run_timestamp": "2026-08-15T15:55:02.752Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Together",
      "accuracy": 0.8721539397539244,
      "accuracy_std_dev": 0.008970575607073516,
      "run_count": 12,
      "total_questions": 2357,
      "avg_cost_per_task_usd": 0.008344311412812899,
      "avg_duration_per_task_ms": 339433.18073822657,
      "avg_output_tokens_per_task": 29494.21552821383,
      "last_run_timestamp": "2026-10-17T08:23:51.334Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "AtlasCloud",
      "accuracy": 0.872053625,
      "accuracy_std_dev": 0.02370397511989021,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.1190935284090909,
      "avg_duration_per_task_ms": 606301.1369949495,
      "avg_output_tokens_per_task": 35231.555555555555,
      "last_run_timestamp": "2026-10-24T16:26:49.397Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Alibaba",
      "accuracy": 0.8720033968503939,
      "accuracy_std_dev": 0.01337304654930955,
      "run_count": 10,
      "total_questions": 1905,
      "avg_cost_per_task_usd": 0.008342944356955382,
      "avg_duration_per_task_ms": 238567.13018372704,
      "avg_output_tokens_per_task": 29657.869816272967,
      "last_run_timestamp": "2026-10-24T16:02:06.915Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Morph",
      "accuracy": 0.871875,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 160,
      "avg_cost_per_task_usd": 0.04507655,
      "avg_duration_per_task_ms": 1504372.61875,
      "avg_output_tokens_per_task": 18693.5375,
      "last_run_timestamp": "2026-08-27T15:00:05.074Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.8717731666666667,
      "accuracy_std_dev": 0.004601771648567064,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.00424808627946128,
      "avg_duration_per_task_ms": 39417.8244949495,
      "avg_output_tokens_per_task": 7731.690235690236,
      "last_run_timestamp": "2026-10-24T08:30:49.056Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "DeepInfra",
      "accuracy": 0.8714859264426126,
      "accuracy_std_dev": 0.00752103815972237,
      "run_count": 8,
      "total_questions": 1577,
      "avg_cost_per_task_usd": 0.06945589727330374,
      "avg_duration_per_task_ms": 276858.71528218134,
      "avg_output_tokens_per_task": 11308.629676601142,
      "last_run_timestamp": "2026-10-10T08:22:14.215Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "auto-routing",
      "accuracy": 0.8713054444444444,
      "accuracy_std_dev": 0.011226478265552096,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.08585465039281706,
      "avg_duration_per_task_ms": 191685.55331088664,
      "avg_output_tokens_per_task": 13992.872615039281,
      "last_run_timestamp": "2026-10-24T08:29:16.805Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "Alibaba",
      "accuracy": 0.8712332981471084,
      "accuracy_std_dev": 0.008224257022992851,
      "run_count": 9,
      "total_questions": 1781,
      "avg_cost_per_task_usd": 0.049519284671532854,
      "avg_duration_per_task_ms": 283250.0583941606,
      "avg_output_tokens_per_task": 21034.03425042111,
      "last_run_timestamp": "2026-10-24T16:29:09.815Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "DigitalOcean",
      "accuracy": 0.871212,
      "accuracy_std_dev": 0.0008420000000000094,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.017874631313131314,
      "avg_duration_per_task_ms": 86992.41414141415,
      "avg_output_tokens_per_task": 29675.651515151516,
      "last_run_timestamp": "2026-10-24T08:40:56.206Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "CoreWeave",
      "accuracy": 0.8711184444444445,
      "accuracy_std_dev": 0.018965626509914724,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.19942039898989897,
      "avg_duration_per_task_ms": 553312.7272727273,
      "avg_output_tokens_per_task": 58346.15095398429,
      "last_run_timestamp": "2026-10-24T16:09:40.215Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "OpenInference",
      "accuracy": 0.8710915714285715,
      "accuracy_std_dev": 0.014505946572326232,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.021487838383838383,
      "avg_duration_per_task_ms": 272335.1349206349,
      "avg_output_tokens_per_task": 35144.1277056277,
      "last_run_timestamp": "2026-10-24T08:13:46.727Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "thinkingmachines/inkling-small-20260730",
      "model_slug": "thinkingmachines/inkling-small",
      "model_name": "Thinking Machines: Inkling Small",
      "provider_name": "Together",
      "accuracy": 0.8710116425661915,
      "accuracy_std_dev": 0.01025225770257459,
      "run_count": 5,
      "total_questions": 982,
      "avg_cost_per_task_usd": 0.025059682281059068,
      "avg_duration_per_task_ms": 328330.6405295316,
      "avg_output_tokens_per_task": 20642.19348268839,
      "last_run_timestamp": "2026-09-12T08:54:29.103Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Fireworks",
      "accuracy": 0.8706863196512894,
      "accuracy_std_dev": 0.010805462536728977,
      "run_count": 14,
      "total_questions": 2753,
      "avg_cost_per_task_usd": 0.047070880130766435,
      "avg_duration_per_task_ms": 208861.94733018524,
      "avg_output_tokens_per_task": 36299.43697784236,
      "last_run_timestamp": "2026-10-24T09:06:03.113Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Parasail",
      "accuracy": 0.8703702499999999,
      "accuracy_std_dev": 0.005050541672682263,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.00817759595959596,
      "avg_duration_per_task_ms": 373579.7626262626,
      "avg_output_tokens_per_task": 28859.406565656565,
      "last_run_timestamp": "2026-10-24T16:12:07.782Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "thinkingmachines/inkling-small-20260730",
      "model_slug": "thinkingmachines/inkling-small",
      "model_name": "Thinking Machines: Inkling Small",
      "provider_name": "BaseTen",
      "accuracy": 0.8703701666666666,
      "accuracy_std_dev": 0.006941288747215641,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.024583196127946128,
      "avg_duration_per_task_ms": 150981.7457912458,
      "avg_output_tokens_per_task": 20283.477272727272,
      "last_run_timestamp": "2026-09-19T08:49:51.735Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Ionstream",
      "accuracy": 0.8699495,
      "accuracy_std_dev": 0.002104500000000009,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.00502325,
      "avg_duration_per_task_ms": 389434.8611111111,
      "avg_output_tokens_per_task": 17944.10101010101,
      "last_run_timestamp": "2026-08-08T10:05:07.443Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Baidu",
      "accuracy": 0.86994925,
      "accuracy_std_dev": 0.017018176890827654,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.03481110921717172,
      "avg_duration_per_task_ms": 505625.7146464646,
      "avg_output_tokens_per_task": 35174.42866161616,
      "last_run_timestamp": "2026-10-24T16:26:47.554Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "DeepInfra",
      "accuracy": 0.8698803797766019,
      "accuracy_std_dev": 0.006846731845194459,
      "run_count": 10,
      "total_questions": 1701,
      "avg_cost_per_task_usd": 0.05467326690182246,
      "avg_duration_per_task_ms": 430735.48677248677,
      "avg_output_tokens_per_task": 18124.479129923573,
      "last_run_timestamp": "2026-10-24T16:13:46.398Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "Together",
      "accuracy": 0.869809111111111,
      "accuracy_std_dev": 0.008765689330115383,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.07596271436588103,
      "avg_duration_per_task_ms": 102049.09147025814,
      "avg_output_tokens_per_task": 12345.802469135802,
      "last_run_timestamp": "2026-10-24T08:49:10.928Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "BaseTen",
      "accuracy": 0.8697483287395923,
      "accuracy_std_dev": 0.025647694972356895,
      "run_count": 18,
      "total_questions": 3483,
      "avg_cost_per_task_usd": 0.03920849038185472,
      "avg_duration_per_task_ms": 78569.80017226528,
      "avg_output_tokens_per_task": 25460.256675279932,
      "last_run_timestamp": "2026-10-24T08:27:04.531Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "CoreWeave",
      "accuracy": 0.8695285000000001,
      "accuracy_std_dev": 0.002525500000000014,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.10538721464646464,
      "avg_duration_per_task_ms": 581981.0101010101,
      "avg_output_tokens_per_task": 34695.7398989899,
      "last_run_timestamp": "2026-08-08T09:03:54.762Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "DeepInfra",
      "accuracy": 0.8694541060070672,
      "accuracy_std_dev": 0.017661177225418892,
      "run_count": 9,
      "total_questions": 1698,
      "avg_cost_per_task_usd": 0.07418639340400471,
      "avg_duration_per_task_ms": 396957.8833922261,
      "avg_output_tokens_per_task": 24048.35217903416,
      "last_run_timestamp": "2026-10-24T08:28:10.076Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Sail Research",
      "accuracy": 0.8692809647058825,
      "accuracy_std_dev": 0.009104176033557385,
      "run_count": 4,
      "total_questions": 765,
      "avg_cost_per_task_usd": 0.0047775359477124175,
      "avg_duration_per_task_ms": 482494.045751634,
      "avg_output_tokens_per_task": 26353.58431372549,
      "last_run_timestamp": "2026-08-22T16:48:48.690Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.1-20251113",
      "model_slug": "openai/gpt-5.1",
      "model_name": "OpenAI: GPT-5.1",
      "provider_name": "OpenAI",
      "accuracy": 0.8692479999999999,
      "accuracy_std_dev": 0.004827531943619518,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.22171807575757574,
      "avg_duration_per_task_ms": 260359.50673400675,
      "avg_output_tokens_per_task": 22076.11111111111,
      "last_run_timestamp": "2026-10-01T16:12:49.000Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "Anthropic",
      "accuracy": 0.8691077499999998,
      "accuracy_std_dev": 0.005877315899924065,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.003982401515151515,
      "avg_duration_per_task_ms": 37577.292929292926,
      "avg_output_tokens_per_task": 7724.75,
      "last_run_timestamp": "2026-10-24T08:16:38.969Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-5-fable-20260609",
      "model_slug": "anthropic/claude-fable-5",
      "model_name": "Anthropic: Claude Fable 5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.8689927272727271,
      "accuracy_std_dev": 0.03917244891523285,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.17977295500459137,
      "avg_duration_per_task_ms": 45320.05463728191,
      "avg_output_tokens_per_task": 3392.5229568411387,
      "last_run_timestamp": "2026-10-24T08:48:02.790Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.868687,
      "accuracy_std_dev": 0.006299080210633947,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.004035915404040404,
      "avg_duration_per_task_ms": 38787.454545454544,
      "avg_output_tokens_per_task": 7838.214646464647,
      "last_run_timestamp": "2026-10-24T08:07:26.994Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.1",
      "model_slug": "minimax/minimax-m2.1",
      "model_name": "MiniMax: MiniMax M2.1",
      "provider_name": "Novita",
      "accuracy": 0.868687,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.024268303030303032,
      "avg_duration_per_task_ms": 252794.202020202,
      "avg_output_tokens_per_task": 20048.777777777777,
      "last_run_timestamp": "2026-08-01T17:57:21.740Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "Azure",
      "accuracy": 0.8686867500000001,
      "accuracy_std_dev": 0.005583575528950976,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.004182061868686869,
      "avg_duration_per_task_ms": 36016.292929292926,
      "avg_output_tokens_per_task": 7729.580176767677,
      "last_run_timestamp": "2026-10-24T08:19:41.880Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "auto-routing",
      "accuracy": 0.8686867500000001,
      "accuracy_std_dev": 0.016015464836760137,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.038159512626262625,
      "avg_duration_per_task_ms": 691231.6849747475,
      "avg_output_tokens_per_task": 34768.117424242424,
      "last_run_timestamp": "2026-10-24T16:16:05.009Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.1-20251113",
      "model_slug": "openai/gpt-5.1",
      "model_name": "OpenAI: GPT-5.1",
      "provider_name": "Azure",
      "accuracy": 0.8686867500000001,
      "accuracy_std_dev": 0.009143975622643609,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.2541581641414141,
      "avg_duration_per_task_ms": 251149.7840909091,
      "avg_output_tokens_per_task": 24116.972222222223,
      "last_run_timestamp": "2026-10-01T16:14:35.237Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "auto-routing",
      "accuracy": 0.8686866666666666,
      "accuracy_std_dev": 0.004955863351178644,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.004029786195286195,
      "avg_duration_per_task_ms": 38652.87037037037,
      "avg_output_tokens_per_task": 7824.787878787879,
      "last_run_timestamp": "2026-10-24T08:01:41.043Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Nebius",
      "accuracy": 0.8686865,
      "accuracy_std_dev": 0.0016835000000000044,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.01774992424242424,
      "avg_duration_per_task_ms": 402650.36616161617,
      "avg_output_tokens_per_task": 29383.83585858586,
      "last_run_timestamp": "2026-10-24T08:28:10.100Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Morph",
      "accuracy": 0.8686612136752137,
      "accuracy_std_dev": 0.02320764839922872,
      "run_count": 3,
      "total_questions": 585,
      "avg_cost_per_task_usd": 0.025878321367521367,
      "avg_duration_per_task_ms": 786906.9829059829,
      "avg_output_tokens_per_task": 23783.8,
      "last_run_timestamp": "2026-08-22T10:03:34.994Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.1",
      "model_slug": "minimax/minimax-m2.1",
      "model_name": "MiniMax: MiniMax M2.1",
      "provider_name": "auto-routing",
      "accuracy": 0.868376,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 195,
      "avg_cost_per_task_usd": 0.027671912820512817,
      "avg_duration_per_task_ms": 285303.4820512821,
      "avg_output_tokens_per_task": 21385.076923076922,
      "last_run_timestamp": "2026-08-01T17:57:47.536Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Z.AI",
      "accuracy": 0.8682876231003039,
      "accuracy_std_dev": 0.028807543557374,
      "run_count": 10,
      "total_questions": 1974,
      "avg_cost_per_task_usd": 0.10277313323201621,
      "avg_duration_per_task_ms": 355978.4974670719,
      "avg_output_tokens_per_task": 23116.26595744681,
      "last_run_timestamp": "2026-10-24T08:57:08.885Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "Modal",
      "accuracy": 0.8682058571428571,
      "accuracy_std_dev": 0.007673379706651383,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.07688844300144301,
      "avg_duration_per_task_ms": 83582.13492063493,
      "avg_output_tokens_per_task": 12574.14357864358,
      "last_run_timestamp": "2026-10-10T08:21:46.324Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "Novita",
      "accuracy": 0.8676766000000001,
      "accuracy_std_dev": 0.004205309600968765,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.07033428585858588,
      "avg_duration_per_task_ms": 232782.28383838382,
      "avg_output_tokens_per_task": 11454.874242424243,
      "last_run_timestamp": "2026-10-24T08:34:00.054Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "thinkingmachines/inkling-small-20260730",
      "model_slug": "thinkingmachines/inkling-small",
      "model_name": "Thinking Machines: Inkling Small",
      "provider_name": "DeepInfra",
      "accuracy": 0.867634625,
      "accuracy_std_dev": 0.007431114989311851,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.025694838383838382,
      "avg_duration_per_task_ms": 248942.36553030304,
      "avg_output_tokens_per_task": 21167.008207070707,
      "last_run_timestamp": "2026-10-08T22:14:16.087Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "Google",
      "accuracy": 0.8675645,
      "accuracy_std_dev": 0.008228295808772726,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.004289666245791245,
      "avg_duration_per_task_ms": 35945.74074074074,
      "avg_output_tokens_per_task": 7815.2925084175085,
      "last_run_timestamp": "2026-10-24T08:43:41.154Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.867003,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.013525050505050505,
      "avg_duration_per_task_ms": 181228.28787878787,
      "avg_output_tokens_per_task": 22673.38888888889,
      "last_run_timestamp": "2026-10-24T08:02:32.162Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "auto-routing",
      "accuracy": 0.8666971818181819,
      "accuracy_std_dev": 0.02628554379335525,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.007686401744719926,
      "avg_duration_per_task_ms": 446263.08815426996,
      "avg_output_tokens_per_task": 25752.568411386594,
      "last_run_timestamp": "2026-10-24T08:15:15.304Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Phala",
      "accuracy": 0.8664308674130818,
      "accuracy_std_dev": 0.027802409922633914,
      "run_count": 9,
      "total_questions": 1697,
      "avg_cost_per_task_usd": 0.011504347672362995,
      "avg_duration_per_task_ms": 435596.48084855627,
      "avg_output_tokens_per_task": 25872.670005892753,
      "last_run_timestamp": "2026-10-24T08:11:43.556Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8660681111111111,
      "accuracy_std_dev": 0.016724420165630353,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.07373856228956228,
      "avg_duration_per_task_ms": 287527.3310886644,
      "avg_output_tokens_per_task": 20979.62962962963,
      "last_run_timestamp": "2026-10-24T16:08:41.781Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "thinkingmachines/inkling-small-20260730",
      "model_slug": "thinkingmachines/inkling-small",
      "model_name": "Thinking Machines: Inkling Small",
      "provider_name": "auto-routing",
      "accuracy": 0.8656005,
      "accuracy_std_dev": 0.00803956708515244,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.02525451262626263,
      "avg_duration_per_task_ms": 233486.3627946128,
      "avg_output_tokens_per_task": 20811.0765993266,
      "last_run_timestamp": "2026-10-01T16:08:04.290Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "DeepInfra",
      "accuracy": 0.8653352747021724,
      "accuracy_std_dev": 0.017080671231962035,
      "run_count": 8,
      "total_questions": 1427,
      "avg_cost_per_task_usd": 0.005114180098107919,
      "avg_duration_per_task_ms": 702127.0336370007,
      "avg_output_tokens_per_task": 28117.81079187106,
      "last_run_timestamp": "2026-10-24T16:15:47.409Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Relace",
      "accuracy": 0.86532,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.01495640909090909,
      "avg_duration_per_task_ms": 120441.0101010101,
      "avg_output_tokens_per_task": 29795.70707070707,
      "last_run_timestamp": "2026-10-24T15:00:02.228Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "NextBit",
      "accuracy": 0.8653195,
      "accuracy_std_dev": 0.0016835000000000044,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.008587179292929294,
      "avg_duration_per_task_ms": 240899.30303030304,
      "avg_output_tokens_per_task": 30387.573232323233,
      "last_run_timestamp": "2026-08-29T18:38:15.907Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Fireworks",
      "accuracy": 0.865101093922652,
      "accuracy_std_dev": 0.005917000000000006,
      "run_count": 2,
      "total_questions": 362,
      "avg_cost_per_task_usd": 0.005541756906077348,
      "avg_duration_per_task_ms": 151652.73480662983,
      "avg_output_tokens_per_task": 19668.124309392264,
      "last_run_timestamp": "2026-08-08T09:25:45.716Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-sonnet-5-20260630",
      "model_slug": "anthropic/claude-sonnet-5",
      "model_name": "Anthropic: Claude Sonnet 5",
      "provider_name": "Azure",
      "accuracy": 0.8645117600000001,
      "accuracy_std_dev": 0.02767954650102491,
      "run_count": 25,
      "total_questions": 4950,
      "avg_cost_per_task_usd": 0.15895250323232327,
      "avg_duration_per_task_ms": 153406.69333333333,
      "avg_output_tokens_per_task": 15378.712121212122,
      "last_run_timestamp": "2026-10-24T08:20:16.287Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Cohere",
      "accuracy": 0.8644780000000001,
      "accuracy_std_dev": 0.01094299999999998,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.05228046464646464,
      "avg_duration_per_task_ms": 195135.57323232322,
      "avg_output_tokens_per_task": 43383.04797979798,
      "last_run_timestamp": "2026-10-24T08:15:15.306Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.6-opus-20260205",
      "model_slug": "anthropic/claude-opus-4.6",
      "model_name": "Anthropic: Claude Opus 4.6",
      "provider_name": "auto-routing",
      "accuracy": 0.86433775,
      "accuracy_std_dev": 0.014059447115285145,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.7497476914983165,
      "avg_duration_per_task_ms": 499048.3042929293,
      "avg_output_tokens_per_task": 29826.423821548822,
      "last_run_timestamp": "2026-10-24T08:11:56.068Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-5-fable-20260609",
      "model_slug": "anthropic/claude-fable-5",
      "model_name": "Anthropic: Claude Fable 5",
      "provider_name": "Anthropic",
      "accuracy": 0.8642836153846152,
      "accuracy_std_dev": 0.035867989623357806,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.18533466744366744,
      "avg_duration_per_task_ms": 45806.41025641026,
      "avg_output_tokens_per_task": 3500.1981351981353,
      "last_run_timestamp": "2026-10-24T08:45:03.898Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-5-fable-20260609",
      "model_slug": "anthropic/claude-fable-5",
      "model_name": "Anthropic: Claude Fable 5",
      "provider_name": "auto-routing",
      "accuracy": 0.8638953076923075,
      "accuracy_std_dev": 0.040435466544539306,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.1779076884226884,
      "avg_duration_per_task_ms": 43919.818181818184,
      "avg_output_tokens_per_task": 3351.4957264957266,
      "last_run_timestamp": "2026-10-24T08:27:26.672Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Nebius",
      "accuracy": 0.863636,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0938164393939394,
      "avg_duration_per_task_ms": 667365.3939393939,
      "avg_output_tokens_per_task": 23446.50505050505,
      "last_run_timestamp": "2026-09-10T01:44:54.935Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.1-20251113",
      "model_slug": "openai/gpt-5.1",
      "model_name": "OpenAI: GPT-5.1",
      "provider_name": "auto-routing",
      "accuracy": 0.863636,
      "accuracy_std_dev": 0.004761657064510171,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.22125699158249157,
      "avg_duration_per_task_ms": 259075.24915824915,
      "avg_output_tokens_per_task": 22030.511784511786,
      "last_run_timestamp": "2026-10-01T16:12:54.241Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.1",
      "model_slug": "minimax/minimax-m2.1",
      "model_name": "MiniMax: MiniMax M2.1",
      "provider_name": "Minimax",
      "accuracy": 0.8635951463414634,
      "accuracy_std_dev": 0.01668149999999996,
      "run_count": 2,
      "total_questions": 369,
      "avg_cost_per_task_usd": 0.047886756097560976,
      "avg_duration_per_task_ms": 401842.43631436315,
      "avg_output_tokens_per_task": 25387.241192411922,
      "last_run_timestamp": "2026-08-01T17:44:01.464Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "Novita",
      "accuracy": 0.8635504804532578,
      "accuracy_std_dev": 0.013176767872113233,
      "run_count": 9,
      "total_questions": 1765,
      "avg_cost_per_task_usd": 0.07539817337110481,
      "avg_duration_per_task_ms": 281436.6181303116,
      "avg_output_tokens_per_task": 20817.862322946177,
      "last_run_timestamp": "2026-10-24T16:12:49.070Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.5-opus-20251124",
      "model_slug": "anthropic/claude-opus-4.5",
      "model_name": "Anthropic: Claude Opus 4.5",
      "provider_name": "auto-routing",
      "accuracy": 0.8634960000000002,
      "accuracy_std_dev": 0.01008821268444846,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.8421511931818182,
      "avg_duration_per_task_ms": 412180.98316498316,
      "avg_output_tokens_per_task": 33494.958333333336,
      "last_run_timestamp": "2026-10-24T08:26:00.557Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Venice",
      "accuracy": 0.8634435114503817,
      "accuracy_std_dev": 0.01366835982067709,
      "run_count": 8,
      "total_questions": 1572,
      "avg_cost_per_task_usd": 0.13531842366412217,
      "avg_duration_per_task_ms": 289599.17048346053,
      "avg_output_tokens_per_task": 28034.88358778626,
      "last_run_timestamp": "2026-10-17T08:23:25.121Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "Fireworks",
      "accuracy": 0.86321525,
      "accuracy_std_dev": 0.010127171825712258,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.022006194444444443,
      "avg_duration_per_task_ms": 265798.13636363635,
      "avg_output_tokens_per_task": 33108.22601010101,
      "last_run_timestamp": "2026-09-19T08:51:58.825Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Relace",
      "accuracy": 0.862963,
      "accuracy_std_dev": 0.009134508372101892,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.06279948181818182,
      "avg_duration_per_task_ms": 441139.7515151515,
      "avg_output_tokens_per_task": 56049.472727272725,
      "last_run_timestamp": "2026-10-24T16:14:38.647Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "GMICloud",
      "accuracy": 0.8629629000000001,
      "accuracy_std_dev": 0.053333789303686284,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.10982828333333332,
      "avg_duration_per_task_ms": 474004.9994949495,
      "avg_output_tokens_per_task": 27746.173737373738,
      "last_run_timestamp": "2026-10-24T16:35:28.009Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.6-opus-20260205",
      "model_slug": "anthropic/claude-opus-4.6",
      "model_name": "Anthropic: Claude Opus 4.6",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.8629348333333335,
      "accuracy_std_dev": 0.008541693195080755,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.8005278598484848,
      "avg_duration_per_task_ms": 459239.8367003367,
      "avg_output_tokens_per_task": 31848.97138047138,
      "last_run_timestamp": "2026-10-24T08:27:05.617Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8626261,
      "accuracy_std_dev": 0.011472416183611877,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.07024428484848486,
      "avg_duration_per_task_ms": 231811.16010101011,
      "avg_output_tokens_per_task": 11437.808585858586,
      "last_run_timestamp": "2026-10-24T08:31:24.553Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Baidu",
      "accuracy": 0.8625841249999999,
      "accuracy_std_dev": 0.006783236993454885,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.004199931818181818,
      "avg_duration_per_task_ms": 250234.78724747474,
      "avg_output_tokens_per_task": 29101.075757575756,
      "last_run_timestamp": "2026-10-24T16:02:06.935Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Phala",
      "accuracy": 0.8625343605313093,
      "accuracy_std_dev": 0.061262981528279976,
      "run_count": 8,
      "total_questions": 1581,
      "avg_cost_per_task_usd": 0.2509589614168248,
      "avg_duration_per_task_ms": 1179158.6533839342,
      "avg_output_tokens_per_task": 55775.307400379505,
      "last_run_timestamp": "2026-10-24T16:03:06.207Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "StreamLake",
      "accuracy": 0.861993011958763,
      "accuracy_std_dev": 0.030079142720378513,
      "run_count": 13,
      "total_questions": 2425,
      "avg_cost_per_task_usd": 0.02418231587628866,
      "avg_duration_per_task_ms": 455044.95463917527,
      "avg_output_tokens_per_task": 23028.103917525772,
      "last_run_timestamp": "2026-10-17T08:18:04.346Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Ionstream",
      "accuracy": 0.8619530000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.008309641414141413,
      "avg_duration_per_task_ms": 730613.3737373737,
      "avg_output_tokens_per_task": 29358.666666666668,
      "last_run_timestamp": "2026-07-31T04:09:58.076Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "auto-routing",
      "accuracy": 0.8619528181818182,
      "accuracy_std_dev": 0.017198547374273336,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.09933105968778695,
      "avg_duration_per_task_ms": 470423.74885215797,
      "avg_output_tokens_per_task": 26261.255280073463,
      "last_run_timestamp": "2026-10-24T16:12:09.459Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Azure",
      "accuracy": 0.8619527499999999,
      "accuracy_std_dev": 0.02717181928832702,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.13016119191919193,
      "avg_duration_per_task_ms": 482692.7720959596,
      "avg_output_tokens_per_task": 34118.55934343435,
      "last_run_timestamp": "2026-10-24T16:05:51.240Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "StreamLake",
      "accuracy": 0.8617843,
      "accuracy_std_dev": 0.016622379589276627,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.09082832878787879,
      "avg_duration_per_task_ms": 184194.31868686868,
      "avg_output_tokens_per_task": 25099.163131313133,
      "last_run_timestamp": "2026-10-24T16:20:59.111Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "CoreWeave",
      "accuracy": 0.861708258064516,
      "accuracy_std_dev": 0.012137624210372552,
      "run_count": 6,
      "total_questions": 1116,
      "avg_cost_per_task_usd": 0.02452295071684588,
      "avg_duration_per_task_ms": 145394.43727598566,
      "avg_output_tokens_per_task": 37540.51523297491,
      "last_run_timestamp": "2026-10-24T08:12:02.536Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-sonnet-5-20260630",
      "model_slug": "anthropic/claude-sonnet-5",
      "model_name": "Anthropic: Claude Sonnet 5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.8616797297297295,
      "accuracy_std_dev": 0.027641986931017132,
      "run_count": 37,
      "total_questions": 7326,
      "avg_cost_per_task_usd": 0.1728344180999181,
      "avg_duration_per_task_ms": 158861.68086268086,
      "avg_output_tokens_per_task": 16129.410865410866,
      "last_run_timestamp": "2026-10-24T08:48:23.815Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "Alibaba",
      "accuracy": 0.8614478,
      "accuracy_std_dev": 0.008075538394435377,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.06555243737373737,
      "avg_duration_per_task_ms": 217588.9297979798,
      "avg_output_tokens_per_task": 10657.188383838384,
      "last_run_timestamp": "2026-10-24T08:11:43.559Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Ambient",
      "accuracy": 0.8613915000000001,
      "accuracy_std_dev": 0.04603966955336814,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.004525512626262627,
      "avg_duration_per_task_ms": 565963.2584175084,
      "avg_output_tokens_per_task": 18978.082491582492,
      "last_run_timestamp": "2026-08-29T10:57:26.955Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.5-opus-20251124",
      "model_slug": "anthropic/claude-opus-4.5",
      "model_name": "Anthropic: Claude Opus 4.5",
      "provider_name": "Anthropic",
      "accuracy": 0.8613514285714288,
      "accuracy_std_dev": 0.011656715536635543,
      "run_count": 14,
      "total_questions": 2772,
      "avg_cost_per_task_usd": 0.842712674963925,
      "avg_duration_per_task_ms": 456401.95995670994,
      "avg_output_tokens_per_task": 33529.13383838384,
      "last_run_timestamp": "2026-10-24T08:15:51.139Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Morph",
      "accuracy": 0.8612792,
      "accuracy_std_dev": 0.03136197207702347,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.005120404040404041,
      "avg_duration_per_task_ms": 530889.794949495,
      "avg_output_tokens_per_task": 17542.636363636364,
      "last_run_timestamp": "2026-09-05T09:22:54.724Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Novita",
      "accuracy": 0.861111,
      "accuracy_std_dev": 0.04797999999999997,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.021780800505050504,
      "avg_duration_per_task_ms": 138470.81818181818,
      "avg_output_tokens_per_task": 20189.02525252525,
      "last_run_timestamp": "2026-10-24T08:12:37.562Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "ModelRun",
      "accuracy": 0.861111,
      "accuracy_std_dev": 0.0025249999999999995,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.01767211111111111,
      "avg_duration_per_task_ms": 125210.21717171717,
      "avg_output_tokens_per_task": 29259.424242424244,
      "last_run_timestamp": "2026-10-24T08:13:46.722Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.5-opus-20251124",
      "model_slug": "anthropic/claude-opus-4.5",
      "model_name": "Anthropic: Claude Opus 4.5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.8609707916666667,
      "accuracy_std_dev": 0.007924941871391254,
      "run_count": 24,
      "total_questions": 4752,
      "avg_cost_per_task_usd": 0.8830173162878786,
      "avg_duration_per_task_ms": 414862.5126262626,
      "avg_output_tokens_per_task": 33448.23737373737,
      "last_run_timestamp": "2026-10-24T08:44:41.418Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "DekaLLM",
      "accuracy": 0.8608923543307087,
      "accuracy_std_dev": 0.017439499999999997,
      "run_count": 2,
      "total_questions": 254,
      "avg_cost_per_task_usd": 0.003837814960629921,
      "avg_duration_per_task_ms": 298144.1771653543,
      "avg_output_tokens_per_task": 11523.807086614173,
      "last_run_timestamp": "2026-10-03T17:04:09.464Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "InferenceNet",
      "accuracy": 0.8605965053212455,
      "accuracy_std_dev": 0.03960090537740342,
      "run_count": 13,
      "total_questions": 2537,
      "avg_cost_per_task_usd": 0.01657229877808435,
      "avg_duration_per_task_ms": 463843.5068979109,
      "avg_output_tokens_per_task": 41792.180134016555,
      "last_run_timestamp": "2026-10-24T08:58:59.550Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "DigitalOcean",
      "accuracy": 0.8602692499999999,
      "accuracy_std_dev": 0.07328871721500406,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.0069893973063973065,
      "avg_duration_per_task_ms": 603844.6805555555,
      "avg_output_tokens_per_task": 28814.18181818182,
      "last_run_timestamp": "2026-10-24T08:23:31.990Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "reflection/beam-1.0-501b-a23b-20261016",
      "model_slug": "reflection/beam-1.0-501b-a23b",
      "model_name": "Reflection: Beam 1.0 501B A23B",
      "provider_name": "Io Net",
      "accuracy": 0.860269,
      "accuracy_std_dev": 0.010100999999999971,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.00987439393939394,
      "avg_duration_per_task_ms": 74831.28787878787,
      "avg_output_tokens_per_task": 9997.058080808081,
      "last_run_timestamp": "2026-10-24T08:14:36.264Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "CoreWeave",
      "accuracy": 0.8601822310030396,
      "accuracy_std_dev": 0.013547575585321513,
      "run_count": 5,
      "total_questions": 987,
      "avg_cost_per_task_usd": 0.007783161094224923,
      "avg_duration_per_task_ms": 526386.7112462006,
      "avg_output_tokens_per_task": 27793.246200607904,
      "last_run_timestamp": "2026-08-29T18:49:16.416Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "auto-routing",
      "accuracy": 0.8601809488313328,
      "accuracy_std_dev": 0.03964118133468496,
      "run_count": 8,
      "total_questions": 1583,
      "avg_cost_per_task_usd": 0.16196978711307647,
      "avg_duration_per_task_ms": 990576.0770688566,
      "avg_output_tokens_per_task": 55234.00947567909,
      "last_run_timestamp": "2026-10-24T16:06:09.255Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-5-fable-20260609",
      "model_slug": "anthropic/claude-fable-5",
      "model_name": "Anthropic: Claude Fable 5",
      "provider_name": "Azure",
      "accuracy": 0.8600102307692308,
      "accuracy_std_dev": 0.03934806390536378,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.18072842657342658,
      "avg_duration_per_task_ms": 45766.77855477855,
      "avg_output_tokens_per_task": 3406.542735042735,
      "last_run_timestamp": "2026-10-24T08:08:40.733Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.5-opus-20251124",
      "model_slug": "anthropic/claude-opus-4.5",
      "model_name": "Anthropic: Claude Opus 4.5",
      "provider_name": "Google",
      "accuracy": 0.8599885833333334,
      "accuracy_std_dev": 0.009617771098842076,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.8533179797979797,
      "avg_duration_per_task_ms": 482512.8122895623,
      "avg_output_tokens_per_task": 33941.89983164983,
      "last_run_timestamp": "2026-10-24T08:14:51.084Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-sonnet-5-20260630",
      "model_slug": "anthropic/claude-sonnet-5",
      "model_name": "Anthropic: Claude Sonnet 5",
      "provider_name": "Anthropic",
      "accuracy": 0.8598806153846154,
      "accuracy_std_dev": 0.030036815288878832,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.15524810489510485,
      "avg_duration_per_task_ms": 152128.8811188811,
      "avg_output_tokens_per_task": 15307.565656565657,
      "last_run_timestamp": "2026-10-24T08:12:37.564Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "InferenceNet",
      "accuracy": 0.8597419723629663,
      "accuracy_std_dev": 0.016505636989518208,
      "run_count": 11,
      "total_questions": 2171,
      "avg_cost_per_task_usd": 0.08549684339014277,
      "avg_duration_per_task_ms": 179079.73422385997,
      "avg_output_tokens_per_task": 24385.910640257945,
      "last_run_timestamp": "2026-10-24T08:41:25.897Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Inceptron",
      "accuracy": 0.8596613057736721,
      "accuracy_std_dev": 0.0311431999096363,
      "run_count": 11,
      "total_questions": 2165,
      "avg_cost_per_task_usd": 0.1269020115473441,
      "avg_duration_per_task_ms": 528998.0508083141,
      "avg_output_tokens_per_task": 43596.739953810626,
      "last_run_timestamp": "2026-10-24T16:12:49.068Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.6-opus-20260205",
      "model_slug": "anthropic/claude-opus-4.6",
      "model_name": "Anthropic: Claude Opus 4.6",
      "provider_name": "Azure",
      "accuracy": 0.8594274166666667,
      "accuracy_std_dev": 0.018601260147538098,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.7585617697811449,
      "avg_duration_per_task_ms": 508230.9494949495,
      "avg_output_tokens_per_task": 30213.497474747473,
      "last_run_timestamp": "2026-10-24T08:10:23.014Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "Moonshot AI",
      "accuracy": 0.8592592,
      "accuracy_std_dev": 0.0373207358952098,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.12732033636363638,
      "avg_duration_per_task_ms": 630447.8414141415,
      "avg_output_tokens_per_task": 42741.22323232323,
      "last_run_timestamp": "2026-08-29T18:25:54.650Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Fireworks",
      "accuracy": 0.8586479490968801,
      "accuracy_std_dev": 0.00913590390487393,
      "run_count": 22,
      "total_questions": 3654,
      "avg_cost_per_task_usd": 0.17185495894909686,
      "avg_duration_per_task_ms": 250875.80268199233,
      "avg_output_tokens_per_task": 31401.503831417624,
      "last_run_timestamp": "2026-09-26T08:54:22.466Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "DigitalOcean",
      "accuracy": 0.8586340031847132,
      "accuracy_std_dev": 0.05344060455646062,
      "run_count": 10,
      "total_questions": 1884,
      "avg_cost_per_task_usd": 0.0115320483014862,
      "avg_duration_per_task_ms": 468303.98089171975,
      "avg_output_tokens_per_task": 23774.996815286624,
      "last_run_timestamp": "2026-10-24T08:32:58.238Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.858586,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.08651329797979797,
      "avg_duration_per_task_ms": 499241.5707070707,
      "avg_output_tokens_per_task": 35858.23737373737,
      "last_run_timestamp": "2026-10-04T03:06:10.317Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Chutes",
      "accuracy": 0.85858575,
      "accuracy_std_dev": 0.021526546714173627,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.1476876622474748,
      "avg_duration_per_task_ms": 1253876.004419192,
      "avg_output_tokens_per_task": 52218.26641414141,
      "last_run_timestamp": "2026-10-24T16:26:15.939Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "InferenceNet",
      "accuracy": 0.8585855,
      "accuracy_std_dev": 0.012710146271778291,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.014605150252525251,
      "avg_duration_per_task_ms": 477497.9494949495,
      "avg_output_tokens_per_task": 26944.89898989899,
      "last_run_timestamp": "2026-10-24T08:15:32.562Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "SiliconFlow",
      "accuracy": 0.858164875,
      "accuracy_std_dev": 0.016185868044667075,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.1437290511363636,
      "avg_duration_per_task_ms": 908116.1098484849,
      "avg_output_tokens_per_task": 45706.13320707071,
      "last_run_timestamp": "2026-10-24T16:14:06.777Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.6-opus-20260205",
      "model_slug": "anthropic/claude-opus-4.6",
      "model_name": "Anthropic: Claude Opus 4.6",
      "provider_name": "Anthropic",
      "accuracy": 0.8577439285714286,
      "accuracy_std_dev": 0.016868073763619186,
      "run_count": 14,
      "total_questions": 2772,
      "avg_cost_per_task_usd": 0.7558037463924961,
      "avg_duration_per_task_ms": 505406.40187590185,
      "avg_output_tokens_per_task": 30123.084054834057,
      "last_run_timestamp": "2026-10-24T08:30:49.044Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.6-opus-20260205",
      "model_slug": "anthropic/claude-opus-4.6",
      "model_name": "Anthropic: Claude Opus 4.6",
      "provider_name": "Google",
      "accuracy": 0.8576738333333334,
      "accuracy_std_dev": 0.01763649745004817,
      "run_count": 24,
      "total_questions": 4752,
      "avg_cost_per_task_usd": 0.8048634583333334,
      "avg_duration_per_task_ms": 484140.92781986535,
      "avg_output_tokens_per_task": 30507.85164141414,
      "last_run_timestamp": "2026-10-24T08:15:15.299Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "DigitalOcean",
      "accuracy": 0.857663467955239,
      "accuracy_std_dev": 0.05434631352575813,
      "run_count": 10,
      "total_questions": 1966,
      "avg_cost_per_task_usd": 0.0781413972533062,
      "avg_duration_per_task_ms": 509131.62817904376,
      "avg_output_tokens_per_task": 21110.80518819939,
      "last_run_timestamp": "2026-10-24T08:57:31.359Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "Minimax",
      "accuracy": 0.8576418156329653,
      "accuracy_std_dev": 0.007837726493074803,
      "run_count": 18,
      "total_questions": 3531,
      "avg_cost_per_task_usd": 0.06304419881053526,
      "avg_duration_per_task_ms": 588364.0832625319,
      "avg_output_tokens_per_task": 34740.37921268762,
      "last_run_timestamp": "2026-10-24T16:03:59.167Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3-ultra-550b-a55b-20260604",
      "model_slug": "nvidia/nemotron-3-ultra-550b-a55b",
      "model_name": "NVIDIA: Nemotron 3 Ultra",
      "provider_name": "BaseTen",
      "accuracy": 0.857418,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 173,
      "avg_cost_per_task_usd": 0.08778269942196533,
      "avg_duration_per_task_ms": 317630.5202312139,
      "avg_output_tokens_per_task": 36575.57225433526,
      "last_run_timestamp": "2026-08-04T03:12:38.954Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.5-opus-20251124",
      "model_slug": "anthropic/claude-opus-4.5",
      "model_name": "Anthropic: Claude Opus 4.5",
      "provider_name": "Azure",
      "accuracy": 0.8573231666666665,
      "accuracy_std_dev": 0.0170736054571246,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.8399580702861953,
      "avg_duration_per_task_ms": 458034.78324915824,
      "avg_output_tokens_per_task": 33422.68813131313,
      "last_run_timestamp": "2026-10-24T08:13:46.758Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Cloudflare",
      "accuracy": 0.8571887570077085,
      "accuracy_std_dev": 0.03653443070150671,
      "run_count": 33,
      "total_questions": 5708,
      "avg_cost_per_task_usd": 0.015801940259285212,
      "avg_duration_per_task_ms": 401037.0720042046,
      "avg_output_tokens_per_task": 17993.8771899089,
      "last_run_timestamp": "2026-10-24T08:49:56.512Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Together",
      "accuracy": 0.8571427108013937,
      "accuracy_std_dev": 0.01011902539224449,
      "run_count": 3,
      "total_questions": 574,
      "avg_cost_per_task_usd": 0.11599589721254357,
      "avg_duration_per_task_ms": 363022.99128919863,
      "avg_output_tokens_per_task": 29391.742160278747,
      "last_run_timestamp": "2026-08-22T10:19:36.962Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.5-opus-20251124",
      "model_slug": "anthropic/claude-opus-4.5",
      "model_name": "Anthropic: Claude Opus 4.5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.8570552727272727,
      "accuracy_std_dev": 0.02678680556845074,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.8339512213039484,
      "avg_duration_per_task_ms": 436211.7764003673,
      "avg_output_tokens_per_task": 33251.8347107438,
      "last_run_timestamp": "2026-10-24T08:08:52.220Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "BaseTen",
      "accuracy": 0.8569032243034418,
      "accuracy_std_dev": 0.036552330331727854,
      "run_count": 22,
      "total_questions": 4271,
      "avg_cost_per_task_usd": 0.0041591236244439235,
      "avg_duration_per_task_ms": 82317.1749004917,
      "avg_output_tokens_per_task": 15615.828377429174,
      "last_run_timestamp": "2026-10-24T08:41:46.943Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Wafer",
      "accuracy": 0.8569025,
      "accuracy_std_dev": 0.005050499999999958,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.018898722222222223,
      "avg_duration_per_task_ms": 164757.63383838383,
      "avg_output_tokens_per_task": 31337.035353535353,
      "last_run_timestamp": "2026-10-24T08:17:56.393Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5-2025-08-07",
      "model_slug": "openai/gpt-5",
      "model_name": "OpenAI: GPT-5",
      "provider_name": "auto-routing",
      "accuracy": 0.8569019999999999,
      "accuracy_std_dev": 0.0033669999999999534,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.20989276262626264,
      "avg_duration_per_task_ms": 263850.72222222225,
      "avg_output_tokens_per_task": 20893.070707070707,
      "last_run_timestamp": "2026-10-01T16:31:36.833Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "BaseTen",
      "accuracy": 0.8568232764406386,
      "accuracy_std_dev": 0.020345851767133397,
      "run_count": 44,
      "total_questions": 8642,
      "avg_cost_per_task_usd": 0.2801212903263133,
      "avg_duration_per_task_ms": 334492.7875491784,
      "avg_output_tokens_per_task": 50394.622193936586,
      "last_run_timestamp": "2026-10-24T16:37:01.548Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Near AI",
      "accuracy": 0.8568129407236336,
      "accuracy_std_dev": 0.01596270457778066,
      "run_count": 8,
      "total_questions": 1299,
      "avg_cost_per_task_usd": 0.004350760585065434,
      "avg_duration_per_task_ms": 410967.9938414165,
      "avg_output_tokens_per_task": 9402.565819861431,
      "last_run_timestamp": "2026-10-24T08:24:10.666Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "AtlasCloud",
      "accuracy": 0.856442089983022,
      "accuracy_std_dev": 0.021498967023132247,
      "run_count": 9,
      "total_questions": 1767,
      "avg_cost_per_task_usd": 0.04329065647990945,
      "avg_duration_per_task_ms": 650523.3395585739,
      "avg_output_tokens_per_task": 36107.27447651386,
      "last_run_timestamp": "2026-10-24T16:04:33.484Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-5-fable-20260609",
      "model_slug": "anthropic/claude-fable-5",
      "model_name": "Anthropic: Claude Fable 5",
      "provider_name": "Google",
      "accuracy": 0.8563635600000002,
      "accuracy_std_dev": 0.050198181182254005,
      "run_count": 25,
      "total_questions": 4950,
      "avg_cost_per_task_usd": 0.18820920787878787,
      "avg_duration_per_task_ms": 41314.649292929294,
      "avg_output_tokens_per_task": 3389.456767676768,
      "last_run_timestamp": "2026-10-24T08:44:43.918Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Crusoe",
      "accuracy": 0.856164284931507,
      "accuracy_std_dev": 0.05295168533540149,
      "run_count": 8,
      "total_questions": 1460,
      "avg_cost_per_task_usd": 0.18168324041095887,
      "avg_duration_per_task_ms": 1626314.2294520547,
      "avg_output_tokens_per_task": 52899.55684931507,
      "last_run_timestamp": "2026-10-24T16:16:05.013Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Azure",
      "accuracy": 0.8558499999999999,
      "accuracy_std_dev": 0.03281670770735542,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.015880464015151512,
      "avg_duration_per_task_ms": 344363.2039141414,
      "avg_output_tokens_per_task": 28437.20138888889,
      "last_run_timestamp": "2026-10-24T16:21:33.540Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Inceptron",
      "accuracy": 0.8558169770378875,
      "accuracy_std_dev": 0.018638102580532955,
      "run_count": 9,
      "total_questions": 1742,
      "avg_cost_per_task_usd": 0.08209278185993112,
      "avg_duration_per_task_ms": 337904.9672789897,
      "avg_output_tokens_per_task": 28132.247990815154,
      "last_run_timestamp": "2026-10-24T08:19:59.729Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "auto-routing",
      "accuracy": 0.8554993333333334,
      "accuracy_std_dev": 0.009394238245979377,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.005048751683501684,
      "avg_duration_per_task_ms": 532489.6868686868,
      "avg_output_tokens_per_task": 30189.478956228955,
      "last_run_timestamp": "2026-10-24T16:02:08.817Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Baidu",
      "accuracy": 0.85530295,
      "accuracy_std_dev": 0.01065364351513132,
      "run_count": 20,
      "total_questions": 3960,
      "avg_cost_per_task_usd": 0.14301361767676768,
      "avg_duration_per_task_ms": 429371.0404040404,
      "avg_output_tokens_per_task": 43957.30757575758,
      "last_run_timestamp": "2026-10-24T16:26:47.544Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "Xiaomi",
      "accuracy": 0.8550657272727274,
      "accuracy_std_dev": 0.012107062717806176,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.02919473599632691,
      "avg_duration_per_task_ms": 757966.051882461,
      "avg_output_tokens_per_task": 33302.242883379244,
      "last_run_timestamp": "2026-10-17T08:19:33.830Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Together",
      "accuracy": 0.8549618244274809,
      "accuracy_std_dev": 0.005347999999999964,
      "run_count": 2,
      "total_questions": 393,
      "avg_cost_per_task_usd": 0.22707390839694658,
      "avg_duration_per_task_ms": 373958.2061068702,
      "avg_output_tokens_per_task": 50795.681933842236,
      "last_run_timestamp": "2026-08-15T08:57:45.830Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Nebius",
      "accuracy": 0.8543647577679137,
      "accuracy_std_dev": 0.08472615418364036,
      "run_count": 8,
      "total_questions": 1577,
      "avg_cost_per_task_usd": 0.11778292010145847,
      "avg_duration_per_task_ms": 1596925.525681674,
      "avg_output_tokens_per_task": 29301.902980342424,
      "last_run_timestamp": "2026-10-24T16:28:09.530Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Minimax",
      "accuracy": 0.8539973475609757,
      "accuracy_std_dev": 0.0065049450612591365,
      "run_count": 10,
      "total_questions": 1968,
      "avg_cost_per_task_usd": 0.054438141260162595,
      "avg_duration_per_task_ms": 424693.43699186994,
      "avg_output_tokens_per_task": 27839.288617886177,
      "last_run_timestamp": "2026-08-29T19:17:08.288Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "GMICloud",
      "accuracy": 0.8536708232911394,
      "accuracy_std_dev": 0.030648058110751494,
      "run_count": 10,
      "total_questions": 1975,
      "avg_cost_per_task_usd": 0.008188765063291139,
      "avg_duration_per_task_ms": 458493.6015189873,
      "avg_output_tokens_per_task": 28024.251139240507,
      "last_run_timestamp": "2026-10-24T08:43:03.068Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "DeepSeek",
      "accuracy": 0.853535,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.028549383838383836,
      "avg_duration_per_task_ms": 582750.7070707071,
      "avg_output_tokens_per_task": 33383.742424242424,
      "last_run_timestamp": "2026-08-08T09:24:32.269Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "DekaLLM",
      "accuracy": 0.8530642646239555,
      "accuracy_std_dev": 0.022329477337983097,
      "run_count": 4,
      "total_questions": 718,
      "avg_cost_per_task_usd": 0.02860370194986072,
      "avg_duration_per_task_ms": 460751.2409470752,
      "avg_output_tokens_per_task": 34791.85933147632,
      "last_run_timestamp": "2026-10-10T08:22:41.628Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-sonnet-5-20260630",
      "model_slug": "anthropic/claude-sonnet-5",
      "model_name": "Anthropic: Claude Sonnet 5",
      "provider_name": "Google",
      "accuracy": 0.8530032199766809,
      "accuracy_std_dev": 0.035466565814003675,
      "run_count": 39,
      "total_questions": 7719,
      "avg_cost_per_task_usd": 0.1653298276978883,
      "avg_duration_per_task_ms": 151135.82432957637,
      "avg_output_tokens_per_task": 15282.314807617568,
      "last_run_timestamp": "2026-10-24T08:32:13.238Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Wafer",
      "accuracy": 0.8528510680564071,
      "accuracy_std_dev": 0.0820596725269595,
      "run_count": 17,
      "total_questions": 3262,
      "avg_cost_per_task_usd": 0.2673039751686082,
      "avg_duration_per_task_ms": 520792.21551195584,
      "avg_output_tokens_per_task": 67728.07817290006,
      "last_run_timestamp": "2026-10-24T09:05:17.850Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Baidu",
      "accuracy": 0.8528355426944971,
      "accuracy_std_dev": 0.013503895836017096,
      "run_count": 8,
      "total_questions": 1581,
      "avg_cost_per_task_usd": 0.0925430898165718,
      "avg_duration_per_task_ms": 296081.5515496521,
      "avg_output_tokens_per_task": 29809.33523086654,
      "last_run_timestamp": "2026-10-24T08:21:56.261Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "DekaLLM",
      "accuracy": 0.8524375810276679,
      "accuracy_std_dev": 0.011217700923094702,
      "run_count": 4,
      "total_questions": 759,
      "avg_cost_per_task_usd": 0.07750415546772069,
      "avg_duration_per_task_ms": 263910.441370224,
      "avg_output_tokens_per_task": 26225.246376811596,
      "last_run_timestamp": "2026-10-24T08:40:56.292Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "auto-routing",
      "accuracy": 0.8524129999999999,
      "accuracy_std_dev": 0.012672927610549278,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.1049078827160494,
      "avg_duration_per_task_ms": 476715.49887766555,
      "avg_output_tokens_per_task": 30743.156565656565,
      "last_run_timestamp": "2026-10-24T08:15:15.299Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Sail Research",
      "accuracy": 0.8522881577868852,
      "accuracy_std_dev": 0.009664339658766143,
      "run_count": 5,
      "total_questions": 976,
      "avg_cost_per_task_usd": 0.09954490881147542,
      "avg_duration_per_task_ms": 487265.8411885246,
      "avg_output_tokens_per_task": 30959.861680327867,
      "last_run_timestamp": "2026-08-29T11:19:17.684Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Venice",
      "accuracy": 0.8521062280219781,
      "accuracy_std_dev": 0.006055500000000047,
      "run_count": 2,
      "total_questions": 364,
      "avg_cost_per_task_usd": 0.005501717032967033,
      "avg_duration_per_task_ms": 539158.467032967,
      "avg_output_tokens_per_task": 15491.851648351649,
      "last_run_timestamp": "2026-10-17T08:17:58.236Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Parasail",
      "accuracy": 0.851852,
      "accuracy_std_dev": 0.006733999999999962,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.01747031565656566,
      "avg_duration_per_task_ms": 122982.43434343435,
      "avg_output_tokens_per_task": 28903.234848484848,
      "last_run_timestamp": "2026-10-24T08:15:51.061Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Ionstream",
      "accuracy": 0.851852,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.16585097979797978,
      "avg_duration_per_task_ms": 389107.60101010103,
      "avg_output_tokens_per_task": 37536.631313131315,
      "last_run_timestamp": "2026-08-04T03:12:10.622Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.6-opus-20260205",
      "model_slug": "anthropic/claude-opus-4.6",
      "model_name": "Anthropic: Claude Opus 4.6",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.8518518181818182,
      "accuracy_std_dev": 0.03385326940520644,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.7072226216712579,
      "avg_duration_per_task_ms": 481789.99357208447,
      "avg_output_tokens_per_task": 28288.861799816346,
      "last_run_timestamp": "2026-10-24T08:43:42.744Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5-2025-08-07",
      "model_slug": "openai/gpt-5",
      "model_name": "OpenAI: GPT-5",
      "provider_name": "Azure",
      "accuracy": 0.8518517499999999,
      "accuracy_std_dev": 0.0020620621201845155,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.21063902146464644,
      "avg_duration_per_task_ms": 251655.29545454544,
      "avg_output_tokens_per_task": 19971.159090909092,
      "last_run_timestamp": "2026-10-01T16:27:58.636Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "auto-routing",
      "accuracy": 0.8516648888888888,
      "accuracy_std_dev": 0.02715089739644159,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.07440295791245792,
      "avg_duration_per_task_ms": 482923.6335578002,
      "avg_output_tokens_per_task": 22694.606621773288,
      "last_run_timestamp": "2026-10-24T16:15:14.403Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Ambient",
      "accuracy": 0.8515965972461275,
      "accuracy_std_dev": 0.006985364039578179,
      "run_count": 9,
      "total_questions": 1743,
      "avg_cost_per_task_usd": 0.09721534480780264,
      "avg_duration_per_task_ms": 672514.3052208836,
      "avg_output_tokens_per_task": 40113.75157773953,
      "last_run_timestamp": "2026-10-24T16:04:55.275Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "Novita",
      "accuracy": 0.8511634577702702,
      "accuracy_std_dev": 0.017198169660005133,
      "run_count": 9,
      "total_questions": 1776,
      "avg_cost_per_task_usd": 0.03812671227477478,
      "avg_duration_per_task_ms": 606913.3012387387,
      "avg_output_tokens_per_task": 35540.335585585584,
      "last_run_timestamp": "2026-10-24T16:36:57.539Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8509267650397275,
      "accuracy_std_dev": 0.029137382315888,
      "run_count": 9,
      "total_questions": 1762,
      "avg_cost_per_task_usd": 0.14751040351872874,
      "avg_duration_per_task_ms": 406196.22701475595,
      "avg_output_tokens_per_task": 33481.99659477866,
      "last_run_timestamp": "2026-10-24T08:12:02.530Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "BaseTen",
      "accuracy": 0.8508249284795883,
      "accuracy_std_dev": 0.01638111560444773,
      "run_count": 33,
      "total_questions": 6222,
      "avg_cost_per_task_usd": 0.1628150337512054,
      "avg_duration_per_task_ms": 188242.73111539698,
      "avg_output_tokens_per_task": 29107.008357441337,
      "last_run_timestamp": "2026-10-24T09:28:03.268Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "Novita",
      "accuracy": 0.8507279006896552,
      "accuracy_std_dev": 0.012037642028375229,
      "run_count": 11,
      "total_questions": 2175,
      "avg_cost_per_task_usd": 0.03107361793103448,
      "avg_duration_per_task_ms": 729282.3756321839,
      "avg_output_tokens_per_task": 32410.26482758621,
      "last_run_timestamp": "2026-10-17T08:19:39.242Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Mara",
      "accuracy": 0.850575,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 145,
      "avg_cost_per_task_usd": 0.015425489655172412,
      "avg_duration_per_task_ms": 62324.744827586204,
      "avg_output_tokens_per_task": 15913.096551724138,
      "last_run_timestamp": "2026-08-01T18:26:19.887Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Crusoe",
      "accuracy": 0.8504104417831005,
      "accuracy_std_dev": 0.011757939662675208,
      "run_count": 9,
      "total_questions": 1503,
      "avg_cost_per_task_usd": 0.006495053226879575,
      "avg_duration_per_task_ms": 421761.22222222225,
      "avg_output_tokens_per_task": 16002.619427811045,
      "last_run_timestamp": "2026-10-24T16:11:33.499Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8504087142857142,
      "accuracy_std_dev": 0.010549160084565759,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.012687737373737374,
      "avg_duration_per_task_ms": 687002.3088023087,
      "avg_output_tokens_per_task": 18864.385281385283,
      "last_run_timestamp": "2026-10-17T16:05:11.655Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Venice",
      "accuracy": 0.8502384166666668,
      "accuracy_std_dev": 0.009974367911955898,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.09800930387205385,
      "avg_duration_per_task_ms": 567891.4086700337,
      "avg_output_tokens_per_task": 30514.193602693602,
      "last_run_timestamp": "2026-10-24T16:13:46.473Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5-2025-08-07",
      "model_slug": "openai/gpt-5",
      "model_name": "OpenAI: GPT-5",
      "provider_name": "OpenAI",
      "accuracy": 0.8501685000000001,
      "accuracy_std_dev": 0.0016835000000000044,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.20853091414141414,
      "avg_duration_per_task_ms": 280266.345959596,
      "avg_output_tokens_per_task": 20755.358585858587,
      "last_run_timestamp": "2026-10-01T16:20:44.419Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "Parasail",
      "accuracy": 0.8501683333333333,
      "accuracy_std_dev": 0.009523284669109108,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.03896970159932659,
      "avg_duration_per_task_ms": 390855.180976431,
      "avg_output_tokens_per_task": 39310.545454545456,
      "last_run_timestamp": "2026-10-24T16:23:26.712Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8501682222222222,
      "accuracy_std_dev": 0.03266346068886207,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.01065054096520763,
      "avg_duration_per_task_ms": 397288.3883277217,
      "avg_output_tokens_per_task": 37715.82323232323,
      "last_run_timestamp": "2026-10-24T16:02:06.909Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Inceptron",
      "accuracy": 0.8499550449236298,
      "accuracy_std_dev": 0.04367521121615224,
      "run_count": 7,
      "total_questions": 1113,
      "avg_cost_per_task_usd": 0.023873053908355795,
      "avg_duration_per_task_ms": 467262.93441150046,
      "avg_output_tokens_per_task": 37562.04132973944,
      "last_run_timestamp": "2026-10-24T08:33:50.644Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "AkashML",
      "accuracy": 0.8497263442622951,
      "accuracy_std_dev": 0.01704,
      "run_count": 2,
      "total_questions": 366,
      "avg_cost_per_task_usd": 0.06129340163934426,
      "avg_duration_per_task_ms": 530265.6393442623,
      "avg_output_tokens_per_task": 25778.054644808744,
      "last_run_timestamp": "2026-08-15T08:47:51.369Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Decart",
      "accuracy": 0.8497037279483036,
      "accuracy_std_dev": 0.02211279185187559,
      "run_count": 16,
      "total_questions": 3095,
      "avg_cost_per_task_usd": 0.2026981760904685,
      "avg_duration_per_task_ms": 259750.08206785138,
      "avg_output_tokens_per_task": 43394.09111470113,
      "last_run_timestamp": "2026-10-24T16:29:51.093Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "GMICloud",
      "accuracy": 0.8492744342105264,
      "accuracy_std_dev": 0.016818664646457527,
      "run_count": 10,
      "total_questions": 1976,
      "avg_cost_per_task_usd": 0.0980292940283401,
      "avg_duration_per_task_ms": 324762.6386639676,
      "avg_output_tokens_per_task": 28156.75556680162,
      "last_run_timestamp": "2026-10-24T08:53:21.957Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "ModelRun",
      "accuracy": 0.8492166318537859,
      "accuracy_std_dev": 0.008781432268570992,
      "run_count": 4,
      "total_questions": 766,
      "avg_cost_per_task_usd": 0.1529455953002611,
      "avg_duration_per_task_ms": 288907.9503916449,
      "avg_output_tokens_per_task": 22996.231070496084,
      "last_run_timestamp": "2026-10-24T08:27:23.476Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Sail Research",
      "accuracy": 0.8489659064484613,
      "accuracy_std_dev": 0.09543669529109948,
      "run_count": 21,
      "total_questions": 4094,
      "avg_cost_per_task_usd": 0.00791240718124084,
      "avg_duration_per_task_ms": 580031.3243771372,
      "avg_output_tokens_per_task": 27584.65388373229,
      "last_run_timestamp": "2026-10-24T08:58:59.998Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "auto-routing",
      "accuracy": 0.8489380769230769,
      "accuracy_std_dev": 0.01772014759645744,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.124409175990676,
      "avg_duration_per_task_ms": 527066.526029526,
      "avg_output_tokens_per_task": 46755.25407925408,
      "last_run_timestamp": "2026-10-24T16:02:47.284Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "GMICloud",
      "accuracy": 0.8485019556313994,
      "accuracy_std_dev": 0.018353041582618763,
      "run_count": 9,
      "total_questions": 1758,
      "avg_cost_per_task_usd": 0.03159626450511945,
      "avg_duration_per_task_ms": 544808.5153583618,
      "avg_output_tokens_per_task": 33468.67861205916,
      "last_run_timestamp": "2026-10-24T16:10:43.276Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "inclusionai/ling-3.0-flash-vl-20260910",
      "model_slug": "inclusionai/ling-3.0-flash-vl",
      "model_name": "inclusionAI: Ling 3.0 Flash VL",
      "provider_name": "auto-routing",
      "accuracy": 0.848485,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0012582575757575758,
      "avg_duration_per_task_ms": 88400.0101010101,
      "avg_output_tokens_per_task": 17501.79797979798,
      "last_run_timestamp": "2026-10-01T08:22:54.644Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-122b-a10b-20260224",
      "model_slug": "qwen/qwen3.5-122b-a10b",
      "model_name": "Qwen: Qwen3.5-122B-A10B",
      "provider_name": "Novita",
      "accuracy": 0.8483339713261648,
      "accuracy_std_dev": 0.009010597309269926,
      "run_count": 13,
      "total_questions": 2511,
      "avg_cost_per_task_usd": 0.06720260175228991,
      "avg_duration_per_task_ms": 196650.21306252488,
      "avg_output_tokens_per_task": 20902.762246117083,
      "last_run_timestamp": "2026-10-24T16:04:55.266Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Phala",
      "accuracy": 0.8482398753606462,
      "accuracy_std_dev": 0.025739734220236125,
      "run_count": 9,
      "total_questions": 1733,
      "avg_cost_per_task_usd": 0.10720182342758222,
      "avg_duration_per_task_ms": 497095.3819965378,
      "avg_output_tokens_per_task": 34429.2602423543,
      "last_run_timestamp": "2026-10-24T08:11:11.432Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8481782975495916,
      "accuracy_std_dev": 0.01581318192794829,
      "run_count": 13,
      "total_questions": 2571,
      "avg_cost_per_task_usd": 0.05179545507584597,
      "avg_duration_per_task_ms": 276145.2582652664,
      "avg_output_tokens_per_task": 45543.73667833528,
      "last_run_timestamp": "2026-10-24T16:18:16.998Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "auto-routing",
      "accuracy": 0.847785277027027,
      "accuracy_std_dev": 0.02236998705607335,
      "run_count": 9,
      "total_questions": 1776,
      "avg_cost_per_task_usd": 0.046796347972972975,
      "avg_duration_per_task_ms": 620353.6756756756,
      "avg_output_tokens_per_task": 38010.02759009009,
      "last_run_timestamp": "2026-10-24T16:27:04.555Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Parasail",
      "accuracy": 0.8477753864168618,
      "accuracy_std_dev": 0.03714670591300797,
      "run_count": 4,
      "total_questions": 427,
      "avg_cost_per_task_usd": 0.10432981733021078,
      "avg_duration_per_task_ms": 293903.51990632317,
      "avg_output_tokens_per_task": 24012.651053864167,
      "last_run_timestamp": "2026-08-22T17:29:49.714Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "DeepInfra",
      "accuracy": 0.8476576319514662,
      "accuracy_std_dev": 0.03635809655963856,
      "run_count": 5,
      "total_questions": 989,
      "avg_cost_per_task_usd": 0.09557547826086957,
      "avg_duration_per_task_ms": 1305294.970677452,
      "avg_output_tokens_per_task": 42689.957532861474,
      "last_run_timestamp": "2026-08-29T18:36:10.818Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Parasail",
      "accuracy": 0.8476458282485877,
      "accuracy_std_dev": 0.010161445779070402,
      "run_count": 9,
      "total_questions": 1770,
      "avg_cost_per_task_usd": 0.007148605084745762,
      "avg_duration_per_task_ms": 569159.6271186441,
      "avg_output_tokens_per_task": 17684.37118644068,
      "last_run_timestamp": "2026-10-24T16:26:58.608Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8476430833333334,
      "accuracy_std_dev": 0.017202829546610125,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.12729099452861953,
      "avg_duration_per_task_ms": 386377.8813131313,
      "avg_output_tokens_per_task": 38419.52356902357,
      "last_run_timestamp": "2026-10-24T16:18:17.002Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-sonnet-5-20260630",
      "model_slug": "anthropic/claude-sonnet-5",
      "model_name": "Anthropic: Claude Sonnet 5",
      "provider_name": "auto-routing",
      "accuracy": 0.8474003453908985,
      "accuracy_std_dev": 0.04393970502657573,
      "run_count": 13,
      "total_questions": 2571,
      "avg_cost_per_task_usd": 0.16651004745235318,
      "avg_duration_per_task_ms": 164543.2010890704,
      "avg_output_tokens_per_task": 16431.475690392843,
      "last_run_timestamp": "2026-10-24T08:13:41.058Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Novita",
      "accuracy": 0.8463488417849898,
      "accuracy_std_dev": 0.018032907906380492,
      "run_count": 10,
      "total_questions": 1972,
      "avg_cost_per_task_usd": 0.010114639452332656,
      "avg_duration_per_task_ms": 449682.4756592292,
      "avg_output_tokens_per_task": 32532.80730223124,
      "last_run_timestamp": "2026-10-24T08:45:13.862Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8462835489864864,
      "accuracy_std_dev": 0.018064668045116323,
      "run_count": 9,
      "total_questions": 1776,
      "avg_cost_per_task_usd": 0.06724871058558558,
      "avg_duration_per_task_ms": 247566.1204954955,
      "avg_output_tokens_per_task": 19807.05518018018,
      "last_run_timestamp": "2026-10-24T08:41:57.620Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "Venice",
      "accuracy": 0.84624,
      "accuracy_std_dev": 0.02002973311521316,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.043265813131313134,
      "avg_duration_per_task_ms": 121558.87037037036,
      "avg_output_tokens_per_task": 34452.29797979798,
      "last_run_timestamp": "2026-10-24T08:13:28.067Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "Novita",
      "accuracy": 0.8459638564102563,
      "accuracy_std_dev": 0.012789089286005938,
      "run_count": 9,
      "total_questions": 1755,
      "avg_cost_per_task_usd": 0.08740559088319089,
      "avg_duration_per_task_ms": 1032079.9236467236,
      "avg_output_tokens_per_task": 44190.14074074074,
      "last_run_timestamp": "2026-10-24T16:33:50.860Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.8459595,
      "accuracy_std_dev": 0.021043499999999993,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.021401732323232326,
      "avg_duration_per_task_ms": 459813.37626262626,
      "avg_output_tokens_per_task": 42563.60858585859,
      "last_run_timestamp": "2026-10-24T08:30:01.288Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "StreamLake",
      "accuracy": 0.8458658888888888,
      "accuracy_std_dev": 0.043929875719895144,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.003920089225589225,
      "avg_duration_per_task_ms": 538039.7317620651,
      "avg_output_tokens_per_task": 39775.50785634119,
      "last_run_timestamp": "2026-10-24T16:14:40.610Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Together",
      "accuracy": 0.8456602006079027,
      "accuracy_std_dev": 0.016336342615469363,
      "run_count": 10,
      "total_questions": 1974,
      "avg_cost_per_task_usd": 0.020697631205673757,
      "avg_duration_per_task_ms": 324636.22441742657,
      "avg_output_tokens_per_task": 41178.20871327254,
      "last_run_timestamp": "2026-10-24T08:56:29.712Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Decart",
      "accuracy": 0.8455742847457628,
      "accuracy_std_dev": 0.019722621271080296,
      "run_count": 9,
      "total_questions": 1770,
      "avg_cost_per_task_usd": 0.10396145480225989,
      "avg_duration_per_task_ms": 199057.26101694914,
      "avg_output_tokens_per_task": 30600.066666666666,
      "last_run_timestamp": "2026-10-24T08:23:31.990Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8455386666666667,
      "accuracy_std_dev": 0.01527369648193331,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.020861409090909087,
      "avg_duration_per_task_ms": 420940.80723905726,
      "avg_output_tokens_per_task": 41492.801346801345,
      "last_run_timestamp": "2026-10-24T08:13:28.065Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "Z.AI",
      "accuracy": 0.8453286747404843,
      "accuracy_std_dev": 0.027756094888949628,
      "run_count": 8,
      "total_questions": 1445,
      "avg_cost_per_task_usd": 0.09002073010380623,
      "avg_duration_per_task_ms": 1465880.0809688582,
      "avg_output_tokens_per_task": 41371.96539792387,
      "last_run_timestamp": "2026-10-24T16:07:10.473Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.8450579454306376,
      "accuracy_std_dev": 0.010092084710428701,
      "run_count": 9,
      "total_questions": 1521,
      "avg_cost_per_task_usd": 0.13361581196581196,
      "avg_duration_per_task_ms": 130442.2596975674,
      "avg_output_tokens_per_task": 31421.66929651545,
      "last_run_timestamp": "2026-10-24T08:11:11.347Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Alibaba",
      "accuracy": 0.8449909073783358,
      "accuracy_std_dev": 0.02024798218992146,
      "run_count": 29,
      "total_questions": 5733,
      "avg_cost_per_task_usd": 0.24276903209488926,
      "avg_duration_per_task_ms": 502111.777603349,
      "avg_output_tokens_per_task": 47287.581719867434,
      "last_run_timestamp": "2026-10-24T16:21:25.073Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-sonnet-5-20260630",
      "model_slug": "anthropic/claude-sonnet-5",
      "model_name": "Anthropic: Claude Sonnet 5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.8449648181818183,
      "accuracy_std_dev": 0.05078720577757961,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.16265579063360883,
      "avg_duration_per_task_ms": 162118.34848484848,
      "avg_output_tokens_per_task": 16092.63269054178,
      "last_run_timestamp": "2026-10-24T08:18:32.507Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Fireworks",
      "accuracy": 0.844913105281815,
      "accuracy_std_dev": 0.019516578349245098,
      "run_count": 15,
      "total_questions": 2821,
      "avg_cost_per_task_usd": 0.13749859695143568,
      "avg_duration_per_task_ms": 278207.19319390284,
      "avg_output_tokens_per_task": 28367.055654023396,
      "last_run_timestamp": "2026-10-24T08:44:43.305Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "StreamLake",
      "accuracy": 0.8448157507598784,
      "accuracy_std_dev": 0.02762329063670728,
      "run_count": 10,
      "total_questions": 1974,
      "avg_cost_per_task_usd": 0.013124052684903749,
      "avg_duration_per_task_ms": 664201.0739614995,
      "avg_output_tokens_per_task": 40826.44782168186,
      "last_run_timestamp": "2026-10-24T08:41:23.787Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Mistral",
      "accuracy": 0.8444763809523811,
      "accuracy_std_dev": 0.010977878186331046,
      "run_count": 21,
      "total_questions": 4158,
      "avg_cost_per_task_usd": 0.14939990764790767,
      "avg_duration_per_task_ms": 163824.38455988455,
      "avg_output_tokens_per_task": 32857.31721981722,
      "last_run_timestamp": "2026-10-24T08:45:58.781Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Together",
      "accuracy": 0.844398036338225,
      "accuracy_std_dev": 0.011789574159373825,
      "run_count": 8,
      "total_questions": 1431,
      "avg_cost_per_task_usd": 0.015440737246680642,
      "avg_duration_per_task_ms": 650403.249475891,
      "avg_output_tokens_per_task": 16520.188679245282,
      "last_run_timestamp": "2026-08-29T18:29:51.235Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "StreamLake",
      "accuracy": 0.8443693333333332,
      "accuracy_std_dev": 0.01866431008338881,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.1411051307519641,
      "avg_duration_per_task_ms": 894934.3092031425,
      "avg_output_tokens_per_task": 55451.94893378227,
      "last_run_timestamp": "2026-10-24T16:03:25.470Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.8441825555555555,
      "accuracy_std_dev": 0.02302992819070732,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.014299253647586982,
      "avg_duration_per_task_ms": 527298.7811447812,
      "avg_output_tokens_per_task": 38122.5847362514,
      "last_run_timestamp": "2026-10-24T08:42:05.088Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Relace",
      "accuracy": 0.8440474857142857,
      "accuracy_std_dev": 0.0037822104436426945,
      "run_count": 5,
      "total_questions": 980,
      "avg_cost_per_task_usd": 0.11554378775510205,
      "avg_duration_per_task_ms": 243142.18265306123,
      "avg_output_tokens_per_task": 31758.279591836734,
      "last_run_timestamp": "2026-10-24T08:42:05.876Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-122b-a10b-20260224",
      "model_slug": "qwen/qwen3.5-122b-a10b",
      "model_name": "Qwen: Qwen3.5-122B-A10B",
      "provider_name": "DeepInfra",
      "accuracy": 0.8440043168316831,
      "accuracy_std_dev": 0.006860781750965057,
      "run_count": 8,
      "total_questions": 1515,
      "avg_cost_per_task_usd": 0.05332531353135314,
      "avg_duration_per_task_ms": 266581.80264026404,
      "avg_output_tokens_per_task": 22128.052145214522,
      "last_run_timestamp": "2026-09-26T09:50:39.842Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Crusoe",
      "accuracy": 0.843855125,
      "accuracy_std_dev": 0.007974456116838023,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.1439108863636364,
      "avg_duration_per_task_ms": 148669.27525252526,
      "avg_output_tokens_per_task": 32529.41856060606,
      "last_run_timestamp": "2026-10-24T08:10:23.024Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Nebius",
      "accuracy": 0.8438239857868021,
      "accuracy_std_dev": 0.012941643289783599,
      "run_count": 5,
      "total_questions": 985,
      "avg_cost_per_task_usd": 0.272832538071066,
      "avg_duration_per_task_ms": 511204.8578680203,
      "avg_output_tokens_per_task": 61801.45989847716,
      "last_run_timestamp": "2026-10-24T16:14:04.272Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-122b-a10b-20260224",
      "model_slug": "qwen/qwen3.5-122b-a10b",
      "model_name": "Qwen: Qwen3.5-122B-A10B",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8434343846153846,
      "accuracy_std_dev": 0.032261314395795763,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.0527346484071484,
      "avg_duration_per_task_ms": 208837.4222999223,
      "avg_output_tokens_per_task": 21871.209401709402,
      "last_run_timestamp": "2026-10-24T16:02:47.280Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Fireworks",
      "accuracy": 0.8427474515050167,
      "accuracy_std_dev": 0.014940270952109593,
      "run_count": 14,
      "total_questions": 2691,
      "avg_cost_per_task_usd": 0.021943730955035302,
      "avg_duration_per_task_ms": 258715.1181716834,
      "avg_output_tokens_per_task": 37901.94314381271,
      "last_run_timestamp": "2026-10-24T08:49:23.400Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "thinkingmachines/inkling-20260715",
      "model_slug": "thinkingmachines/inkling",
      "model_name": "Thinking Machines: Inkling",
      "provider_name": "Together",
      "accuracy": 0.8425437868421053,
      "accuracy_std_dev": 0.00573439306339873,
      "run_count": 6,
      "total_questions": 1140,
      "avg_cost_per_task_usd": 0.08756267807017544,
      "avg_duration_per_task_ms": 282086.36140350875,
      "avg_output_tokens_per_task": 21480.616666666665,
      "last_run_timestamp": "2026-10-01T16:06:12.801Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Novita",
      "accuracy": 0.8423346063291139,
      "accuracy_std_dev": 0.054296645253950386,
      "run_count": 12,
      "total_questions": 2370,
      "avg_cost_per_task_usd": 0.07867724514767932,
      "avg_duration_per_task_ms": 564357.3624472574,
      "avg_output_tokens_per_task": 43556.894092827,
      "last_run_timestamp": "2026-10-24T16:27:11.126Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8419978532110092,
      "accuracy_std_dev": 0.010540078834619807,
      "run_count": 5,
      "total_questions": 981,
      "avg_cost_per_task_usd": 0.027336418960244645,
      "avg_duration_per_task_ms": 516242.43323139654,
      "avg_output_tokens_per_task": 22594.290519877675,
      "last_run_timestamp": "2026-08-29T18:16:39.840Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Ambient",
      "accuracy": 0.8419939695817491,
      "accuracy_std_dev": 0.05773343437894112,
      "run_count": 4,
      "total_questions": 789,
      "avg_cost_per_task_usd": 0.08744185044359949,
      "avg_duration_per_task_ms": 697734.2648922687,
      "avg_output_tokens_per_task": 26121.35234474018,
      "last_run_timestamp": "2026-08-29T11:31:24.094Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "auto-routing",
      "accuracy": 0.84196125,
      "accuracy_std_dev": 0.04164286996518729,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.10050041856060607,
      "avg_duration_per_task_ms": 1121158.3566919193,
      "avg_output_tokens_per_task": 39757.59027777778,
      "last_run_timestamp": "2026-10-24T16:02:28.490Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "SambaNova",
      "accuracy": 0.8415218367346938,
      "accuracy_std_dev": 0.019781995922842507,
      "run_count": 9,
      "total_questions": 1323,
      "avg_cost_per_task_usd": 0.017032966742252457,
      "avg_duration_per_task_ms": 90861.69312169313,
      "avg_output_tokens_per_task": 14612.785336356765,
      "last_run_timestamp": "2026-10-24T16:27:17.205Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Sail Research",
      "accuracy": 0.8415046100309946,
      "accuracy_std_dev": 0.012823272459877284,
      "run_count": 18,
      "total_questions": 3549,
      "avg_cost_per_task_usd": 0.11274710115525499,
      "avg_duration_per_task_ms": 322682.1735700197,
      "avg_output_tokens_per_task": 34551.41730064807,
      "last_run_timestamp": "2026-10-24T08:41:42.263Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Together",
      "accuracy": 0.84141817433047,
      "accuracy_std_dev": 0.011968232237469333,
      "run_count": 10,
      "total_questions": 1979,
      "avg_cost_per_task_usd": 0.14348533804951996,
      "avg_duration_per_task_ms": 205361.71248105104,
      "avg_output_tokens_per_task": 32378.852956038405,
      "last_run_timestamp": "2026-10-24T08:12:37.564Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.5-flash-lite-20260721",
      "model_slug": "google/gemini-3.5-flash-lite",
      "model_name": "Google: Gemini 3.5 Flash Lite",
      "provider_name": "Google AI Studio",
      "accuracy": 0.8413977096774194,
      "accuracy_std_dev": 0.007842951293996439,
      "run_count": 4,
      "total_questions": 744,
      "avg_cost_per_task_usd": 0.03688066532258065,
      "avg_duration_per_task_ms": 47685.283602150535,
      "avg_output_tokens_per_task": 14664.099462365592,
      "last_run_timestamp": "2026-10-01T19:13:29.958Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.5-flash-lite-20260721",
      "model_slug": "google/gemini-3.5-flash-lite",
      "model_name": "Google: Gemini 3.5 Flash Lite",
      "provider_name": "auto-routing",
      "accuracy": 0.8411896666666667,
      "accuracy_std_dev": 0.008729468916008341,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.036658867003367,
      "avg_duration_per_task_ms": 51092.77104377104,
      "avg_output_tokens_per_task": 14574.808080808081,
      "last_run_timestamp": "2026-10-01T19:12:26.032Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Inceptron",
      "accuracy": 0.8409475076142132,
      "accuracy_std_dev": 0.005443376341940744,
      "run_count": 3,
      "total_questions": 591,
      "avg_cost_per_task_usd": 0.020375204737732657,
      "avg_duration_per_task_ms": 653417.3384094755,
      "avg_output_tokens_per_task": 23451.566835871403,
      "last_run_timestamp": "2026-08-15T14:57:08.773Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "Venice",
      "accuracy": 0.8408155555555555,
      "accuracy_std_dev": 0.0506301697894998,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.09469512233445566,
      "avg_duration_per_task_ms": 391927.05611672276,
      "avg_output_tokens_per_task": 21250.56509539843,
      "last_run_timestamp": "2026-10-24T16:06:30.903Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.4-mini-20260317",
      "model_slug": "openai/gpt-5.4-mini",
      "model_name": "OpenAI: GPT-5.4 Mini",
      "provider_name": "OpenAI",
      "accuracy": 0.84048825,
      "accuracy_std_dev": 0.004667830993887862,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.06197197222222223,
      "avg_duration_per_task_ms": 84839.70707070707,
      "avg_output_tokens_per_task": 13643.420454545454,
      "last_run_timestamp": "2026-10-02T23:24:30.140Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.5-flash-lite-20260721",
      "model_slug": "google/gemini-3.5-flash-lite",
      "model_name": "Google: Gemini 3.5 Flash Lite",
      "provider_name": "Google",
      "accuracy": 0.8402775531250001,
      "accuracy_std_dev": 0.007466846475587956,
      "run_count": 10,
      "total_questions": 1920,
      "avg_cost_per_task_usd": 0.039408469791666664,
      "avg_duration_per_task_ms": 51791.722395833334,
      "avg_output_tokens_per_task": 14757.431770833333,
      "last_run_timestamp": "2026-10-01T19:13:57.987Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "thinkingmachines/inkling-20260715",
      "model_slug": "thinkingmachines/inkling",
      "model_name": "Thinking Machines: Inkling",
      "provider_name": "DeepInfra",
      "accuracy": 0.8402425498181818,
      "accuracy_std_dev": 0.00976753775827742,
      "run_count": 8,
      "total_questions": 1375,
      "avg_cost_per_task_usd": 0.0796483170909091,
      "avg_duration_per_task_ms": 509723.63709090906,
      "avg_output_tokens_per_task": 19683.845818181817,
      "last_run_timestamp": "2026-10-08T22:17:54.321Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Makora",
      "accuracy": 0.8398882325043376,
      "accuracy_std_dev": 0.015516849481773015,
      "run_count": 10,
      "total_questions": 1729,
      "avg_cost_per_task_usd": 0.14290184152689414,
      "avg_duration_per_task_ms": 273124.987854251,
      "avg_output_tokens_per_task": 33153.25390399075,
      "last_run_timestamp": "2026-10-24T08:53:46.038Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Reka",
      "accuracy": 0.839837363066202,
      "accuracy_std_dev": 0.014816504049501316,
      "run_count": 15,
      "total_questions": 2870,
      "avg_cost_per_task_usd": 0.07680571289198605,
      "avg_duration_per_task_ms": 389476.49477351917,
      "avg_output_tokens_per_task": 22208.951219512193,
      "last_run_timestamp": "2026-10-24T08:10:23.015Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Io Net",
      "accuracy": 0.8397870000000001,
      "accuracy_std_dev": 0.06305207943174174,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.013501392255892255,
      "avg_duration_per_task_ms": 269245.1767676768,
      "avg_output_tokens_per_task": 24293.242424242424,
      "last_run_timestamp": "2026-10-01T17:44:47.331Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Venice",
      "accuracy": 0.839646375,
      "accuracy_std_dev": 0.01029199634348825,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.006103154040404041,
      "avg_duration_per_task_ms": 431518.17424242425,
      "avg_output_tokens_per_task": 25837.31818181818,
      "last_run_timestamp": "2026-10-24T16:03:25.398Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Cloudflare",
      "accuracy": 0.8396166498586238,
      "accuracy_std_dev": 0.03493831082667691,
      "run_count": 11,
      "total_questions": 2122,
      "avg_cost_per_task_usd": 0.029913213949104622,
      "avg_duration_per_task_ms": 288209.38784165884,
      "avg_output_tokens_per_task": 9239.83930254477,
      "last_run_timestamp": "2026-10-24T08:08:54.186Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Together",
      "accuracy": 0.8389520859671303,
      "accuracy_std_dev": 0.014418680686632029,
      "run_count": 12,
      "total_questions": 2373,
      "avg_cost_per_task_usd": 0.22912267003792672,
      "avg_duration_per_task_ms": 279595.97260851244,
      "avg_output_tokens_per_task": 51927.226717235564,
      "last_run_timestamp": "2026-10-24T16:04:55.201Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "DigitalOcean",
      "accuracy": 0.8389448333333334,
      "accuracy_std_dev": 0.03599784248820413,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.004737409090909091,
      "avg_duration_per_task_ms": 1034495.3484848485,
      "avg_output_tokens_per_task": 27462.325757575756,
      "last_run_timestamp": "2026-10-24T16:19:29.603Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Novita",
      "accuracy": 0.838522452928095,
      "accuracy_std_dev": 0.010501612316687394,
      "run_count": 8,
      "total_questions": 1349,
      "avg_cost_per_task_usd": 0.006855791697553743,
      "avg_duration_per_task_ms": 622066.7916975537,
      "avg_output_tokens_per_task": 16928.25722757598,
      "last_run_timestamp": "2026-10-24T16:21:33.540Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "Z.AI",
      "accuracy": 0.838520908319185,
      "accuracy_std_dev": 0.021628456230475167,
      "run_count": 9,
      "total_questions": 1767,
      "avg_cost_per_task_usd": 0.09386730616864743,
      "avg_duration_per_task_ms": 1055070.8788907754,
      "avg_output_tokens_per_task": 42969.125636672325,
      "last_run_timestamp": "2026-10-24T16:12:09.457Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Modal",
      "accuracy": 0.8383968759493671,
      "accuracy_std_dev": 0.006763999999999992,
      "run_count": 2,
      "total_questions": 395,
      "avg_cost_per_task_usd": 0.033454222784810125,
      "avg_duration_per_task_ms": 81357.47848101266,
      "avg_output_tokens_per_task": 14534.177215189873,
      "last_run_timestamp": "2026-10-24T09:12:44.274Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "StreamLake",
      "accuracy": 0.83813855538695,
      "accuracy_std_dev": 0.02473346097415401,
      "run_count": 10,
      "total_questions": 1977,
      "avg_cost_per_task_usd": 0.1589729180576631,
      "avg_duration_per_task_ms": 941860.1360647446,
      "avg_output_tokens_per_task": 52231.05108750632,
      "last_run_timestamp": "2026-10-24T16:05:13.521Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Alibaba",
      "accuracy": 0.8378745924484235,
      "accuracy_std_dev": 0.026005006584645496,
      "run_count": 13,
      "total_questions": 2569,
      "avg_cost_per_task_usd": 0.08716310782405605,
      "avg_duration_per_task_ms": 387520.7785130401,
      "avg_output_tokens_per_task": 21670.57571039315,
      "last_run_timestamp": "2026-10-24T16:33:45.711Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "InferenceNet",
      "accuracy": 0.8378226666666666,
      "accuracy_std_dev": 0.00975199857579064,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.15828378956228956,
      "avg_duration_per_task_ms": 446608.7861952862,
      "avg_output_tokens_per_task": 56298.93097643097,
      "last_run_timestamp": "2026-10-10T08:30:28.864Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Novita",
      "accuracy": 0.8377103,
      "accuracy_std_dev": 0.023183391999662194,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.10627496717171718,
      "avg_duration_per_task_ms": 386419.75505050505,
      "avg_output_tokens_per_task": 35856.45909090909,
      "last_run_timestamp": "2026-10-24T08:25:29.458Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "StreamLake",
      "accuracy": 0.837692958038147,
      "accuracy_std_dev": 0.020571358270420545,
      "run_count": 10,
      "total_questions": 1835,
      "avg_cost_per_task_usd": 0.07314428610354223,
      "avg_duration_per_task_ms": 380364.9722070845,
      "avg_output_tokens_per_task": 24128.798910081743,
      "last_run_timestamp": "2026-10-24T16:08:29.374Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8376859309031556,
      "accuracy_std_dev": 0.027367300407603226,
      "run_count": 10,
      "total_questions": 1838,
      "avg_cost_per_task_usd": 0.011182803590859632,
      "avg_duration_per_task_ms": 390045.14526659413,
      "avg_output_tokens_per_task": 22148.7992383025,
      "last_run_timestamp": "2026-10-24T08:39:40.816Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Nebius",
      "accuracy": 0.8375419999999999,
      "accuracy_std_dev": 0.012922002863333565,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.15197074579124578,
      "avg_duration_per_task_ms": 226641.9739057239,
      "avg_output_tokens_per_task": 34291.69781144781,
      "last_run_timestamp": "2026-10-24T08:43:12.480Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Mistral",
      "accuracy": 0.83744096,
      "accuracy_std_dev": 0.012056350192259705,
      "run_count": 25,
      "total_questions": 4950,
      "avg_cost_per_task_usd": 0.2619175503030303,
      "avg_duration_per_task_ms": 272403.29393939395,
      "avg_output_tokens_per_task": 56992.821616161615,
      "last_run_timestamp": "2026-10-24T16:27:11.141Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Chutes",
      "accuracy": 0.8372897388724037,
      "accuracy_std_dev": 0.028847852416095617,
      "run_count": 12,
      "total_questions": 2359,
      "avg_cost_per_task_usd": 0.06617999915218314,
      "avg_duration_per_task_ms": 1053779.2886816447,
      "avg_output_tokens_per_task": 33011.97880457821,
      "last_run_timestamp": "2026-10-24T16:02:06.932Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8371210833333332,
      "accuracy_std_dev": 0.012487611784340083,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.0845222638888889,
      "avg_duration_per_task_ms": 727103.742003367,
      "avg_output_tokens_per_task": 48945.89772727273,
      "last_run_timestamp": "2026-10-24T16:11:13.852Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Modal",
      "accuracy": 0.8368686,
      "accuracy_std_dev": 0.01362896250783603,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.1397023909090909,
      "avg_duration_per_task_ms": 182914.46161616163,
      "avg_output_tokens_per_task": 31522.08787878788,
      "last_run_timestamp": "2026-10-24T08:24:39.068Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "ModelRun",
      "accuracy": 0.8367790607594936,
      "accuracy_std_dev": 0.029638370939918782,
      "run_count": 12,
      "total_questions": 2370,
      "avg_cost_per_task_usd": 0.017527371729957806,
      "avg_duration_per_task_ms": 145536.47299578058,
      "avg_output_tokens_per_task": 18246.05611814346,
      "last_run_timestamp": "2026-10-24T16:35:32.863Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.8367000000000001,
      "accuracy_std_dev": 0.006734000000000018,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.18573769191919193,
      "avg_duration_per_task_ms": 281534.297979798,
      "avg_output_tokens_per_task": 34913.7095959596,
      "last_run_timestamp": "2026-10-24T08:56:15.847Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "DeepInfra",
      "accuracy": 0.8366667090476191,
      "accuracy_std_dev": 0.01978861195373632,
      "run_count": 12,
      "total_questions": 2100,
      "avg_cost_per_task_usd": 0.09574838047619047,
      "avg_duration_per_task_ms": 613293.1757142857,
      "avg_output_tokens_per_task": 29845.292857142857,
      "last_run_timestamp": "2026-10-24T16:10:16.903Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "AtlasCloud",
      "accuracy": 0.836653454183267,
      "accuracy_std_dev": 0.007833999999999952,
      "run_count": 2,
      "total_questions": 251,
      "avg_cost_per_task_usd": 0.06977503187250997,
      "avg_duration_per_task_ms": 237256.35059760956,
      "avg_output_tokens_per_task": 17511.450199203187,
      "last_run_timestamp": "2026-08-15T08:47:41.523Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Nebius",
      "accuracy": 0.836139,
      "accuracy_std_dev": 0.05678065343641852,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.005011018518518518,
      "avg_duration_per_task_ms": 214818.47643097644,
      "avg_output_tokens_per_task": 17529.722222222223,
      "last_run_timestamp": "2026-09-26T11:06:41.988Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "Reka",
      "accuracy": 0.836139,
      "accuracy_std_dev": 0.0007933738084913017,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.05002690909090909,
      "avg_duration_per_task_ms": 537584.335016835,
      "avg_output_tokens_per_task": 49932.06734006734,
      "last_run_timestamp": "2026-09-30T11:44:08.104Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Reka",
      "accuracy": 0.8360894826189674,
      "accuracy_std_dev": 0.027487401181389745,
      "run_count": 15,
      "total_questions": 2963,
      "avg_cost_per_task_usd": 0.02392725244684441,
      "avg_duration_per_task_ms": 312624.6992912589,
      "avg_output_tokens_per_task": 17783.592642591968,
      "last_run_timestamp": "2026-10-24T08:22:23.704Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Z.AI",
      "accuracy": 0.83599875,
      "accuracy_std_dev": 0.06733100699544627,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.20365952861952863,
      "avg_duration_per_task_ms": 537295.401936027,
      "avg_output_tokens_per_task": 46280.556818181816,
      "last_run_timestamp": "2026-10-24T16:48:27.857Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "inclusionai/ling-3.0-flash-vl-20260910",
      "model_slug": "inclusionai/ling-3.0-flash-vl",
      "model_name": "inclusionAI: Ling 3.0 Flash VL",
      "provider_name": "Novita",
      "accuracy": 0.8358585,
      "accuracy_std_dev": 0.01430950000000003,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.0011451691919191918,
      "avg_duration_per_task_ms": 116299.33333333333,
      "avg_output_tokens_per_task": 18355.219696969696,
      "last_run_timestamp": "2026-10-01T08:23:48.135Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Mara",
      "accuracy": 0.835384616923077,
      "accuracy_std_dev": 0.032642666915557006,
      "run_count": 5,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.09615046153846153,
      "avg_duration_per_task_ms": 334242.9476923077,
      "avg_output_tokens_per_task": 39819.60307692308,
      "last_run_timestamp": "2026-10-24T16:40:24.303Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Relace",
      "accuracy": 0.8353175277777779,
      "accuracy_std_dev": 0.01836981616741691,
      "run_count": 9,
      "total_questions": 1764,
      "avg_cost_per_task_usd": 0.014329623015873016,
      "avg_duration_per_task_ms": 445894.7193877551,
      "avg_output_tokens_per_task": 39122.69671201814,
      "last_run_timestamp": "2026-10-24T08:17:53.301Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "inclusionai/ling-3.0-flash-vl-20260910",
      "model_slug": "inclusionai/ling-3.0-flash-vl",
      "model_name": "inclusionAI: Ling 3.0 Flash VL",
      "provider_name": "DeepInfra",
      "accuracy": 0.8351849999999998,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 180,
      "avg_cost_per_task_usd": 0.0030483999999999997,
      "avg_duration_per_task_ms": 288802.22222222225,
      "avg_output_tokens_per_task": 16703.5,
      "last_run_timestamp": "2026-10-01T08:43:11.415Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.4-mini-20260317",
      "model_slug": "openai/gpt-5.4-mini",
      "model_name": "OpenAI: GPT-5.4 Mini",
      "provider_name": "auto-routing",
      "accuracy": 0.8350166666666667,
      "accuracy_std_dev": 0.005991526757754575,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.06194655555555556,
      "avg_duration_per_task_ms": 84396.87373737374,
      "avg_output_tokens_per_task": 13638.5,
      "last_run_timestamp": "2026-10-01T19:12:44.401Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "SiliconFlow",
      "accuracy": 0.834789191724138,
      "accuracy_std_dev": 0.03295145245372225,
      "run_count": 11,
      "total_questions": 2175,
      "avg_cost_per_task_usd": 0.10863842022988504,
      "avg_duration_per_task_ms": 1052937.216091954,
      "avg_output_tokens_per_task": 33937.42436781609,
      "last_run_timestamp": "2026-10-24T16:17:57.470Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-122b-a10b-20260224",
      "model_slug": "qwen/qwen3.5-122b-a10b",
      "model_name": "Qwen: Qwen3.5-122B-A10B",
      "provider_name": "auto-routing",
      "accuracy": 0.834628230769231,
      "accuracy_std_dev": 0.032424071580691764,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.05403845415695415,
      "avg_duration_per_task_ms": 253079.37917637918,
      "avg_output_tokens_per_task": 22985.918026418025,
      "last_run_timestamp": "2026-10-24T16:06:30.902Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "AkashML",
      "accuracy": 0.8346143259031515,
      "accuracy_std_dev": 0.012531384063964916,
      "run_count": 16,
      "total_questions": 2602,
      "avg_cost_per_task_usd": 0.1179007190622598,
      "avg_duration_per_task_ms": 273695.7221368178,
      "avg_output_tokens_per_task": 27899.308224442735,
      "last_run_timestamp": "2026-10-24T08:12:02.631Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Phala",
      "accuracy": 0.8341984152249134,
      "accuracy_std_dev": 0.029002188381032377,
      "run_count": 3,
      "total_questions": 578,
      "avg_cost_per_task_usd": 0.005492470588235294,
      "avg_duration_per_task_ms": 956260.6712802768,
      "avg_output_tokens_per_task": 15404.778546712803,
      "last_run_timestamp": "2026-08-29T18:08:44.008Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "DekaLLM",
      "accuracy": 0.8340433944004868,
      "accuracy_std_dev": 0.008747099804372552,
      "run_count": 9,
      "total_questions": 1643,
      "avg_cost_per_task_usd": 0.04404786731588557,
      "avg_duration_per_task_ms": 94664.89896530737,
      "avg_output_tokens_per_task": 15175.068167985393,
      "last_run_timestamp": "2026-10-24T08:10:23.016Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "thinkingmachines/inkling-20260715",
      "model_slug": "thinkingmachines/inkling",
      "model_name": "Thinking Machines: Inkling",
      "provider_name": "auto-routing",
      "accuracy": 0.8340364430379746,
      "accuracy_std_dev": 0.01069546220626094,
      "run_count": 6,
      "total_questions": 1185,
      "avg_cost_per_task_usd": 0.0954248565400844,
      "avg_duration_per_task_ms": 293770.05400843883,
      "avg_output_tokens_per_task": 23403.932489451476,
      "last_run_timestamp": "2026-10-01T16:11:05.832Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Cloudflare",
      "accuracy": 0.8339663204791117,
      "accuracy_std_dev": 0.017426212050589154,
      "run_count": 18,
      "total_questions": 3423,
      "avg_cost_per_task_usd": 0.25302901986561493,
      "avg_duration_per_task_ms": 374119.30879345606,
      "avg_output_tokens_per_task": 43927.57055214724,
      "last_run_timestamp": "2026-10-24T16:03:06.205Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8338943333333333,
      "accuracy_std_dev": 0.011690806112307073,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.03773407239057239,
      "avg_duration_per_task_ms": 276396.1228956229,
      "avg_output_tokens_per_task": 28290.345117845118,
      "last_run_timestamp": "2026-09-12T09:56:44.384Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Alibaba",
      "accuracy": 0.8337216923076924,
      "accuracy_std_dev": 0.011740948902712052,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.2507070532245532,
      "avg_duration_per_task_ms": 380155.1087801088,
      "avg_output_tokens_per_task": 40506.53496503497,
      "last_run_timestamp": "2026-10-24T08:43:07.995Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Parasail",
      "accuracy": 0.8336574603112842,
      "accuracy_std_dev": 0.021426026057470033,
      "run_count": 13,
      "total_questions": 2570,
      "avg_cost_per_task_usd": 0.024801823735408558,
      "avg_duration_per_task_ms": 393313.6081712062,
      "avg_output_tokens_per_task": 46122.12840466926,
      "last_run_timestamp": "2026-10-24T08:48:13.246Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Alibaba",
      "accuracy": 0.8334863636363636,
      "accuracy_std_dev": 0.011783391626836715,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.0389523406795225,
      "avg_duration_per_task_ms": 226530.9031221304,
      "avg_output_tokens_per_task": 14515.32966023875,
      "last_run_timestamp": "2026-10-24T08:08:52.224Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-flash-20260826",
      "model_slug": "qwen/qwen3.8-flash",
      "model_name": "Qwen: Qwen3.8 Flash",
      "provider_name": "Makora",
      "accuracy": 0.8333329999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 156,
      "avg_cost_per_task_usd": 0.003982064102564103,
      "avg_duration_per_task_ms": 34178.44871794872,
      "avg_output_tokens_per_task": 8341.692307692309,
      "last_run_timestamp": "2026-09-09T15:03:02.824Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Cloudflare",
      "accuracy": 0.8330725023474179,
      "accuracy_std_dev": 0.010168177287347969,
      "run_count": 7,
      "total_questions": 1278,
      "avg_cost_per_task_usd": 0.029710525039123633,
      "avg_duration_per_task_ms": 655555.5469483568,
      "avg_output_tokens_per_task": 42323.18544600939,
      "last_run_timestamp": "2026-10-03T09:40:53.626Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Friendli",
      "accuracy": 0.8328383188836104,
      "accuracy_std_dev": 0.012861707810491917,
      "run_count": 9,
      "total_questions": 1684,
      "avg_cost_per_task_usd": 0.2188719726840855,
      "avg_duration_per_task_ms": 450599.8604513064,
      "avg_output_tokens_per_task": 49626.254156769595,
      "last_run_timestamp": "2026-10-24T16:08:01.591Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Parasail",
      "accuracy": 0.8327628225898459,
      "accuracy_std_dev": 0.014159242972544667,
      "run_count": 9,
      "total_questions": 1753,
      "avg_cost_per_task_usd": 0.15039315516257845,
      "avg_duration_per_task_ms": 277652.9520821449,
      "avg_output_tokens_per_task": 33995.16029663434,
      "last_run_timestamp": "2026-10-24T08:12:02.633Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Parasail",
      "accuracy": 0.832178375487224,
      "accuracy_std_dev": 0.011051436981414083,
      "run_count": 12,
      "total_questions": 2309,
      "avg_cost_per_task_usd": 0.2290666569943699,
      "avg_duration_per_task_ms": 296861.6738847986,
      "avg_output_tokens_per_task": 51892.169770463406,
      "last_run_timestamp": "2026-10-24T16:31:42.371Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "StreamLake",
      "accuracy": 0.8320383846153845,
      "accuracy_std_dev": 0.008064443483478893,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.07931912276612277,
      "avg_duration_per_task_ms": 701901.0648795649,
      "avg_output_tokens_per_task": 57457.6013986014,
      "last_run_timestamp": "2026-10-24T16:24:24.498Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8319781768292682,
      "accuracy_std_dev": 0.02156774450887252,
      "run_count": 5,
      "total_questions": 984,
      "avg_cost_per_task_usd": 0.07742283333333333,
      "avg_duration_per_task_ms": 694188.4644308944,
      "avg_output_tokens_per_task": 41828.17479674797,
      "last_run_timestamp": "2026-10-04T00:07:08.217Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Crusoe",
      "accuracy": 0.83176317025572,
      "accuracy_std_dev": 0.014297695836545667,
      "run_count": 8,
      "total_questions": 1486,
      "avg_cost_per_task_usd": 0.023221829744279946,
      "avg_duration_per_task_ms": 299260.8001345895,
      "avg_output_tokens_per_task": 46221.84320323015,
      "last_run_timestamp": "2026-10-24T08:22:23.703Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.2-exp",
      "model_slug": "deepseek/deepseek-v3.2-exp",
      "model_name": "DeepSeek: DeepSeek V3.2 Exp",
      "provider_name": "AtlasCloud",
      "accuracy": 0.83165,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.009051732323232323,
      "avg_duration_per_task_ms": 428104.83838383836,
      "avg_output_tokens_per_task": 21594.323232323233,
      "last_run_timestamp": "2026-08-01T20:03:34.768Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "DeepInfra",
      "accuracy": 0.8313810512445096,
      "accuracy_std_dev": 0.012942947508356888,
      "run_count": 7,
      "total_questions": 1366,
      "avg_cost_per_task_usd": 0.027591674963396776,
      "avg_duration_per_task_ms": 601596.977306003,
      "avg_output_tokens_per_task": 14650.56149341142,
      "last_run_timestamp": "2026-10-24T08:48:04.402Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Novita",
      "accuracy": 0.831313,
      "accuracy_std_dev": 0.01944119231940264,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.025460874747474747,
      "avg_duration_per_task_ms": 242198.30808080808,
      "avg_output_tokens_per_task": 21221.607070707072,
      "last_run_timestamp": "2026-08-29T18:12:43.943Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Io Net",
      "accuracy": 0.8312629347826086,
      "accuracy_std_dev": 0.044536486139877485,
      "run_count": 4,
      "total_questions": 644,
      "avg_cost_per_task_usd": 0.005694647515527951,
      "avg_duration_per_task_ms": 270388.4906832298,
      "avg_output_tokens_per_task": 13521.194099378881,
      "last_run_timestamp": "2026-08-22T09:56:05.550Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Cerebras",
      "accuracy": 0.83122875,
      "accuracy_std_dev": 0.007839878231675557,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.022249369949494953,
      "avg_duration_per_task_ms": 11972.136363636364,
      "avg_output_tokens_per_task": 14384.287878787878,
      "last_run_timestamp": "2026-10-24T08:21:05.266Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Parasail",
      "accuracy": 0.8311627944186045,
      "accuracy_std_dev": 0.006786821620057476,
      "run_count": 11,
      "total_questions": 2150,
      "avg_cost_per_task_usd": 0.03491511023255814,
      "avg_duration_per_task_ms": 201279.6446511628,
      "avg_output_tokens_per_task": 14261.299069767441,
      "last_run_timestamp": "2026-10-24T08:41:23.420Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "BaseTen",
      "accuracy": 0.8310956896816264,
      "accuracy_std_dev": 0.010867420886961873,
      "run_count": 16,
      "total_questions": 2607,
      "avg_cost_per_task_usd": 0.020108517453011125,
      "avg_duration_per_task_ms": 239467.13693901035,
      "avg_output_tokens_per_task": 40022.99923283468,
      "last_run_timestamp": "2026-10-24T08:13:46.749Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.4-mini-20260317",
      "model_slug": "openai/gpt-5.4-mini",
      "model_name": "OpenAI: GPT-5.4 Mini",
      "provider_name": "Azure",
      "accuracy": 0.8310885000000001,
      "accuracy_std_dev": 0.012955632388399007,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.06418621717171719,
      "avg_duration_per_task_ms": 117116.24831649831,
      "avg_output_tokens_per_task": 13483.090067340067,
      "last_run_timestamp": "2026-10-01T19:14:05.661Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Chutes",
      "accuracy": 0.830874032193159,
      "accuracy_std_dev": 0.029241810305408605,
      "run_count": 8,
      "total_questions": 1491,
      "avg_cost_per_task_usd": 0.00674413145539906,
      "avg_duration_per_task_ms": 717174.3158953722,
      "avg_output_tokens_per_task": 18114.10865191147,
      "last_run_timestamp": "2026-10-24T16:05:36.255Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8306676666666667,
      "accuracy_std_dev": 0.02403995285052688,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.13645198905723904,
      "avg_duration_per_task_ms": 551462.0761784512,
      "avg_output_tokens_per_task": 43445.10816498317,
      "last_run_timestamp": "2026-10-24T16:02:47.376Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Wafer",
      "accuracy": 0.8302065714285713,
      "accuracy_std_dev": 0.0046260801628883914,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.05203393939393939,
      "avg_duration_per_task_ms": 277192.08946608944,
      "avg_output_tokens_per_task": 15155.578643578643,
      "last_run_timestamp": "2026-10-24T08:29:16.815Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "Venice",
      "accuracy": 0.8301606153846154,
      "accuracy_std_dev": 0.025057780433164603,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.038417987567987574,
      "avg_duration_per_task_ms": 349233.05205905205,
      "avg_output_tokens_per_task": 39432.61577311577,
      "last_run_timestamp": "2026-10-24T16:01:48.224Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Friendli",
      "accuracy": 0.8301281231808731,
      "accuracy_std_dev": 0.02299658002943048,
      "run_count": 10,
      "total_questions": 1924,
      "avg_cost_per_task_usd": 0.021267546777546775,
      "avg_duration_per_task_ms": 232718.9979209979,
      "avg_output_tokens_per_task": 42313.036902286905,
      "last_run_timestamp": "2026-10-24T08:44:41.331Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Friendli",
      "accuracy": 0.8298259166666667,
      "accuracy_std_dev": 0.01956609097162035,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.2451282104377105,
      "avg_duration_per_task_ms": 378368.86447811447,
      "avg_output_tokens_per_task": 55545.628367003366,
      "last_run_timestamp": "2026-10-24T16:24:45.221Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "CoreWeave",
      "accuracy": 0.8294438755401903,
      "accuracy_std_dev": 0.015719092947537823,
      "run_count": 15,
      "total_questions": 2314,
      "avg_cost_per_task_usd": 0.038434746326706994,
      "avg_duration_per_task_ms": 267856.29127052723,
      "avg_output_tokens_per_task": 30639.18668971478,
      "last_run_timestamp": "2026-10-24T16:05:40.120Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "AkashML",
      "accuracy": 0.8286576957919346,
      "accuracy_std_dev": 0.020416737880752586,
      "run_count": 18,
      "total_questions": 3422,
      "avg_cost_per_task_usd": 0.04187343191116307,
      "avg_duration_per_task_ms": 431990.98655756866,
      "avg_output_tokens_per_task": 45586.554061952076,
      "last_run_timestamp": "2026-10-24T16:01:48.241Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "CoreWeave",
      "accuracy": 0.828497570984456,
      "accuracy_std_dev": 0.015230029187102692,
      "run_count": 5,
      "total_questions": 965,
      "avg_cost_per_task_usd": 0.11257541554404144,
      "avg_duration_per_task_ms": 438124.0186528497,
      "avg_output_tokens_per_task": 31261.672538860104,
      "last_run_timestamp": "2026-08-29T10:53:09.444Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Friendli",
      "accuracy": 0.8284511,
      "accuracy_std_dev": 0.007394059013153757,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.14032680555555557,
      "avg_duration_per_task_ms": 177948.25353535355,
      "avg_output_tokens_per_task": 34732.438888888886,
      "last_run_timestamp": "2026-10-24T08:41:54.142Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "StreamLake",
      "accuracy": 0.8283759292307693,
      "accuracy_std_dev": 0.023115878020096917,
      "run_count": 5,
      "total_questions": 975,
      "avg_cost_per_task_usd": 0.0696730758974359,
      "avg_duration_per_task_ms": 827348.3805128206,
      "avg_output_tokens_per_task": 39830.013333333336,
      "last_run_timestamp": "2026-08-30T00:37:59.094Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Decart",
      "accuracy": 0.82828275,
      "accuracy_std_dev": 0.004610363346581274,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.0223316047979798,
      "avg_duration_per_task_ms": 615131.3409090909,
      "avg_output_tokens_per_task": 56159.15404040404,
      "last_run_timestamp": "2026-10-17T08:18:07.611Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "CoreWeave",
      "accuracy": 0.8280900921348314,
      "accuracy_std_dev": 0.00897116480731462,
      "run_count": 5,
      "total_questions": 890,
      "avg_cost_per_task_usd": 0.02632537640449438,
      "avg_duration_per_task_ms": 122831.4988764045,
      "avg_output_tokens_per_task": 20925.239325842696,
      "last_run_timestamp": "2026-08-29T18:29:13.946Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "CoreWeave",
      "accuracy": 0.8279570000000002,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 186,
      "avg_cost_per_task_usd": 0.024709930107526885,
      "avg_duration_per_task_ms": 419969.8602150538,
      "avg_output_tokens_per_task": 20901.392473118278,
      "last_run_timestamp": "2026-08-01T18:03:43.857Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Cloudflare",
      "accuracy": 0.8276692169459963,
      "accuracy_std_dev": 0.017541679518523267,
      "run_count": 11,
      "total_questions": 2148,
      "avg_cost_per_task_usd": 0.14409020577281192,
      "avg_duration_per_task_ms": 292895.9767225326,
      "avg_output_tokens_per_task": 32531.135009310987,
      "last_run_timestamp": "2026-10-24T08:12:02.536Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-122b-a10b-20260224",
      "model_slug": "qwen/qwen3.5-122b-a10b",
      "model_name": "Qwen: Qwen3.5-122B-A10B",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8273709166666667,
      "accuracy_std_dev": 0.056014389335030584,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.051560792508417504,
      "avg_duration_per_task_ms": 531412.6561447812,
      "avg_output_tokens_per_task": 24685.848905723906,
      "last_run_timestamp": "2026-10-24T16:11:29.497Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Modal",
      "accuracy": 0.8273257868177137,
      "accuracy_std_dev": 0.013461365335284555,
      "run_count": 10,
      "total_questions": 1942,
      "avg_cost_per_task_usd": 0.0351326035015448,
      "avg_duration_per_task_ms": 424285.9001029866,
      "avg_output_tokens_per_task": 48592.330072090626,
      "last_run_timestamp": "2026-10-17T08:18:01.015Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Morph",
      "accuracy": 0.8272187207100591,
      "accuracy_std_dev": 0.0496186451996,
      "run_count": 13,
      "total_questions": 2535,
      "avg_cost_per_task_usd": 0.10137412702169626,
      "avg_duration_per_task_ms": 538485.309270217,
      "avg_output_tokens_per_task": 29237.781459566075,
      "last_run_timestamp": "2026-10-24T08:19:59.641Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "NextBit",
      "accuracy": 0.8271604444444445,
      "accuracy_std_dev": 0.023434861676623345,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.0269165101010101,
      "avg_duration_per_task_ms": 568738.2138047138,
      "avg_output_tokens_per_task": 46021.95230078563,
      "last_run_timestamp": "2026-10-17T08:17:59.868Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "CoreWeave",
      "accuracy": 0.8269359,
      "accuracy_std_dev": 0.017098920278485433,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.1309031328282828,
      "avg_duration_per_task_ms": 311479.3813131313,
      "avg_output_tokens_per_task": 54002.43080808081,
      "last_run_timestamp": "2026-10-24T16:04:33.470Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "BaseTen",
      "accuracy": 0.8269358,
      "accuracy_std_dev": 0.006767606146932625,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.04578402121212121,
      "avg_duration_per_task_ms": 60291.652525252524,
      "avg_output_tokens_per_task": 11285.754545454545,
      "last_run_timestamp": "2026-09-26T09:19:03.901Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "Google",
      "accuracy": 0.826809625,
      "accuracy_std_dev": 0.013025065527066464,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.09912519444444444,
      "avg_duration_per_task_ms": 360768.5435606061,
      "avg_output_tokens_per_task": 45556.78724747475,
      "last_run_timestamp": "2026-10-17T16:06:16.399Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Ionstream",
      "accuracy": 0.8267677,
      "accuracy_std_dev": 0.010999569410208776,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.03662119747474748,
      "avg_duration_per_task_ms": 290125.5409090909,
      "avg_output_tokens_per_task": 15502.71515151515,
      "last_run_timestamp": "2026-10-24T08:13:28.065Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.2-exp",
      "model_slug": "deepseek/deepseek-v3.2-exp",
      "model_name": "DeepSeek: DeepSeek V3.2 Exp",
      "provider_name": "Novita",
      "accuracy": 0.8265989999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.009182863636363637,
      "avg_duration_per_task_ms": 852550.7929292929,
      "avg_output_tokens_per_task": 21914.838383838385,
      "last_run_timestamp": "2026-08-01T22:57:01.346Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8265763479729729,
      "accuracy_std_dev": 0.010997465889922129,
      "run_count": 9,
      "total_questions": 1776,
      "avg_cost_per_task_usd": 0.00803203153153153,
      "avg_duration_per_task_ms": 497969.4971846847,
      "avg_output_tokens_per_task": 20675.66554054054,
      "last_run_timestamp": "2026-10-24T16:12:11.567Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Crusoe",
      "accuracy": 0.8264543613323679,
      "accuracy_std_dev": 0.018188360323715465,
      "run_count": 7,
      "total_questions": 1381,
      "avg_cost_per_task_usd": 0.20616153874004342,
      "avg_duration_per_task_ms": 315117.1817523534,
      "avg_output_tokens_per_task": 46875.9225199131,
      "last_run_timestamp": "2026-09-12T08:53:53.147Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Novita",
      "accuracy": 0.8264309000000002,
      "accuracy_std_dev": 0.009654922303674958,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.04312889949494949,
      "avg_duration_per_task_ms": 229184.7,
      "avg_output_tokens_per_task": 14260.256565656566,
      "last_run_timestamp": "2026-10-24T08:44:44.600Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "Phala",
      "accuracy": 0.8263402307692308,
      "accuracy_std_dev": 0.013738089499657219,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.06289175718725717,
      "avg_duration_per_task_ms": 535684.9638694639,
      "avg_output_tokens_per_task": 49516.1418026418,
      "last_run_timestamp": "2026-10-24T16:07:10.471Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Crusoe",
      "accuracy": 0.8261094014447884,
      "accuracy_std_dev": 0.026551028658038853,
      "run_count": 5,
      "total_questions": 969,
      "avg_cost_per_task_usd": 0.2076026150670795,
      "avg_duration_per_task_ms": 857894.439628483,
      "avg_output_tokens_per_task": 47560.28689370485,
      "last_run_timestamp": "2026-08-29T18:53:23.077Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "Baidu",
      "accuracy": 0.8258501862244899,
      "accuracy_std_dev": 0.010007618232626584,
      "run_count": 10,
      "total_questions": 1960,
      "avg_cost_per_task_usd": 0.007764268367346939,
      "avg_duration_per_task_ms": 349312.62346938776,
      "avg_output_tokens_per_task": 19101.2056122449,
      "last_run_timestamp": "2026-10-24T16:01:08.902Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Cohere",
      "accuracy": 0.8257575,
      "accuracy_std_dev": 0.017676499999999984,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.02807809090909091,
      "avg_duration_per_task_ms": 230919.77777777778,
      "avg_output_tokens_per_task": 55916.378787878784,
      "last_run_timestamp": "2026-10-24T08:15:10.265Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8254771111111111,
      "accuracy_std_dev": 0.00965472805813521,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.047990214365881044,
      "avg_duration_per_task_ms": 442130.5162738496,
      "avg_output_tokens_per_task": 26578.690796857463,
      "last_run_timestamp": "2026-10-24T16:31:34.472Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Morph",
      "accuracy": 0.8253999912733042,
      "accuracy_std_dev": 0.022060427402287506,
      "run_count": 13,
      "total_questions": 2521,
      "avg_cost_per_task_usd": 0.027058457754859183,
      "avg_duration_per_task_ms": 873701.1269337564,
      "avg_output_tokens_per_task": 48650.64220547402,
      "last_run_timestamp": "2026-10-24T08:26:00.559Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "CoreWeave",
      "accuracy": 0.8253968696428573,
      "accuracy_std_dev": 0.014772865576229071,
      "run_count": 9,
      "total_questions": 1680,
      "avg_cost_per_task_usd": 0.024977942261904765,
      "avg_duration_per_task_ms": 272190.5416666667,
      "avg_output_tokens_per_task": 49745.9630952381,
      "last_run_timestamp": "2026-10-24T08:27:40.063Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "auto-routing",
      "accuracy": 0.8252562970922881,
      "accuracy_std_dev": 0.043641090936944585,
      "run_count": 12,
      "total_questions": 2373,
      "avg_cost_per_task_usd": 0.08621131184155077,
      "avg_duration_per_task_ms": 759710.0198061526,
      "avg_output_tokens_per_task": 31352.24863042562,
      "last_run_timestamp": "2026-10-24T16:13:03.130Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Mancer 2",
      "accuracy": 0.8247287777777778,
      "accuracy_std_dev": 0.07543645506101702,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.012084622895622895,
      "avg_duration_per_task_ms": 775310.8434343435,
      "avg_output_tokens_per_task": 26356.983164983165,
      "last_run_timestamp": "2026-10-24T16:04:55.172Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "OpenInference",
      "accuracy": 0.8246474711042312,
      "accuracy_std_dev": 0.04610697461436829,
      "run_count": 10,
      "total_questions": 1938,
      "avg_cost_per_task_usd": 0.019926908668730647,
      "avg_duration_per_task_ms": 787644.0030959753,
      "avg_output_tokens_per_task": 45616.08255933953,
      "last_run_timestamp": "2026-10-24T08:22:23.706Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "auto-routing",
      "accuracy": 0.8244948000000002,
      "accuracy_std_dev": 0.0601779527199123,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.028505976262626263,
      "avg_duration_per_task_ms": 700158.9757575757,
      "avg_output_tokens_per_task": 29700.529292929292,
      "last_run_timestamp": "2026-10-17T08:18:04.911Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "DeepInfra",
      "accuracy": 0.8241368042686101,
      "accuracy_std_dev": 0.01320757262179545,
      "run_count": 10,
      "total_questions": 1921,
      "avg_cost_per_task_usd": 0.024530483081728265,
      "avg_duration_per_task_ms": 442142.1447162936,
      "avg_output_tokens_per_task": 24482.630921395106,
      "last_run_timestamp": "2026-10-24T16:11:33.501Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "AkashML",
      "accuracy": 0.8241029543178973,
      "accuracy_std_dev": 0.01722442160378035,
      "run_count": 17,
      "total_questions": 3196,
      "avg_cost_per_task_usd": 0.030665748122653304,
      "avg_duration_per_task_ms": 264011.0794743429,
      "avg_output_tokens_per_task": 13973.099186483103,
      "last_run_timestamp": "2026-10-24T08:58:05.316Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "CoreWeave",
      "accuracy": 0.8240257172774869,
      "accuracy_std_dev": 0.000479814547507715,
      "run_count": 3,
      "total_questions": 573,
      "avg_cost_per_task_usd": 0.1285148184991274,
      "avg_duration_per_task_ms": 295843.96684118675,
      "avg_output_tokens_per_task": 31965.242582897034,
      "last_run_timestamp": "2026-10-24T08:50:12.983Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "Xiaomi",
      "accuracy": 0.8239975454545455,
      "accuracy_std_dev": 0.022862830634275574,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.010785916437098255,
      "avg_duration_per_task_ms": 717852.0280073462,
      "avg_output_tokens_per_task": 38282.08126721763,
      "last_run_timestamp": "2026-10-17T08:02:36.531Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Venice",
      "accuracy": 0.8239540000000001,
      "accuracy_std_dev": 0.009474601205327877,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.14372687085137084,
      "avg_duration_per_task_ms": 420021.29581529583,
      "avg_output_tokens_per_task": 28768.17316017316,
      "last_run_timestamp": "2026-10-24T08:48:18.601Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "StreamLake",
      "accuracy": 0.8239056,
      "accuracy_std_dev": 0.013153034579138006,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.02469372222222222,
      "avg_duration_per_task_ms": 472106.2909090909,
      "avg_output_tokens_per_task": 23003.946464646466,
      "last_run_timestamp": "2026-08-29T18:09:03.379Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "SiliconFlow",
      "accuracy": 0.823871008431703,
      "accuracy_std_dev": 0.014025260116423268,
      "run_count": 9,
      "total_questions": 1779,
      "avg_cost_per_task_usd": 0.19393739516582348,
      "avg_duration_per_task_ms": 941267.5373805509,
      "avg_output_tokens_per_task": 51620.3125351321,
      "last_run_timestamp": "2026-10-24T16:12:49.071Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "Venice",
      "accuracy": 0.8236976925031766,
      "accuracy_std_dev": 0.11602625064229809,
      "run_count": 8,
      "total_questions": 1574,
      "avg_cost_per_task_usd": 0.12381996124523506,
      "avg_duration_per_task_ms": 755214.6099110546,
      "avg_output_tokens_per_task": 38337.82973316391,
      "last_run_timestamp": "2026-10-24T16:03:25.472Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.1-terminus",
      "model_slug": "deepseek/deepseek-v3.1-terminus",
      "model_name": "DeepSeek: DeepSeek V3.1 Terminus",
      "provider_name": "Novita",
      "accuracy": 0.8235687999999998,
      "accuracy_std_dev": 0.003085731900214278,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.018559439393939392,
      "avg_duration_per_task_ms": 535812.6131313131,
      "avg_output_tokens_per_task": 18445.026262626263,
      "last_run_timestamp": "2026-08-29T18:35:57.732Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "Alibaba",
      "accuracy": 0.823296396946565,
      "accuracy_std_dev": 0.009206241083574164,
      "run_count": 6,
      "total_questions": 1179,
      "avg_cost_per_task_usd": 0.023398153519932147,
      "avg_duration_per_task_ms": 341450.7421543681,
      "avg_output_tokens_per_task": 20864.225614927906,
      "last_run_timestamp": "2026-10-04T00:17:30.035Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "Novita",
      "accuracy": 0.823232125,
      "accuracy_std_dev": 0.02934086116850314,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.09752844696969695,
      "avg_duration_per_task_ms": 1429250.3825757576,
      "avg_output_tokens_per_task": 44988.86111111111,
      "last_run_timestamp": "2026-10-24T16:04:55.201Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "GLM 5.3",
      "provider_name": "digitalocean",
      "accuracy": 0.823232,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.025912994949494946,
      "avg_duration_per_task_ms": 162312.898989899,
      "avg_output_tokens_per_task": 9036.257575757576,
      "last_run_timestamp": "2026-10-01T23:50:45.024Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-2.5-pro",
      "model_slug": "google/gemini-2.5-pro",
      "model_name": "Google: Gemini 2.5 Pro",
      "provider_name": "Google AI Studio",
      "accuracy": 0.8223905000000001,
      "accuracy_std_dev": 0.0244105,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.2415952904040404,
      "avg_duration_per_task_ms": 181953.0101010101,
      "avg_output_tokens_per_task": 24066.49494949495,
      "last_run_timestamp": "2026-10-01T19:12:31.045Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Io Net",
      "accuracy": 0.8223551976047904,
      "accuracy_std_dev": 0.029631373343505716,
      "run_count": 6,
      "total_questions": 1169,
      "avg_cost_per_task_usd": 0.06724194183062446,
      "avg_duration_per_task_ms": 236148.74337040205,
      "avg_output_tokens_per_task": 20314.131736526946,
      "last_run_timestamp": "2026-10-17T08:26:17.960Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "Novita",
      "accuracy": 0.8223140000000001,
      "accuracy_std_dev": 0.020822165248512558,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.012249293847566574,
      "avg_duration_per_task_ms": 720758.7061524334,
      "avg_output_tokens_per_task": 36235.862258953166,
      "last_run_timestamp": "2026-10-17T08:04:21.262Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Chutes",
      "accuracy": 0.8223139999999999,
      "accuracy_std_dev": 0.02011718801965576,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.03311863039485767,
      "avg_duration_per_task_ms": 262977.6432506887,
      "avg_output_tokens_per_task": 14545.10927456382,
      "last_run_timestamp": "2026-10-24T08:29:59.056Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "AtlasCloud",
      "accuracy": 0.82211,
      "accuracy_std_dev": 0.057194952006274134,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.09893642424242423,
      "avg_duration_per_task_ms": 769490.925925926,
      "avg_output_tokens_per_task": 39542.70622895623,
      "last_run_timestamp": "2026-10-03T17:58:22.995Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "Phala",
      "accuracy": 0.821895294117647,
      "accuracy_std_dev": 0.07409118108715027,
      "run_count": 3,
      "total_questions": 408,
      "avg_cost_per_task_usd": 0.13934752450980392,
      "avg_duration_per_task_ms": 1455873.1495098039,
      "avg_output_tokens_per_task": 48235.80147058824,
      "last_run_timestamp": "2026-08-29T18:25:52.753Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Alibaba",
      "accuracy": 0.8214756245059289,
      "accuracy_std_dev": 0.05273064972335267,
      "run_count": 9,
      "total_questions": 1771,
      "avg_cost_per_task_usd": 0.18877710615471485,
      "avg_duration_per_task_ms": 607666.5719932242,
      "avg_output_tokens_per_task": 44926.648785996615,
      "last_run_timestamp": "2026-10-24T16:01:45.799Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "DeepInfra",
      "accuracy": 0.8214198161925602,
      "accuracy_std_dev": 0.04954945097072217,
      "run_count": 8,
      "total_questions": 1371,
      "avg_cost_per_task_usd": 0.04645164551422319,
      "avg_duration_per_task_ms": 608936.6484318016,
      "avg_output_tokens_per_task": 28638.388767323122,
      "last_run_timestamp": "2026-09-26T09:56:00.145Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Makora",
      "accuracy": 0.8213506405228759,
      "accuracy_std_dev": 0.02248216924587128,
      "run_count": 3,
      "total_questions": 459,
      "avg_cost_per_task_usd": 0.08758356862745097,
      "avg_duration_per_task_ms": 304618.9760348584,
      "avg_output_tokens_per_task": 34720.03050108933,
      "last_run_timestamp": "2026-08-29T10:57:24.797Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "CoreWeave",
      "accuracy": 0.8212682499999999,
      "accuracy_std_dev": 0.0185418798135689,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.04273052441077441,
      "avg_duration_per_task_ms": 239740.20075757575,
      "avg_output_tokens_per_task": 14238.065656565657,
      "last_run_timestamp": "2026-10-24T08:50:39.623Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta/muse-glimmer-30b-20260810",
      "model_slug": "meta/muse-glimmer-30b",
      "model_name": "Meta: Muse Glimmer 30B",
      "provider_name": "Parasail",
      "accuracy": 0.8212070710659898,
      "accuracy_std_dev": 0.011946281848340899,
      "run_count": 3,
      "total_questions": 591,
      "avg_cost_per_task_usd": 0.01807649915397631,
      "avg_duration_per_task_ms": 302940.0592216582,
      "avg_output_tokens_per_task": 16413.382402707277,
      "last_run_timestamp": "2026-08-29T09:59:35.603Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Io Net",
      "accuracy": 0.8210525603238865,
      "accuracy_std_dev": 0.07599126782917703,
      "run_count": 7,
      "total_questions": 1235,
      "avg_cost_per_task_usd": 0.007044106882591094,
      "avg_duration_per_task_ms": 525854.9595141701,
      "avg_output_tokens_per_task": 14201.662348178137,
      "last_run_timestamp": "2026-10-03T10:22:02.091Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "auto-routing",
      "accuracy": 0.8210063161094224,
      "accuracy_std_dev": 0.031028311643400763,
      "run_count": 5,
      "total_questions": 987,
      "avg_cost_per_task_usd": 0.025833400202634246,
      "avg_duration_per_task_ms": 497222.336372847,
      "avg_output_tokens_per_task": 24022.691995947316,
      "last_run_timestamp": "2026-08-29T18:05:36.579Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.5-sonnet-20250929",
      "model_slug": "anthropic/claude-sonnet-4.5",
      "model_name": "Anthropic: Claude Sonnet 4.5",
      "provider_name": "auto-routing",
      "accuracy": 0.8202537692307693,
      "accuracy_std_dev": 0.0131593300691875,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.31460015734265734,
      "avg_duration_per_task_ms": 289943.24397824396,
      "avg_output_tokens_per_task": 20783.278943278943,
      "last_run_timestamp": "2026-10-24T08:12:02.536Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "DeepInfra",
      "accuracy": 0.8199838489742471,
      "accuracy_std_dev": 0.060002328561629244,
      "run_count": 24,
      "total_questions": 4582,
      "avg_cost_per_task_usd": 0.007503320384111742,
      "avg_duration_per_task_ms": 496893.82627673505,
      "avg_output_tokens_per_task": 17430.922959406373,
      "last_run_timestamp": "2026-10-24T16:15:47.415Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "StreamLake",
      "accuracy": 0.819865,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.13854771717171718,
      "avg_duration_per_task_ms": 540824.1212121212,
      "avg_output_tokens_per_task": 31240.651515151516,
      "last_run_timestamp": "2026-09-07T06:00:57.184Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "DigitalOcean",
      "accuracy": 0.8197662921174653,
      "accuracy_std_dev": 0.015949041727953438,
      "run_count": 10,
      "total_questions": 1941,
      "avg_cost_per_task_usd": 0.018771416280267902,
      "avg_duration_per_task_ms": 575656.4595569294,
      "avg_output_tokens_per_task": 20084.957753735187,
      "last_run_timestamp": "2026-10-24T16:01:29.842Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "SambaNova",
      "accuracy": 0.8192149207920791,
      "accuracy_std_dev": 0.018823182524040232,
      "run_count": 6,
      "total_questions": 909,
      "avg_cost_per_task_usd": 0.11224348184818483,
      "avg_duration_per_task_ms": 401079.6094609461,
      "avg_output_tokens_per_task": 46532.80198019802,
      "last_run_timestamp": "2026-10-17T16:06:46.637Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "DigitalOcean",
      "accuracy": 0.8191437142857143,
      "accuracy_std_dev": 0.03986796895724072,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.04288789971139971,
      "avg_duration_per_task_ms": 992008.3816738817,
      "avg_output_tokens_per_task": 26956.146464646463,
      "last_run_timestamp": "2026-10-17T08:20:47.476Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "CoreWeave",
      "accuracy": 0.8191170000000002,
      "accuracy_std_dev": 0.01945730630551585,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.006372274410774412,
      "avg_duration_per_task_ms": 464617.35016835015,
      "avg_output_tokens_per_task": 18796.420314253646,
      "last_run_timestamp": "2026-10-24T16:31:34.552Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "Darkbloom",
      "accuracy": 0.8184965756676558,
      "accuracy_std_dev": 0.008315500000000031,
      "run_count": 2,
      "total_questions": 337,
      "avg_cost_per_task_usd": 0.010153243323442137,
      "avg_duration_per_task_ms": 1501083.5370919881,
      "avg_output_tokens_per_task": 28895.703264094955,
      "last_run_timestamp": "2026-10-24T08:27:40.065Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "auto-routing",
      "accuracy": 0.8183501,
      "accuracy_std_dev": 0.014134440253862193,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.033925831313131316,
      "avg_duration_per_task_ms": 215707.2409090909,
      "avg_output_tokens_per_task": 13837.719191919192,
      "last_run_timestamp": "2026-10-24T08:16:59.511Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "StreamLake",
      "accuracy": 0.8181818,
      "accuracy_std_dev": 0.02137456277353997,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.1023312393939394,
      "avg_duration_per_task_ms": 562383.997979798,
      "avg_output_tokens_per_task": 37995.12222222222,
      "last_run_timestamp": "2026-08-29T18:26:14.942Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-5-fable-20260609",
      "model_slug": "anthropic/claude-fable-5",
      "model_name": "Anthropic: Claude Fable 5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.8181816666666667,
      "accuracy_std_dev": 0.013255807590972678,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.17135338047138046,
      "avg_duration_per_task_ms": 45814.95117845118,
      "avg_output_tokens_per_task": 3234.355218855219,
      "last_run_timestamp": "2026-08-15T09:15:35.213Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Mancer 2",
      "accuracy": 0.818143387341772,
      "accuracy_std_dev": 0.008051034079545293,
      "run_count": 8,
      "total_questions": 1580,
      "avg_cost_per_task_usd": 0.03837104746835443,
      "avg_duration_per_task_ms": 216584.66329113924,
      "avg_output_tokens_per_task": 14732.209493670885,
      "last_run_timestamp": "2026-10-24T08:26:33.259Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "thinkingmachines/inkling-20260715",
      "model_slug": "thinkingmachines/inkling",
      "model_name": "Thinking Machines: Inkling",
      "provider_name": "BaseTen",
      "accuracy": 0.818035862576065,
      "accuracy_std_dev": 0.01278056208505712,
      "run_count": 10,
      "total_questions": 1972,
      "avg_cost_per_task_usd": 0.0909173985801217,
      "avg_duration_per_task_ms": 165913.36257606492,
      "avg_output_tokens_per_task": 22308.299188640973,
      "last_run_timestamp": "2026-09-19T10:14:07.966Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta/muse-glimmer-30b-20260810",
      "model_slug": "meta/muse-glimmer-30b",
      "model_name": "Meta: Muse Glimmer 30B",
      "provider_name": "Together",
      "accuracy": 0.8178077777777777,
      "accuracy_std_dev": 0.009999885885768644,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.024910318742985414,
      "avg_duration_per_task_ms": 190389.1571268238,
      "avg_output_tokens_per_task": 16514.1734006734,
      "last_run_timestamp": "2026-10-03T10:10:35.343Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "StreamLake",
      "accuracy": 0.8177376144306652,
      "accuracy_std_dev": 0.03151916867911054,
      "run_count": 9,
      "total_questions": 1774,
      "avg_cost_per_task_usd": 0.013423886696730551,
      "avg_duration_per_task_ms": 855848.4718151072,
      "avg_output_tokens_per_task": 39679.556933483655,
      "last_run_timestamp": "2026-10-17T08:19:03.202Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "DigitalOcean",
      "accuracy": 0.8174118475609755,
      "accuracy_std_dev": 0.019551780804826967,
      "run_count": 5,
      "total_questions": 984,
      "avg_cost_per_task_usd": 0.02269767581300813,
      "avg_duration_per_task_ms": 492334.5833333333,
      "avg_output_tokens_per_task": 22647.941056910568,
      "last_run_timestamp": "2026-08-29T18:16:26.161Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Phala",
      "accuracy": 0.8173077377622378,
      "accuracy_std_dev": 0.04183386114142466,
      "run_count": 3,
      "total_questions": 572,
      "avg_cost_per_task_usd": 0.00918055944055944,
      "avg_duration_per_task_ms": 556054.2482517483,
      "avg_output_tokens_per_task": 20397.216783216783,
      "last_run_timestamp": "2026-08-29T19:08:46.030Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "Venice",
      "accuracy": 0.8172464444444444,
      "accuracy_std_dev": 0.007234812308263586,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.027580057800224465,
      "avg_duration_per_task_ms": 186978.31425364758,
      "avg_output_tokens_per_task": 23463.768237934906,
      "last_run_timestamp": "2026-10-24T16:20:59.116Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8170044527027026,
      "accuracy_std_dev": 0.0041701025034031,
      "run_count": 9,
      "total_questions": 1776,
      "avg_cost_per_task_usd": 0.010224835585585587,
      "avg_duration_per_task_ms": 768325.4386261262,
      "avg_output_tokens_per_task": 18282.219594594593,
      "last_run_timestamp": "2026-10-24T16:12:04.706Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta/muse-glimmer-30b-20260810",
      "model_slug": "meta/muse-glimmer-30b",
      "model_name": "Meta: Muse Glimmer 30B",
      "provider_name": "Phala",
      "accuracy": 0.816967559819413,
      "accuracy_std_dev": 0.022532849378432182,
      "run_count": 9,
      "total_questions": 1772,
      "avg_cost_per_task_usd": 0.021389000564334085,
      "avg_duration_per_task_ms": 297135.54796839727,
      "avg_output_tokens_per_task": 17946.336343115123,
      "last_run_timestamp": "2026-10-03T09:25:02.390Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "auto-routing",
      "accuracy": 0.8167087499999999,
      "accuracy_std_dev": 0.01709464425448801,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.007263619318181818,
      "avg_duration_per_task_ms": 512210.79671717173,
      "avg_output_tokens_per_task": 18196.01452020202,
      "last_run_timestamp": "2026-10-24T16:06:30.906Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Darkbloom",
      "accuracy": 0.8166852222222222,
      "accuracy_std_dev": 0.009167637558011412,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.03448444051627385,
      "avg_duration_per_task_ms": 613993.6262626263,
      "avg_output_tokens_per_task": 16126.99775533109,
      "last_run_timestamp": "2026-10-24T09:01:52.623Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "auto-routing",
      "accuracy": 0.8166289763513515,
      "accuracy_std_dev": 0.02933294963006619,
      "run_count": 9,
      "total_questions": 1776,
      "avg_cost_per_task_usd": 0.08447106756756757,
      "avg_duration_per_task_ms": 968983.0568693693,
      "avg_output_tokens_per_task": 41870.689752252256,
      "last_run_timestamp": "2026-10-24T16:15:14.401Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.5-sonnet-20250929",
      "model_slug": "anthropic/claude-sonnet-4.5",
      "model_name": "Anthropic: Claude Sonnet 4.5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.8162661034482758,
      "accuracy_std_dev": 0.01119088698017683,
      "run_count": 29,
      "total_questions": 5742,
      "avg_cost_per_task_usd": 0.33315613949843254,
      "avg_duration_per_task_ms": 297815.49094392196,
      "avg_output_tokens_per_task": 21074.363636363636,
      "last_run_timestamp": "2026-10-24T08:56:14.444Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Friendli",
      "accuracy": 0.8160686215384615,
      "accuracy_std_dev": 0.03359599240742859,
      "run_count": 5,
      "total_questions": 975,
      "avg_cost_per_task_usd": 0.026009656410256408,
      "avg_duration_per_task_ms": 463494.8851282051,
      "avg_output_tokens_per_task": 21626.916923076922,
      "last_run_timestamp": "2026-08-29T19:19:06.625Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.1-flash-lite-20260507",
      "model_slug": "google/gemini-3.1-flash-lite",
      "model_name": "Google: Gemini 3.1 Flash Lite",
      "provider_name": "auto-routing",
      "accuracy": 0.815937,
      "accuracy_std_dev": 0.021042110493009022,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.04137441245791246,
      "avg_duration_per_task_ms": 82874.20875420875,
      "avg_output_tokens_per_task": 27465.196969696968,
      "last_run_timestamp": "2026-10-01T19:23:01.327Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "Novita",
      "accuracy": 0.815488,
      "accuracy_std_dev": 0.010410666606898895,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.01878690909090909,
      "avg_duration_per_task_ms": 548800.3747474748,
      "avg_output_tokens_per_task": 18588.853535353537,
      "last_run_timestamp": "2026-08-30T00:16:38.394Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "BaseTen",
      "accuracy": 0.815286480254777,
      "accuracy_std_dev": 0.006665997576319691,
      "run_count": 4,
      "total_questions": 785,
      "avg_cost_per_task_usd": 0.08723442165605096,
      "avg_duration_per_task_ms": 185900.4152866242,
      "avg_output_tokens_per_task": 24915.47388535032,
      "last_run_timestamp": "2026-08-29T18:58:14.477Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "auto-routing",
      "accuracy": 0.8152233565365026,
      "accuracy_std_dev": 0.024345312062738052,
      "run_count": 9,
      "total_questions": 1767,
      "avg_cost_per_task_usd": 0.21931891850594232,
      "avg_duration_per_task_ms": 991109.1279003961,
      "avg_output_tokens_per_task": 59192.10016977929,
      "last_run_timestamp": "2026-10-24T16:04:12.636Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Z.AI",
      "accuracy": 0.8151274695945945,
      "accuracy_std_dev": 0.04695932867392988,
      "run_count": 9,
      "total_questions": 1776,
      "avg_cost_per_task_usd": 0.2281274650900901,
      "avg_duration_per_task_ms": 1121263.125563063,
      "avg_output_tokens_per_task": 52145.57432432433,
      "last_run_timestamp": "2026-10-24T16:06:30.902Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.2-exp",
      "model_slug": "deepseek/deepseek-v3.2-exp",
      "model_name": "DeepSeek: DeepSeek V3.2 Exp",
      "provider_name": "SiliconFlow",
      "accuracy": 0.814815,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.006498378787878787,
      "avg_duration_per_task_ms": 589595.8333333334,
      "avg_output_tokens_per_task": 15681.459595959595,
      "last_run_timestamp": "2026-08-01T18:25:37.046Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.2-exp",
      "model_slug": "deepseek/deepseek-v3.2-exp",
      "model_name": "DeepSeek: DeepSeek V3.2 Exp",
      "provider_name": "auto-routing",
      "accuracy": 0.814815,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 189,
      "avg_cost_per_task_usd": 0.007393566137566137,
      "avg_duration_per_task_ms": 549965.8571428572,
      "avg_output_tokens_per_task": 17741.814814814814,
      "last_run_timestamp": "2026-08-01T18:07:34.003Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Parasail",
      "accuracy": 0.8147209999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 197,
      "avg_cost_per_task_usd": 0.024774192893401017,
      "avg_duration_per_task_ms": 272533.6497461929,
      "avg_output_tokens_per_task": 20619.17766497462,
      "last_run_timestamp": "2026-08-01T18:21:26.176Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta/muse-glimmer-30b-20260810",
      "model_slug": "meta/muse-glimmer-30b",
      "model_name": "Meta: Muse Glimmer 30B",
      "provider_name": "Fireworks",
      "accuracy": 0.8143338571428571,
      "accuracy_std_dev": 0.014248444761532711,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.026212479076479073,
      "avg_duration_per_task_ms": 329590.34992784995,
      "avg_output_tokens_per_task": 17511.52380952381,
      "last_run_timestamp": "2026-09-19T09:05:52.633Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.5-sonnet-20250929",
      "model_slug": "anthropic/claude-sonnet-4.5",
      "model_name": "Anthropic: Claude Sonnet 4.5",
      "provider_name": "Google",
      "accuracy": 0.8142967307692308,
      "accuracy_std_dev": 0.010500659783511362,
      "run_count": 26,
      "total_questions": 5148,
      "avg_cost_per_task_usd": 0.36391435470085476,
      "avg_duration_per_task_ms": 358484.0462315462,
      "avg_output_tokens_per_task": 22920.8277000777,
      "last_run_timestamp": "2026-10-24T08:44:41.412Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Tenstorrent",
      "accuracy": 0.8138046,
      "accuracy_std_dev": 0.01492933679169976,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.010332651515151515,
      "avg_duration_per_task_ms": 579885.2626262626,
      "avg_output_tokens_per_task": 12515.30505050505,
      "last_run_timestamp": "2026-10-24T08:57:37.499Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Morph",
      "accuracy": 0.8136272825651302,
      "accuracy_std_dev": 0.008773473238499465,
      "run_count": 3,
      "total_questions": 499,
      "avg_cost_per_task_usd": 0.0068851422845691386,
      "avg_duration_per_task_ms": 1090136,
      "avg_output_tokens_per_task": 17404.21643286573,
      "last_run_timestamp": "2026-08-15T10:05:32.709Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-2.5-pro",
      "model_slug": "google/gemini-2.5-pro",
      "model_name": "Google: Gemini 2.5 Pro",
      "provider_name": "auto-routing",
      "accuracy": 0.8131313333333333,
      "accuracy_std_dev": 0.02212166805545088,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.2550713585858586,
      "avg_duration_per_task_ms": 227479.37878787878,
      "avg_output_tokens_per_task": 25491.23232323232,
      "last_run_timestamp": "2026-10-01T19:12:42.150Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.2-chat-20251211",
      "model_slug": "openai/gpt-5.2-chat",
      "model_name": "OpenAI: GPT-5.2 Chat",
      "provider_name": "OpenAI",
      "accuracy": 0.813131,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.037260742424242425,
      "avg_duration_per_task_ms": 35156.99494949495,
      "avg_output_tokens_per_task": 2566.8030303030305,
      "last_run_timestamp": "2026-08-01T08:32:23.484Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "DigitalOcean",
      "accuracy": 0.8130656097285068,
      "accuracy_std_dev": 0.09667022957112152,
      "run_count": 9,
      "total_questions": 1768,
      "avg_cost_per_task_usd": 0.061109888574660626,
      "avg_duration_per_task_ms": 1704990.4128959277,
      "avg_output_tokens_per_task": 19426.679864253394,
      "last_run_timestamp": "2026-10-24T16:10:43.374Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "inclusionai/ling-3.1-flash-20261002",
      "model_slug": "inclusionai/ling-3.1-flash",
      "model_name": "inclusionAI: Ling 3.1 Flash",
      "provider_name": "Novita",
      "accuracy": 0.813051,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 189,
      "avg_cost_per_task_usd": 0.00766168253968254,
      "avg_duration_per_task_ms": 356938.3703703704,
      "avg_output_tokens_per_task": 33904.37566137566,
      "last_run_timestamp": "2026-10-14T18:00:03.925Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Tenstorrent",
      "accuracy": 0.8127462577873253,
      "accuracy_std_dev": 0.009366182671718485,
      "run_count": 5,
      "total_questions": 931,
      "avg_cost_per_task_usd": 0.043591290010741135,
      "avg_duration_per_task_ms": 179665.51987110634,
      "avg_output_tokens_per_task": 17266.0977443609,
      "last_run_timestamp": "2026-10-24T08:11:11.333Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "StreamLake",
      "accuracy": 0.8124578000000001,
      "accuracy_std_dev": 0.006784287240381237,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.00604990505050505,
      "avg_duration_per_task_ms": 545364.5707070708,
      "avg_output_tokens_per_task": 18416.73232323232,
      "last_run_timestamp": "2026-08-29T19:09:30.030Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Novita",
      "accuracy": 0.8122996179966044,
      "accuracy_std_dev": 0.02424420574075381,
      "run_count": 9,
      "total_questions": 1767,
      "avg_cost_per_task_usd": 0.26123569835880023,
      "avg_duration_per_task_ms": 965363.1103565365,
      "avg_output_tokens_per_task": 59428.235427277876,
      "last_run_timestamp": "2026-10-24T16:00:53.339Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Near AI",
      "accuracy": 0.8122604367816092,
      "accuracy_std_dev": 0.03327589059946405,
      "run_count": 4,
      "total_questions": 783,
      "avg_cost_per_task_usd": 0.024626680715197958,
      "avg_duration_per_task_ms": 153695.50446998724,
      "avg_output_tokens_per_task": 14302.206896551725,
      "last_run_timestamp": "2026-10-24T08:18:45.292Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "DigitalOcean",
      "accuracy": 0.812079,
      "accuracy_std_dev": 0.1279844797709863,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.15244413194444442,
      "avg_duration_per_task_ms": 1980931.956439394,
      "avg_output_tokens_per_task": 55523.080808080806,
      "last_run_timestamp": "2026-10-24T16:14:24.978Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3-ultra-550b-a55b-20260604",
      "model_slug": "nvidia/nemotron-3-ultra-550b-a55b",
      "model_name": "NVIDIA: Nemotron 3 Ultra",
      "provider_name": "Together",
      "accuracy": 0.812016,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 172,
      "avg_cost_per_task_usd": 0.14135672674418603,
      "avg_duration_per_task_ms": 305120.88372093026,
      "avg_output_tokens_per_task": 39843.74418604651,
      "last_run_timestamp": "2026-08-04T03:11:59.879Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Ionstream",
      "accuracy": 0.8118685,
      "accuracy_std_dev": 0.0618254994258033,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.042041156565656566,
      "avg_duration_per_task_ms": 1127766.4457070706,
      "avg_output_tokens_per_task": 24710.378787878788,
      "last_run_timestamp": "2026-08-29T19:26:41.546Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "auto-routing",
      "accuracy": 0.8114476666666668,
      "accuracy_std_dev": 0.024344661016329636,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.00885003591470258,
      "avg_duration_per_task_ms": 567443.1509539842,
      "avg_output_tokens_per_task": 18949.043209876545,
      "last_run_timestamp": "2026-10-24T16:00:55.675Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.5-sonnet-20250929",
      "model_slug": "anthropic/claude-sonnet-4.5",
      "model_name": "Anthropic: Claude Sonnet 4.5",
      "provider_name": "Anthropic",
      "accuracy": 0.8113425625,
      "accuracy_std_dev": 0.01663880630096684,
      "run_count": 16,
      "total_questions": 3168,
      "avg_cost_per_task_usd": 0.34052509564393935,
      "avg_duration_per_task_ms": 354477.9971590909,
      "avg_output_tokens_per_task": 22529.18023989899,
      "last_run_timestamp": "2026-10-24T08:55:49.941Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Reka",
      "accuracy": 0.8111930073800738,
      "accuracy_std_dev": 0.03224095604623893,
      "run_count": 7,
      "total_questions": 813,
      "avg_cost_per_task_usd": 0.05198734809348094,
      "avg_duration_per_task_ms": 309460.76998769987,
      "avg_output_tokens_per_task": 20725.47355473555,
      "last_run_timestamp": "2026-10-24T16:02:08.817Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.1-flash-lite-20260507",
      "model_slug": "google/gemini-3.1-flash-lite",
      "model_name": "Google: Gemini 3.1 Flash Lite",
      "provider_name": "Google AI Studio",
      "accuracy": 0.8103253333333333,
      "accuracy_std_dev": 0.011527714209195522,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.04107354882154882,
      "avg_duration_per_task_ms": 75169.15151515152,
      "avg_output_tokens_per_task": 27258.996632996634,
      "last_run_timestamp": "2026-10-01T19:16:58.595Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-122b-a10b-20260224",
      "model_slug": "qwen/qwen3.5-122b-a10b",
      "model_name": "Qwen: Qwen3.5-122B-A10B",
      "provider_name": "Alibaba",
      "accuracy": 0.8099945799011532,
      "accuracy_std_dev": 0.10451583165401959,
      "run_count": 13,
      "total_questions": 2428,
      "avg_cost_per_task_usd": 0.042656735584843494,
      "avg_duration_per_task_ms": 161924.82866556838,
      "avg_output_tokens_per_task": 20413.041598023065,
      "last_run_timestamp": "2026-10-24T16:12:07.781Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Relace",
      "accuracy": 0.8091166153846153,
      "accuracy_std_dev": 0.03183613126622014,
      "run_count": 3,
      "total_questions": 585,
      "avg_cost_per_task_usd": 0.19830280854700855,
      "avg_duration_per_task_ms": 1060541.5008547008,
      "avg_output_tokens_per_task": 55592.3264957265,
      "last_run_timestamp": "2026-10-03T16:58:49.489Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "Crusoe",
      "accuracy": 0.8089225,
      "accuracy_std_dev": 0.07491550000000002,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.023557068181818182,
      "avg_duration_per_task_ms": 701295.6338383838,
      "avg_output_tokens_per_task": 58706.843434343435,
      "last_run_timestamp": "2026-10-24T08:01:41.139Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "StreamLake",
      "accuracy": 0.8089224,
      "accuracy_std_dev": 0.023427325085036924,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.11485773383838385,
      "avg_duration_per_task_ms": 769661.1737373738,
      "avg_output_tokens_per_task": 59732.906565656565,
      "last_run_timestamp": "2026-10-24T16:03:25.400Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.5-sonnet-20250929",
      "model_slug": "anthropic/claude-sonnet-4.5",
      "model_name": "Anthropic: Claude Sonnet 4.5",
      "provider_name": "Azure",
      "accuracy": 0.8082103076923077,
      "accuracy_std_dev": 0.02014208867935159,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.3422723123543123,
      "avg_duration_per_task_ms": 356747.24864024867,
      "avg_output_tokens_per_task": 22654.46153846154,
      "last_run_timestamp": "2026-10-24T08:14:36.266Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2-0905",
      "model_slug": "moonshotai/kimi-k2-0905",
      "model_name": "MoonshotAI: Kimi K2 0905",
      "provider_name": "Novita",
      "accuracy": 0.808081,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.016894853535353537,
      "avg_duration_per_task_ms": 156500.61111111112,
      "avg_output_tokens_per_task": 6566.772727272727,
      "last_run_timestamp": "2026-08-01T17:44:11.144Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5-mini-2025-08-07",
      "model_slug": "openai/gpt-5-mini",
      "model_name": "OpenAI: GPT-5 Mini",
      "provider_name": "auto-routing",
      "accuracy": 0.808081,
      "accuracy_std_dev": 0.018027342695657245,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.044340436026936034,
      "avg_duration_per_task_ms": 217888.1734006734,
      "avg_output_tokens_per_task": 22075.030303030304,
      "last_run_timestamp": "2026-10-01T19:19:28.473Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Venice",
      "accuracy": 0.8069390466155811,
      "accuracy_std_dev": 0.09137699015060836,
      "run_count": 8,
      "total_questions": 1566,
      "avg_cost_per_task_usd": 0.0966928748403576,
      "avg_duration_per_task_ms": 758869.1143039592,
      "avg_output_tokens_per_task": 28380.814814814814,
      "last_run_timestamp": "2026-10-24T16:10:43.300Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "DeepInfra",
      "accuracy": 0.806699224509804,
      "accuracy_std_dev": 0.027467920997381046,
      "run_count": 7,
      "total_questions": 1020,
      "avg_cost_per_task_usd": 0.07044824803921569,
      "avg_duration_per_task_ms": 1518236.9009803922,
      "avg_output_tokens_per_task": 40306.42941176471,
      "last_run_timestamp": "2026-10-24T16:26:52.183Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "ModelRun",
      "accuracy": 0.8066778333333334,
      "accuracy_std_dev": 0.00898293258382558,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.04676386700336701,
      "avg_duration_per_task_ms": 57672.9595959596,
      "avg_output_tokens_per_task": 15383.77946127946,
      "last_run_timestamp": "2026-10-24T08:45:03.799Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8065842222222221,
      "accuracy_std_dev": 0.02588009608765339,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.039878716610549945,
      "avg_duration_per_task_ms": 183894.81369248035,
      "avg_output_tokens_per_task": 22053.50448933782,
      "last_run_timestamp": "2026-10-24T16:05:36.261Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-2.5-pro",
      "model_slug": "google/gemini-2.5-pro",
      "model_name": "Google: Gemini 2.5 Pro",
      "provider_name": "Google",
      "accuracy": 0.806210222222222,
      "accuracy_std_dev": 0.03697273211969953,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.25474633557800225,
      "avg_duration_per_task_ms": 231596.29349046017,
      "avg_output_tokens_per_task": 25466.63243546577,
      "last_run_timestamp": "2026-10-01T19:17:25.275Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta/muse-glimmer-30b-20260810",
      "model_slug": "meta/muse-glimmer-30b",
      "model_name": "Meta: Muse Glimmer 30B",
      "provider_name": "auto-routing",
      "accuracy": 0.8059763749999999,
      "accuracy_std_dev": 0.01367028747811745,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.022531703914141412,
      "avg_duration_per_task_ms": 251092.985479798,
      "avg_output_tokens_per_task": 17458.415404040403,
      "last_run_timestamp": "2026-10-03T09:13:36.277Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "OpenInference",
      "accuracy": 0.8058380621468926,
      "accuracy_std_dev": 0.060531258531109376,
      "run_count": 5,
      "total_questions": 885,
      "avg_cost_per_task_usd": 0.012335033898305085,
      "avg_duration_per_task_ms": 403352.9514124294,
      "avg_output_tokens_per_task": 16083.899435028248,
      "last_run_timestamp": "2026-10-17T16:05:00.316Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-thinking-2507",
      "model_slug": "qwen/qwen3-235b-a22b-thinking-2507",
      "model_name": "Qwen: Qwen3 235B A22B Thinking 2507",
      "provider_name": "DeepInfra",
      "accuracy": 0.8057394503311259,
      "accuracy_std_dev": 0.027315871855937524,
      "run_count": 4,
      "total_questions": 604,
      "avg_cost_per_task_usd": 0.04681624668874173,
      "avg_duration_per_task_ms": 685385.5794701987,
      "avg_output_tokens_per_task": 20286.42715231788,
      "last_run_timestamp": "2026-08-22T16:06:14.894Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Chutes",
      "accuracy": 0.805556,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 192,
      "avg_cost_per_task_usd": 0.16585933333333333,
      "avg_duration_per_task_ms": 688743.8177083334,
      "avg_output_tokens_per_task": 42426.447916666664,
      "last_run_timestamp": "2026-07-31T00:21:35.641Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "AtlasCloud",
      "accuracy": 0.805345125,
      "accuracy_std_dev": 0.036272944774437245,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.19822186489898988,
      "avg_duration_per_task_ms": 832606.8560606061,
      "avg_output_tokens_per_task": 49537.894570707074,
      "last_run_timestamp": "2026-10-24T16:31:00.679Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Chutes",
      "accuracy": 0.8049668265895952,
      "accuracy_std_dev": 0.04314684443397195,
      "run_count": 8,
      "total_questions": 1557,
      "avg_cost_per_task_usd": 0.14483265382145152,
      "avg_duration_per_task_ms": 1590993.8619139371,
      "avg_output_tokens_per_task": 47443.71419396275,
      "last_run_timestamp": "2026-10-24T16:00:53.339Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.5-sonnet-20250929",
      "model_slug": "anthropic/claude-sonnet-4.5",
      "model_name": "Anthropic: Claude Sonnet 4.5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.8048667272727271,
      "accuracy_std_dev": 0.020423284626269588,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.33759371303948577,
      "avg_duration_per_task_ms": 354437.65059687785,
      "avg_output_tokens_per_task": 22371.59963269054,
      "last_run_timestamp": "2026-10-24T08:45:54.077Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.3-chat-20260303",
      "model_slug": "openai/gpt-5.3-chat",
      "model_name": "OpenAI: GPT-5.3 Chat",
      "provider_name": "auto-routing",
      "accuracy": 0.804714,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.024081237373737373,
      "avg_duration_per_task_ms": 24354.015151515152,
      "avg_output_tokens_per_task": 1625.5555555555557,
      "last_run_timestamp": "2026-08-01T08:38:11.900Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Alibaba",
      "accuracy": 0.8045734999999999,
      "accuracy_std_dev": 0.09801590415888298,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.09265821506734007,
      "avg_duration_per_task_ms": 514476.9250841751,
      "avg_output_tokens_per_task": 34180.35227272727,
      "last_run_timestamp": "2026-10-24T16:04:33.487Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "DekaLLM",
      "accuracy": 0.8043401184631804,
      "accuracy_std_dev": 0.024791902944308233,
      "run_count": 5,
      "total_questions": 937,
      "avg_cost_per_task_usd": 0.03799079615795091,
      "avg_duration_per_task_ms": 223986.25720384205,
      "avg_output_tokens_per_task": 37938.2166488794,
      "last_run_timestamp": "2026-10-24T16:04:55.218Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "Friendli",
      "accuracy": 0.802714625,
      "accuracy_std_dev": 0.02322035202111231,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.028997014520202022,
      "avg_duration_per_task_ms": 309649.94507575757,
      "avg_output_tokens_per_task": 19185.933712121212,
      "last_run_timestamp": "2026-10-17T16:09:25.569Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.1-flash-lite-20260507",
      "model_slug": "google/gemini-3.1-flash-lite",
      "model_name": "Google: Gemini 3.1 Flash Lite",
      "provider_name": "Google",
      "accuracy": 0.8020951111111111,
      "accuracy_std_dev": 0.027492606765554845,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.04329651571268237,
      "avg_duration_per_task_ms": 84651.98765432098,
      "avg_output_tokens_per_task": 26957.542087542086,
      "last_run_timestamp": "2026-10-01T19:25:58.919Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Makora",
      "accuracy": 0.8011755539419088,
      "accuracy_std_dev": 0.08103010676902754,
      "run_count": 8,
      "total_questions": 1446,
      "avg_cost_per_task_usd": 0.018679138312586448,
      "avg_duration_per_task_ms": 300257.92392807745,
      "avg_output_tokens_per_task": 39516.01521438451,
      "last_run_timestamp": "2026-10-24T08:32:39.573Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5-mini-2025-08-07",
      "model_slug": "openai/gpt-5-mini",
      "model_name": "OpenAI: GPT-5 Mini",
      "provider_name": "OpenAI",
      "accuracy": 0.8007853333333332,
      "accuracy_std_dev": 0.008398773614178595,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0439676531986532,
      "avg_duration_per_task_ms": 201662.1632996633,
      "avg_output_tokens_per_task": 21888.12962962963,
      "last_run_timestamp": "2026-10-01T19:26:25.385Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "auto-routing",
      "accuracy": 0.8005984444444443,
      "accuracy_std_dev": 0.04083905003278416,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.026883831649831647,
      "avg_duration_per_task_ms": 331939.1515151515,
      "avg_output_tokens_per_task": 25611.442199775534,
      "last_run_timestamp": "2026-10-24T16:03:25.477Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5-mini-2025-08-07",
      "model_slug": "openai/gpt-5-mini",
      "model_name": "OpenAI: GPT-5 Mini",
      "provider_name": "Azure",
      "accuracy": 0.7996631666666666,
      "accuracy_std_dev": 0.024569953550469384,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.045980551346801346,
      "avg_duration_per_task_ms": 194150.27525252526,
      "avg_output_tokens_per_task": 21808.128787878788,
      "last_run_timestamp": "2026-10-01T19:34:33.810Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7995168260869566,
      "accuracy_std_dev": 0.015725089308490422,
      "run_count": 5,
      "total_questions": 966,
      "avg_cost_per_task_usd": 0.1727357877846791,
      "avg_duration_per_task_ms": 863549.4824016563,
      "avg_output_tokens_per_task": 55062.07867494824,
      "last_run_timestamp": "2026-08-29T19:16:33.097Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "DigitalOcean",
      "accuracy": 0.798727888888889,
      "accuracy_std_dev": 0.10517670417967454,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.05109221941638608,
      "avg_duration_per_task_ms": 1276720.367003367,
      "avg_output_tokens_per_task": 27966.93265993266,
      "last_run_timestamp": "2026-10-24T16:17:39.036Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "AtlasCloud",
      "accuracy": 0.798541,
      "accuracy_std_dev": 0.005382643216859184,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.013865464646464647,
      "avg_duration_per_task_ms": 310914.7727272727,
      "avg_output_tokens_per_task": 14252.202861952863,
      "last_run_timestamp": "2026-10-04T00:00:36.942Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.2-chat-20251211",
      "model_slug": "openai/gpt-5.2-chat",
      "model_name": "OpenAI: GPT-5.2 Chat",
      "provider_name": "auto-routing",
      "accuracy": 0.79798,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.03602157070707071,
      "avg_duration_per_task_ms": 32858.843434343435,
      "avg_output_tokens_per_task": 2479.818181818182,
      "last_run_timestamp": "2026-08-01T08:32:36.352Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Phala",
      "accuracy": 0.7979798333333333,
      "accuracy_std_dev": 0.04819829830819293,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.1916180345117845,
      "avg_duration_per_task_ms": 1503115.5151515151,
      "avg_output_tokens_per_task": 46476.21548821549,
      "last_run_timestamp": "2026-10-03T17:00:45.414Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Io Net",
      "accuracy": 0.7976190000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 98,
      "avg_cost_per_task_usd": 0.012265704081632654,
      "avg_duration_per_task_ms": 285854.5612244898,
      "avg_output_tokens_per_task": 10545.989795918367,
      "last_run_timestamp": "2026-10-03T17:16:45.102Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "GMICloud",
      "accuracy": 0.7975339581632652,
      "accuracy_std_dev": 0.14314198310659945,
      "run_count": 10,
      "total_questions": 1960,
      "avg_cost_per_task_usd": 0.01657647346938776,
      "avg_duration_per_task_ms": 547902.756122449,
      "avg_output_tokens_per_task": 26479.933673469386,
      "last_run_timestamp": "2026-10-17T08:10:43.825Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.1-terminus",
      "model_slug": "deepseek/deepseek-v3.1-terminus",
      "model_name": "DeepSeek: DeepSeek V3.1 Terminus",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7969696000000001,
      "accuracy_std_dev": 0.010464901654578508,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.013516212121212122,
      "avg_duration_per_task_ms": 270529.88181818184,
      "avg_output_tokens_per_task": 14086.99090909091,
      "last_run_timestamp": "2026-08-29T19:12:57.906Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2-thinking-20251106",
      "model_slug": "moonshotai/kimi-k2-thinking",
      "model_name": "MoonshotAI: Kimi K2 Thinking",
      "provider_name": "Google",
      "accuracy": 0.796581,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 195,
      "avg_cost_per_task_usd": 0.06800715384615384,
      "avg_duration_per_task_ms": 384890.6564102564,
      "avg_output_tokens_per_task": 29421.871794871793,
      "last_run_timestamp": "2026-08-01T17:52:55.742Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-r1-0528",
      "model_slug": "deepseek/deepseek-r1-0528",
      "model_name": "DeepSeek: R1 0528",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7964659214659687,
      "accuracy_std_dev": 0.005460143336030668,
      "run_count": 8,
      "total_questions": 1528,
      "avg_cost_per_task_usd": 0.050215497382198944,
      "avg_duration_per_task_ms": 1089140.4594240838,
      "avg_output_tokens_per_task": 22865.37238219895,
      "last_run_timestamp": "2026-10-17T16:06:10.469Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "Novita",
      "accuracy": 0.7956228,
      "accuracy_std_dev": 0.021522407276138952,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.006844225252525253,
      "avg_duration_per_task_ms": 363476.91515151516,
      "avg_output_tokens_per_task": 16652.98888888889,
      "last_run_timestamp": "2026-08-29T19:20:50.265Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Io Net",
      "accuracy": 0.7950336666666666,
      "accuracy_std_dev": 0.03307482054205114,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.03994933501683502,
      "avg_duration_per_task_ms": 251252.56818181818,
      "avg_output_tokens_per_task": 12806.098484848484,
      "last_run_timestamp": "2026-09-19T09:05:41.015Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-thinking-2507",
      "model_slug": "qwen/qwen3-235b-a22b-thinking-2507",
      "model_name": "Qwen: Qwen3 235B A22B Thinking 2507",
      "provider_name": "Alibaba",
      "accuracy": 0.7950335,
      "accuracy_std_dev": 0.010057183614213304,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.053936401515151514,
      "avg_duration_per_task_ms": 418610.5353535354,
      "avg_output_tokens_per_task": 23400.56818181818,
      "last_run_timestamp": "2026-08-22T17:18:01.115Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7950015744680852,
      "accuracy_std_dev": 0.020697025299303277,
      "run_count": 5,
      "total_questions": 987,
      "avg_cost_per_task_usd": 0.028578175278622088,
      "avg_duration_per_task_ms": 466139.9655521783,
      "avg_output_tokens_per_task": 23874.156028368794,
      "last_run_timestamp": "2026-08-29T18:09:57.600Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "auto-routing",
      "accuracy": 0.7944022499999999,
      "accuracy_std_dev": 0.03928997983185408,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.04325068055555555,
      "avg_duration_per_task_ms": 495166.12626262626,
      "avg_output_tokens_per_task": 42549.15993265993,
      "last_run_timestamp": "2026-10-24T16:02:47.095Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.1-terminus",
      "model_slug": "deepseek/deepseek-v3.1-terminus",
      "model_name": "DeepSeek: DeepSeek V3.1 Terminus",
      "provider_name": "StreamLake",
      "accuracy": 0.7942762,
      "accuracy_std_dev": 0.008274731256059032,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.014654871717171718,
      "avg_duration_per_task_ms": 281996.1565656566,
      "avg_output_tokens_per_task": 14013.025252525253,
      "last_run_timestamp": "2026-08-29T16:10:33.070Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7940515555555555,
      "accuracy_std_dev": 0.013561145494972105,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.017693217171717172,
      "avg_duration_per_task_ms": 59941.32828282828,
      "avg_output_tokens_per_task": 5742.835016835017,
      "last_run_timestamp": "2026-10-24T16:02:28.570Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Venice",
      "accuracy": 0.7929315464581417,
      "accuracy_std_dev": 0.02021084770060026,
      "run_count": 11,
      "total_questions": 2174,
      "avg_cost_per_task_usd": 0.06395397148114075,
      "avg_duration_per_task_ms": 189643.35372585096,
      "avg_output_tokens_per_task": 20031.46780128795,
      "last_run_timestamp": "2026-10-24T08:53:01.970Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "Io Net",
      "accuracy": 0.7927813197452228,
      "accuracy_std_dev": 0.008711832686065535,
      "run_count": 4,
      "total_questions": 785,
      "avg_cost_per_task_usd": 0.011660411464968153,
      "avg_duration_per_task_ms": 735190.7324840764,
      "avg_output_tokens_per_task": 39161.06751592357,
      "last_run_timestamp": "2026-10-24T16:38:44.156Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "Venice",
      "accuracy": 0.79250825,
      "accuracy_std_dev": 0.011190804686028344,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.004598987794612795,
      "avg_duration_per_task_ms": 248268.39141414143,
      "avg_output_tokens_per_task": 30111.32239057239,
      "last_run_timestamp": "2026-10-17T16:03:34.733Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta/muse-glimmer-30b-20260810",
      "model_slug": "meta/muse-glimmer-30b",
      "model_name": "Meta: Muse Glimmer 30B",
      "provider_name": "DeepInfra",
      "accuracy": 0.7920900579268293,
      "accuracy_std_dev": 0.011512971285033247,
      "run_count": 10,
      "total_questions": 1968,
      "avg_cost_per_task_usd": 0.01824498475609756,
      "avg_duration_per_task_ms": 181543.2469512195,
      "avg_output_tokens_per_task": 15051.83943089431,
      "last_run_timestamp": "2026-10-08T22:29:11.504Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "GMICloud",
      "accuracy": 0.79194725,
      "accuracy_std_dev": 0.03294298282417921,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.10781298148148147,
      "avg_duration_per_task_ms": 400979.42887205386,
      "avg_output_tokens_per_task": 29714.260101010103,
      "last_run_timestamp": "2026-10-24T16:36:23.874Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7919261222410866,
      "accuracy_std_dev": 0.024709971729827782,
      "run_count": 9,
      "total_questions": 1767,
      "avg_cost_per_task_usd": 0.29162476117713637,
      "avg_duration_per_task_ms": 950106.2246745897,
      "avg_output_tokens_per_task": 73483.63723825694,
      "last_run_timestamp": "2026-10-24T16:02:08.814Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "DeepInfra",
      "accuracy": 0.7914427791218638,
      "accuracy_std_dev": 0.014961655425666921,
      "run_count": 12,
      "total_questions": 2232,
      "avg_cost_per_task_usd": 0.004660359318996416,
      "avg_duration_per_task_ms": 1335489.0049283155,
      "avg_output_tokens_per_task": 30608.047491039426,
      "last_run_timestamp": "2026-10-24T16:07:10.479Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-r1-0528",
      "model_slug": "deepseek/deepseek-r1-0528",
      "model_name": "DeepSeek: R1 0528",
      "provider_name": "DeepInfra",
      "accuracy": 0.7902960850403475,
      "accuracy_std_dev": 0.020316089842022662,
      "run_count": 9,
      "total_questions": 1611,
      "avg_cost_per_task_usd": 0.040685735567970206,
      "avg_duration_per_task_ms": 737944.181874612,
      "avg_output_tokens_per_task": 18801.04593420236,
      "last_run_timestamp": "2026-10-17T16:06:15.029Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Phala",
      "accuracy": 0.7893355839160839,
      "accuracy_std_dev": 0.10239971274711347,
      "run_count": 9,
      "total_questions": 1716,
      "avg_cost_per_task_usd": 0.07681546794871794,
      "avg_duration_per_task_ms": 733179.0168997669,
      "avg_output_tokens_per_task": 27895.818764568765,
      "last_run_timestamp": "2026-10-10T16:29:40.552Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "Together",
      "accuracy": 0.7891638401060072,
      "accuracy_std_dev": 0.014821929173009913,
      "run_count": 12,
      "total_questions": 2264,
      "avg_cost_per_task_usd": 0.008643914752650177,
      "avg_duration_per_task_ms": 377329.0070671378,
      "avg_output_tokens_per_task": 34148.09540636042,
      "last_run_timestamp": "2026-10-24T16:13:05.629Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "Groq",
      "accuracy": 0.7887112722044729,
      "accuracy_std_dev": 0.11519716273394331,
      "run_count": 9,
      "total_questions": 1565,
      "avg_cost_per_task_usd": 0.03999107923322684,
      "avg_duration_per_task_ms": 140844.86581469647,
      "avg_output_tokens_per_task": 29636.458785942494,
      "last_run_timestamp": "2026-10-24T16:27:33.561Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "Mara",
      "accuracy": 0.7885339041353384,
      "accuracy_std_dev": 0.023230108247703007,
      "run_count": 5,
      "total_questions": 532,
      "avg_cost_per_task_usd": 0.034030992481203004,
      "avg_duration_per_task_ms": 116832.8590225564,
      "avg_output_tokens_per_task": 35347.6954887218,
      "last_run_timestamp": "2026-08-29T19:44:46.543Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7878787777777778,
      "accuracy_std_dev": 0.04255992961506759,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.17480242087542086,
      "avg_duration_per_task_ms": 541272.2081930415,
      "avg_output_tokens_per_task": 54419.242424242424,
      "last_run_timestamp": "2026-10-24T16:04:55.367Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "Venice",
      "accuracy": 0.7854672791234141,
      "accuracy_std_dev": 0.12208478471357521,
      "run_count": 5,
      "total_questions": 867,
      "avg_cost_per_task_usd": 0.023352209919261825,
      "avg_duration_per_task_ms": 590708.8500576701,
      "avg_output_tokens_per_task": 19635.643598615916,
      "last_run_timestamp": "2026-10-17T08:18:07.707Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-r1-0528",
      "model_slug": "deepseek/deepseek-r1-0528",
      "model_name": "DeepSeek: R1 0528",
      "provider_name": "StreamLake",
      "accuracy": 0.7851099187817259,
      "accuracy_std_dev": 0.024182763553029427,
      "run_count": 9,
      "total_questions": 1773,
      "avg_cost_per_task_usd": 0.05363871517202482,
      "avg_duration_per_task_ms": 470093.5600676819,
      "avg_output_tokens_per_task": 23310.749576988157,
      "last_run_timestamp": "2026-10-17T16:03:34.700Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Nebius",
      "accuracy": 0.7849183946015424,
      "accuracy_std_dev": 0.0724374841967196,
      "run_count": 8,
      "total_questions": 1556,
      "avg_cost_per_task_usd": 0.1867961233933162,
      "avg_duration_per_task_ms": 1640932.23907455,
      "avg_output_tokens_per_task": 43985.5205655527,
      "last_run_timestamp": "2026-10-24T16:01:08.930Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "Parasail",
      "accuracy": 0.7842683649508387,
      "accuracy_std_dev": 0.12512206342943477,
      "run_count": 9,
      "total_questions": 1729,
      "avg_cost_per_task_usd": 0.024085618276460385,
      "avg_duration_per_task_ms": 354469.05031810293,
      "avg_output_tokens_per_task": 24486.457489878543,
      "last_run_timestamp": "2026-10-24T16:05:36.229Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.1-terminus",
      "model_slug": "deepseek/deepseek-v3.1-terminus",
      "model_name": "DeepSeek: DeepSeek V3.1 Terminus",
      "provider_name": "DeepInfra",
      "accuracy": 0.7839820659898478,
      "accuracy_std_dev": 0.002844524682028023,
      "run_count": 3,
      "total_questions": 591,
      "avg_cost_per_task_usd": 0.013566983079526225,
      "avg_duration_per_task_ms": 498549.5752961083,
      "avg_output_tokens_per_task": 14135.162436548224,
      "last_run_timestamp": "2026-08-15T14:56:55.269Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "Parasail",
      "accuracy": 0.7838284348819498,
      "accuracy_std_dev": 0.08896072103931436,
      "run_count": 11,
      "total_questions": 1313,
      "avg_cost_per_task_usd": 0.006210692307692306,
      "avg_duration_per_task_ms": 295622.3008377761,
      "avg_output_tokens_per_task": 24966.161462300075,
      "last_run_timestamp": "2026-10-24T16:40:33.762Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "CoreWeave",
      "accuracy": 0.7836699999999999,
      "accuracy_std_dev": 0.010838720902394346,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.021049996969696963,
      "avg_duration_per_task_ms": 263607.26313131314,
      "avg_output_tokens_per_task": 12518.16212121212,
      "last_run_timestamp": "2026-10-24T16:18:15.010Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Wafer",
      "accuracy": 0.7832240326797386,
      "accuracy_std_dev": 0.07801749999999996,
      "run_count": 2,
      "total_questions": 153,
      "avg_cost_per_task_usd": 0.07358333333333333,
      "avg_duration_per_task_ms": 500220.6274509804,
      "avg_output_tokens_per_task": 27733.385620915033,
      "last_run_timestamp": "2026-08-15T15:24:07.556Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Mistral",
      "accuracy": 0.7819865,
      "accuracy_std_dev": 0.010942500000000022,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.028098762626262628,
      "avg_duration_per_task_ms": 185634.4823232323,
      "avg_output_tokens_per_task": 55958.419191919194,
      "last_run_timestamp": "2026-10-24T08:34:00.019Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "Cloudflare",
      "accuracy": 0.7816675629587804,
      "accuracy_std_dev": 0.02991313237433426,
      "run_count": 9,
      "total_questions": 1771,
      "avg_cost_per_task_usd": 0.011191632411067194,
      "avg_duration_per_task_ms": 582007.9429700734,
      "avg_output_tokens_per_task": 37062.36589497459,
      "last_run_timestamp": "2026-10-24T16:18:35.530Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "Venice",
      "accuracy": 0.7808153280721533,
      "accuracy_std_dev": 0.05722357962705003,
      "run_count": 9,
      "total_questions": 1774,
      "avg_cost_per_task_usd": 0.007362833145434047,
      "avg_duration_per_task_ms": 871339.3737316799,
      "avg_output_tokens_per_task": 18677.131341600903,
      "last_run_timestamp": "2026-10-24T16:21:54.634Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Nebius",
      "accuracy": 0.7805836666666667,
      "accuracy_std_dev": 0.009152126540986119,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.03482366666666667,
      "avg_duration_per_task_ms": 531300.7171717172,
      "avg_output_tokens_per_task": 69410.62962962964,
      "last_run_timestamp": "2026-10-24T08:41:52.312Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "SambaNova",
      "accuracy": 0.7805031924528302,
      "accuracy_std_dev": 0.020942354059550135,
      "run_count": 3,
      "total_questions": 265,
      "avg_cost_per_task_usd": 0.09049491698113207,
      "avg_duration_per_task_ms": 131957.70943396227,
      "avg_output_tokens_per_task": 37606.645283018865,
      "last_run_timestamp": "2026-08-15T14:54:40.470Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "auto-routing",
      "accuracy": 0.780302875,
      "accuracy_std_dev": 0.044334001808537145,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.012861900883838383,
      "avg_duration_per_task_ms": 776646.4172979798,
      "avg_output_tokens_per_task": 38634.71654040404,
      "last_run_timestamp": "2026-10-17T08:18:06.210Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Phala",
      "accuracy": 0.7786195454545455,
      "accuracy_std_dev": 0.04555204390657037,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.21966554178145087,
      "avg_duration_per_task_ms": 867448.9150596878,
      "avg_output_tokens_per_task": 73058.02662993572,
      "last_run_timestamp": "2026-10-17T16:06:46.266Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Venice",
      "accuracy": 0.7785086368421053,
      "accuracy_std_dev": 0.010062970594585871,
      "run_count": 4,
      "total_questions": 760,
      "avg_cost_per_task_usd": 0.002249036842105263,
      "avg_duration_per_task_ms": 780019.0210526315,
      "avg_output_tokens_per_task": 14812.598684210527,
      "last_run_timestamp": "2026-10-24T16:06:32.425Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "Google",
      "accuracy": 0.7784476091370558,
      "accuracy_std_dev": 0.018688411382040342,
      "run_count": 8,
      "total_questions": 1576,
      "avg_cost_per_task_usd": 0.022516517131979694,
      "avg_duration_per_task_ms": 871719.5628172589,
      "avg_output_tokens_per_task": 38655.122461928935,
      "last_run_timestamp": "2026-10-24T16:37:21.981Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "Darkbloom",
      "accuracy": 0.7780583333333333,
      "accuracy_std_dev": 0.018220714966457528,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.021326852693602693,
      "avg_duration_per_task_ms": 472582.55723905726,
      "avg_output_tokens_per_task": 28353.299663299662,
      "last_run_timestamp": "2026-10-24T16:05:13.522Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "auto-routing",
      "accuracy": 0.7779681626712327,
      "accuracy_std_dev": 0.05316331589603613,
      "run_count": 9,
      "total_questions": 1752,
      "avg_cost_per_task_usd": 0.15604383333333335,
      "avg_duration_per_task_ms": 982840.4474885844,
      "avg_output_tokens_per_task": 64456.35901826484,
      "last_run_timestamp": "2026-10-24T16:03:59.070Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "DeepInfra",
      "accuracy": 0.7774583256704981,
      "accuracy_std_dev": 0.04658687436660674,
      "run_count": 4,
      "total_questions": 522,
      "avg_cost_per_task_usd": 0.1387438639846743,
      "avg_duration_per_task_ms": 1171411.5785440614,
      "avg_output_tokens_per_task": 40017.36590038314,
      "last_run_timestamp": "2026-08-22T17:30:52.646Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.4-nano-20260317",
      "model_slug": "openai/gpt-5.4-nano",
      "model_name": "OpenAI: GPT-5.4 Nano",
      "provider_name": "auto-routing",
      "accuracy": 0.7772166666666667,
      "accuracy_std_dev": 0.003459479954495376,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.013080006734006736,
      "avg_duration_per_task_ms": 67697.39225589225,
      "avg_output_tokens_per_task": 10341.513468013469,
      "last_run_timestamp": "2026-10-01T19:12:45.924Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "DekaLLM",
      "accuracy": 0.7769359999999998,
      "accuracy_std_dev": 0.005389920933248147,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.00513745202020202,
      "avg_duration_per_task_ms": 836559.9427609427,
      "avg_output_tokens_per_task": 28379.09006734007,
      "last_run_timestamp": "2026-10-24T16:31:50.394Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "DeepInfra",
      "accuracy": 0.7768158778004073,
      "accuracy_std_dev": 0.02614019988234382,
      "run_count": 6,
      "total_questions": 982,
      "avg_cost_per_task_usd": 0.010268270875763747,
      "avg_duration_per_task_ms": 1330121.382892057,
      "avg_output_tokens_per_task": 39680.64867617108,
      "last_run_timestamp": "2026-10-24T08:12:02.532Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "auto-routing",
      "accuracy": 0.7767955833333333,
      "accuracy_std_dev": 0.0196863031719109,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.005944730218855219,
      "avg_duration_per_task_ms": 663975.4595959596,
      "avg_output_tokens_per_task": 36414.121632996634,
      "last_run_timestamp": "2026-10-24T16:20:08.201Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.4-nano-20260317",
      "model_slug": "openai/gpt-5.4-nano",
      "model_name": "OpenAI: GPT-5.4 Nano",
      "provider_name": "OpenAI",
      "accuracy": 0.77567325,
      "accuracy_std_dev": 0.010934638822910419,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.013401992424242422,
      "avg_duration_per_task_ms": 67621.48106060606,
      "avg_output_tokens_per_task": 10598.125,
      "last_run_timestamp": "2026-10-02T23:20:38.343Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-thinking-2507",
      "model_slug": "qwen/qwen3-235b-a22b-thinking-2507",
      "model_name": "Qwen: Qwen3 235B A22B Thinking 2507",
      "provider_name": "auto-routing",
      "accuracy": 0.7752524999999999,
      "accuracy_std_dev": 0.021901537394667075,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.05750532828282828,
      "avg_duration_per_task_ms": 632332.7689393939,
      "avg_output_tokens_per_task": 23028.97095959596,
      "last_run_timestamp": "2026-08-22T17:09:41.940Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "DeepInfra",
      "accuracy": 0.7747842799160252,
      "accuracy_std_dev": 0.014649286504468402,
      "run_count": 10,
      "total_questions": 1429,
      "avg_cost_per_task_usd": 0.04802459482155353,
      "avg_duration_per_task_ms": 974559.5045486355,
      "avg_output_tokens_per_task": 24459.735479356194,
      "last_run_timestamp": "2026-10-24T16:01:45.820Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7741301666666666,
      "accuracy_std_dev": 0.007235617536802828,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.019828150673400675,
      "avg_duration_per_task_ms": 721854.4562289562,
      "avg_output_tokens_per_task": 49359.81902356903,
      "last_run_timestamp": "2026-10-10T16:25:17.654Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "inclusionai/ling-3.0-flash-20260723",
      "model_slug": "inclusionai/ling-3.0-flash",
      "model_name": "inclusionAI: Ling 3.0 Flash",
      "provider_name": "DeepInfra",
      "accuracy": 0.7741139703903096,
      "accuracy_std_dev": 0.09718478875928833,
      "run_count": 4,
      "total_questions": 743,
      "avg_cost_per_task_usd": 0.004475604306864065,
      "avg_duration_per_task_ms": 312870.95827725437,
      "avg_output_tokens_per_task": 22926.07267833109,
      "last_run_timestamp": "2026-10-08T22:27:01.555Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-r1-0528",
      "model_slug": "deepseek/deepseek-r1-0528",
      "model_name": "DeepSeek: R1 0528",
      "provider_name": "auto-routing",
      "accuracy": 0.7718855,
      "accuracy_std_dev": 0.03473679068005565,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.05189419633838384,
      "avg_duration_per_task_ms": 871715.6748737374,
      "avg_output_tokens_per_task": 23212.847222222223,
      "last_run_timestamp": "2026-10-17T16:03:24.588Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Mara",
      "accuracy": 0.7715289560632689,
      "accuracy_std_dev": 0.034441084737222544,
      "run_count": 9,
      "total_questions": 1138,
      "avg_cost_per_task_usd": 0.012996073813708259,
      "avg_duration_per_task_ms": 52972.56414762742,
      "avg_output_tokens_per_task": 17226.330404217926,
      "last_run_timestamp": "2026-10-24T16:14:06.770Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4-sonnet-20250522",
      "model_slug": "anthropic/claude-sonnet-4",
      "model_name": "Anthropic: Claude Sonnet 4",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7714368913043479,
      "accuracy_std_dev": 0.014261246953428234,
      "run_count": 3,
      "total_questions": 552,
      "avg_cost_per_task_usd": 0.23206160869565215,
      "avg_duration_per_task_ms": 164099.60869565216,
      "avg_output_tokens_per_task": 15307.03079710145,
      "last_run_timestamp": "2026-09-01T13:33:46.786Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "BaseTen",
      "accuracy": 0.7714323076923078,
      "accuracy_std_dev": 0.009526863648769632,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.014571612276612276,
      "avg_duration_per_task_ms": 136246.84032634032,
      "avg_output_tokens_per_task": 29062.332167832166,
      "last_run_timestamp": "2026-10-24T16:26:32.305Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "DeepInfra",
      "accuracy": 0.7707066962474647,
      "accuracy_std_dev": 0.0069177655641109856,
      "run_count": 10,
      "total_questions": 1972,
      "avg_cost_per_task_usd": 0.011308173427991888,
      "avg_duration_per_task_ms": 972932.3559837728,
      "avg_output_tokens_per_task": 11764.48985801217,
      "last_run_timestamp": "2026-10-24T16:05:13.521Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.4-nano-20260317",
      "model_slug": "openai/gpt-5.4-nano",
      "model_name": "OpenAI: GPT-5.4 Nano",
      "provider_name": "Azure",
      "accuracy": 0.770202,
      "accuracy_std_dev": 0.008625232557251217,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.013833271043771043,
      "avg_duration_per_task_ms": 119359.86616161616,
      "avg_output_tokens_per_task": 10422.487373737375,
      "last_run_timestamp": "2026-10-01T19:20:04.986Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "CoreWeave",
      "accuracy": 0.769679306122449,
      "accuracy_std_dev": 0.017686462624278497,
      "run_count": 5,
      "total_questions": 686,
      "avg_cost_per_task_usd": 0.009246491253644315,
      "avg_duration_per_task_ms": 700873.1428571428,
      "avg_output_tokens_per_task": 30693.99271137026,
      "last_run_timestamp": "2026-10-24T16:14:24.978Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Sail Research",
      "accuracy": 0.7691307272727272,
      "accuracy_std_dev": 0.12910889092269848,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.016461994031221304,
      "avg_duration_per_task_ms": 677663.191919192,
      "avg_output_tokens_per_task": 35541.38842975206,
      "last_run_timestamp": "2026-10-24T08:57:59.951Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Parasail",
      "accuracy": 0.768994582345191,
      "accuracy_std_dev": 0.016269118990725185,
      "run_count": 8,
      "total_questions": 1518,
      "avg_cost_per_task_usd": 0.021888522397891963,
      "avg_duration_per_task_ms": 236418.71277997366,
      "avg_output_tokens_per_task": 29098.49209486166,
      "last_run_timestamp": "2026-10-24T16:34:51.218Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "Phala",
      "accuracy": 0.7689713726812815,
      "accuracy_std_dev": 0.07621573205272868,
      "run_count": 6,
      "total_questions": 1186,
      "avg_cost_per_task_usd": 0.018058317875210794,
      "avg_duration_per_task_ms": 1377820.0623946036,
      "avg_output_tokens_per_task": 17446.093591905566,
      "last_run_timestamp": "2026-10-24T16:01:29.840Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Xiaomi",
      "accuracy": 0.7685185,
      "accuracy_std_dev": 0.018461144134370444,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.01486243308080808,
      "avg_duration_per_task_ms": 1091396.6527777778,
      "avg_output_tokens_per_task": 52782.86363636364,
      "last_run_timestamp": "2026-10-24T08:54:50.797Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Morph",
      "accuracy": 0.7677873143148243,
      "accuracy_std_dev": 0.10954529486385985,
      "run_count": 10,
      "total_questions": 1963,
      "avg_cost_per_task_usd": 0.15105290779419256,
      "avg_duration_per_task_ms": 869942.4396332145,
      "avg_output_tokens_per_task": 39571.91288843607,
      "last_run_timestamp": "2026-10-24T16:19:11.498Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.1-terminus",
      "model_slug": "deepseek/deepseek-v3.1-terminus",
      "model_name": "DeepSeek: DeepSeek V3.1 Terminus",
      "provider_name": "auto-routing",
      "accuracy": 0.7676768,
      "accuracy_std_dev": 0.018594973916625974,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.015932259595959595,
      "avg_duration_per_task_ms": 441129.9383838384,
      "avg_output_tokens_per_task": 16024.338383838383,
      "last_run_timestamp": "2026-08-29T18:21:38.666Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "AkashML",
      "accuracy": 0.7676733942992873,
      "accuracy_std_dev": 0.016408032276093717,
      "run_count": 15,
      "total_questions": 2947,
      "avg_cost_per_task_usd": 0.00717780895826264,
      "avg_duration_per_task_ms": 441831.9945707499,
      "avg_output_tokens_per_task": 27893.447234475738,
      "last_run_timestamp": "2026-10-24T16:03:06.208Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4-sonnet-20250522",
      "model_slug": "anthropic/claude-sonnet-4",
      "model_name": "Anthropic: Claude Sonnet 4",
      "provider_name": "Google",
      "accuracy": 0.7672559999999999,
      "accuracy_std_dev": 0.008469894233892938,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.27037940151515155,
      "avg_duration_per_task_ms": 263755.6675084175,
      "avg_output_tokens_per_task": 17837.74579124579,
      "last_run_timestamp": "2026-09-01T13:23:02.059Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "DekaLLM",
      "accuracy": 0.7671445564924115,
      "accuracy_std_dev": 0.01261567820104104,
      "run_count": 6,
      "total_questions": 1186,
      "avg_cost_per_task_usd": 0.019929450252951095,
      "avg_duration_per_task_ms": 1212404.510961214,
      "avg_output_tokens_per_task": 60253.559021922425,
      "last_run_timestamp": "2026-10-24T16:08:41.765Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4-sonnet-20250522",
      "model_slug": "anthropic/claude-sonnet-4",
      "model_name": "Anthropic: Claude Sonnet 4",
      "provider_name": "auto-routing",
      "accuracy": 0.7668349999999999,
      "accuracy_std_dev": 0.01935999999999999,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.2670539015151515,
      "avg_duration_per_task_ms": 177095.6616161616,
      "avg_output_tokens_per_task": 17616.689393939392,
      "last_run_timestamp": "2026-09-01T13:19:59.945Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "SambaNova",
      "accuracy": 0.7666666510144927,
      "accuracy_std_dev": 0.008075638536845108,
      "run_count": 9,
      "total_questions": 1725,
      "avg_cost_per_task_usd": 0.02488580347826087,
      "avg_duration_per_task_ms": 72284.5831884058,
      "avg_output_tokens_per_task": 26060.85043478261,
      "last_run_timestamp": "2026-10-24T16:10:43.382Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Crusoe",
      "accuracy": 0.76638602247191,
      "accuracy_std_dev": 0.0119351260850762,
      "run_count": 3,
      "total_questions": 356,
      "avg_cost_per_task_usd": 0.002961258426966292,
      "avg_duration_per_task_ms": 69864.89325842696,
      "avg_output_tokens_per_task": 3595.182584269663,
      "last_run_timestamp": "2026-08-15T14:54:48.654Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "GMICloud",
      "accuracy": 0.76599325,
      "accuracy_std_dev": 0.026856781949584005,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.012383484848484849,
      "avg_duration_per_task_ms": 910556.1414141414,
      "avg_output_tokens_per_task": 45554.207070707074,
      "last_run_timestamp": "2026-10-24T08:15:51.063Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Cerebras",
      "accuracy": 0.7659344612403101,
      "accuracy_std_dev": 0.01074604027072297,
      "run_count": 4,
      "total_questions": 774,
      "avg_cost_per_task_usd": 0.004193177002583979,
      "avg_duration_per_task_ms": 4103.266149870801,
      "avg_output_tokens_per_task": 2323.702842377261,
      "last_run_timestamp": "2026-08-29T18:58:30.919Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Reka",
      "accuracy": 0.7651274895635675,
      "accuracy_std_dev": 0.012168813200753945,
      "run_count": 16,
      "total_questions": 3162,
      "avg_cost_per_task_usd": 0.01713221790006325,
      "avg_duration_per_task_ms": 104832.72896900696,
      "avg_output_tokens_per_task": 5987.342504743833,
      "last_run_timestamp": "2026-10-24T09:05:10.210Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7646065075987841,
      "accuracy_std_dev": 0.016409967770839785,
      "run_count": 5,
      "total_questions": 987,
      "avg_cost_per_task_usd": 0.056628046605876385,
      "avg_duration_per_task_ms": 583978.3080040527,
      "avg_output_tokens_per_task": 26120.428571428572,
      "last_run_timestamp": "2026-08-29T18:26:43.144Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2-0905",
      "model_slug": "moonshotai/kimi-k2-0905",
      "model_name": "MoonshotAI: Kimi K2 0905",
      "provider_name": "auto-routing",
      "accuracy": 0.76431,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.017299227272727275,
      "avg_duration_per_task_ms": 73383.17171717172,
      "avg_output_tokens_per_task": 5944.843434343435,
      "last_run_timestamp": "2026-08-01T17:45:27.258Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "Parasail",
      "accuracy": 0.7632574999999999,
      "accuracy_std_dev": 0.011447194448859535,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.020826034090909093,
      "avg_duration_per_task_ms": 1150567.0700757576,
      "avg_output_tokens_per_task": 51976.968434343435,
      "last_run_timestamp": "2026-10-24T16:20:24.814Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "DeepInfra",
      "accuracy": 0.7631440720576461,
      "accuracy_std_dev": 0.033247584998763444,
      "run_count": 8,
      "total_questions": 1249,
      "avg_cost_per_task_usd": 0.06387943394715773,
      "avg_duration_per_task_ms": 833616.5316253003,
      "avg_output_tokens_per_task": 53539.40272217774,
      "last_run_timestamp": "2026-08-29T19:07:57.500Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Relace",
      "accuracy": 0.7628913204419889,
      "accuracy_std_dev": 0.046742499999999965,
      "run_count": 2,
      "total_questions": 362,
      "avg_cost_per_task_usd": 0.015507096685082874,
      "avg_duration_per_task_ms": 476532.59392265195,
      "avg_output_tokens_per_task": 53949.68508287293,
      "last_run_timestamp": "2026-10-24T08:33:50.635Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Crusoe",
      "accuracy": 0.7626265,
      "accuracy_std_dev": 0.007714766182458135,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.0075155000000000005,
      "avg_duration_per_task_ms": 129316.82575757576,
      "avg_output_tokens_per_task": 29867.227272727272,
      "last_run_timestamp": "2026-10-24T16:12:49.068Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-coder-next-2025-02-03",
      "model_slug": "qwen/qwen3-coder-next",
      "model_name": "Qwen: Qwen3 Coder Next",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.7626260000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.022027883838383836,
      "avg_duration_per_task_ms": 90904.42929292929,
      "avg_output_tokens_per_task": 14522.767676767677,
      "last_run_timestamp": "2026-10-04T03:02:47.748Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "InferenceNet",
      "accuracy": 0.7626259999999999,
      "accuracy_std_dev": 0.01346799999999998,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.01717197474747475,
      "avg_duration_per_task_ms": 622307.4974747475,
      "avg_output_tokens_per_task": 61038.34090909091,
      "last_run_timestamp": "2026-10-03T10:05:09.589Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Near AI",
      "accuracy": 0.762346,
      "accuracy_std_dev": 0.0014306672103136615,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.004426390572390572,
      "avg_duration_per_task_ms": 502309.91414141416,
      "avg_output_tokens_per_task": 29382.90404040404,
      "last_run_timestamp": "2026-10-24T16:30:53.943Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "DeepInfra",
      "accuracy": 0.7616160933333332,
      "accuracy_std_dev": 0.0330626227537687,
      "run_count": 9,
      "total_questions": 1650,
      "avg_cost_per_task_usd": 0.0073882921212121205,
      "avg_duration_per_task_ms": 943448.4084848484,
      "avg_output_tokens_per_task": 19044.426666666666,
      "last_run_timestamp": "2026-10-24T16:00:43.926Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "inclusionai/ling-3.0-flash-20260723",
      "model_slug": "inclusionai/ling-3.0-flash",
      "model_name": "inclusionAI: Ling 3.0 Flash",
      "provider_name": "Novita",
      "accuracy": 0.7605754720812183,
      "accuracy_std_dev": 0.010447569992586773,
      "run_count": 4,
      "total_questions": 788,
      "avg_cost_per_task_usd": 0.0025078527918781725,
      "avg_duration_per_task_ms": 100160.06345177664,
      "avg_output_tokens_per_task": 31388.34771573604,
      "last_run_timestamp": "2026-10-01T16:04:09.046Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7603815555555556,
      "accuracy_std_dev": 0.014085160481956509,
      "run_count": 18,
      "total_questions": 3564,
      "avg_cost_per_task_usd": 0.01835896689113356,
      "avg_duration_per_task_ms": 156375.88019079686,
      "avg_output_tokens_per_task": 30374.710718294053,
      "last_run_timestamp": "2026-10-24T16:26:58.593Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "GMICloud",
      "accuracy": 0.7603039230769231,
      "accuracy_std_dev": 0.042288769102708845,
      "run_count": 9,
      "total_questions": 1755,
      "avg_cost_per_task_usd": 0.27848958746438746,
      "avg_duration_per_task_ms": 886078.2860398861,
      "avg_output_tokens_per_task": 74626.49743589743,
      "last_run_timestamp": "2026-10-24T16:00:51.134Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Nebius",
      "accuracy": 0.7602694000000001,
      "accuracy_std_dev": 0.013238766658567559,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.01848461414141414,
      "avg_duration_per_task_ms": 133542.18232323232,
      "avg_output_tokens_per_task": 30573.586868686867,
      "last_run_timestamp": "2026-10-24T16:05:36.263Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Groq",
      "accuracy": 0.7596333333333335,
      "accuracy_std_dev": 0.006337831034527698,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.017320402918069584,
      "avg_duration_per_task_ms": 74089.44051627385,
      "avg_output_tokens_per_task": 28627.265432098764,
      "last_run_timestamp": "2026-10-24T16:15:45.667Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "Alibaba",
      "accuracy": 0.7592592222222222,
      "accuracy_std_dev": 0.13656078835349622,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.02930702805836139,
      "avg_duration_per_task_ms": 133789.24130190795,
      "avg_output_tokens_per_task": 22446.539842873175,
      "last_run_timestamp": "2026-10-24T16:02:08.877Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "mistralai/mistral-small-2603",
      "model_slug": "mistralai/mistral-small-2603",
      "model_name": "Mistral: Mistral Small 4",
      "provider_name": "Venice",
      "accuracy": 0.759259,
      "accuracy_std_dev": 0.003367000000000009,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.018210444444444442,
      "avg_duration_per_task_ms": 254627.46717171717,
      "avg_output_tokens_per_task": 24079.401515151516,
      "last_run_timestamp": "2026-09-01T12:59:55.618Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-2.5-flash",
      "model_slug": "google/gemini-2.5-flash",
      "model_name": "Google: Gemini 2.5 Flash",
      "provider_name": "Google AI Studio",
      "accuracy": 0.758698,
      "accuracy_std_dev": 0.012009662443216293,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.056970961279461274,
      "avg_duration_per_task_ms": 99388.35016835017,
      "avg_output_tokens_per_task": 22699.07239057239,
      "last_run_timestamp": "2026-10-01T19:11:42.323Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "CoreWeave",
      "accuracy": 0.757883010135135,
      "accuracy_std_dev": 0.030369548845212157,
      "run_count": 9,
      "total_questions": 1776,
      "avg_cost_per_task_usd": 0.004753131193693694,
      "avg_duration_per_task_ms": 685197.0304054054,
      "avg_output_tokens_per_task": 28398.651463963964,
      "last_run_timestamp": "2026-10-24T16:26:55.134Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "mistralai/mistral-small-2603",
      "model_slug": "mistralai/mistral-small-2603",
      "model_name": "Mistral: Mistral Small 4",
      "provider_name": "auto-routing",
      "accuracy": 0.7575759999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.014556641414141414,
      "avg_duration_per_task_ms": 173179.91919191918,
      "avg_output_tokens_per_task": 23980.964646464647,
      "last_run_timestamp": "2026-09-01T13:30:26.223Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-coder-next-2025-02-03",
      "model_slug": "qwen/qwen3-coder-next",
      "model_name": "Qwen: Qwen3 Coder Next",
      "provider_name": "Alibaba",
      "accuracy": 0.7575756666666668,
      "accuracy_std_dev": 0.0064470965471977405,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.022524166666666668,
      "avg_duration_per_task_ms": 96254.67171717172,
      "avg_output_tokens_per_task": 14855.659932659933,
      "last_run_timestamp": "2026-10-03T16:56:57.458Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "GMICloud",
      "accuracy": 0.7569795193929174,
      "accuracy_std_dev": 0.12787048553193514,
      "run_count": 9,
      "total_questions": 1779,
      "avg_cost_per_task_usd": 0.006095398538504777,
      "avg_duration_per_task_ms": 396590.173693086,
      "avg_output_tokens_per_task": 18629.006745362563,
      "last_run_timestamp": "2026-10-24T16:01:05.275Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Together",
      "accuracy": 0.7566855118644068,
      "accuracy_std_dev": 0.014912476720874281,
      "run_count": 9,
      "total_questions": 1770,
      "avg_cost_per_task_usd": 0.016834027683615818,
      "avg_duration_per_task_ms": 248403.6107344633,
      "avg_output_tokens_per_task": 28301.17288135593,
      "last_run_timestamp": "2026-10-24T16:36:35.542Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "auto-routing",
      "accuracy": 0.7564534444444445,
      "accuracy_std_dev": 0.04085365377950671,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.05544543434343434,
      "avg_duration_per_task_ms": 998107.7586980921,
      "avg_output_tokens_per_task": 28448.752525252527,
      "last_run_timestamp": "2026-10-24T16:02:06.908Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Baidu",
      "accuracy": 0.7557063684210527,
      "accuracy_std_dev": 0.055628184844211205,
      "run_count": 9,
      "total_questions": 1767,
      "avg_cost_per_task_usd": 0.2860427289190719,
      "avg_duration_per_task_ms": 984270.246179966,
      "avg_output_tokens_per_task": 83930.79456706283,
      "last_run_timestamp": "2026-10-24T16:08:29.474Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "OpenInference",
      "accuracy": 0.7547698333333334,
      "accuracy_std_dev": 0.0766049855914889,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.004288400673400673,
      "avg_duration_per_task_ms": 447675.2971380471,
      "avg_output_tokens_per_task": 15291.021885521886,
      "last_run_timestamp": "2026-10-17T08:26:03.926Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "Baidu",
      "accuracy": 0.754543774030354,
      "accuracy_std_dev": 0.030380293532339533,
      "run_count": 9,
      "total_questions": 1779,
      "avg_cost_per_task_usd": 0.17601991231028669,
      "avg_duration_per_task_ms": 1058551.6509274873,
      "avg_output_tokens_per_task": 79756.01405283867,
      "last_run_timestamp": "2026-10-24T16:01:48.224Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "DeepInfra",
      "accuracy": 0.7544054846967168,
      "accuracy_std_dev": 0.03260817390833722,
      "run_count": 11,
      "total_questions": 1797,
      "avg_cost_per_task_usd": 0.045709284919309964,
      "avg_duration_per_task_ms": 262995.0717863105,
      "avg_output_tokens_per_task": 17224.77796327212,
      "last_run_timestamp": "2026-10-24T08:41:25.770Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-thinking-2507",
      "model_slug": "qwen/qwen3-235b-a22b-thinking-2507",
      "model_name": "Qwen: Qwen3 235B A22B Thinking 2507",
      "provider_name": "Venice",
      "accuracy": 0.753788,
      "accuracy_std_dev": 0.02000377094449942,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.07918905176767678,
      "avg_duration_per_task_ms": 855570.0353535353,
      "avg_output_tokens_per_task": 22639.340909090908,
      "last_run_timestamp": "2026-08-22T17:38:59.682Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "mistralai/mistral-small-2603",
      "model_slug": "mistralai/mistral-small-2603",
      "model_name": "Mistral: Mistral Small 4",
      "provider_name": "Mistral",
      "accuracy": 0.7528056666666667,
      "accuracy_std_dev": 0.004601995207395269,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.015185508417508417,
      "avg_duration_per_task_ms": 152622.4537037037,
      "avg_output_tokens_per_task": 24359.508417508416,
      "last_run_timestamp": "2026-09-17T02:17:49.284Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.7525250000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.008515843434343435,
      "avg_duration_per_task_ms": 110884.25757575757,
      "avg_output_tokens_per_task": 8312.732323232323,
      "last_run_timestamp": "2026-10-04T03:06:11.710Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Phala",
      "accuracy": 0.7516834000000001,
      "accuracy_std_dev": 0.0580270664952141,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.01393037222222222,
      "avg_duration_per_task_ms": 97444.4191919192,
      "avg_output_tokens_per_task": 5974.968181818182,
      "last_run_timestamp": "2026-10-24T08:48:00.943Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "Google",
      "accuracy": 0.75084367721519,
      "accuracy_std_dev": 0.024667232906226022,
      "run_count": 4,
      "total_questions": 790,
      "avg_cost_per_task_usd": 0.020310108860759495,
      "avg_duration_per_task_ms": 155074.11012658227,
      "avg_output_tokens_per_task": 11752.967088607595,
      "last_run_timestamp": "2026-08-22T17:12:08.919Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Morph",
      "accuracy": 0.7508415000000002,
      "accuracy_std_dev": 0.0622895,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.06288509848484848,
      "avg_duration_per_task_ms": 493754.3712121212,
      "avg_output_tokens_per_task": 26228.825757575756,
      "last_run_timestamp": "2026-08-15T10:07:37.662Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "GMICloud",
      "accuracy": 0.75076231097561,
      "accuracy_std_dev": 0.10261490128027215,
      "run_count": 10,
      "total_questions": 1968,
      "avg_cost_per_task_usd": 0.006507009654471546,
      "avg_duration_per_task_ms": 566183.218495935,
      "avg_output_tokens_per_task": 27810.678353658535,
      "last_run_timestamp": "2026-10-10T08:22:53.947Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-coder-next-2025-02-03",
      "model_slug": "qwen/qwen3-coder-next",
      "model_name": "Qwen: Qwen3 Coder Next",
      "provider_name": "StreamLake",
      "accuracy": 0.7493987142857143,
      "accuracy_std_dev": 0.00923961159842752,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.013690483405483407,
      "avg_duration_per_task_ms": 96167.38383838384,
      "avg_output_tokens_per_task": 15049.160173160173,
      "last_run_timestamp": "2026-10-03T16:54:39.591Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "inclusionai/ling-3.0-flash-20260723",
      "model_slug": "inclusionai/ling-3.0-flash",
      "model_name": "inclusionAI: Ling 3.0 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.7487311624365482,
      "accuracy_std_dev": 0.019042500000000018,
      "run_count": 2,
      "total_questions": 394,
      "avg_cost_per_task_usd": 0.00260698730964467,
      "avg_duration_per_task_ms": 166191.14467005077,
      "avg_output_tokens_per_task": 30980.81725888325,
      "last_run_timestamp": "2026-10-01T16:04:24.420Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "NextBit",
      "accuracy": 0.7485971111111113,
      "accuracy_std_dev": 0.012497660327210088,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.017153025252525253,
      "avg_duration_per_task_ms": 1346127.6537598204,
      "avg_output_tokens_per_task": 55391.88832772166,
      "last_run_timestamp": "2026-10-17T16:05:03.311Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "Wafer",
      "accuracy": 0.7481013645569621,
      "accuracy_std_dev": 0.004003499999999993,
      "run_count": 2,
      "total_questions": 395,
      "avg_cost_per_task_usd": 0.010130237974683544,
      "avg_duration_per_task_ms": 298779.6734177215,
      "avg_output_tokens_per_task": 60721.07341772152,
      "last_run_timestamp": "2026-10-10T08:39:12.933Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-coder-next-2025-02-03",
      "model_slug": "qwen/qwen3-coder-next",
      "model_name": "Qwen: Qwen3 Coder Next",
      "provider_name": "auto-routing",
      "accuracy": 0.7477555,
      "accuracy_std_dev": 0.009932117124930308,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.01480825,
      "avg_duration_per_task_ms": 108491.74326599327,
      "avg_output_tokens_per_task": 14487.641414141413,
      "last_run_timestamp": "2026-10-03T16:43:33.673Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Io Net",
      "accuracy": 0.7477012353448277,
      "accuracy_std_dev": 0.024896992942589145,
      "run_count": 6,
      "total_questions": 1160,
      "avg_cost_per_task_usd": 0.016306832758620687,
      "avg_duration_per_task_ms": 316542.88879310346,
      "avg_output_tokens_per_task": 52791.42327586207,
      "last_run_timestamp": "2026-10-24T08:12:52.465Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2-0905",
      "model_slug": "moonshotai/kimi-k2-0905",
      "model_name": "MoonshotAI: Kimi K2 0905",
      "provider_name": "Groq",
      "accuracy": 0.7474749999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.018268570707070705,
      "avg_duration_per_task_ms": 31005.68686868687,
      "avg_output_tokens_per_task": 5873.606060606061,
      "last_run_timestamp": "2026-08-01T17:45:49.025Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "Makora",
      "accuracy": 0.7468014,
      "accuracy_std_dev": 0.009196272649285712,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.015463989898989897,
      "avg_duration_per_task_ms": 342566.1424242424,
      "avg_output_tokens_per_task": 63071.326262626266,
      "last_run_timestamp": "2026-10-24T16:16:25.413Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "Io Net",
      "accuracy": 0.7465941264367815,
      "accuracy_std_dev": 0.10532480903970348,
      "run_count": 8,
      "total_questions": 1566,
      "avg_cost_per_task_usd": 0.04816395977011494,
      "avg_duration_per_task_ms": 468882.4540229885,
      "avg_output_tokens_per_task": 36470.09642401022,
      "last_run_timestamp": "2026-09-19T17:18:00.087Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "DeepInfra",
      "accuracy": 0.7465043071718538,
      "accuracy_std_dev": 0.027768881506803414,
      "run_count": 6,
      "total_questions": 739,
      "avg_cost_per_task_usd": 0.053211952638700946,
      "avg_duration_per_task_ms": 424838.5182679296,
      "avg_output_tokens_per_task": 15554.012178619756,
      "last_run_timestamp": "2026-09-19T08:53:38.031Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7462993119730186,
      "accuracy_std_dev": 0.0800254867207247,
      "run_count": 9,
      "total_questions": 1779,
      "avg_cost_per_task_usd": 0.00489628442945475,
      "avg_duration_per_task_ms": 596052.9263631253,
      "avg_output_tokens_per_task": 32164.360314783586,
      "last_run_timestamp": "2026-10-17T16:03:21.011Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "StreamLake",
      "accuracy": 0.7460317142857142,
      "accuracy_std_dev": 0.007918207277521612,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.006722455266955267,
      "avg_duration_per_task_ms": 164422.91197691197,
      "avg_output_tokens_per_task": 7810.999278499278,
      "last_run_timestamp": "2026-10-03T16:58:32.590Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Novita",
      "accuracy": 0.7459349573170732,
      "accuracy_std_dev": 0.02985061357091343,
      "run_count": 5,
      "total_questions": 984,
      "avg_cost_per_task_usd": 0.014618331300813008,
      "avg_duration_per_task_ms": 1032360.968495935,
      "avg_output_tokens_per_task": 51870.760162601626,
      "last_run_timestamp": "2026-10-24T08:20:32.397Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7441247969543147,
      "accuracy_std_dev": 0.013743685198646926,
      "run_count": 9,
      "total_questions": 1773,
      "avg_cost_per_task_usd": 0.1143873355893965,
      "avg_duration_per_task_ms": 753346.6830231247,
      "avg_output_tokens_per_task": 44665.73547659334,
      "last_run_timestamp": "2026-10-24T16:39:40.802Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-coder-next-2025-02-03",
      "model_slug": "qwen/qwen3-coder-next",
      "model_name": "Qwen: Qwen3 Coder Next",
      "provider_name": "Parasail",
      "accuracy": 0.7435209222520109,
      "accuracy_std_dev": 0.004720778410977948,
      "run_count": 6,
      "total_questions": 1119,
      "avg_cost_per_task_usd": 0.008935632707774798,
      "avg_duration_per_task_ms": 139189.39052725647,
      "avg_output_tokens_per_task": 11357.61126005362,
      "last_run_timestamp": "2026-10-03T17:03:40.665Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Nebius",
      "accuracy": 0.743266,
      "accuracy_std_dev": 0.006355106254029135,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.005297381944444444,
      "avg_duration_per_task_ms": 155629.39141414143,
      "avg_output_tokens_per_task": 8558.156565656565,
      "last_run_timestamp": "2026-10-17T16:06:15.944Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.742214,
      "accuracy_std_dev": 0.015132597149861616,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.01671199116161616,
      "avg_duration_per_task_ms": 766495.9797979798,
      "avg_output_tokens_per_task": 59349.530303030304,
      "last_run_timestamp": "2026-10-24T08:26:00.562Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-coder-next-2025-02-03",
      "model_slug": "qwen/qwen3-coder-next",
      "model_name": "Qwen: Qwen3 Coder Next",
      "provider_name": "Novita",
      "accuracy": 0.7404603333333333,
      "accuracy_std_dev": 0.009642303833743399,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.023280140572390574,
      "avg_duration_per_task_ms": 99017.15824915825,
      "avg_output_tokens_per_task": 15411.767676767677,
      "last_run_timestamp": "2026-10-03T16:58:08.462Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "Novita",
      "accuracy": 0.7395181320432556,
      "accuracy_std_dev": 0.058919001127620475,
      "run_count": 9,
      "total_questions": 1757,
      "avg_cost_per_task_usd": 0.2870167250996016,
      "avg_duration_per_task_ms": 1318827.1798520205,
      "avg_output_tokens_per_task": 91069.35685828116,
      "last_run_timestamp": "2026-10-24T16:02:28.413Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-2.5-flash",
      "model_slug": "google/gemini-2.5-flash",
      "model_name": "Google: Gemini 2.5 Flash",
      "provider_name": "Google",
      "accuracy": 0.738309,
      "accuracy_std_dev": 0.008136215650479735,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.059655349046015715,
      "avg_duration_per_task_ms": 125601.6632996633,
      "avg_output_tokens_per_task": 23853.966891133558,
      "last_run_timestamp": "2026-10-01T19:15:31.968Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "Novita",
      "accuracy": 0.737284008888889,
      "accuracy_std_dev": 0.05166831288604067,
      "run_count": 7,
      "total_questions": 1350,
      "avg_cost_per_task_usd": 0.017425666666666666,
      "avg_duration_per_task_ms": 990522.262962963,
      "avg_output_tokens_per_task": 44050.16814814815,
      "last_run_timestamp": "2026-10-24T16:13:05.630Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Alibaba",
      "accuracy": 0.7359708333333334,
      "accuracy_std_dev": 0.009394253179766615,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.004828659932659932,
      "avg_duration_per_task_ms": 106903.15909090909,
      "avg_output_tokens_per_task": 7872.062289562289,
      "last_run_timestamp": "2026-10-03T16:59:36.480Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "Mara",
      "accuracy": 0.7349895701863354,
      "accuracy_std_dev": 0.017162011317797845,
      "run_count": 6,
      "total_questions": 805,
      "avg_cost_per_task_usd": 0.015730058385093167,
      "avg_duration_per_task_ms": 68792.66708074535,
      "avg_output_tokens_per_task": 9043.119254658384,
      "last_run_timestamp": "2026-10-17T16:03:44.043Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Parasail",
      "accuracy": 0.7349291052009457,
      "accuracy_std_dev": 0.007543459813017969,
      "run_count": 9,
      "total_questions": 1692,
      "avg_cost_per_task_usd": 0.006669232269503548,
      "avg_duration_per_task_ms": 243172.80910165486,
      "avg_output_tokens_per_task": 8248.872340425532,
      "last_run_timestamp": "2026-10-24T16:02:47.278Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "Venice",
      "accuracy": 0.7341193589990375,
      "accuracy_std_dev": 0.14473932746638518,
      "run_count": 11,
      "total_questions": 2078,
      "avg_cost_per_task_usd": 0.10010441241578441,
      "avg_duration_per_task_ms": 669272.0486044274,
      "avg_output_tokens_per_task": 29290.59720885467,
      "last_run_timestamp": "2026-10-24T16:11:49.038Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Google",
      "accuracy": 0.7340817802024036,
      "accuracy_std_dev": 0.01834012481629673,
      "run_count": 16,
      "total_questions": 3162,
      "avg_cost_per_task_usd": 0.006654911448450348,
      "avg_duration_per_task_ms": 123137.50853889943,
      "avg_output_tokens_per_task": 6968.909550917141,
      "last_run_timestamp": "2026-10-17T16:03:44.090Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Novita",
      "accuracy": 0.7333332203389831,
      "accuracy_std_dev": 0.014287186233444833,
      "run_count": 9,
      "total_questions": 1770,
      "avg_cost_per_task_usd": 0.006560344632768362,
      "avg_duration_per_task_ms": 335200.51129943505,
      "avg_output_tokens_per_task": 26048.953107344634,
      "last_run_timestamp": "2026-10-24T16:16:05.012Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-coder-next-2025-02-03",
      "model_slug": "qwen/qwen3-coder-next",
      "model_name": "Qwen: Qwen3 Coder Next",
      "provider_name": "Ionstream",
      "accuracy": 0.7308918662420382,
      "accuracy_std_dev": 0.007761999999999991,
      "run_count": 2,
      "total_questions": 314,
      "avg_cost_per_task_usd": 0.007396050955414012,
      "avg_duration_per_task_ms": 197815.93949044586,
      "avg_output_tokens_per_task": 9168.713375796178,
      "last_run_timestamp": "2026-08-08T16:46:11.805Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.5-haiku-20251001",
      "model_slug": "anthropic/claude-haiku-4.5",
      "model_name": "Anthropic: Claude Haiku 4.5",
      "provider_name": "Anthropic",
      "accuracy": 0.7307692307692307,
      "accuracy_std_dev": 0.014845665656780083,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.18598931313131314,
      "avg_duration_per_task_ms": 298695.4405594406,
      "avg_output_tokens_per_task": 37027.50971250971,
      "last_run_timestamp": "2026-10-24T08:28:43.171Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.5-haiku-20251001",
      "model_slug": "anthropic/claude-haiku-4.5",
      "model_name": "Anthropic: Claude Haiku 4.5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7307283421052632,
      "accuracy_std_dev": 0.012478804160563647,
      "run_count": 38,
      "total_questions": 7524,
      "avg_cost_per_task_usd": 0.20430948484848477,
      "avg_duration_per_task_ms": 243835.08971291865,
      "avg_output_tokens_per_task": 38154.19364699628,
      "last_run_timestamp": "2026-10-24T08:29:16.808Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "Io Net",
      "accuracy": 0.7299411591650359,
      "accuracy_std_dev": 0.0992559733114335,
      "run_count": 9,
      "total_questions": 1533,
      "avg_cost_per_task_usd": 0.00431450815394651,
      "avg_duration_per_task_ms": 554566.0965427267,
      "avg_output_tokens_per_task": 26096.888454011743,
      "last_run_timestamp": "2026-10-24T08:13:46.683Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Venice",
      "accuracy": 0.729798,
      "accuracy_std_dev": 0.06649999628571418,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.12289258193041527,
      "avg_duration_per_task_ms": 609051.913580247,
      "avg_output_tokens_per_task": 36908.911335578,
      "last_run_timestamp": "2026-10-24T16:25:38.398Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "Z.AI",
      "accuracy": 0.7294042813688214,
      "accuracy_std_dev": 0.0505548117787454,
      "run_count": 8,
      "total_questions": 1578,
      "avg_cost_per_task_usd": 0.3084836787072243,
      "avg_duration_per_task_ms": 1347611.5551330799,
      "avg_output_tokens_per_task": 97858.03675538657,
      "last_run_timestamp": "2026-10-24T16:28:00.946Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Venice",
      "accuracy": 0.7293771250000001,
      "accuracy_std_dev": 0.010428876514724617,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.00597866792929293,
      "avg_duration_per_task_ms": 411352.74242424243,
      "avg_output_tokens_per_task": 7899.722222222223,
      "last_run_timestamp": "2026-10-24T16:07:46.992Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Novita",
      "accuracy": 0.7281143333333333,
      "accuracy_std_dev": 0.016544932312409815,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.004527595117845118,
      "avg_duration_per_task_ms": 163776.3367003367,
      "avg_output_tokens_per_task": 7680.468013468014,
      "last_run_timestamp": "2026-10-03T16:57:10.991Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.5-haiku-20251001",
      "model_slug": "anthropic/claude-haiku-4.5",
      "model_name": "Anthropic: Claude Haiku 4.5",
      "provider_name": "auto-routing",
      "accuracy": 0.7277907692307694,
      "accuracy_std_dev": 0.009063148715488058,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.19022068764568764,
      "avg_duration_per_task_ms": 243477.47397047398,
      "avg_output_tokens_per_task": 37853.29875679876,
      "last_run_timestamp": "2026-10-24T08:26:35.486Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "Phala",
      "accuracy": 0.7271808509532062,
      "accuracy_std_dev": 0.03540222199908109,
      "run_count": 6,
      "total_questions": 1154,
      "avg_cost_per_task_usd": 0.011152869150779895,
      "avg_duration_per_task_ms": 504501.5831889082,
      "avg_output_tokens_per_task": 55506.62998266898,
      "last_run_timestamp": "2026-10-10T08:24:27.616Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "auto-routing",
      "accuracy": 0.7265246666666667,
      "accuracy_std_dev": 0.015018083728921232,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.005456846240179574,
      "avg_duration_per_task_ms": 293364.7452300786,
      "avg_output_tokens_per_task": 7944.652076318743,
      "last_run_timestamp": "2026-10-24T16:01:05.272Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "auto-routing",
      "accuracy": 0.7265245555555556,
      "accuracy_std_dev": 0.023902048992182463,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.0058103249158249146,
      "avg_duration_per_task_ms": 345990.65993265994,
      "avg_output_tokens_per_task": 18519.231200897866,
      "last_run_timestamp": "2026-10-24T16:02:47.367Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.5-haiku-20251001",
      "model_slug": "anthropic/claude-haiku-4.5",
      "model_name": "Anthropic: Claude Haiku 4.5",
      "provider_name": "Azure",
      "accuracy": 0.7264923098256736,
      "accuracy_std_dev": 0.012443953888007772,
      "run_count": 13,
      "total_questions": 2524,
      "avg_cost_per_task_usd": 0.18690665134706813,
      "avg_duration_per_task_ms": 345005.80625990493,
      "avg_output_tokens_per_task": 37190.982171156895,
      "last_run_timestamp": "2026-10-24T08:44:41.412Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Tenstorrent",
      "accuracy": 0.7264309999999999,
      "accuracy_std_dev": 0.021044000000000007,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.1303771111111111,
      "avg_duration_per_task_ms": 318087.7601010101,
      "avg_output_tokens_per_task": 43267.416666666664,
      "last_run_timestamp": "2026-10-04T05:57:29.940Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-2.5-flash",
      "model_slug": "google/gemini-2.5-flash",
      "model_name": "Google: Gemini 2.5 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.7261503333333333,
      "accuracy_std_dev": 0.004418546090690373,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.059926168350168355,
      "avg_duration_per_task_ms": 123042.70033670033,
      "avg_output_tokens_per_task": 23940.483164983165,
      "last_run_timestamp": "2026-10-01T19:12:41.089Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Krea",
      "accuracy": 0.7234044255319149,
      "accuracy_std_dev": 0.1917994784299478,
      "run_count": 3,
      "total_questions": 517,
      "avg_cost_per_task_usd": 0.033565591876208896,
      "avg_duration_per_task_ms": 242772.43907156674,
      "avg_output_tokens_per_task": 37102.30560928433,
      "last_run_timestamp": "2026-09-29T20:57:20.673Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "DeepInfra",
      "accuracy": 0.7221381,
      "accuracy_std_dev": 0.010360325819683453,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.004400843939393938,
      "avg_duration_per_task_ms": 655428.28989899,
      "avg_output_tokens_per_task": 7876.879292929293,
      "last_run_timestamp": "2026-10-24T16:01:45.799Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "ModelRun",
      "accuracy": 0.7219408075949366,
      "accuracy_std_dev": 0.025598999999999983,
      "run_count": 2,
      "total_questions": 395,
      "avg_cost_per_task_usd": 0.27939689620253166,
      "avg_duration_per_task_ms": 450249.90379746835,
      "avg_output_tokens_per_task": 45941.00253164557,
      "last_run_timestamp": "2026-08-15T08:16:15.061Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "CoreWeave",
      "accuracy": 0.721866038034188,
      "accuracy_std_dev": 0.03876582276336764,
      "run_count": 12,
      "total_questions": 2340,
      "avg_cost_per_task_usd": 0.010606732051282052,
      "avg_duration_per_task_ms": 258545.00683760684,
      "avg_output_tokens_per_task": 48059.278205128205,
      "last_run_timestamp": "2026-10-24T08:11:11.359Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.5-haiku-20251001",
      "model_slug": "anthropic/claude-haiku-4.5",
      "model_name": "Anthropic: Claude Haiku 4.5",
      "provider_name": "Google",
      "accuracy": 0.72179058974359,
      "accuracy_std_dev": 0.013298167116268404,
      "run_count": 39,
      "total_questions": 7722,
      "avg_cost_per_task_usd": 0.1972910025900026,
      "avg_duration_per_task_ms": 268583.0019425019,
      "avg_output_tokens_per_task": 36795.75433825434,
      "last_run_timestamp": "2026-10-24T09:06:24.922Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "GMICloud",
      "accuracy": 0.7208904075342465,
      "accuracy_std_dev": 0.09870737973249176,
      "run_count": 9,
      "total_questions": 1752,
      "avg_cost_per_task_usd": 0.13571237785388127,
      "avg_duration_per_task_ms": 977847.7089041095,
      "avg_output_tokens_per_task": 72379.5450913242,
      "last_run_timestamp": "2026-10-24T16:26:47.558Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Cohere",
      "accuracy": 0.719697,
      "accuracy_std_dev": 0.004209000000000018,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.021745989898989897,
      "avg_duration_per_task_ms": 204822.55555555556,
      "avg_output_tokens_per_task": 77279.37878787878,
      "last_run_timestamp": "2026-10-24T08:14:03.073Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Makora",
      "accuracy": 0.7174523333333331,
      "accuracy_std_dev": 0.013163488932987655,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.018911546296296298,
      "avg_duration_per_task_ms": 531218.9570707071,
      "avg_output_tokens_per_task": 67191.76851851853,
      "last_run_timestamp": "2026-10-24T09:17:16.298Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Friendli",
      "accuracy": 0.7163299999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.005710287878787878,
      "avg_duration_per_task_ms": 197152.55555555556,
      "avg_output_tokens_per_task": 6940.146464646465,
      "last_run_timestamp": "2026-08-01T17:52:16.799Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-next-80b-a3b-instruct-2509",
      "model_slug": "qwen/qwen3-next-80b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Next 80B A3B Instruct",
      "provider_name": "Novita",
      "accuracy": 0.7135241666666668,
      "accuracy_std_dev": 0.005271695534223852,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.015626640572390573,
      "avg_duration_per_task_ms": 96955.32070707071,
      "avg_output_tokens_per_task": 10336.517676767677,
      "last_run_timestamp": "2026-10-03T16:46:22.205Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-next-80b-a3b-instruct-2509",
      "model_slug": "qwen/qwen3-next-80b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Next 80B A3B Instruct",
      "provider_name": "Alibaba",
      "accuracy": 0.7135240000000002,
      "accuracy_std_dev": 0.009836496361340617,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.007830127104377103,
      "avg_duration_per_task_ms": 79042.44696969698,
      "avg_output_tokens_per_task": 9942.877946127946,
      "last_run_timestamp": "2026-10-03T16:40:12.993Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Darkbloom",
      "accuracy": 0.713213,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 111,
      "avg_cost_per_task_usd": 0.007665369369369369,
      "avg_duration_per_task_ms": 1401051.936936937,
      "avg_output_tokens_per_task": 27183.8018018018,
      "last_run_timestamp": "2026-10-03T09:34:00.571Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "DeepInfra",
      "accuracy": 0.712284535775862,
      "accuracy_std_dev": 0.11489540719980705,
      "run_count": 13,
      "total_questions": 2320,
      "avg_cost_per_task_usd": 0.032706028879310345,
      "avg_duration_per_task_ms": 1027452.9099137931,
      "avg_output_tokens_per_task": 34342.02112068966,
      "last_run_timestamp": "2026-10-24T16:14:06.772Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-next-80b-a3b-instruct-2509",
      "model_slug": "qwen/qwen3-next-80b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Next 80B A3B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.7121211666666666,
      "accuracy_std_dev": 0.009066017803068557,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.011008058080808082,
      "avg_duration_per_task_ms": 95816.33922558922,
      "avg_output_tokens_per_task": 10395.001683501683,
      "last_run_timestamp": "2026-10-03T16:59:33.949Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-thinking-2507",
      "model_slug": "qwen/qwen3-235b-a22b-thinking-2507",
      "model_name": "Qwen: Qwen3 235B A22B Thinking 2507",
      "provider_name": "Novita",
      "accuracy": 0.71170025,
      "accuracy_std_dev": 0.018456257615982147,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.0649322058080808,
      "avg_duration_per_task_ms": 1066951.6603535353,
      "avg_output_tokens_per_task": 21770.531565656565,
      "last_run_timestamp": "2026-08-22T17:45:06.645Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Io Net",
      "accuracy": 0.7112403023255814,
      "accuracy_std_dev": 0.09623599999999999,
      "run_count": 2,
      "total_questions": 344,
      "avg_cost_per_task_usd": 0.06576875,
      "avg_duration_per_task_ms": 744084.4970930233,
      "avg_output_tokens_per_task": 27095.636627906977,
      "last_run_timestamp": "2026-08-15T09:29:41.209Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7107745999999999,
      "accuracy_std_dev": 0.05091182812510271,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.006678358585858586,
      "avg_duration_per_task_ms": 125158.29696969697,
      "avg_output_tokens_per_task": 7409.80808080808,
      "last_run_timestamp": "2026-08-29T18:01:46.940Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Mancer 2",
      "accuracy": 0.7106481250000001,
      "accuracy_std_dev": 0.043384424386631835,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.007522804924242423,
      "avg_duration_per_task_ms": 781654.2998737374,
      "avg_output_tokens_per_task": 26137.825126262625,
      "last_run_timestamp": "2026-10-24T16:14:06.772Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Friendli",
      "accuracy": 0.7100635555555554,
      "accuracy_std_dev": 0.054318885093514825,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.006609520763187431,
      "avg_duration_per_task_ms": 110734.31762065095,
      "avg_output_tokens_per_task": 16252.387766554433,
      "last_run_timestamp": "2026-10-24T16:28:53.260Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-next-80b-a3b-instruct-2509",
      "model_slug": "qwen/qwen3-next-80b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Next 80B A3B Instruct",
      "provider_name": "Parasail",
      "accuracy": 0.709650171648164,
      "accuracy_std_dev": 0.01649939376664086,
      "run_count": 6,
      "total_questions": 1171,
      "avg_cost_per_task_usd": 0.010711521776259607,
      "avg_duration_per_task_ms": 103546.38514090521,
      "avg_output_tokens_per_task": 9904.411614005125,
      "last_run_timestamp": "2026-10-03T16:43:17.105Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "DeepInfra",
      "accuracy": 0.7096191251428572,
      "accuracy_std_dev": 0.04936102018827993,
      "run_count": 18,
      "total_questions": 3500,
      "avg_cost_per_task_usd": 0.007894835428571428,
      "avg_duration_per_task_ms": 520799.64857142855,
      "avg_output_tokens_per_task": 25638.9,
      "last_run_timestamp": "2026-10-24T16:12:04.707Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_slug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_name": "Qwen: Qwen3 VL 235B A22B Instruct",
      "provider_name": "Novita",
      "accuracy": 0.7093155,
      "accuracy_std_dev": 0.008890530969332882,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.008779646464646466,
      "avg_duration_per_task_ms": 132997.85521885523,
      "avg_output_tokens_per_task": 5690.617845117845,
      "last_run_timestamp": "2026-10-03T16:45:33.688Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "Google",
      "accuracy": 0.707752053820598,
      "accuracy_std_dev": 0.25911579823070297,
      "run_count": 8,
      "total_questions": 1505,
      "avg_cost_per_task_usd": 0.025081889036544854,
      "avg_duration_per_task_ms": 430774.81063122925,
      "avg_output_tokens_per_task": 14989.863122923589,
      "last_run_timestamp": "2026-10-17T16:03:44.233Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_slug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_name": "Qwen: Qwen3 VL 235B A22B Instruct",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.707071,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.008867828282828283,
      "avg_duration_per_task_ms": 116045.65151515152,
      "avg_output_tokens_per_task": 5749.39898989899,
      "last_run_timestamp": "2026-10-04T03:05:59.018Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "Parasail",
      "accuracy": 0.705263,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 95,
      "avg_cost_per_task_usd": 0.10938622105263157,
      "avg_duration_per_task_ms": 1384847.8,
      "avg_output_tokens_per_task": 40599.36842105263,
      "last_run_timestamp": "2026-08-01T18:16:39.158Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "auto-routing",
      "accuracy": 0.7048612500000001,
      "accuracy_std_dev": 0.04887384717246537,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.016058294823232323,
      "avg_duration_per_task_ms": 999836.755050505,
      "avg_output_tokens_per_task": 49417.84027777778,
      "last_run_timestamp": "2026-10-24T16:05:13.521Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5-nano-2025-08-07",
      "model_slug": "openai/gpt-5-nano",
      "model_name": "OpenAI: GPT-5 Nano",
      "provider_name": "auto-routing",
      "accuracy": 0.7045455,
      "accuracy_std_dev": 0.004208500000000004,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.01336774494949495,
      "avg_duration_per_task_ms": 231791.09848484848,
      "avg_output_tokens_per_task": 33323.15909090909,
      "last_run_timestamp": "2026-10-01T19:27:22.032Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_slug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_name": "Qwen: Qwen3 VL 235B A22B Instruct",
      "provider_name": "Alibaba",
      "accuracy": 0.7045453333333335,
      "accuracy_std_dev": 0.009510928024587757,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.00597795707070707,
      "avg_duration_per_task_ms": 110020.64814814815,
      "avg_output_tokens_per_task": 5546.97138047138,
      "last_run_timestamp": "2026-10-03T17:02:24.058Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Venice",
      "accuracy": 0.7032012317327765,
      "accuracy_std_dev": 0.20342432055346774,
      "run_count": 5,
      "total_questions": 958,
      "avg_cost_per_task_usd": 0.08802485803757831,
      "avg_duration_per_task_ms": 1349048.8903966597,
      "avg_output_tokens_per_task": 33413.259916492694,
      "last_run_timestamp": "2026-10-24T16:13:05.627Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_slug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_name": "Qwen: Qwen3 VL 235B A22B Instruct",
      "provider_name": "Parasail",
      "accuracy": 0.7031426666666668,
      "accuracy_std_dev": 0.014373068248947025,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.011147152356902357,
      "avg_duration_per_task_ms": 198021.7255892256,
      "avg_output_tokens_per_task": 5803.353535353535,
      "last_run_timestamp": "2026-10-03T16:59:57.847Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "auto-routing",
      "accuracy": 0.7030114672897196,
      "accuracy_std_dev": 0.02694021764221664,
      "run_count": 10,
      "total_questions": 1926,
      "avg_cost_per_task_usd": 0.009145764278296989,
      "avg_duration_per_task_ms": 363422.58099688473,
      "avg_output_tokens_per_task": 44741.07476635514,
      "last_run_timestamp": "2026-10-24T08:01:41.230Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "GMICloud",
      "accuracy": 0.7028622499999999,
      "accuracy_std_dev": 0.08200476941122087,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.0070032777777777775,
      "avg_duration_per_task_ms": 172461.12373737374,
      "avg_output_tokens_per_task": 7212.60101010101,
      "last_run_timestamp": "2026-10-03T16:56:36.098Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-next-80b-a3b-instruct-2509",
      "model_slug": "qwen/qwen3-next-80b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Next 80B A3B Instruct",
      "provider_name": "DeepInfra",
      "accuracy": 0.7027678642599278,
      "accuracy_std_dev": 0.009533289144800297,
      "run_count": 7,
      "total_questions": 1385,
      "avg_cost_per_task_usd": 0.010358495306859207,
      "avg_duration_per_task_ms": 145686.42599277978,
      "avg_output_tokens_per_task": 9350.431768953069,
      "last_run_timestamp": "2026-10-08T22:40:37.311Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-next-80b-a3b-instruct-2509",
      "model_slug": "qwen/qwen3-next-80b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Next 80B A3B Instruct",
      "provider_name": "Google",
      "accuracy": 0.7011786666666667,
      "accuracy_std_dev": 0.0268613535999626,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.010966026094276094,
      "avg_duration_per_task_ms": 64637.51430976431,
      "avg_output_tokens_per_task": 9084.513468013469,
      "last_run_timestamp": "2026-10-03T16:36:16.895Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Venice",
      "accuracy": 0.6995251282660333,
      "accuracy_std_dev": 0.037727051631422225,
      "run_count": 5,
      "total_questions": 842,
      "avg_cost_per_task_usd": 0.03918485154394299,
      "avg_duration_per_task_ms": 785269.9073634205,
      "avg_output_tokens_per_task": 41981.14845605701,
      "last_run_timestamp": "2026-08-29T18:51:07.939Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3-nano-30b-a3b",
      "model_slug": "nvidia/nemotron-3-nano-30b-a3b",
      "model_name": "NVIDIA: Nemotron 3 Nano 30B A3B",
      "provider_name": "DeepInfra",
      "accuracy": 0.6948820649606299,
      "accuracy_std_dev": 0.027729042227588305,
      "run_count": 8,
      "total_questions": 1524,
      "avg_cost_per_task_usd": 0.015564356299212597,
      "avg_duration_per_task_ms": 765563.3963254594,
      "avg_output_tokens_per_task": 77634.76837270342,
      "last_run_timestamp": "2026-09-26T16:34:21.219Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_slug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_name": "Qwen: Qwen3 VL 235B A22B Instruct",
      "provider_name": "DeepInfra",
      "accuracy": 0.6938431428571429,
      "accuracy_std_dev": 0.009541538965844799,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.005521870129870129,
      "avg_duration_per_task_ms": 387910.27056277054,
      "avg_output_tokens_per_task": 6090.331168831169,
      "last_run_timestamp": "2026-10-08T23:09:15.796Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.6-sonnet-20260217",
      "model_slug": "anthropic/claude-sonnet-4.6",
      "model_name": "Anthropic: Claude Sonnet 4.6",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.6932966363636364,
      "accuracy_std_dev": 0.13147150091889792,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.609068142332415,
      "avg_duration_per_task_ms": 632474.5367309458,
      "avg_output_tokens_per_task": 40521.938016528926,
      "last_run_timestamp": "2026-10-24T08:08:54.166Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "DeepInfra",
      "accuracy": 0.6932378040665435,
      "accuracy_std_dev": 0.046335214744527585,
      "run_count": 11,
      "total_questions": 2164,
      "avg_cost_per_task_usd": 0.0074276700554528645,
      "avg_duration_per_task_ms": 281567.51894639555,
      "avg_output_tokens_per_task": 40195.87939001848,
      "last_run_timestamp": "2026-10-24T08:42:05.963Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_slug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_name": "Qwen: Qwen3 VL 235B A22B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.6927610000000001,
      "accuracy_std_dev": 0.01993714402817014,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.009704229797979799,
      "avg_duration_per_task_ms": 259526.76094276094,
      "avg_output_tokens_per_task": 6456.316498316498,
      "last_run_timestamp": "2026-10-03T17:02:48.284Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "DekaLLM",
      "accuracy": 0.6899941197301855,
      "accuracy_std_dev": 0.005866752821337055,
      "run_count": 3,
      "total_questions": 593,
      "avg_cost_per_task_usd": 0.0068884283305227665,
      "avg_duration_per_task_ms": 1677214.3709949409,
      "avg_output_tokens_per_task": 49011.02360876897,
      "last_run_timestamp": "2026-10-24T16:08:06.921Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5-nano-2025-08-07",
      "model_slug": "openai/gpt-5-nano",
      "model_name": "OpenAI: GPT-5 Nano",
      "provider_name": "Azure",
      "accuracy": 0.689815,
      "accuracy_std_dev": 0.0034448118235979193,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.013852137626262627,
      "avg_duration_per_task_ms": 234308.42045454544,
      "avg_output_tokens_per_task": 32893.11994949495,
      "last_run_timestamp": "2026-10-01T19:27:10.142Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "DeepInfra",
      "accuracy": 0.6855188289251547,
      "accuracy_std_dev": 0.09527891619753961,
      "run_count": 9,
      "total_questions": 1777,
      "avg_cost_per_task_usd": 0.011474473832301631,
      "avg_duration_per_task_ms": 906865.2909397861,
      "avg_output_tokens_per_task": 33586.68429938098,
      "last_run_timestamp": "2026-10-24T16:17:39.034Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403:free",
      "model_slug": null,
      "model_name": "Gemma 4 26B A4B IT (free)",
      "provider_name": "Darkbloom",
      "accuracy": 0.684722,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 120,
      "avg_cost_per_task_usd": null,
      "avg_duration_per_task_ms": 382535.98333333334,
      "avg_output_tokens_per_task": 6994.291666666667,
      "last_run_timestamp": "2026-08-04T05:36:23.551Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "SambaNova",
      "accuracy": 0.6844864918238994,
      "accuracy_std_dev": 0.03067807249818805,
      "run_count": 7,
      "total_questions": 795,
      "avg_cost_per_task_usd": 0.00916637610062893,
      "avg_duration_per_task_ms": 54426.96603773585,
      "avg_output_tokens_per_task": 5936.947169811321,
      "last_run_timestamp": "2026-10-24T16:45:45.392Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.6-sonnet-20260217",
      "model_slug": "anthropic/claude-sonnet-4.6",
      "model_name": "Anthropic: Claude Sonnet 4.6",
      "provider_name": "Google",
      "accuracy": 0.6844233356431357,
      "accuracy_std_dev": 0.1414237811051766,
      "run_count": 35,
      "total_questions": 6927,
      "avg_cost_per_task_usd": 0.6621923740435977,
      "avg_duration_per_task_ms": 602080.901255955,
      "avg_output_tokens_per_task": 41272.465136422696,
      "last_run_timestamp": "2026-10-24T08:43:07.831Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5-nano-2025-08-07",
      "model_slug": "openai/gpt-5-nano",
      "model_name": "OpenAI: GPT-5 Nano",
      "provider_name": "OpenAI",
      "accuracy": 0.6843435000000001,
      "accuracy_std_dev": 0.005892499999999967,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.01356670707070707,
      "avg_duration_per_task_ms": 229761.702020202,
      "avg_output_tokens_per_task": 33820.56313131313,
      "last_run_timestamp": "2026-10-01T19:28:34.261Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.6-sonnet-20260217",
      "model_slug": "anthropic/claude-sonnet-4.6",
      "model_name": "Anthropic: Claude Sonnet 4.6",
      "provider_name": "auto-routing",
      "accuracy": 0.6830809999999999,
      "accuracy_std_dev": 0.14005659745795623,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.6164722958754209,
      "avg_duration_per_task_ms": 635625.3400673401,
      "avg_output_tokens_per_task": 40941.11069023569,
      "last_run_timestamp": "2026-10-24T08:26:35.484Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.6-sonnet-20260217",
      "model_slug": "anthropic/claude-sonnet-4.6",
      "model_name": "Anthropic: Claude Sonnet 4.6",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.6825664722222222,
      "accuracy_std_dev": 0.136152835465126,
      "run_count": 36,
      "total_questions": 7128,
      "avg_cost_per_task_usd": 0.6585870970819304,
      "avg_duration_per_task_ms": 454569.8709315376,
      "avg_output_tokens_per_task": 40985.43560606061,
      "last_run_timestamp": "2026-10-24T08:24:10.735Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "SiliconFlow",
      "accuracy": 0.6807458630136987,
      "accuracy_std_dev": 0.019503295687310546,
      "run_count": 9,
      "total_questions": 1752,
      "avg_cost_per_task_usd": 0.0021889891552511415,
      "avg_duration_per_task_ms": 195459.59874429225,
      "avg_output_tokens_per_task": 4065.609589041096,
      "last_run_timestamp": "2026-10-24T16:39:13.265Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.6-sonnet-20260217",
      "model_slug": "anthropic/claude-sonnet-4.6",
      "model_name": "Anthropic: Claude Sonnet 4.6",
      "provider_name": "Azure",
      "accuracy": 0.6805556666666668,
      "accuracy_std_dev": 0.14593486187813,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.6194684494949494,
      "avg_duration_per_task_ms": 636905.944023569,
      "avg_output_tokens_per_task": 41202.27062289562,
      "last_run_timestamp": "2026-10-24T08:14:03.137Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "Darkbloom",
      "accuracy": 0.6805556666666668,
      "accuracy_std_dev": 0.06224365077839043,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.007968234006734005,
      "avg_duration_per_task_ms": 1481788.1506734006,
      "avg_output_tokens_per_task": 62936.50420875421,
      "last_run_timestamp": "2026-10-24T16:26:47.574Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "AkashML",
      "accuracy": 0.6797738037518039,
      "accuracy_std_dev": 0.015747258010206083,
      "run_count": 10,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.004640227272727273,
      "avg_duration_per_task_ms": 817645.9841269841,
      "avg_output_tokens_per_task": 46248.668831168834,
      "last_run_timestamp": "2026-10-24T16:05:13.521Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_slug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_name": "Qwen: Qwen3 VL 235B A22B Instruct",
      "provider_name": "Venice",
      "accuracy": 0.6748038333333333,
      "accuracy_std_dev": 0.04521916160735648,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.01522573063973064,
      "avg_duration_per_task_ms": 286760.11026936024,
      "avg_output_tokens_per_task": 8293.19696969697,
      "last_run_timestamp": "2026-10-03T17:36:48.437Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "Venice",
      "accuracy": 0.6746166405451448,
      "accuracy_std_dev": 0.03659361862233009,
      "run_count": 9,
      "total_questions": 1761,
      "avg_cost_per_task_usd": 0.027729125496876776,
      "avg_duration_per_task_ms": 864050.6172629189,
      "avg_output_tokens_per_task": 18372.465644520158,
      "last_run_timestamp": "2026-10-24T16:22:30.498Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Phala",
      "accuracy": 0.6735395649484536,
      "accuracy_std_dev": 0.02076995482325371,
      "run_count": 5,
      "total_questions": 485,
      "avg_cost_per_task_usd": 0.003951113402061856,
      "avg_duration_per_task_ms": 479082.6536082474,
      "avg_output_tokens_per_task": 26380.544329896908,
      "last_run_timestamp": "2026-08-29T17:24:11.557Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-chat-v3",
      "model_slug": "deepseek/deepseek-chat",
      "model_name": "DeepSeek: DeepSeek V3",
      "provider_name": "StreamLake",
      "accuracy": 0.6689113333333334,
      "accuracy_std_dev": 0.0067804647005611455,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0026409175084175086,
      "avg_duration_per_task_ms": 47935.90572390572,
      "avg_output_tokens_per_task": 2383.6902356902356,
      "last_run_timestamp": "2026-10-02T15:53:31.869Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.6-20251208",
      "model_slug": "z-ai/glm-4.6v",
      "model_name": "Z.ai: GLM 4.6V",
      "provider_name": "auto-routing",
      "accuracy": 0.66835,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.013946772727272728,
      "avg_duration_per_task_ms": 324889.0353535354,
      "avg_output_tokens_per_task": 15315.242424242424,
      "last_run_timestamp": "2026-10-01T08:41:30.639Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Venice",
      "accuracy": 0.66813975,
      "accuracy_std_dev": 0.0169491097609137,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.005768640151515151,
      "avg_duration_per_task_ms": 418288.6957070707,
      "avg_output_tokens_per_task": 15810.410353535353,
      "last_run_timestamp": "2026-10-24T16:11:51.309Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "OpenInference",
      "accuracy": 0.6662457499999999,
      "accuracy_std_dev": 0.017551141520923935,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.005718719696969698,
      "avg_duration_per_task_ms": 566951.7815656565,
      "avg_output_tokens_per_task": 16218.064393939394,
      "last_run_timestamp": "2026-08-29T18:40:21.422Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-4-maverick-17b-128e-instruct",
      "model_slug": "meta-llama/llama-4-maverick",
      "model_name": "Meta: Llama 4 Maverick",
      "provider_name": "DigitalOcean",
      "accuracy": 0.6661054999999999,
      "accuracy_std_dev": 0.006849761936252494,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.002126188552188552,
      "avg_duration_per_task_ms": 243791.89646464647,
      "avg_output_tokens_per_task": 3027.095117845118,
      "last_run_timestamp": "2026-10-05T20:44:48.611Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-4-maverick-17b-128e-instruct",
      "model_slug": "meta-llama/llama-4-maverick",
      "model_name": "Meta: Llama 4 Maverick",
      "provider_name": "auto-routing",
      "accuracy": 0.6656563999999999,
      "accuracy_std_dev": 0.00579280583482652,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.0027868222222222224,
      "avg_duration_per_task_ms": 87457.03737373737,
      "avg_output_tokens_per_task": 2951.019191919192,
      "last_run_timestamp": "2026-08-29T19:49:56.943Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 VL 30B A3B Instruct",
      "provider_name": "Alibaba",
      "accuracy": 0.6651514999999998,
      "accuracy_std_dev": 0.008019320822239275,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.005088861616161617,
      "avg_duration_per_task_ms": 104484.55757575757,
      "avg_output_tokens_per_task": 9583.160606060606,
      "last_run_timestamp": "2026-10-03T09:23:08.256Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "DeepInfra",
      "accuracy": 0.6649444988518944,
      "accuracy_std_dev": 0.01638003285131017,
      "run_count": 10,
      "total_questions": 1742,
      "avg_cost_per_task_usd": 0.006743279563719862,
      "avg_duration_per_task_ms": 475467.2680826636,
      "avg_output_tokens_per_task": 47990.47588978186,
      "last_run_timestamp": "2026-10-24T16:02:28.575Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "AtlasCloud",
      "accuracy": 0.6637718184210527,
      "accuracy_std_dev": 0.20207857374716895,
      "run_count": 10,
      "total_questions": 1900,
      "avg_cost_per_task_usd": 0.012164477368421052,
      "avg_duration_per_task_ms": 311613.41736842104,
      "avg_output_tokens_per_task": 14454.243684210527,
      "last_run_timestamp": "2026-10-17T08:17:57.529Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-chat-v3-0324",
      "model_slug": "deepseek/deepseek-chat-v3-0324",
      "model_name": "DeepSeek: DeepSeek V3 0324",
      "provider_name": "SiliconFlow",
      "accuracy": 0.66372075,
      "accuracy_std_dev": 0.00844903779654822,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.00248682702020202,
      "avg_duration_per_task_ms": 87766.55681818182,
      "avg_output_tokens_per_task": 2304.1906565656564,
      "last_run_timestamp": "2026-08-29T20:54:16.286Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.6-20251208",
      "model_slug": "z-ai/glm-4.6v",
      "model_name": "Z.ai: GLM 4.6V",
      "provider_name": "Novita",
      "accuracy": 0.6633000000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.013461267676767677,
      "avg_duration_per_task_ms": 327894.71717171714,
      "avg_output_tokens_per_task": 14738.742424242424,
      "last_run_timestamp": "2026-10-01T08:27:45.371Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "auto-routing",
      "accuracy": 0.6619902222222221,
      "accuracy_std_dev": 0.13296183010404633,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.014990915263748598,
      "avg_duration_per_task_ms": 410516.86419753084,
      "avg_output_tokens_per_task": 13130.794051627385,
      "last_run_timestamp": "2026-10-24T16:02:06.940Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-chat-v3",
      "model_slug": "deepseek/deepseek-chat",
      "model_name": "DeepSeek: DeepSeek V3",
      "provider_name": "Novita",
      "accuracy": 0.661616,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.003362974747474748,
      "avg_duration_per_task_ms": 65871.15656565657,
      "avg_output_tokens_per_task": 2362.121212121212,
      "last_run_timestamp": "2026-08-01T09:30:53.884Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4.1-2025-04-14",
      "model_slug": "openai/gpt-4.1",
      "model_name": "OpenAI: GPT-4.1",
      "provider_name": "OpenAI",
      "accuracy": 0.6607744999999999,
      "accuracy_std_dev": 0.011572094721786573,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.015565636363636365,
      "avg_duration_per_task_ms": 14159.460858585859,
      "avg_output_tokens_per_task": 1751.3977272727273,
      "last_run_timestamp": "2026-10-02T23:17:48.072Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Parasail",
      "accuracy": 0.6601731428571429,
      "accuracy_std_dev": 0.023160494075833857,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.008067838383838384,
      "avg_duration_per_task_ms": 808632.0974025974,
      "avg_output_tokens_per_task": 55153.242424242424,
      "last_run_timestamp": "2026-10-24T16:14:24.977Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.6-20251208",
      "model_slug": "z-ai/glm-4.6v",
      "model_name": "Z.ai: GLM 4.6V",
      "provider_name": "Z.AI",
      "accuracy": 0.659933,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.014984520202020202,
      "avg_duration_per_task_ms": 348987.1767676768,
      "avg_output_tokens_per_task": 16461.863636363636,
      "last_run_timestamp": "2026-10-01T08:28:22.621Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.6594651666666667,
      "accuracy_std_dev": 0.015979149081259345,
      "run_count": 18,
      "total_questions": 3564,
      "avg_cost_per_task_usd": 0.008953686026936029,
      "avg_duration_per_task_ms": 317080.76767676766,
      "avg_output_tokens_per_task": 59355.69332210999,
      "last_run_timestamp": "2026-10-24T16:22:30.492Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 VL 30B A3B Instruct",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.658249,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0066776414141414145,
      "avg_duration_per_task_ms": 101359.9898989899,
      "avg_output_tokens_per_task": 9307.358585858587,
      "last_run_timestamp": "2026-10-04T03:07:35.168Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "ibm-granite/granite-4.2-8b-20260831",
      "model_slug": "ibm-granite/granite-4.2-8b",
      "model_name": "IBM: Granite 4.2 8B",
      "provider_name": "auto-routing",
      "accuracy": 0.65736,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 197,
      "avg_cost_per_task_usd": 0.014864365482233502,
      "avg_duration_per_task_ms": 1717906.416243655,
      "avg_output_tokens_per_task": 64415.350253807104,
      "last_run_timestamp": "2026-10-01T08:48:38.304Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.6-sonnet-20260217",
      "model_slug": "anthropic/claude-sonnet-4.6",
      "model_name": "Anthropic: Claude Sonnet 4.6",
      "provider_name": "Anthropic",
      "accuracy": 0.6570467857142857,
      "accuracy_std_dev": 0.14261200486794257,
      "run_count": 14,
      "total_questions": 2772,
      "avg_cost_per_task_usd": 0.6177520173160173,
      "avg_duration_per_task_ms": 633109.20995671,
      "avg_output_tokens_per_task": 41054.18795093795,
      "last_run_timestamp": "2026-10-24T08:12:02.408Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "CoreWeave",
      "accuracy": 0.6567339999999999,
      "accuracy_std_dev": 0.02088421366487137,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.007585557575757577,
      "avg_duration_per_task_ms": 623991.9545454546,
      "avg_output_tokens_per_task": 58619.2005050505,
      "last_run_timestamp": "2026-10-24T16:15:47.411Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "DigitalOcean",
      "accuracy": 0.6552547006369427,
      "accuracy_std_dev": 0.08012703749983525,
      "run_count": 4,
      "total_questions": 628,
      "avg_cost_per_task_usd": 0.06530196178343949,
      "avg_duration_per_task_ms": 1671084.4410828026,
      "avg_output_tokens_per_task": 24249.23566878981,
      "last_run_timestamp": "2026-08-22T17:09:42.218Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Phala",
      "accuracy": 0.6532027662178702,
      "accuracy_std_dev": 0.21133166802603598,
      "run_count": 9,
      "total_questions": 1634,
      "avg_cost_per_task_usd": 0.008997689106487149,
      "avg_duration_per_task_ms": 186886.9828641371,
      "avg_output_tokens_per_task": 14831.725826193391,
      "last_run_timestamp": "2026-10-24T16:26:15.957Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "Venice",
      "accuracy": 0.6529368014874142,
      "accuracy_std_dev": 0.054041396139361146,
      "run_count": 9,
      "total_questions": 1748,
      "avg_cost_per_task_usd": 0.05780651773455378,
      "avg_duration_per_task_ms": 1127057.8415331808,
      "avg_output_tokens_per_task": 24934.28718535469,
      "last_run_timestamp": "2026-10-24T16:10:43.378Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "ibm-granite/granite-4.2-8b-20260831",
      "model_slug": "ibm-granite/granite-4.2-8b",
      "model_name": "IBM: Granite 4.2 8B",
      "provider_name": "DeepInfra",
      "accuracy": 0.6523569999999999,
      "accuracy_std_dev": 0.0025249999999999995,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.01652235606060606,
      "avg_duration_per_task_ms": 1917916.494949495,
      "avg_output_tokens_per_task": 65966.15909090909,
      "last_run_timestamp": "2026-10-09T00:15:37.020Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "NextBit",
      "accuracy": 0.651601,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 177,
      "avg_cost_per_task_usd": 0.03645816384180791,
      "avg_duration_per_task_ms": 1143037.5875706214,
      "avg_output_tokens_per_task": 29354.60451977401,
      "last_run_timestamp": "2026-08-01T23:27:37.975Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-r1-0528",
      "model_slug": "deepseek/deepseek-r1-0528",
      "model_name": "DeepSeek: R1 0528",
      "provider_name": "Novita",
      "accuracy": 0.651515,
      "accuracy_std_dev": 0.1905456270048375,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.058297676767676776,
      "avg_duration_per_task_ms": 1098090.5791245792,
      "avg_output_tokens_per_task": 23174.20202020202,
      "last_run_timestamp": "2026-10-17T16:03:44.041Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "DeepInfra",
      "accuracy": 0.6510548691983122,
      "accuracy_std_dev": 0.12013379799957409,
      "run_count": 9,
      "total_questions": 1185,
      "avg_cost_per_task_usd": 0.0951064641350211,
      "avg_duration_per_task_ms": 1230349.5139240506,
      "avg_output_tokens_per_task": 27086.752742616034,
      "last_run_timestamp": "2026-10-24T16:06:30.901Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-4-maverick-17b-128e-instruct",
      "model_slug": "meta-llama/llama-4-maverick",
      "model_name": "Meta: Llama 4 Maverick",
      "provider_name": "Parasail",
      "accuracy": 0.6505050000000001,
      "accuracy_std_dev": 0.008356827340564098,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.003142527272727273,
      "avg_duration_per_task_ms": 40777.49797979798,
      "avg_output_tokens_per_task": 2977.0545454545454,
      "last_run_timestamp": "2026-08-29T19:51:51.198Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4.1-mini-2025-04-14",
      "model_slug": "openai/gpt-4.1-mini",
      "model_name": "OpenAI: GPT-4.1 Mini",
      "provider_name": "auto-routing",
      "accuracy": 0.6492706666666667,
      "accuracy_std_dev": 0.011195005920300148,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.004075954545454546,
      "avg_duration_per_task_ms": 28668.510101010103,
      "avg_output_tokens_per_task": 2354.649831649832,
      "last_run_timestamp": "2026-10-01T19:12:31.553Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-chat-v3-0324",
      "model_slug": "deepseek/deepseek-chat-v3-0324",
      "model_name": "DeepSeek: DeepSeek V3 0324",
      "provider_name": "Crusoe",
      "accuracy": 0.6488281909758657,
      "accuracy_std_dev": 0.006868775133893942,
      "run_count": 5,
      "total_questions": 953,
      "avg_cost_per_task_usd": 0.003697750262329486,
      "avg_duration_per_task_ms": 59831.95383001049,
      "avg_output_tokens_per_task": 2242.5666316894017,
      "last_run_timestamp": "2026-08-29T16:10:45.295Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-4-maverick-17b-128e-instruct",
      "model_slug": "meta-llama/llama-4-maverick",
      "model_name": "Meta: Llama 4 Maverick",
      "provider_name": "DeepInfra",
      "accuracy": 0.6478116,
      "accuracy_std_dev": 0.01301437818107341,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.0023702696969696967,
      "avg_duration_per_task_ms": 65368.43434343435,
      "avg_output_tokens_per_task": 2772.5838383838386,
      "last_run_timestamp": "2026-08-29T16:11:23.388Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "ibm-granite/granite-4.2-8b-20260831",
      "model_slug": "ibm-granite/granite-4.2-8b",
      "model_name": "IBM: Granite 4.2 8B",
      "provider_name": "CoreWeave",
      "accuracy": 0.647493,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 113,
      "avg_cost_per_task_usd": 0.005189893805309735,
      "avg_duration_per_task_ms": 619234.2654867256,
      "avg_output_tokens_per_task": 34365.4778761062,
      "last_run_timestamp": "2026-10-01T08:28:32.911Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4.1-mini-2025-04-14",
      "model_slug": "openai/gpt-4.1-mini",
      "model_name": "OpenAI: GPT-4.1 Mini",
      "provider_name": "OpenAI",
      "accuracy": 0.6473065000000001,
      "accuracy_std_dev": 0.009707729665065862,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.00413114393939394,
      "avg_duration_per_task_ms": 27377.72601010101,
      "avg_output_tokens_per_task": 2388.9305555555557,
      "last_run_timestamp": "2026-10-02T23:15:53.868Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "Parasail",
      "accuracy": 0.6470586811145511,
      "accuracy_std_dev": 0.06899576470516142,
      "run_count": 4,
      "total_questions": 646,
      "avg_cost_per_task_usd": 0.006973526315789473,
      "avg_duration_per_task_ms": 477220.6222910217,
      "avg_output_tokens_per_task": 36135.41950464396,
      "last_run_timestamp": "2026-08-29T12:30:09.617Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-chat-v3-0324",
      "model_slug": "deepseek/deepseek-chat-v3-0324",
      "model_name": "DeepSeek: DeepSeek V3 0324",
      "provider_name": "GMICloud",
      "accuracy": 0.646465,
      "accuracy_std_dev": 0.004761657064510171,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0029047154882154884,
      "avg_duration_per_task_ms": 53981.06734006734,
      "avg_output_tokens_per_task": 2363.939393939394,
      "last_run_timestamp": "2026-08-29T16:33:11.860Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4.1-2025-04-14",
      "model_slug": "openai/gpt-4.1",
      "model_name": "OpenAI: GPT-4.1",
      "provider_name": "auto-routing",
      "accuracy": 0.6464646666666667,
      "accuracy_std_dev": 0.013537798302858865,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.015368191919191919,
      "avg_duration_per_task_ms": 14060.520202020201,
      "avg_output_tokens_per_task": 1727.3535353535353,
      "last_run_timestamp": "2026-10-01T19:12:45.033Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-4-maverick-17b-128e-instruct",
      "model_slug": "meta-llama/llama-4-maverick",
      "model_name": "Meta: Llama 4 Maverick",
      "provider_name": "Novita",
      "accuracy": 0.6454544,
      "accuracy_std_dev": 0.002519632084253531,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.002477851515151515,
      "avg_duration_per_task_ms": 59827.33737373738,
      "avg_output_tokens_per_task": 2673.8050505050505,
      "last_run_timestamp": "2026-08-29T17:03:33.297Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "auto-routing",
      "accuracy": 0.6451553333333333,
      "accuracy_std_dev": 0.022132801022715396,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.007395829966329966,
      "avg_duration_per_task_ms": 612621.9057239058,
      "avg_output_tokens_per_task": 44969.978114478115,
      "last_run_timestamp": "2026-10-24T16:13:05.627Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "Mancer 2",
      "accuracy": 0.6439055298507463,
      "accuracy_std_dev": 0.20314360210271282,
      "run_count": 7,
      "total_questions": 1340,
      "avg_cost_per_task_usd": 0.07648400820895522,
      "avg_duration_per_task_ms": 1782783.4514925373,
      "avg_output_tokens_per_task": 36935.7671641791,
      "last_run_timestamp": "2026-10-24T16:17:39.031Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4.1-2025-04-14",
      "model_slug": "openai/gpt-4.1",
      "model_name": "OpenAI: GPT-4.1",
      "provider_name": "Azure",
      "accuracy": 0.643659,
      "accuracy_std_dev": 0.015752669318774726,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.015761657407407408,
      "avg_duration_per_task_ms": 18961.11531986532,
      "avg_output_tokens_per_task": 1683.8872053872053,
      "last_run_timestamp": "2026-10-01T19:14:40.644Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "auto-routing",
      "accuracy": 0.6428171666666667,
      "accuracy_std_dev": 0.012039053318494584,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.002693207070707071,
      "avg_duration_per_task_ms": 142395.0521885522,
      "avg_output_tokens_per_task": 9633.525252525253,
      "last_run_timestamp": "2026-10-03T17:16:11.892Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "Alibaba",
      "accuracy": 0.6419754999999999,
      "accuracy_std_dev": 0.013802963036367705,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.004449612794612795,
      "avg_duration_per_task_ms": 74990.68265993266,
      "avg_output_tokens_per_task": 8364.945286195287,
      "last_run_timestamp": "2026-10-03T16:51:25.541Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 VL 30B A3B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.6419754444444444,
      "accuracy_std_dev": 0.014018014458475362,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.007092590347923681,
      "avg_duration_per_task_ms": 241593.06509539843,
      "avg_output_tokens_per_task": 11377.103254769921,
      "last_run_timestamp": "2026-10-03T10:20:18.247Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "Ambient",
      "accuracy": 0.641414,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.00284089898989899,
      "avg_duration_per_task_ms": 209352.64646464647,
      "avg_output_tokens_per_task": 9266.040404040405,
      "last_run_timestamp": "2026-10-09T15:00:06.492Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4.1-mini-2025-04-14",
      "model_slug": "openai/gpt-4.1-mini",
      "model_name": "OpenAI: GPT-4.1 Mini",
      "provider_name": "Azure",
      "accuracy": 0.6405723333333334,
      "accuracy_std_dev": 0.0030349496792460334,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.004210166666666667,
      "avg_duration_per_task_ms": 30431.28367003367,
      "avg_output_tokens_per_task": 2308.9638047138046,
      "last_run_timestamp": "2026-10-01T19:13:29.560Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "Darkbloom",
      "accuracy": 0.6399558039735098,
      "accuracy_std_dev": 0.01859859452183148,
      "run_count": 8,
      "total_questions": 1510,
      "avg_cost_per_task_usd": 0.005625637086092715,
      "avg_duration_per_task_ms": 835153.6013245033,
      "avg_output_tokens_per_task": 34992.65165562914,
      "last_run_timestamp": "2026-10-24T08:14:51.111Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "SiliconFlow",
      "accuracy": 0.6368495215189873,
      "accuracy_std_dev": 0.014462691735980549,
      "run_count": 6,
      "total_questions": 1185,
      "avg_cost_per_task_usd": 0.0026009772151898737,
      "avg_duration_per_task_ms": 238051.80337552744,
      "avg_output_tokens_per_task": 8445.996624472575,
      "last_run_timestamp": "2026-10-03T17:03:50.406Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Darkbloom",
      "accuracy": 0.6353816666666667,
      "accuracy_std_dev": 0.011788772294951762,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.0010992390572390571,
      "avg_duration_per_task_ms": 380904.83164983167,
      "avg_output_tokens_per_task": 11589.25,
      "last_run_timestamp": "2026-10-24T16:24:45.215Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-32b-04-28",
      "model_slug": "qwen/qwen3-32b",
      "model_name": "Qwen: Qwen3 32B",
      "provider_name": "Groq",
      "accuracy": 0.6342374433962265,
      "accuracy_std_dev": 0.011539125582122754,
      "run_count": 5,
      "total_questions": 848,
      "avg_cost_per_task_usd": 0.007657488207547171,
      "avg_duration_per_task_ms": 34742.110849056604,
      "avg_output_tokens_per_task": 12687.265330188678,
      "last_run_timestamp": "2026-08-29T16:23:45.123Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "StreamLake",
      "accuracy": 0.6334780000000001,
      "accuracy_std_dev": 0.013155152146592608,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.0016263744588744587,
      "avg_duration_per_task_ms": 75470.49134199134,
      "avg_output_tokens_per_task": 8221.989177489177,
      "last_run_timestamp": "2026-10-03T16:47:32.559Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3-nano-30b-a3b",
      "model_slug": "nvidia/nemotron-3-nano-30b-a3b",
      "model_name": "NVIDIA: Nemotron 3 Nano 30B A3B",
      "provider_name": "Crusoe",
      "accuracy": 0.6325426993295513,
      "accuracy_std_dev": 0.010534965676735725,
      "run_count": 10,
      "total_questions": 1939,
      "avg_cost_per_task_usd": 0.01766013512119649,
      "avg_duration_per_task_ms": 478451.96183599794,
      "avg_output_tokens_per_task": 88282.73388344508,
      "last_run_timestamp": "2026-10-03T17:08:02.072Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "inclusionai/ling-3.0-flash-fin-20260827",
      "model_slug": "inclusionai/ling-3.0-flash-fin",
      "model_name": "inclusionAI: Ling 3.0 Flash Fin",
      "provider_name": "Novita",
      "accuracy": 0.632155,
      "accuracy_std_dev": 0.005892000000000008,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.005966434343434343,
      "avg_duration_per_task_ms": 235473.01515151514,
      "avg_output_tokens_per_task": 48249.11363636364,
      "last_run_timestamp": "2026-10-01T08:24:52.614Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Groq",
      "accuracy": 0.6310882796803652,
      "accuracy_std_dev": 0.015991023651789744,
      "run_count": 9,
      "total_questions": 1752,
      "avg_cost_per_task_usd": 0.01677861757990868,
      "avg_duration_per_task_ms": 87974.94634703196,
      "avg_output_tokens_per_task": 55697.56792237443,
      "last_run_timestamp": "2026-10-24T16:45:19.455Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "CoreWeave",
      "accuracy": 0.6308680607902736,
      "accuracy_std_dev": 0.011456259937693471,
      "run_count": 5,
      "total_questions": 987,
      "avg_cost_per_task_usd": 0.0034108257345491387,
      "avg_duration_per_task_ms": 140057.4062816616,
      "avg_output_tokens_per_task": 11472.941236068895,
      "last_run_timestamp": "2026-08-29T18:20:19.448Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "DekaLLM",
      "accuracy": 0.6306148156028369,
      "accuracy_std_dev": 0.05281625698168657,
      "run_count": 3,
      "total_questions": 282,
      "avg_cost_per_task_usd": 0.0037584255319148938,
      "avg_duration_per_task_ms": 208452.59574468085,
      "avg_output_tokens_per_task": 12308.468085106382,
      "last_run_timestamp": "2026-10-23T09:29:14.171Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 VL 30B A3B Instruct",
      "provider_name": "Novita",
      "accuracy": 0.6299665,
      "accuracy_std_dev": 0.033651413370763474,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.0068110621212121225,
      "avg_duration_per_task_ms": 109012.8005050505,
      "avg_output_tokens_per_task": 9497.993434343434,
      "last_run_timestamp": "2026-10-03T10:32:36.729Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "Venice",
      "accuracy": 0.6292411856287424,
      "accuracy_std_dev": 0.0117815,
      "run_count": 2,
      "total_questions": 334,
      "avg_cost_per_task_usd": 0.0071094161676646716,
      "avg_duration_per_task_ms": 112258.81736526947,
      "avg_output_tokens_per_task": 28232.077844311378,
      "last_run_timestamp": "2026-08-15T08:02:44.624Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "Nebius",
      "accuracy": 0.6289855826086956,
      "accuracy_std_dev": 0.01629778854411147,
      "run_count": 7,
      "total_questions": 1380,
      "avg_cost_per_task_usd": 0.0038922746376811594,
      "avg_duration_per_task_ms": 225279.24782608697,
      "avg_output_tokens_per_task": 12705.184057971015,
      "last_run_timestamp": "2026-10-23T08:08:11.470Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3-nano-30b-a3b",
      "model_slug": "nvidia/nemotron-3-nano-30b-a3b",
      "model_name": "NVIDIA: Nemotron 3 Nano 30B A3B",
      "provider_name": "auto-routing",
      "accuracy": 0.6285795064570466,
      "accuracy_std_dev": 0.020919173358428876,
      "run_count": 9,
      "total_questions": 1781,
      "avg_cost_per_task_usd": 0.01718929421673217,
      "avg_duration_per_task_ms": 631165.4817518248,
      "avg_output_tokens_per_task": 83912.1010668164,
      "last_run_timestamp": "2026-10-03T17:40:51.448Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "SiliconFlow",
      "accuracy": 0.6280247450759707,
      "accuracy_std_dev": 0.010380266086410858,
      "run_count": 9,
      "total_questions": 1777,
      "avg_cost_per_task_usd": 0.0018582673044456948,
      "avg_duration_per_task_ms": 228947.18570624647,
      "avg_output_tokens_per_task": 10187.530106921778,
      "last_run_timestamp": "2026-10-24T16:03:25.396Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3-nano-30b-a3b",
      "model_slug": "nvidia/nemotron-3-nano-30b-a3b",
      "model_name": "NVIDIA: Nemotron 3 Nano 30B A3B",
      "provider_name": "Nebius",
      "accuracy": 0.627104375,
      "accuracy_std_dev": 0.008333089267155095,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.024497946969696968,
      "avg_duration_per_task_ms": 1060409.1414141415,
      "avg_output_tokens_per_task": 101874.34027777778,
      "last_run_timestamp": "2026-10-23T09:03:34.804Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-chat-v3",
      "model_slug": "deepseek/deepseek-chat",
      "model_name": "DeepSeek: DeepSeek V3",
      "provider_name": "auto-routing",
      "accuracy": 0.6212125000000001,
      "accuracy_std_dev": 0.011784500000000031,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.0023548611111111114,
      "avg_duration_per_task_ms": 64938.90151515151,
      "avg_output_tokens_per_task": 2165.318181818182,
      "last_run_timestamp": "2026-10-01T19:21:21.945Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.621212,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.00552190909090909,
      "avg_duration_per_task_ms": 174055.64141414143,
      "avg_output_tokens_per_task": 10415.944444444445,
      "last_run_timestamp": "2026-10-04T03:09:40.101Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Fireworks",
      "accuracy": 0.6182949425287356,
      "accuracy_std_dev": 0.0013475000000000015,
      "run_count": 2,
      "total_questions": 348,
      "avg_cost_per_task_usd": 0.011105244252873564,
      "avg_duration_per_task_ms": 345775.4454022989,
      "avg_output_tokens_per_task": 38606.68103448276,
      "last_run_timestamp": "2026-08-22T17:46:40.285Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 VL 30B A3B Instruct",
      "provider_name": "SiliconFlow",
      "accuracy": 0.6170940030769231,
      "accuracy_std_dev": 0.014320214565431623,
      "run_count": 5,
      "total_questions": 975,
      "avg_cost_per_task_usd": 0.015717948717948717,
      "avg_duration_per_task_ms": 780521.6543589743,
      "avg_output_tokens_per_task": 16712.84717948718,
      "last_run_timestamp": "2026-10-03T11:06:38.068Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-14b-04-28",
      "model_slug": "qwen/qwen3-14b",
      "model_name": "Qwen: Qwen3 14B",
      "provider_name": "DeepInfra",
      "accuracy": 0.6128760749682338,
      "accuracy_std_dev": 0.007225322047320813,
      "run_count": 4,
      "total_questions": 787,
      "avg_cost_per_task_usd": 0.0033133367217280814,
      "avg_duration_per_task_ms": 404332.3011435832,
      "avg_output_tokens_per_task": 13398.312579415502,
      "last_run_timestamp": "2026-10-08T22:16:42.056Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Google",
      "accuracy": 0.6121724426966293,
      "accuracy_std_dev": 0.10417909241135824,
      "run_count": 9,
      "total_questions": 1780,
      "avg_cost_per_task_usd": 0.009812219101123598,
      "avg_duration_per_task_ms": 327793.9264044944,
      "avg_output_tokens_per_task": 27904.43820224719,
      "last_run_timestamp": "2026-10-24T16:05:40.119Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "Novita",
      "accuracy": 0.6099888333333333,
      "accuracy_std_dev": 0.01106802965627677,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.003627024410774411,
      "avg_duration_per_task_ms": 33220.15151515151,
      "avg_output_tokens_per_task": 2140.8367003367002,
      "last_run_timestamp": "2026-10-03T17:03:29.436Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "Google",
      "accuracy": 0.6099860008438818,
      "accuracy_std_dev": 0.004922906311880752,
      "run_count": 6,
      "total_questions": 1185,
      "avg_cost_per_task_usd": 0.0040446168776371305,
      "avg_duration_per_task_ms": 27156.491983122363,
      "avg_output_tokens_per_task": 2148.8616033755275,
      "last_run_timestamp": "2026-10-03T17:03:50.598Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-30b-a3b-04-28",
      "model_slug": "qwen/qwen3-30b-a3b",
      "model_name": "Qwen: Qwen3 30B A3B",
      "provider_name": "auto-routing",
      "accuracy": 0.609428,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.007801348484848485,
      "avg_duration_per_task_ms": 208074.1717171717,
      "avg_output_tokens_per_task": 15081.5,
      "last_run_timestamp": "2026-08-01T18:08:49.019Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-30b-a3b-04-28",
      "model_slug": "qwen/qwen3-30b-a3b",
      "model_name": "Qwen: Qwen3 30B A3B",
      "provider_name": "DeepInfra",
      "accuracy": 0.608586,
      "accuracy_std_dev": 0.0008420000000000094,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.007727368686868687,
      "avg_duration_per_task_ms": 263769.8611111111,
      "avg_output_tokens_per_task": 15259.444444444445,
      "last_run_timestamp": "2026-10-08T22:57:53.933Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "Darkbloom",
      "accuracy": 0.6080220220713072,
      "accuracy_std_dev": 0.18476388731221863,
      "run_count": 12,
      "total_questions": 2356,
      "avg_cost_per_task_usd": 0.0260303675721562,
      "avg_duration_per_task_ms": 717454.6184210526,
      "avg_output_tokens_per_task": 36816.12648556876,
      "last_run_timestamp": "2026-10-24T16:13:05.631Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "CoreWeave",
      "accuracy": 0.607744,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.003953611111111112,
      "avg_duration_per_task_ms": 57224.242424242424,
      "avg_output_tokens_per_task": 2094.1161616161617,
      "last_run_timestamp": "2026-08-01T17:44:33.510Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "DeepInfra",
      "accuracy": 0.607708238128011,
      "accuracy_std_dev": 0.09072663406949609,
      "run_count": 9,
      "total_questions": 1453,
      "avg_cost_per_task_usd": 0.05455010392291811,
      "avg_duration_per_task_ms": 329922.3234686855,
      "avg_output_tokens_per_task": 20685.869924294562,
      "last_run_timestamp": "2026-10-24T16:28:50.076Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-32b-04-28",
      "model_slug": "qwen/qwen3-32b",
      "model_name": "Qwen: Qwen3 32B",
      "provider_name": "auto-routing",
      "accuracy": 0.6062142522796352,
      "accuracy_std_dev": 0.009864001918085761,
      "run_count": 5,
      "total_questions": 987,
      "avg_cost_per_task_usd": 0.008205017223910841,
      "avg_duration_per_task_ms": 297243.1023302938,
      "avg_output_tokens_per_task": 17225.689969604864,
      "last_run_timestamp": "2026-08-29T19:52:22.642Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "Mara",
      "accuracy": 0.6061323018867925,
      "accuracy_std_dev": 0.029270000000000018,
      "run_count": 2,
      "total_questions": 212,
      "avg_cost_per_task_usd": 0.034763094339622644,
      "avg_duration_per_task_ms": 71660.68867924529,
      "avg_output_tokens_per_task": 7461.462264150943,
      "last_run_timestamp": "2026-10-17T16:06:00.921Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 VL 30B A3B Instruct",
      "provider_name": "DeepInfra",
      "accuracy": 0.6059645129698252,
      "accuracy_std_dev": 0.012733602453351516,
      "run_count": 10,
      "total_questions": 1889,
      "avg_cost_per_task_usd": 0.008509499205929062,
      "avg_duration_per_task_ms": 554697.4531498147,
      "avg_output_tokens_per_task": 13980.221281101112,
      "last_run_timestamp": "2026-10-08T22:50:36.237Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-32b-04-28",
      "model_slug": "qwen/qwen3-32b",
      "model_name": "Qwen: Qwen3 32B",
      "provider_name": "Nebius",
      "accuracy": 0.6050506,
      "accuracy_std_dev": 0.015451550117706637,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.005770865656565657,
      "avg_duration_per_task_ms": 636718.6202020202,
      "avg_output_tokens_per_task": 19268.987878787877,
      "last_run_timestamp": "2026-08-29T19:06:01.882Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-30b-a3b-04-28",
      "model_slug": "qwen/qwen3-30b-a3b",
      "model_name": "Qwen: Qwen3 30B A3B",
      "provider_name": "Alibaba",
      "accuracy": 0.604377,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.00791619191919192,
      "avg_duration_per_task_ms": 146373.4797979798,
      "avg_output_tokens_per_task": 15059.80303030303,
      "last_run_timestamp": "2026-08-01T17:51:29.206Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "DeepInfra",
      "accuracy": 0.6042256518448439,
      "accuracy_std_dev": 0.02010368015853481,
      "run_count": 6,
      "total_questions": 1057,
      "avg_cost_per_task_usd": 0.0021907710501419112,
      "avg_duration_per_task_ms": 51585.36707663198,
      "avg_output_tokens_per_task": 2058.464522232734,
      "last_run_timestamp": "2026-10-08T23:15:07.689Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "auto-routing",
      "accuracy": 0.6038161666666666,
      "accuracy_std_dev": 0.007251665174672309,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.0035006599326599325,
      "avg_duration_per_task_ms": 56534.7601010101,
      "avg_output_tokens_per_task": 2119.1574074074074,
      "last_run_timestamp": "2026-10-03T16:57:23.266Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "Alibaba",
      "accuracy": 0.6035354999999999,
      "accuracy_std_dev": 0.0052117790868377945,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.010990611952861952,
      "avg_duration_per_task_ms": 32954.444444444445,
      "avg_output_tokens_per_task": 2093.7045454545455,
      "last_run_timestamp": "2026-10-03T17:43:44.972Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.5-air",
      "model_slug": "z-ai/glm-4.5-air",
      "model_name": "Z.ai: GLM 4.5 Air",
      "provider_name": "SiliconFlow",
      "accuracy": 0.602339,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 171,
      "avg_cost_per_task_usd": 0.0316934269005848,
      "avg_duration_per_task_ms": 1131594.5263157894,
      "avg_output_tokens_per_task": 37899.070175438595,
      "last_run_timestamp": "2026-08-01T17:59:38.872Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "Darkbloom",
      "accuracy": 0.6017236343772759,
      "accuracy_std_dev": 0.07577236111713366,
      "run_count": 7,
      "total_questions": 1373,
      "avg_cost_per_task_usd": 0.012778518572469047,
      "avg_duration_per_task_ms": 1389201.3306627823,
      "avg_output_tokens_per_task": 56525.64020393299,
      "last_run_timestamp": "2026-10-24T16:33:52.420Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "DigitalOcean",
      "accuracy": 0.6016879113924051,
      "accuracy_std_dev": 0.05594315696758628,
      "run_count": 4,
      "total_questions": 790,
      "avg_cost_per_task_usd": 0.033461643037974685,
      "avg_duration_per_task_ms": 2683325.098734177,
      "avg_output_tokens_per_task": 25317.36075949367,
      "last_run_timestamp": "2026-08-22T18:19:05.094Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Nebius",
      "accuracy": 0.599327,
      "accuracy_std_dev": 0.0033669999999999534,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.014991068181818182,
      "avg_duration_per_task_ms": 284433.01515151514,
      "avg_output_tokens_per_task": 12302.075757575758,
      "last_run_timestamp": "2026-08-15T15:26:03.892Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.599327,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.012440297979797978,
      "avg_duration_per_task_ms": 230792.08080808082,
      "avg_output_tokens_per_task": 41242.71717171717,
      "last_run_timestamp": "2026-10-04T03:37:16.537Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-14b-04-28",
      "model_slug": "qwen/qwen3-14b",
      "model_name": "Qwen: Qwen3 14B",
      "provider_name": "NextBit",
      "accuracy": 0.5973626666666666,
      "accuracy_std_dev": 0.014397638598819675,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.003072445286195286,
      "avg_duration_per_task_ms": 655562.0867003367,
      "avg_output_tokens_per_task": 13595.005050505051,
      "last_run_timestamp": "2026-10-07T13:51:26.485Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "Venice",
      "accuracy": 0.5962405000000001,
      "accuracy_std_dev": 0.005702057428390789,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.003268709595959596,
      "avg_duration_per_task_ms": 65099.85437710438,
      "avg_output_tokens_per_task": 2080.245791245791,
      "last_run_timestamp": "2026-10-03T16:59:44.383Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "DeepInfra",
      "accuracy": 0.5948373333333333,
      "accuracy_std_dev": 0.010316724200162651,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.04306378282828283,
      "avg_duration_per_task_ms": 401108.6296296296,
      "avg_output_tokens_per_task": 21434.68181818182,
      "last_run_timestamp": "2026-08-22T10:29:25.897Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3-nano-30b-a3b",
      "model_slug": "nvidia/nemotron-3-nano-30b-a3b",
      "model_name": "NVIDIA: Nemotron 3 Nano 30B A3B",
      "provider_name": "Novita",
      "accuracy": 0.5942281016850668,
      "accuracy_std_dev": 0.00965016140081236,
      "run_count": 9,
      "total_questions": 1721,
      "avg_cost_per_task_usd": 0.009055742010459035,
      "avg_duration_per_task_ms": 523063.0418361418,
      "avg_output_tokens_per_task": 45085.088901801275,
      "last_run_timestamp": "2026-10-03T17:04:03.802Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-2.5-flash-lite",
      "model_slug": "google/gemini-2.5-flash-lite",
      "model_name": "Google: Gemini 2.5 Flash Lite",
      "provider_name": "Google AI Studio",
      "accuracy": 0.591751,
      "accuracy_std_dev": 0.006627755628164546,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.017204867003367003,
      "avg_duration_per_task_ms": 118615.4696969697,
      "avg_output_tokens_per_task": 42873.59259259259,
      "last_run_timestamp": "2026-10-01T19:14:48.696Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-14b-04-28",
      "model_slug": "qwen/qwen3-14b",
      "model_name": "Qwen: Qwen3 14B",
      "provider_name": "auto-routing",
      "accuracy": 0.591751,
      "accuracy_std_dev": 0.0025249999999999995,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.0029208080808080807,
      "avg_duration_per_task_ms": 521963.51515151514,
      "avg_output_tokens_per_task": 11688.588383838383,
      "last_run_timestamp": "2026-10-01T16:32:22.405Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Novita",
      "accuracy": 0.5906884318766067,
      "accuracy_std_dev": 0.010206288841526426,
      "run_count": 6,
      "total_questions": 1167,
      "avg_cost_per_task_usd": 0.006399965724078834,
      "avg_duration_per_task_ms": 683293.4078834619,
      "avg_output_tokens_per_task": 42407.72150814053,
      "last_run_timestamp": "2026-10-03T16:12:07.492Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Cerebras",
      "accuracy": 0.5905352222222223,
      "accuracy_std_dev": 0.008575970638654383,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.010281984287317622,
      "avg_duration_per_task_ms": 11349.746913580248,
      "avg_output_tokens_per_task": 13254.841189674524,
      "last_run_timestamp": "2026-10-24T16:20:41.333Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "CoreWeave",
      "accuracy": 0.5872781331360947,
      "accuracy_std_dev": 0.004687499999999956,
      "run_count": 2,
      "total_questions": 338,
      "avg_cost_per_task_usd": 0.09507673668639052,
      "avg_duration_per_task_ms": 191708.50591715975,
      "avg_output_tokens_per_task": 21455.86094674556,
      "last_run_timestamp": "2026-08-01T18:26:20.700Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.7-flash-20260119",
      "model_slug": "z-ai/glm-4.7-flash",
      "model_name": "Z.ai: GLM 4.7 Flash",
      "provider_name": "Cloudflare",
      "accuracy": 0.5841753333333334,
      "accuracy_std_dev": 0.033949349496304376,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.02777477777777778,
      "avg_duration_per_task_ms": 1412449.8375420875,
      "avg_output_tokens_per_task": 69325.78367003366,
      "last_run_timestamp": "2026-10-13T21:28:41.390Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.7-flash-20260119",
      "model_slug": "z-ai/glm-4.7-flash",
      "model_name": "Z.ai: GLM 4.7 Flash",
      "provider_name": "Novita",
      "accuracy": 0.5816397652439025,
      "accuracy_std_dev": 0.007030422274657478,
      "run_count": 5,
      "total_questions": 984,
      "avg_cost_per_task_usd": 0.025769568089430896,
      "avg_duration_per_task_ms": 1126074.3231707318,
      "avg_output_tokens_per_task": 67814.54166666667,
      "last_run_timestamp": "2026-08-30T00:32:50.853Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Google",
      "accuracy": 0.5774414999999999,
      "accuracy_std_dev": 0.005050500000000013,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.010289136363636365,
      "avg_duration_per_task_ms": 331432.74242424243,
      "avg_output_tokens_per_task": 40883.856060606064,
      "last_run_timestamp": "2026-10-17T16:06:05.579Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.5-air",
      "model_slug": "z-ai/glm-4.5-air",
      "model_name": "Z.ai: GLM 4.5 Air",
      "provider_name": "auto-routing",
      "accuracy": 0.575269,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 186,
      "avg_cost_per_task_usd": 0.03680247311827957,
      "avg_duration_per_task_ms": 1058115.6344086023,
      "avg_output_tokens_per_task": 47809.01612903226,
      "last_run_timestamp": "2026-08-01T18:25:56.196Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "SambaNova",
      "accuracy": 0.5710451276332094,
      "accuracy_std_dev": 0.03546539526778179,
      "run_count": 8,
      "total_questions": 807,
      "avg_cost_per_task_usd": 0.03254549814126394,
      "avg_duration_per_task_ms": 62701.02726146221,
      "avg_output_tokens_per_task": 6973.700123915737,
      "last_run_timestamp": "2026-10-24T16:01:29.840Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "DigitalOcean",
      "accuracy": 0.5683499999999999,
      "accuracy_std_dev": 0.1383388370104361,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.0030159525252525248,
      "avg_duration_per_task_ms": 227529.94494949494,
      "avg_output_tokens_per_task": 7208.462626262626,
      "last_run_timestamp": "2026-10-24T16:03:25.478Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "DeepInfra",
      "accuracy": 0.5621250282776349,
      "accuracy_std_dev": 0.06329693423460571,
      "run_count": 4,
      "total_questions": 389,
      "avg_cost_per_task_usd": 0.04400786118251928,
      "avg_duration_per_task_ms": 513471.6503856041,
      "avg_output_tokens_per_task": 21046.47557840617,
      "last_run_timestamp": "2026-08-29T19:07:49.385Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-32b-04-28",
      "model_slug": "qwen/qwen3-32b",
      "model_name": "Qwen: Qwen3 32B",
      "provider_name": "DeepInfra",
      "accuracy": 0.5572391666666667,
      "accuracy_std_dev": 0.009959776661094785,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.004621723905723906,
      "avg_duration_per_task_ms": 448955.2752525252,
      "avg_output_tokens_per_task": 16279.704545454546,
      "last_run_timestamp": "2026-10-08T23:30:03.872Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-32b-04-28",
      "model_slug": "qwen/qwen3-32b",
      "model_name": "Qwen: Qwen3 32B",
      "provider_name": "Tenstorrent",
      "accuracy": 0.5512932188317349,
      "accuracy_std_dev": 0.02565312085410966,
      "run_count": 6,
      "total_questions": 1147,
      "avg_cost_per_task_usd": 0.005740845684394072,
      "avg_duration_per_task_ms": 730028.2885789014,
      "avg_output_tokens_per_task": 14883.85701830863,
      "last_run_timestamp": "2026-10-13T09:46:38.737Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-2.5-flash-lite",
      "model_slug": "google/gemini-2.5-flash-lite",
      "model_name": "Google: Gemini 2.5 Flash Lite",
      "provider_name": "Google",
      "accuracy": 0.5505053333333333,
      "accuracy_std_dev": 0.016890890559378122,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.017397131313131312,
      "avg_duration_per_task_ms": 129348.1271043771,
      "avg_output_tokens_per_task": 43400.656565656565,
      "last_run_timestamp": "2026-10-01T19:27:01.395Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-2.5-flash-lite",
      "model_slug": "google/gemini-2.5-flash-lite",
      "model_name": "Google: Gemini 2.5 Flash Lite",
      "provider_name": "auto-routing",
      "accuracy": 0.5488216666666668,
      "accuracy_std_dev": 0.006298902038362641,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0171891531986532,
      "avg_duration_per_task_ms": 121387.12121212122,
      "avg_output_tokens_per_task": 42892.78114478115,
      "last_run_timestamp": "2026-10-01T19:15:36.386Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4o-2024-05-13",
      "model_slug": "openai/gpt-4o-2024-05-13",
      "model_name": "OpenAI: GPT-4o (2024-05-13)",
      "provider_name": "OpenAI",
      "accuracy": 0.542088,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.023943484848484846,
      "avg_duration_per_task_ms": 10793.646464646465,
      "avg_output_tokens_per_task": 1334.611111111111,
      "last_run_timestamp": "2026-08-01T08:53:34.329Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-vl-8b-instruct",
      "model_slug": "qwen/qwen3-vl-8b-instruct",
      "model_name": "Qwen: Qwen3 VL 8B Instruct",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.538721,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.005473,
      "avg_duration_per_task_ms": 117831.5808080808,
      "avg_output_tokens_per_task": 11819.661616161617,
      "last_run_timestamp": "2026-10-04T03:06:12.419Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-chat-v3",
      "model_slug": "deepseek/deepseek-chat",
      "model_name": "DeepSeek: DeepSeek V3",
      "provider_name": "DeepInfra",
      "accuracy": 0.535985,
      "accuracy_std_dev": 0.011835000612589788,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.0018682651515151514,
      "avg_duration_per_task_ms": 92043.92676767676,
      "avg_output_tokens_per_task": 1839.9911616161617,
      "last_run_timestamp": "2026-10-08T22:10:22.349Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-vl-8b-instruct",
      "model_slug": "qwen/qwen3-vl-8b-instruct",
      "model_name": "Qwen: Qwen3 VL 8B Instruct",
      "provider_name": "Alibaba",
      "accuracy": 0.53367,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.00486480303030303,
      "avg_duration_per_task_ms": 109732.54545454546,
      "avg_output_tokens_per_task": 10482.964646464647,
      "last_run_timestamp": "2026-08-01T17:48:17.532Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-14b-04-28",
      "model_slug": "qwen/qwen3-14b",
      "model_name": "Qwen: Qwen3 14B",
      "provider_name": "Alibaba",
      "accuracy": 0.5331086666666667,
      "accuracy_std_dev": 0.05754091454106565,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.005119237373737374,
      "avg_duration_per_task_ms": 101325.84511784512,
      "avg_output_tokens_per_task": 5420.193602693603,
      "last_run_timestamp": "2026-10-01T16:05:56.758Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.5-air",
      "model_slug": "z-ai/glm-4.5-air",
      "model_name": "Z.ai: GLM 4.5 Air",
      "provider_name": "Novita",
      "accuracy": 0.532164,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 171,
      "avg_cost_per_task_usd": 0.03933312280701755,
      "avg_duration_per_task_ms": 1068622.9473684211,
      "avg_output_tokens_per_task": 51714.59064327485,
      "last_run_timestamp": "2026-08-01T18:07:49.346Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4o-2024-08-06",
      "model_slug": "openai/gpt-4o-2024-08-06",
      "model_name": "OpenAI: GPT-4o (2024-08-06)",
      "provider_name": "OpenAI",
      "accuracy": 0.5280586666666666,
      "accuracy_std_dev": 0.004827144451490511,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.01799902861952862,
      "avg_duration_per_task_ms": 14939.786195286195,
      "avg_output_tokens_per_task": 1607.080808080808,
      "last_run_timestamp": "2026-10-02T23:16:52.382Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.7-flash-20260119",
      "model_slug": "z-ai/glm-4.7-flash",
      "model_name": "Z.ai: GLM 4.7 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.5238096808510638,
      "accuracy_std_dev": 0.00755476843854264,
      "run_count": 5,
      "total_questions": 987,
      "avg_cost_per_task_usd": 0.0166975390070922,
      "avg_duration_per_task_ms": 736031.547112462,
      "avg_output_tokens_per_task": 42415.456940222895,
      "last_run_timestamp": "2026-08-30T00:11:29.373Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-chat-v3-0324",
      "model_slug": "deepseek/deepseek-chat-v3-0324",
      "model_name": "DeepSeek: DeepSeek V3 0324",
      "provider_name": "DeepInfra",
      "accuracy": 0.5230117451776649,
      "accuracy_std_dev": 0.08115411707264149,
      "run_count": 5,
      "total_questions": 985,
      "avg_cost_per_task_usd": 0.0036350548223350256,
      "avg_duration_per_task_ms": 131062.38578680203,
      "avg_output_tokens_per_task": 3887.226395939086,
      "last_run_timestamp": "2026-08-29T20:34:10.117Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4o",
      "model_slug": "openai/gpt-4o",
      "model_name": "OpenAI: GPT-4o",
      "provider_name": "OpenAI",
      "accuracy": 0.522126,
      "accuracy_std_dev": 0.02474951208754282,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.01514130735930736,
      "avg_duration_per_task_ms": 13491.472582972583,
      "avg_output_tokens_per_task": 1320.8239538239538,
      "last_run_timestamp": "2026-10-02T23:17:37.980Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Coder 30B A3B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.5202023333333333,
      "accuracy_std_dev": 0.0023810642345154617,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0035723956228956226,
      "avg_duration_per_task_ms": 90157.08922558922,
      "avg_output_tokens_per_task": 2721.4242424242425,
      "last_run_timestamp": "2026-08-15T08:46:42.897Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4o-2024-08-06",
      "model_slug": "openai/gpt-4o-2024-08-06",
      "model_name": "OpenAI: GPT-4o (2024-08-06)",
      "provider_name": "auto-routing",
      "accuracy": 0.5202019999999999,
      "accuracy_std_dev": 0.0033669999999999534,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.017511833333333334,
      "avg_duration_per_task_ms": 16177.118686868687,
      "avg_output_tokens_per_task": 1559.209595959596,
      "last_run_timestamp": "2026-10-01T16:04:12.703Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Coder 30B A3B Instruct",
      "provider_name": "Alibaba",
      "accuracy": 0.5190796666666666,
      "accuracy_std_dev": 0.008837261503932561,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.004016927609427609,
      "avg_duration_per_task_ms": 28916.838383838385,
      "avg_output_tokens_per_task": 2584.1296296296296,
      "last_run_timestamp": "2026-08-15T09:25:42.974Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4o",
      "model_slug": "openai/gpt-4o",
      "model_name": "OpenAI: GPT-4o",
      "provider_name": "Azure",
      "accuracy": 0.5168355,
      "accuracy_std_dev": 0.005050499999999958,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.017194939393939394,
      "avg_duration_per_task_ms": 13199.247474747475,
      "avg_output_tokens_per_task": 1523.2777777777778,
      "last_run_timestamp": "2026-10-01T16:28:34.286Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Coder 30B A3B Instruct",
      "provider_name": "DigitalOcean",
      "accuracy": 0.5157126666666666,
      "accuracy_std_dev": 0.02305568803763809,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.004710382154882156,
      "avg_duration_per_task_ms": 211036.34006734006,
      "avg_output_tokens_per_task": 2694.722222222222,
      "last_run_timestamp": "2026-08-15T09:11:29.276Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4o-2024-08-06",
      "model_slug": "openai/gpt-4o-2024-08-06",
      "model_name": "OpenAI: GPT-4o (2024-08-06)",
      "provider_name": "Azure",
      "accuracy": 0.5151519999999999,
      "accuracy_std_dev": 0.006733999999999962,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.018041959595959597,
      "avg_duration_per_task_ms": 13498.479797979799,
      "avg_output_tokens_per_task": 1611.5353535353536,
      "last_run_timestamp": "2026-10-01T16:26:08.836Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Coder 30B A3B Instruct",
      "provider_name": "SiliconFlow",
      "accuracy": 0.5134683333333333,
      "accuracy_std_dev": 0.00836132925370652,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0008388316498316498,
      "avg_duration_per_task_ms": 89120.7138047138,
      "avg_output_tokens_per_task": 2792.7188552188554,
      "last_run_timestamp": "2026-08-15T08:49:03.156Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-vl-8b-instruct",
      "model_slug": "qwen/qwen3-vl-8b-instruct",
      "model_name": "Qwen: Qwen3 VL 8B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.513468,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0047852878787878785,
      "avg_duration_per_task_ms": 118138.42929292929,
      "avg_output_tokens_per_task": 10671.39898989899,
      "last_run_timestamp": "2026-08-01T18:08:24.732Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.5-air",
      "model_slug": "z-ai/glm-4.5-air",
      "model_name": "Z.ai: GLM 4.5 Air",
      "provider_name": "Z.AI",
      "accuracy": 0.506823,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 171,
      "avg_cost_per_task_usd": 0.03426515789473684,
      "avg_duration_per_task_ms": 1056421.3684210526,
      "avg_output_tokens_per_task": 40494.5730994152,
      "last_run_timestamp": "2026-08-01T18:05:37.080Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4o-2024-05-13",
      "model_slug": "openai/gpt-4o-2024-05-13",
      "model_name": "OpenAI: GPT-4o (2024-05-13)",
      "provider_name": "auto-routing",
      "accuracy": 0.506734,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0237055303030303,
      "avg_duration_per_task_ms": 11478.515151515152,
      "avg_output_tokens_per_task": 1318.7474747474748,
      "last_run_timestamp": "2026-08-01T08:34:23.935Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-chat-v3-0324",
      "model_slug": "deepseek/deepseek-chat-v3-0324",
      "model_name": "DeepSeek: DeepSeek V3 0324",
      "provider_name": "auto-routing",
      "accuracy": 0.5065361568627451,
      "accuracy_std_dev": 0.1358455723807,
      "run_count": 5,
      "total_questions": 918,
      "avg_cost_per_task_usd": 0.004705715686274509,
      "avg_duration_per_task_ms": 112349.47167755991,
      "avg_output_tokens_per_task": 4227.696078431372,
      "last_run_timestamp": "2026-08-29T20:57:58.443Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4o",
      "model_slug": "openai/gpt-4o",
      "model_name": "OpenAI: GPT-4o",
      "provider_name": "auto-routing",
      "accuracy": 0.5056116666666668,
      "accuracy_std_dev": 0.007570922855827376,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.012942734006734005,
      "avg_duration_per_task_ms": 12612.308080808081,
      "avg_output_tokens_per_task": 1100.3198653198654,
      "last_run_timestamp": "2026-10-01T16:03:45.472Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "inclusionai/ling-3.0-flash-fin-20260827",
      "model_slug": "inclusionai/ling-3.0-flash-fin",
      "model_name": "inclusionAI: Ling 3.0 Flash Fin",
      "provider_name": "auto-routing",
      "accuracy": 0.505051,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.009755954545454545,
      "avg_duration_per_task_ms": 700759.0555555555,
      "avg_output_tokens_per_task": 64762.606060606064,
      "last_run_timestamp": "2026-10-01T08:24:17.043Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4o-2024-05-13",
      "model_slug": "openai/gpt-4o-2024-05-13",
      "model_name": "OpenAI: GPT-4o (2024-05-13)",
      "provider_name": "Azure",
      "accuracy": 0.505051,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.02292151515151515,
      "avg_duration_per_task_ms": 12797.353535353535,
      "avg_output_tokens_per_task": 1266.479797979798,
      "last_run_timestamp": "2026-08-01T08:47:47.980Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4.1-nano-2025-04-14",
      "model_slug": "openai/gpt-4.1-nano",
      "model_name": "OpenAI: GPT-4.1 Nano",
      "provider_name": "auto-routing",
      "accuracy": 0.5,
      "accuracy_std_dev": 0.015120700182200542,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0008362104377104377,
      "avg_duration_per_task_ms": 19874.548821548822,
      "avg_output_tokens_per_task": 1898.5521885521885,
      "last_run_timestamp": "2026-10-01T19:13:37.127Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "Groq",
      "accuracy": 0.4979423333333333,
      "accuracy_std_dev": 0.013034980185809425,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.0023551829405162737,
      "avg_duration_per_task_ms": 10287.485970819303,
      "avg_output_tokens_per_task": 2334.3832772166106,
      "last_run_timestamp": "2026-10-24T16:02:08.875Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4.1-nano-2025-04-14",
      "model_slug": "openai/gpt-4.1-nano",
      "model_name": "OpenAI: GPT-4.1 Nano",
      "provider_name": "Azure",
      "accuracy": 0.49764320000000006,
      "accuracy_std_dev": 0.01852187333290023,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.0008495191919191919,
      "avg_duration_per_task_ms": 13033.880808080808,
      "avg_output_tokens_per_task": 1851.6353535353535,
      "last_run_timestamp": "2026-10-01T19:12:29.219Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4.1-nano-2025-04-14",
      "model_slug": "openai/gpt-4.1-nano",
      "model_name": "OpenAI: GPT-4.1 Nano",
      "provider_name": "OpenAI",
      "accuracy": 0.4966332500000001,
      "accuracy_std_dev": 0.014773079898501197,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.0008230315656565656,
      "avg_duration_per_task_ms": 17739.39898989899,
      "avg_output_tokens_per_task": 1863.909090909091,
      "last_run_timestamp": "2026-10-02T23:16:50.042Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.7-flash-20260119",
      "model_slug": "z-ai/glm-4.7-flash",
      "model_name": "Z.ai: GLM 4.7 Flash",
      "provider_name": "DeepInfra",
      "accuracy": 0.49197393750000007,
      "accuracy_std_dev": 0.015797164320219007,
      "run_count": 5,
      "total_questions": 976,
      "avg_cost_per_task_usd": 0.01112520286885246,
      "avg_duration_per_task_ms": 456396.28176229505,
      "avg_output_tokens_per_task": 27747.69262295082,
      "last_run_timestamp": "2026-08-30T01:15:14.669Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.7-flash-20260119",
      "model_slug": "z-ai/glm-4.7-flash",
      "model_name": "Z.ai: GLM 4.7 Flash",
      "provider_name": "Venice",
      "accuracy": 0.48962583163265305,
      "accuracy_std_dev": 0.02041486033848873,
      "run_count": 5,
      "total_questions": 980,
      "avg_cost_per_task_usd": 0.011151638775510205,
      "avg_duration_per_task_ms": 495714.47040816327,
      "avg_output_tokens_per_task": 27939.333673469388,
      "last_run_timestamp": "2026-08-30T00:23:23.003Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "SambaNova",
      "accuracy": 0.48512888888888883,
      "accuracy_std_dev": 0.005611738895574061,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.002352908529741863,
      "avg_duration_per_task_ms": 15334.566217732885,
      "avg_output_tokens_per_task": 2211.6778900112236,
      "last_run_timestamp": "2026-10-24T16:21:36.007Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "CoreWeave",
      "accuracy": 0.4820423333333333,
      "accuracy_std_dev": 0.009490110618720696,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.002550328282828283,
      "avg_duration_per_task_ms": 48799.52581369248,
      "avg_output_tokens_per_task": 2783.7373737373737,
      "last_run_timestamp": "2026-10-24T16:03:59.174Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "AkashML",
      "accuracy": 0.48202820260989004,
      "accuracy_std_dev": 0.015459058724256144,
      "run_count": 15,
      "total_questions": 2912,
      "avg_cost_per_task_usd": 0.0012292771291208791,
      "avg_duration_per_task_ms": 58443.023008241755,
      "avg_output_tokens_per_task": 2247.816620879121,
      "last_run_timestamp": "2026-10-24T16:04:33.482Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "Cerebras",
      "accuracy": 0.48091999999999996,
      "accuracy_std_dev": 0.004418715424192887,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.048067654882154885,
      "avg_duration_per_task_ms": 46868.4898989899,
      "avg_output_tokens_per_task": 16842.27946127946,
      "last_run_timestamp": "2026-08-15T14:46:10.940Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "Parasail",
      "accuracy": 0.4783013333333333,
      "accuracy_std_dev": 0.014050459771045853,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.001328496071829405,
      "avg_duration_per_task_ms": 46533.534792368126,
      "avg_output_tokens_per_task": 2380.4786756453423,
      "last_run_timestamp": "2026-10-24T16:03:59.176Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "Nebius",
      "accuracy": 0.47777759999999997,
      "accuracy_std_dev": 0.010931021024588695,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.0010385141414141412,
      "avg_duration_per_task_ms": 206806.0606060606,
      "avg_output_tokens_per_task": 2333.3515151515153,
      "last_run_timestamp": "2026-08-29T20:09:29.346Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen-2.5-72b-instruct",
      "model_slug": "qwen/qwen-2.5-72b-instruct",
      "model_name": "Qwen2.5 72B Instruct",
      "provider_name": "DeepInfra",
      "accuracy": 0.4765793529411765,
      "accuracy_std_dev": 0.006944000000000006,
      "run_count": 2,
      "total_questions": 306,
      "avg_cost_per_task_usd": 0.0011232875816993464,
      "avg_duration_per_task_ms": 126677.51307189543,
      "avg_output_tokens_per_task": 2069.3627450980393,
      "last_run_timestamp": "2026-10-01T16:04:45.916Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.47325066666666665,
      "accuracy_std_dev": 0.009636497738632366,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.0014022805836139168,
      "avg_duration_per_task_ms": 75040.13580246913,
      "avg_output_tokens_per_task": 2343.6004489337824,
      "last_run_timestamp": "2026-10-24T16:04:55.367Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "Together",
      "accuracy": 0.472642875,
      "accuracy_std_dev": 0.00832212530603661,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.0031359400252525256,
      "avg_duration_per_task_ms": 59854.52904040404,
      "avg_output_tokens_per_task": 2209.2967171717173,
      "last_run_timestamp": "2026-10-17T16:05:55.476Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "Novita",
      "accuracy": 0.4711285466816648,
      "accuracy_std_dev": 0.0145633743402015,
      "run_count": 9,
      "total_questions": 1778,
      "avg_cost_per_task_usd": 0.0012865393700787402,
      "avg_duration_per_task_ms": 87597.63948256469,
      "avg_output_tokens_per_task": 2944.271653543307,
      "last_run_timestamp": "2026-10-24T16:02:28.569Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.469697,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.006110727272727273,
      "avg_duration_per_task_ms": 42879.368686868685,
      "avg_output_tokens_per_task": 2318.919191919192,
      "last_run_timestamp": "2026-10-04T03:02:23.034Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "Crusoe",
      "accuracy": 0.4664644357937311,
      "accuracy_std_dev": 0.0072384874663150465,
      "run_count": 5,
      "total_questions": 989,
      "avg_cost_per_task_usd": 0.0021660940343781595,
      "avg_duration_per_task_ms": 35847.60161779575,
      "avg_output_tokens_per_task": 2625.268958543984,
      "last_run_timestamp": "2026-08-29T16:31:42.504Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "DeepInfra",
      "accuracy": 0.46456220000000004,
      "accuracy_std_dev": 0.016514098623903146,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.0007998439393939393,
      "avg_duration_per_task_ms": 160078.42626262625,
      "avg_output_tokens_per_task": 2248.7131313131313,
      "last_run_timestamp": "2026-10-24T16:22:30.499Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen2.5-vl-72b-instruct",
      "model_slug": "qwen/qwen2.5-vl-72b-instruct",
      "model_name": "Qwen: Qwen2.5 VL 72B Instruct",
      "provider_name": "Parasail",
      "accuracy": 0.462542,
      "accuracy_std_dev": 0.005471000000000004,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.001783989898989899,
      "avg_duration_per_task_ms": 27644.777777777777,
      "avg_output_tokens_per_task": 1397.5252525252524,
      "last_run_timestamp": "2026-09-01T13:03:42.840Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "Ionstream",
      "accuracy": 0.46128619685039374,
      "accuracy_std_dev": 0.075159,
      "run_count": 2,
      "total_questions": 254,
      "avg_cost_per_task_usd": 0.00731248031496063,
      "avg_duration_per_task_ms": 1016477.8543307086,
      "avg_output_tokens_per_task": 20586.740157480315,
      "last_run_timestamp": "2026-08-08T10:11:59.748Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen-2.5-72b-instruct",
      "model_slug": "qwen/qwen-2.5-72b-instruct",
      "model_name": "Qwen2.5 72B Instruct",
      "provider_name": "Novita",
      "accuracy": 0.461279,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0011232171717171719,
      "avg_duration_per_task_ms": 136788.7626262626,
      "avg_output_tokens_per_task": 2036.2272727272727,
      "last_run_timestamp": "2026-10-01T16:27:00.375Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen-2.5-72b-instruct",
      "model_slug": "qwen/qwen-2.5-72b-instruct",
      "model_name": "Qwen2.5 72B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.45959600000000006,
      "accuracy_std_dev": 0.01346799999999998,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.0010278964646464647,
      "avg_duration_per_task_ms": 96379.95959595959,
      "avg_output_tokens_per_task": 1848.388888888889,
      "last_run_timestamp": "2026-10-01T16:47:11.191Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-vl-8b-instruct",
      "model_slug": "qwen/qwen3-vl-8b-instruct",
      "model_name": "Qwen: Qwen3 VL 8B Instruct",
      "provider_name": "Parasail",
      "accuracy": 0.45862900000000006,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 141,
      "avg_cost_per_task_usd": 0.0011888297872340425,
      "avg_duration_per_task_ms": 278701.0070921986,
      "avg_output_tokens_per_task": 7719.879432624113,
      "last_run_timestamp": "2026-08-01T18:13:08.069Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen2.5-vl-72b-instruct",
      "model_slug": "qwen/qwen2.5-vl-72b-instruct",
      "model_name": "Qwen: Qwen2.5 VL 72B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.4444445,
      "accuracy_std_dev": 0.015151500000000012,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.0015239469696969696,
      "avg_duration_per_task_ms": 59255.792929292926,
      "avg_output_tokens_per_task": 1592.9924242424242,
      "last_run_timestamp": "2026-09-01T13:21:35.512Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen2.5-vl-72b-instruct",
      "model_slug": "qwen/qwen2.5-vl-72b-instruct",
      "model_name": "Qwen: Qwen2.5 VL 72B Instruct",
      "provider_name": "Nebius",
      "accuracy": 0.4402355,
      "accuracy_std_dev": 0.002525500000000014,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.0013545,
      "avg_duration_per_task_ms": 68001.7196969697,
      "avg_output_tokens_per_task": 1535.1843434343434,
      "last_run_timestamp": "2026-09-01T12:41:25.955Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4o-mini",
      "model_slug": "openai/gpt-4o-mini",
      "model_name": "OpenAI: GPT-4o-mini",
      "provider_name": "auto-routing",
      "accuracy": 0.4320986666666667,
      "accuracy_std_dev": 0.00781631474687543,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0010060690235690235,
      "avg_duration_per_task_ms": 20918.510101010103,
      "avg_output_tokens_per_task": 1483.5286195286196,
      "last_run_timestamp": "2026-10-01T19:16:07.087Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4o-mini",
      "model_slug": "openai/gpt-4o-mini",
      "model_name": "OpenAI: GPT-4o-mini",
      "provider_name": "Azure",
      "accuracy": 0.4315373333333334,
      "accuracy_std_dev": 0.010588221658469791,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.0010723358585858589,
      "avg_duration_per_task_ms": 21765.176767676767,
      "avg_output_tokens_per_task": 1505.858585858586,
      "last_run_timestamp": "2026-10-01T19:05:18.003Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4o-mini",
      "model_slug": "openai/gpt-4o-mini",
      "model_name": "OpenAI: GPT-4o-mini",
      "provider_name": "OpenAI",
      "accuracy": 0.4305555,
      "accuracy_std_dev": 0.007469952225416168,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.0010262436868686867,
      "avg_duration_per_task_ms": 17602.7702020202,
      "avg_output_tokens_per_task": 1517.9292929292928,
      "last_run_timestamp": "2026-10-02T23:19:57.879Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "DigitalOcean",
      "accuracy": 0.422559,
      "accuracy_std_dev": 0.09221941654915557,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.019149646464646466,
      "avg_duration_per_task_ms": 3630846.4646464647,
      "avg_output_tokens_per_task": 17674.203703703704,
      "last_run_timestamp": "2026-08-22T17:37:26.215Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "Cloudflare",
      "accuracy": 0.41010556441717794,
      "accuracy_std_dev": 0.09310346585944049,
      "run_count": 10,
      "total_questions": 1956,
      "avg_cost_per_task_usd": 0.004216284764826176,
      "avg_duration_per_task_ms": 41336.05981595092,
      "avg_output_tokens_per_task": 1775.0184049079755,
      "last_run_timestamp": "2026-10-24T16:02:47.168Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-3-12b-it",
      "model_slug": "google/gemma-3-12b-it",
      "model_name": "Google: Gemma 3 12B",
      "provider_name": "DeepInfra",
      "accuracy": 0.405724,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0003259848484848485,
      "avg_duration_per_task_ms": 29056.348484848484,
      "avg_output_tokens_per_task": 1917.1515151515152,
      "last_run_timestamp": "2026-10-08T22:01:16.852Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "tencent/hy-mt2-30b-a3b-20260521",
      "model_slug": "tencent/hy-mt2-30b-a3b",
      "model_name": "Tencent: Hy-MT2-30B-A3B",
      "provider_name": "Reka",
      "accuracy": 0.388889,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.00023147979797979797,
      "avg_duration_per_task_ms": 8604.353535353535,
      "avg_output_tokens_per_task": 710.540404040404,
      "last_run_timestamp": "2026-09-24T06:00:03.613Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "amazon/nova-micro-v1",
      "model_slug": "amazon/nova-micro-v1",
      "model_name": "Amazon: Nova Micro 1.0",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.38720499999999997,
      "accuracy_std_dev": 0.010647388881786938,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.00015740101010101008,
      "avg_duration_per_task_ms": 6591.242424242424,
      "avg_output_tokens_per_task": 937.0363636363636,
      "last_run_timestamp": "2026-10-01T19:39:32.831Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-3-12b-it",
      "model_slug": "google/gemma-3-12b-it",
      "model_name": "Google: Gemma 3 12B",
      "provider_name": "NextBit",
      "accuracy": 0.37878799999999996,
      "accuracy_std_dev": 0.01683499999999999,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.000344469696969697,
      "avg_duration_per_task_ms": 46431.74494949495,
      "avg_output_tokens_per_task": 2036.358585858586,
      "last_run_timestamp": "2026-10-07T13:35:26.756Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen-2.5-7b-instruct",
      "model_slug": "qwen/qwen-2.5-7b-instruct",
      "model_name": "Qwen: Qwen2.5 7B Instruct",
      "provider_name": "Phala",
      "accuracy": 0.36363599999999996,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.00041437878787878784,
      "avg_duration_per_task_ms": 34981.92424242424,
      "avg_output_tokens_per_task": 1665.6666666666667,
      "last_run_timestamp": "2026-08-01T08:55:19.318Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Reka",
      "accuracy": 0.346801,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.008616373737373737,
      "avg_duration_per_task_ms": 1179181.5303030303,
      "avg_output_tokens_per_task": 32995.232323232325,
      "last_run_timestamp": "2026-10-03T00:19:24.096Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen-2.5-7b-instruct",
      "model_slug": "qwen/qwen-2.5-7b-instruct",
      "model_name": "Qwen: Qwen2.5 7B Instruct",
      "provider_name": "Together",
      "accuracy": 0.3417510000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0007054343434343434,
      "avg_duration_per_task_ms": 14064.868686868687,
      "avg_output_tokens_per_task": 1539,
      "last_run_timestamp": "2026-08-01T08:06:23.600Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen-2.5-7b-instruct",
      "model_slug": "qwen/qwen-2.5-7b-instruct",
      "model_name": "Qwen: Qwen2.5 7B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.328283,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0005318181818181818,
      "avg_duration_per_task_ms": 36388.32323232323,
      "avg_output_tokens_per_task": 2041.050505050505,
      "last_run_timestamp": "2026-08-01T08:36:11.581Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "Venice",
      "accuracy": 0.32428450000000003,
      "accuracy_std_dev": 0.09813129033340996,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.008669871843434344,
      "avg_duration_per_task_ms": 575303.3933080808,
      "avg_output_tokens_per_task": 21440.64835858586,
      "last_run_timestamp": "2026-10-24T16:36:16.779Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "sao10k/l3-lunaris-8b",
      "model_slug": "sao10k/l3-lunaris-8b",
      "model_name": "Sao10K: Llama 3 8B Lunaris",
      "provider_name": "DeepInfra",
      "accuracy": 0.2891415,
      "accuracy_std_dev": 0.003232916137174003,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.00006112626262626263,
      "avg_duration_per_task_ms": 10017.040404040405,
      "avg_output_tokens_per_task": 577.7083333333334,
      "last_run_timestamp": "2026-10-08T22:00:27.722Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.1-8b-instruct",
      "model_slug": "meta-llama/llama-3.1-8b-instruct",
      "model_name": "Meta: Llama 3.1 8B Instruct",
      "provider_name": "Groq",
      "accuracy": 0.28554066666666666,
      "accuracy_std_dev": 0.017098526187299817,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.002770008978675645,
      "avg_duration_per_task_ms": 51158.87991021325,
      "avg_output_tokens_per_task": 34564.77048260382,
      "last_run_timestamp": "2026-10-24T16:02:28.572Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.1-8b-instruct",
      "model_slug": "meta-llama/llama-3.1-8b-instruct",
      "model_name": "Meta: Llama 3.1 8B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.28226711111111114,
      "accuracy_std_dev": 0.020967071598868986,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.0013839943883277217,
      "avg_duration_per_task_ms": 137971.6234567901,
      "avg_output_tokens_per_task": 19125.452300785633,
      "last_run_timestamp": "2026-10-24T16:08:06.919Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Coder 30B A3B Instruct",
      "provider_name": "Novita",
      "accuracy": 0.275533,
      "accuracy_std_dev": 0.009152576176501717,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0007405521885521885,
      "avg_duration_per_task_ms": 48360.410774410775,
      "avg_output_tokens_per_task": 2532.1531986531986,
      "last_run_timestamp": "2026-08-15T09:28:07.208Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "sao10k/l3-lunaris-8b",
      "model_slug": "sao10k/l3-lunaris-8b",
      "model_name": "Sao10K: Llama 3 8B Lunaris",
      "provider_name": "Novita",
      "accuracy": 0.2739361560283688,
      "accuracy_std_dev": 0.008153796102974918,
      "run_count": 3,
      "total_questions": 564,
      "avg_cost_per_task_usd": 0.00006610283687943263,
      "avg_duration_per_task_ms": 8879.929078014184,
      "avg_output_tokens_per_task": 551.2074468085107,
      "last_run_timestamp": "2026-10-01T16:01:04.877Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.1-8b-instruct",
      "model_slug": "meta-llama/llama-3.1-8b-instruct",
      "model_name": "Meta: Llama 3.1 8B Instruct",
      "provider_name": "CoreWeave",
      "accuracy": 0.2735093284457478,
      "accuracy_std_dev": 0.01586648965587537,
      "run_count": 9,
      "total_questions": 1705,
      "avg_cost_per_task_usd": 0.009129268035190615,
      "avg_duration_per_task_ms": 373704.25923753664,
      "avg_output_tokens_per_task": 40779.12316715543,
      "last_run_timestamp": "2026-10-24T16:10:43.379Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "sao10k/l3-lunaris-8b",
      "model_slug": "sao10k/l3-lunaris-8b",
      "model_name": "Sao10K: Llama 3 8B Lunaris",
      "provider_name": "auto-routing",
      "accuracy": 0.273148,
      "accuracy_std_dev": 0.006113657170630353,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.00006268181818181819,
      "avg_duration_per_task_ms": 9553.181818181818,
      "avg_output_tokens_per_task": 586.439393939394,
      "last_run_timestamp": "2026-10-01T20:43:57.173Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-3-4b-it",
      "model_slug": "google/gemma-3-4b-it",
      "model_name": "Google: Gemma 3 4B",
      "provider_name": "Parasail",
      "accuracy": 0.271044,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.00011306565656565657,
      "avg_duration_per_task_ms": 10846.11616161616,
      "avg_output_tokens_per_task": 1804.4646464646464,
      "last_run_timestamp": "2026-10-15T03:00:03.049Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "sao10k/l3-lunaris-8b",
      "model_slug": "sao10k/l3-lunaris-8b",
      "model_name": "Sao10K: Llama 3 8B Lunaris",
      "provider_name": "Parasail",
      "accuracy": 0.2700681612244898,
      "accuracy_std_dev": 0.016752437809465226,
      "run_count": 5,
      "total_questions": 980,
      "avg_cost_per_task_usd": 0.00005762346938775511,
      "avg_duration_per_task_ms": 9734.870408163266,
      "avg_output_tokens_per_task": 576.8010204081633,
      "last_run_timestamp": "2026-10-01T20:47:49.128Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.1-8b-instruct",
      "model_slug": "meta-llama/llama-3.1-8b-instruct",
      "model_name": "Meta: Llama 3.1 8B Instruct",
      "provider_name": "DeepInfra",
      "accuracy": 0.2677151506849315,
      "accuracy_std_dev": 0.012439954983841373,
      "run_count": 10,
      "total_questions": 1971,
      "avg_cost_per_task_usd": 0.0003856813800101471,
      "avg_duration_per_task_ms": 392968.5474378488,
      "avg_output_tokens_per_task": 9238.72602739726,
      "last_run_timestamp": "2026-10-24T16:01:29.844Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "AkashML",
      "accuracy": 0.26430966666666666,
      "accuracy_std_dev": 0.1964870751683739,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.014085552188552189,
      "avg_duration_per_task_ms": 155316.38383838383,
      "avg_output_tokens_per_task": 16743.055555555555,
      "last_run_timestamp": "2026-08-15T15:24:35.389Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.1-8b-instruct",
      "model_slug": "meta-llama/llama-3.1-8b-instruct",
      "model_name": "Meta: Llama 3.1 8B Instruct",
      "provider_name": "Novita",
      "accuracy": 0.2575758888888889,
      "accuracy_std_dev": 0.014825798696449854,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.00042795959595959596,
      "avg_duration_per_task_ms": 75249.49046015713,
      "avg_output_tokens_per_task": 8235.553310886644,
      "last_run_timestamp": "2026-10-24T16:02:28.577Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.2-3b-instruct",
      "model_slug": "meta-llama/llama-3.2-3b-instruct",
      "model_name": "Meta: Llama 3.2 3B Instruct",
      "provider_name": "Parasail",
      "accuracy": 0.21734899999999996,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 171,
      "avg_cost_per_task_usd": 0.0003523450292397661,
      "avg_duration_per_task_ms": 72727.70760233919,
      "avg_output_tokens_per_task": 4269.023391812865,
      "last_run_timestamp": "2026-08-01T08:05:56.858Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.1-8b-instruct",
      "model_slug": "meta-llama/llama-3.1-8b-instruct",
      "model_name": "Meta: Llama 3.1 8B Instruct",
      "provider_name": "Cloudflare",
      "accuracy": 0.19163883333333334,
      "accuracy_std_dev": 0.1344497289267339,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.0010863872053872054,
      "avg_duration_per_task_ms": 449223.712962963,
      "avg_output_tokens_per_task": 3508.121212121212,
      "last_run_timestamp": "2026-10-24T16:03:19.773Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-chat-v3-0324",
      "model_slug": "deepseek/deepseek-chat-v3-0324",
      "model_name": "DeepSeek: DeepSeek V3 0324",
      "provider_name": "Novita",
      "accuracy": 0.1861954,
      "accuracy_std_dev": 0.12390642146491036,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.001524880808080808,
      "avg_duration_per_task_ms": 49007.062626262625,
      "avg_output_tokens_per_task": 1318.2202020202021,
      "last_run_timestamp": "2026-08-29T18:43:09.114Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.2-3b-instruct",
      "model_slug": "meta-llama/llama-3.2-3b-instruct",
      "model_name": "Meta: Llama 3.2 3B Instruct",
      "provider_name": "Cloudflare",
      "accuracy": 0.17340049999999999,
      "accuracy_std_dev": 0.11279449999999999,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.00031717929292929297,
      "avg_duration_per_task_ms": 9334.631313131313,
      "avg_output_tokens_per_task": 867.929292929293,
      "last_run_timestamp": "2026-10-13T18:56:05.001Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "inclusionai/ling-3.0-flash-fin-20260827",
      "model_slug": "inclusionai/ling-3.0-flash-fin",
      "model_name": "inclusionAI: Ling 3.0 Flash Fin",
      "provider_name": "DeepInfra",
      "accuracy": 0.17174282233502539,
      "accuracy_std_dev": 0.01788397279378631,
      "run_count": 3,
      "total_questions": 591,
      "avg_cost_per_task_usd": 0.019897788494077834,
      "avg_duration_per_task_ms": 1784246.4196277496,
      "avg_output_tokens_per_task": 110294.06598984772,
      "last_run_timestamp": "2026-10-08T23:13:58.126Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.2-3b-instruct",
      "model_slug": "meta-llama/llama-3.2-3b-instruct",
      "model_name": "Meta: Llama 3.2 3B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.116162,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0001537979797979798,
      "avg_duration_per_task_ms": 29722.909090909092,
      "avg_output_tokens_per_task": 1836.7929292929293,
      "last_run_timestamp": "2026-08-01T09:03:02.970Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.1-flash-lite-image-20260630",
      "model_slug": "google/gemini-3.1-flash-lite-image",
      "model_name": "Google: Nano Banana 2 Lite (Gemini 3.1 Flash Lite Image)",
      "provider_name": "Google AI Studio",
      "accuracy": 0.1161615,
      "accuracy_std_dev": 0.0016835000000000044,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.017850737373737373,
      "avg_duration_per_task_ms": 46853.982323232325,
      "avg_output_tokens_per_task": 11776.444444444445,
      "last_run_timestamp": "2026-10-01T16:00:58.389Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.1-flash-lite-image-20260630",
      "model_slug": "google/gemini-3.1-flash-lite-image",
      "model_name": "Google: Nano Banana 2 Lite (Gemini 3.1 Flash Lite Image)",
      "provider_name": "Google",
      "accuracy": 0.112795,
      "accuracy_std_dev": 0.006733999999999997,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.018038613636363634,
      "avg_duration_per_task_ms": 21125.79292929293,
      "avg_output_tokens_per_task": 11795.222222222223,
      "last_run_timestamp": "2026-10-01T16:00:58.364Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.1-flash-lite-image-20260630",
      "model_slug": "google/gemini-3.1-flash-lite-image",
      "model_name": "Google: Nano Banana 2 Lite (Gemini 3.1 Flash Lite Image)",
      "provider_name": "auto-routing",
      "accuracy": 0.102694,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.017984406565656568,
      "avg_duration_per_task_ms": 21554.590909090908,
      "avg_output_tokens_per_task": 11812.823232323231,
      "last_run_timestamp": "2026-10-01T16:02:13.023Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.1-terminus",
      "model_slug": "deepseek/deepseek-v3.1-terminus",
      "model_name": "DeepSeek: DeepSeek V3.1 Terminus",
      "provider_name": "SiliconFlow",
      "accuracy": 0.0025255000000000004,
      "accuracy_std_dev": 0.0008415000000000001,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.0034195530303030306,
      "avg_duration_per_task_ms": 159765.32575757575,
      "avg_output_tokens_per_task": 3219.0782828282827,
      "last_run_timestamp": "2026-08-22T17:29:06.480Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "SiliconFlow",
      "accuracy": 0.001684,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.0034544065656565656,
      "avg_duration_per_task_ms": 164367.952020202,
      "avg_output_tokens_per_task": 3253.931818181818,
      "last_run_timestamp": "2026-08-15T14:54:58.071Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "BaseTen",
      "accuracy": 0.8368423578947368,
      "accuracy_std_dev": 0.014073500000000017,
      "run_count": 2,
      "total_questions": 95,
      "avg_cost_per_task_usd": 0.008221831578947368,
      "avg_duration_per_task_ms": 31565.694736842106,
      "avg_output_tokens_per_task": 9295.157894736842,
      "last_run_timestamp": "2026-10-10T11:02:59.339Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "Parasail",
      "accuracy": 0.809028,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 48,
      "avg_cost_per_task_usd": 0.06453804166666667,
      "avg_duration_per_task_ms": 453773,
      "avg_output_tokens_per_task": 8614.875,
      "last_run_timestamp": "2026-08-01T17:45:43.989Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.7-flash-20260813",
      "model_slug": "google/gemini-3.7-flash",
      "model_name": "Google: Gemini 3.7 Flash",
      "provider_name": "Google AI Studio",
      "accuracy": 0.806667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.20567228,
      "avg_duration_per_task_ms": 69455.26,
      "avg_output_tokens_per_task": 15737.28,
      "last_run_timestamp": "2026-08-15T00:15:11.753Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-5-fable-20260609",
      "model_slug": "anthropic/claude-fable-5",
      "model_name": "Anthropic: Claude Fable 5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.8013334,
      "accuracy_std_dev": 0.021249926537284785,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.8758712919999999,
      "avg_duration_per_task_ms": 139145.43,
      "avg_output_tokens_per_task": 5724.488,
      "last_run_timestamp": "2026-10-24T11:41:42.405Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-5-fable-20260609",
      "model_slug": "anthropic/claude-fable-5",
      "model_name": "Anthropic: Claude Fable 5",
      "provider_name": "Azure",
      "accuracy": 0.8011110833333334,
      "accuracy_std_dev": 0.022580570421855795,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.8756809116666666,
      "avg_duration_per_task_ms": 131059.94833333333,
      "avg_output_tokens_per_task": 5774.293333333333,
      "last_run_timestamp": "2026-10-24T11:46:02.314Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-5-fable-20260609",
      "model_slug": "anthropic/claude-fable-5",
      "model_name": "Anthropic: Claude Fable 5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.8,
      "accuracy_std_dev": 0.033110611481316166,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.7322859599999999,
      "avg_duration_per_task_ms": 146639.07333333333,
      "avg_output_tokens_per_task": 5472.34,
      "last_run_timestamp": "2026-08-15T08:09:14.229Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "DeepSeek",
      "accuracy": 0.8,
      "accuracy_std_dev": 0.013332999999999984,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.015587,
      "avg_duration_per_task_ms": 194697.61,
      "avg_output_tokens_per_task": 11724.04,
      "last_run_timestamp": "2026-08-08T08:29:14.821Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Reka",
      "accuracy": 0.7996450638297874,
      "accuracy_std_dev": 0.011700000000000044,
      "run_count": 2,
      "total_questions": 94,
      "avg_cost_per_task_usd": 0.06506034042553191,
      "avg_duration_per_task_ms": 304476.60638297873,
      "avg_output_tokens_per_task": 17465.68085106383,
      "last_run_timestamp": "2026-10-24T19:25:11.376Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-5-fable-20260609",
      "model_slug": "anthropic/claude-fable-5",
      "model_name": "Anthropic: Claude Fable 5",
      "provider_name": "auto-routing",
      "accuracy": 0.7994444166666667,
      "accuracy_std_dev": 0.026311211407745082,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.9154123233333333,
      "avg_duration_per_task_ms": 123961.04,
      "avg_output_tokens_per_task": 5773.983333333334,
      "last_run_timestamp": "2026-10-24T11:40:21.012Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Tenstorrent",
      "accuracy": 0.7983334999999999,
      "accuracy_std_dev": 0.03444005258779377,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.04458143,
      "avg_duration_per_task_ms": 93369.73,
      "avg_output_tokens_per_task": 5128.475,
      "last_run_timestamp": "2026-10-24T12:30:27.997Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Sail Research",
      "accuracy": 0.7977776666666667,
      "accuracy_std_dev": 0.01911642045177097,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.008363246666666668,
      "avg_duration_per_task_ms": 272879.14,
      "avg_output_tokens_per_task": 12625.026666666667,
      "last_run_timestamp": "2026-08-22T16:20:38.832Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-5-fable-20260609",
      "model_slug": "anthropic/claude-fable-5",
      "model_name": "Anthropic: Claude Fable 5",
      "provider_name": "Anthropic",
      "accuracy": 0.7944445,
      "accuracy_std_dev": 0.01651785042764341,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.8238463366666665,
      "avg_duration_per_task_ms": 131103.42666666667,
      "avg_output_tokens_per_task": 5759.988333333334,
      "last_run_timestamp": "2026-10-24T12:06:18.636Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-5-fable-20260609",
      "model_slug": "anthropic/claude-fable-5",
      "model_name": "Anthropic: Claude Fable 5",
      "provider_name": "Google",
      "accuracy": 0.7940580869565217,
      "accuracy_std_dev": 0.02103502135862134,
      "run_count": 23,
      "total_questions": 1150,
      "avg_cost_per_task_usd": 0.9401787339130434,
      "avg_duration_per_task_ms": 94460.84695652174,
      "avg_output_tokens_per_task": 5691.410434782609,
      "last_run_timestamp": "2026-10-24T11:52:13.306Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7936708987341772,
      "accuracy_std_dev": 0.03037925132800345,
      "run_count": 8,
      "total_questions": 395,
      "avg_cost_per_task_usd": 0.12511070886075948,
      "avg_duration_per_task_ms": 2933394.1088607595,
      "avg_output_tokens_per_task": 27206.47088607595,
      "last_run_timestamp": "2026-10-17T19:41:14.105Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "inclusionai/ling-3.1-flash-20261002",
      "model_slug": "inclusionai/ling-3.1-flash",
      "model_name": "inclusionAI: Ling 3.1 Flash",
      "provider_name": "Novita",
      "accuracy": 0.793333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.00828808,
      "avg_duration_per_task_ms": 196026.56,
      "avg_output_tokens_per_task": 16303,
      "last_run_timestamp": "2026-10-14T19:21:35.384Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "inclusionai/ling-3.1-flash-20261002",
      "model_slug": "inclusionai/ling-3.1-flash",
      "model_name": "inclusionAI: Ling 3.1 Flash",
      "provider_name": "DeepInfra",
      "accuracy": 0.793333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.03146312,
      "avg_duration_per_task_ms": 636210.72,
      "avg_output_tokens_per_task": 16745.44,
      "last_run_timestamp": "2026-10-14T19:17:59.742Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "Xiaomi",
      "accuracy": 0.7920001999999999,
      "accuracy_std_dev": 0.014847359662916514,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.05966906,
      "avg_duration_per_task_ms": 918022.212,
      "avg_output_tokens_per_task": 41767.624,
      "last_run_timestamp": "2026-10-24T12:47:38.794Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "Phala",
      "accuracy": 0.7916668169642858,
      "accuracy_std_dev": 0.03139197384682907,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.1966235267857143,
      "avg_duration_per_task_ms": 790382.3013392857,
      "avg_output_tokens_per_task": 33130.86830357143,
      "last_run_timestamp": "2026-10-24T19:22:19.879Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-fable-5.1-20260831",
      "model_slug": "anthropic/claude-fable-5.1",
      "model_name": "Anthropic: Claude Fable 5.1",
      "provider_name": "auto-routing",
      "accuracy": 0.79166675,
      "accuracy_std_dev": 0.012801703448662584,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.71616581,
      "avg_duration_per_task_ms": 111830.815,
      "avg_output_tokens_per_task": 6900.955,
      "last_run_timestamp": "2026-10-01T08:03:01.529Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Crusoe",
      "accuracy": 0.7911647630522088,
      "accuracy_std_dev": 0.020527481807567145,
      "run_count": 5,
      "total_questions": 249,
      "avg_cost_per_task_usd": 0.028665044176706823,
      "avg_duration_per_task_ms": 359430.156626506,
      "avg_output_tokens_per_task": 12014.867469879518,
      "last_run_timestamp": "2026-08-29T17:06:21.366Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-fable-5.1-20260831",
      "model_slug": "anthropic/claude-fable-5.1",
      "model_name": "Anthropic: Claude Fable 5.1",
      "provider_name": "Anthropic",
      "accuracy": 0.7906666,
      "accuracy_std_dev": 0.009977824303925176,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.814870324,
      "avg_duration_per_task_ms": 182126.652,
      "avg_output_tokens_per_task": 7046.908,
      "last_run_timestamp": "2026-10-01T08:05:22.427Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "Novita",
      "accuracy": 0.7883330000000001,
      "accuracy_std_dev": 0.02598076211353313,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.06002512,
      "avg_duration_per_task_ms": 888300.44,
      "avg_output_tokens_per_task": 41523.41,
      "last_run_timestamp": "2026-10-24T12:33:32.948Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-opus-5-20260723",
      "model_slug": "anthropic/claude-opus-5",
      "model_name": "Anthropic: Claude Opus 5",
      "provider_name": "auto-routing",
      "accuracy": 0.78777775,
      "accuracy_std_dev": 0.020875937181058474,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.48297491666666664,
      "avg_duration_per_task_ms": 122364.275,
      "avg_output_tokens_per_task": 7096.316666666667,
      "last_run_timestamp": "2026-10-24T11:55:24.959Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Chutes",
      "accuracy": 0.7876542314814815,
      "accuracy_std_dev": 0.022797342804780508,
      "run_count": 11,
      "total_questions": 540,
      "avg_cost_per_task_usd": 0.07938424444444445,
      "avg_duration_per_task_ms": 1140818.088888889,
      "avg_output_tokens_per_task": 26072.82222222222,
      "last_run_timestamp": "2026-10-24T19:36:21.865Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Wafer",
      "accuracy": 0.787288561983471,
      "accuracy_std_dev": 0.015937097625777535,
      "run_count": 17,
      "total_questions": 847,
      "avg_cost_per_task_usd": 0.09573880283353008,
      "avg_duration_per_task_ms": 142932.90082644628,
      "avg_output_tokens_per_task": 11874.43919716647,
      "last_run_timestamp": "2026-10-24T13:24:07.217Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Alibaba",
      "accuracy": 0.786667,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.029090889999999998,
      "avg_duration_per_task_ms": 149301.66,
      "avg_output_tokens_per_task": 12176.77,
      "last_run_timestamp": "2026-10-24T11:38:31.986Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-4-argon-20261013",
      "model_slug": "google/gemini-4-argon",
      "model_name": "Google: Gemini 4 Argon",
      "provider_name": "Google",
      "accuracy": 0.786667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.49828548,
      "avg_duration_per_task_ms": 348417.68,
      "avg_output_tokens_per_task": 21971.04,
      "last_run_timestamp": "2026-10-14T15:16:05.578Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-sonnet-5.5-20260928",
      "model_slug": "anthropic/claude-sonnet-5.5",
      "model_name": "Anthropic: Claude Sonnet 5.5",
      "provider_name": "auto-routing",
      "accuracy": 0.78666675,
      "accuracy_std_dev": 0.024944159861328273,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.15998305999999998,
      "avg_duration_per_task_ms": 73714.275,
      "avg_output_tokens_per_task": 6299.305,
      "last_run_timestamp": "2026-10-24T11:26:51.227Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "Darkbloom",
      "accuracy": 0.7862320000000002,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 46,
      "avg_cost_per_task_usd": 0.029582717391304348,
      "avg_duration_per_task_ms": 2435445.304347826,
      "avg_output_tokens_per_task": 108721.58695652174,
      "last_run_timestamp": "2026-10-03T16:12:07.492Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "BaseTen",
      "accuracy": 0.7860963903743317,
      "accuracy_std_dev": 0.025611788150185825,
      "run_count": 4,
      "total_questions": 187,
      "avg_cost_per_task_usd": 0.0982251550802139,
      "avg_duration_per_task_ms": 69951.20320855615,
      "avg_output_tokens_per_task": 7732.208556149732,
      "last_run_timestamp": "2026-08-29T17:15:26.910Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "DeepInfra",
      "accuracy": 0.7860360641891891,
      "accuracy_std_dev": 0.025554613476174253,
      "run_count": 6,
      "total_questions": 296,
      "avg_cost_per_task_usd": 0.09440423986486486,
      "avg_duration_per_task_ms": 381879.5878378378,
      "avg_output_tokens_per_task": 16484.91891891892,
      "last_run_timestamp": "2026-10-10T11:07:21.694Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Together",
      "accuracy": 0.78583325,
      "accuracy_std_dev": 0.016811747713652493,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.07747623,
      "avg_duration_per_task_ms": 47946.0625,
      "avg_output_tokens_per_task": 4735.27,
      "last_run_timestamp": "2026-10-24T11:58:13.073Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-sonnet-5.5-20260928",
      "model_slug": "anthropic/claude-sonnet-5.5",
      "model_name": "Anthropic: Claude Sonnet 5.5",
      "provider_name": "Azure",
      "accuracy": 0.7853333,
      "accuracy_std_dev": 0.011469724399914758,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.17582804400000004,
      "avg_duration_per_task_ms": 67798.288,
      "avg_output_tokens_per_task": 6240.81,
      "last_run_timestamp": "2026-10-24T12:30:32.152Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Fireworks",
      "accuracy": 0.7842432787356322,
      "accuracy_std_dev": 0.019450509073790118,
      "run_count": 14,
      "total_questions": 696,
      "avg_cost_per_task_usd": 0.09254919540229886,
      "avg_duration_per_task_ms": 69468.46551724138,
      "avg_output_tokens_per_task": 4697.962643678161,
      "last_run_timestamp": "2026-10-24T13:01:18.254Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-sonnet-5.5-20260928",
      "model_slug": "anthropic/claude-sonnet-5.5",
      "model_name": "Anthropic: Claude Sonnet 5.5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7840001999999999,
      "accuracy_std_dev": 0.014966487007978884,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.174428044,
      "avg_duration_per_task_ms": 71541.076,
      "avg_output_tokens_per_task": 6317.768,
      "last_run_timestamp": "2026-10-24T11:35:39.965Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "Alibaba",
      "accuracy": 0.7837036666666667,
      "accuracy_std_dev": 0.021513329521743285,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.18632613555555552,
      "avg_duration_per_task_ms": 421396.91333333333,
      "avg_output_tokens_per_task": 30066.795555555556,
      "last_run_timestamp": "2026-10-24T18:46:07.800Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "auto-routing",
      "accuracy": 0.783333375,
      "accuracy_std_dev": 0.023804813940343558,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.19070967249999998,
      "avg_duration_per_task_ms": 778974.4025,
      "avg_output_tokens_per_task": 30803.295,
      "last_run_timestamp": "2026-10-24T19:20:47.328Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "ModelRun",
      "accuracy": 0.7833331666666667,
      "accuracy_std_dev": 0.018358760147103848,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.06827201333333334,
      "avg_duration_per_task_ms": 62356.823333333334,
      "avg_output_tokens_per_task": 12364.343333333334,
      "last_run_timestamp": "2026-10-24T11:44:08.040Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-sonnet-5.5-20260928",
      "model_slug": "anthropic/claude-sonnet-5.5",
      "model_name": "Anthropic: Claude Sonnet 5.5",
      "provider_name": "Google",
      "accuracy": 0.7831111333333334,
      "accuracy_std_dev": 0.01851801018780244,
      "run_count": 15,
      "total_questions": 750,
      "avg_cost_per_task_usd": 0.181601536,
      "avg_duration_per_task_ms": 65069.36666666667,
      "avg_output_tokens_per_task": 6347.210666666667,
      "last_run_timestamp": "2026-10-24T12:24:09.591Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.7-flash-20260813",
      "model_slug": "google/gemini-3.7-flash",
      "model_name": "Google: Gemini 3.7 Flash",
      "provider_name": "Google",
      "accuracy": 0.7831052054794521,
      "accuracy_std_dev": 0.017155537965592597,
      "run_count": 3,
      "total_questions": 146,
      "avg_cost_per_task_usd": 0.14992966438356164,
      "avg_duration_per_task_ms": 183129.3904109589,
      "avg_output_tokens_per_task": 16313.164383561643,
      "last_run_timestamp": "2026-10-15T21:26:28.195Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "GMICloud",
      "accuracy": 0.782963111111111,
      "accuracy_std_dev": 0.025936403174964765,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.07150440444444443,
      "avg_duration_per_task_ms": 328376.76444444444,
      "avg_output_tokens_per_task": 17192.448888888888,
      "last_run_timestamp": "2026-10-24T19:51:50.739Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Novita",
      "accuracy": 0.7826668,
      "accuracy_std_dev": 0.01717857063204037,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.01173768,
      "avg_duration_per_task_ms": 599702.892,
      "avg_output_tokens_per_task": 22568,
      "last_run_timestamp": "2026-10-24T14:27:48.875Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.78166675,
      "accuracy_std_dev": 0.015183827290492351,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.014457230000000001,
      "avg_duration_per_task_ms": 401220.695,
      "avg_output_tokens_per_task": 21060.795,
      "last_run_timestamp": "2026-10-24T13:36:02.112Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Novita",
      "accuracy": 0.7809524285714285,
      "accuracy_std_dev": 0.025053077443455044,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.08778324,
      "avg_duration_per_task_ms": 269653.90285714285,
      "avg_output_tokens_per_task": 20106.945714285714,
      "last_run_timestamp": "2026-10-24T12:40:07.007Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Crusoe",
      "accuracy": 0.7809524285714285,
      "accuracy_std_dev": 0.011507741653303821,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.08243106857142857,
      "avg_duration_per_task_ms": 42328.47714285714,
      "avg_output_tokens_per_task": 4763.78,
      "last_run_timestamp": "2026-10-24T12:26:24.296Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Z.AI",
      "accuracy": 0.7807406666666666,
      "accuracy_std_dev": 0.01923075098204262,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.08175265555555557,
      "avg_duration_per_task_ms": 137863.5,
      "avg_output_tokens_per_task": 4670.331111111111,
      "last_run_timestamp": "2026-10-24T13:21:36.785Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "BaseTen",
      "accuracy": 0.7804090721205597,
      "accuracy_std_dev": 0.030493720143653286,
      "run_count": 19,
      "total_questions": 929,
      "avg_cost_per_task_usd": 0.07802858557588804,
      "avg_duration_per_task_ms": 76989.98815931109,
      "avg_output_tokens_per_task": 4254.709364908504,
      "last_run_timestamp": "2026-10-24T12:32:51.108Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "amazon/nova-micro-v1",
      "model_slug": "amazon/nova-micro-v1",
      "model_name": "Amazon: Nova Micro 1.0",
      "provider_name": "auto-routing",
      "accuracy": 0.7800003333333333,
      "accuracy_std_dev": 0.018856180831641284,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 1.2757919799999997,
      "avg_duration_per_task_ms": 102078.39333333333,
      "avg_output_tokens_per_task": 5685.653333333334,
      "last_run_timestamp": "2026-10-01T19:04:24.509Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "auto-routing",
      "accuracy": 0.7800001111111111,
      "accuracy_std_dev": 0.0235177999322709,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.06914864222222221,
      "avg_duration_per_task_ms": 267980.70666666667,
      "avg_output_tokens_per_task": 14178.462222222222,
      "last_run_timestamp": "2026-10-24T12:19:13.995Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "ModelRun",
      "accuracy": 0.78,
      "accuracy_std_dev": 0.005443582705045132,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.12296124000000001,
      "avg_duration_per_task_ms": 145422.42,
      "avg_output_tokens_per_task": 4596.633333333333,
      "last_run_timestamp": "2026-10-24T13:02:59.045Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Io Net",
      "accuracy": 0.7799998,
      "accuracy_std_dev": 0.01633001326882497,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.055028052,
      "avg_duration_per_task_ms": 146838.836,
      "avg_output_tokens_per_task": 4580.112,
      "last_run_timestamp": "2026-10-17T10:59:26.084Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-opus-5-20260723",
      "model_slug": "anthropic/claude-opus-5",
      "model_name": "Anthropic: Claude Opus 5",
      "provider_name": "Google",
      "accuracy": 0.7794594324324324,
      "accuracy_std_dev": 0.024130720890140577,
      "run_count": 37,
      "total_questions": 1850,
      "avg_cost_per_task_usd": 0.5428561410810812,
      "avg_duration_per_task_ms": 112001.87405405406,
      "avg_output_tokens_per_task": 7201.864864864865,
      "last_run_timestamp": "2026-10-24T11:58:58.417Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-opus-5-20260723",
      "model_slug": "anthropic/claude-opus-5",
      "model_name": "Anthropic: Claude Opus 5",
      "provider_name": "Azure",
      "accuracy": 0.7794445,
      "accuracy_std_dev": 0.022926342067223316,
      "run_count": 24,
      "total_questions": 1200,
      "avg_cost_per_task_usd": 0.47765851166666656,
      "avg_duration_per_task_ms": 106942.50333333333,
      "avg_output_tokens_per_task": 7221.848333333333,
      "last_run_timestamp": "2026-10-24T11:37:51.375Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "auto-routing",
      "accuracy": 0.7793333,
      "accuracy_std_dev": 0.02393506864853326,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.09867859,
      "avg_duration_per_task_ms": 301600.814,
      "avg_output_tokens_per_task": 20737.49,
      "last_run_timestamp": "2026-10-24T11:44:08.040Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Nebius",
      "accuracy": 0.7792640401337794,
      "accuracy_std_dev": 0.028913483122066055,
      "run_count": 6,
      "total_questions": 299,
      "avg_cost_per_task_usd": 0.1886598294314381,
      "avg_duration_per_task_ms": 56557.29765886287,
      "avg_output_tokens_per_task": 4653.71237458194,
      "last_run_timestamp": "2026-10-24T12:16:13.900Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-opus-5-20260723",
      "model_slug": "anthropic/claude-opus-5",
      "model_name": "Anthropic: Claude Opus 5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7790740277777776,
      "accuracy_std_dev": 0.020193614874352558,
      "run_count": 36,
      "total_questions": 1800,
      "avg_cost_per_task_usd": 0.4918987627777778,
      "avg_duration_per_task_ms": 122390.91666666667,
      "avg_output_tokens_per_task": 7175.807222222223,
      "last_run_timestamp": "2026-10-24T11:54:38.630Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Alibaba",
      "accuracy": 0.7788095,
      "accuracy_std_dev": 0.023891421292858595,
      "run_count": 14,
      "total_questions": 700,
      "avg_cost_per_task_usd": 0.13300575857142857,
      "avg_duration_per_task_ms": 319690.7614285714,
      "avg_output_tokens_per_task": 19596.084285714285,
      "last_run_timestamp": "2026-10-24T13:53:01.669Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Near AI",
      "accuracy": 0.77833325,
      "accuracy_std_dev": 0.011902275693643617,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.0336993,
      "avg_duration_per_task_ms": 220470.06,
      "avg_output_tokens_per_task": 12832.385,
      "last_run_timestamp": "2026-10-24T11:48:43.119Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7781817272727273,
      "accuracy_std_dev": 0.03929598310743943,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.1061598909090909,
      "avg_duration_per_task_ms": 180685.46181818182,
      "avg_output_tokens_per_task": 6783.716363636364,
      "last_run_timestamp": "2026-10-24T11:43:03.112Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Ambient",
      "accuracy": 0.777778,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 48,
      "avg_cost_per_task_usd": 0.012714583333333333,
      "avg_duration_per_task_ms": 516568.7291666667,
      "avg_output_tokens_per_task": 14230.208333333334,
      "last_run_timestamp": "2026-08-03T17:20:47.207Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-opus-5-20260723",
      "model_slug": "anthropic/claude-opus-5",
      "model_name": "Anthropic: Claude Opus 5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.777436,
      "accuracy_std_dev": 0.01913318832652994,
      "run_count": 13,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.446252443076923,
      "avg_duration_per_task_ms": 126382.57538461538,
      "avg_output_tokens_per_task": 7217.507692307692,
      "last_run_timestamp": "2026-10-24T11:43:26.286Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.8-opus-20260528",
      "model_slug": "anthropic/claude-opus-4.8",
      "model_name": "Anthropic: Claude Opus 4.8",
      "provider_name": "Anthropic",
      "accuracy": 0.7773333333333333,
      "accuracy_std_dev": 0.022016936243618362,
      "run_count": 15,
      "total_questions": 750,
      "avg_cost_per_task_usd": 0.49882911999999996,
      "avg_duration_per_task_ms": 126443.04666666666,
      "avg_output_tokens_per_task": 8076.794666666667,
      "last_run_timestamp": "2026-10-24T12:26:01.097Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7771429999999999,
      "accuracy_std_dev": 0.017036713725699246,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.15415246857142859,
      "avg_duration_per_task_ms": 315516.36,
      "avg_output_tokens_per_task": 19966.505714285715,
      "last_run_timestamp": "2026-10-24T14:03:17.639Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7771428571428571,
      "accuracy_std_dev": 0.018120213341763423,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.09363316857142856,
      "avg_duration_per_task_ms": 130500.70857142858,
      "avg_output_tokens_per_task": 5167.157142857143,
      "last_run_timestamp": "2026-10-24T13:07:35.631Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Friendli",
      "accuracy": 0.7770371111111111,
      "accuracy_std_dev": 0.019903844298942444,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.0686149911111111,
      "avg_duration_per_task_ms": 43266.05111111111,
      "avg_output_tokens_per_task": 4744.831111111111,
      "last_run_timestamp": "2026-10-24T12:49:42.912Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "DeepSeek",
      "accuracy": 0.7766665,
      "accuracy_std_dev": 0.0033334999999999892,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.00755077,
      "avg_duration_per_task_ms": 173908.24,
      "avg_output_tokens_per_task": 19171.69,
      "last_run_timestamp": "2026-08-08T08:27:33.509Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "DeepSeek",
      "accuracy": 0.7766665,
      "accuracy_std_dev": 0.019148513142887433,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.008114253333333333,
      "avg_duration_per_task_ms": 195047.99,
      "avg_output_tokens_per_task": 21425.716666666667,
      "last_run_timestamp": "2026-08-08T08:30:38.880Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "InferenceNet",
      "accuracy": 0.7766665,
      "accuracy_std_dev": 0.016666499999999973,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.017346550000000002,
      "avg_duration_per_task_ms": 335231.13,
      "avg_output_tokens_per_task": 20113.88,
      "last_run_timestamp": "2026-10-03T08:17:48.643Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "DeepInfra",
      "accuracy": 0.7766553164983165,
      "accuracy_std_dev": 0.020416356422812885,
      "run_count": 6,
      "total_questions": 297,
      "avg_cost_per_task_usd": 0.0789071414141414,
      "avg_duration_per_task_ms": 233953.09764309763,
      "avg_output_tokens_per_task": 11808.565656565657,
      "last_run_timestamp": "2026-10-24T19:28:54.737Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.7-opus-20260416",
      "model_slug": "anthropic/claude-opus-4.7",
      "model_name": "Anthropic: Claude Opus 4.7",
      "provider_name": "Anthropic",
      "accuracy": 0.7764444,
      "accuracy_std_dev": 0.017017152714442756,
      "run_count": 15,
      "total_questions": 750,
      "avg_cost_per_task_usd": 0.43634274400000006,
      "avg_duration_per_task_ms": 93543.88666666667,
      "avg_output_tokens_per_task": 5699.424,
      "last_run_timestamp": "2026-10-24T12:19:07.891Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "StreamLake",
      "accuracy": 0.7759999000000001,
      "accuracy_std_dev": 0.02174607193242035,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.08178571999999999,
      "avg_duration_per_task_ms": 336667.87,
      "avg_output_tokens_per_task": 20194.232,
      "last_run_timestamp": "2026-10-24T12:55:53.240Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Wafer",
      "accuracy": 0.7758335,
      "accuracy_std_dev": 0.014883874571159226,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.12466751749999999,
      "avg_duration_per_task_ms": 254510.0675,
      "avg_output_tokens_per_task": 23726.7775,
      "last_run_timestamp": "2026-10-24T12:19:24.121Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7755444170854271,
      "accuracy_std_dev": 0.02087028870637634,
      "run_count": 8,
      "total_questions": 398,
      "avg_cost_per_task_usd": 0.15095997738693467,
      "avg_duration_per_task_ms": 341068.3417085427,
      "avg_output_tokens_per_task": 20195.442211055277,
      "last_run_timestamp": "2026-10-24T13:15:20.470Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "DeepInfra",
      "accuracy": 0.7753625384615385,
      "accuracy_std_dev": 0.017323958686038136,
      "run_count": 6,
      "total_questions": 299,
      "avg_cost_per_task_usd": 0.06710363879598663,
      "avg_duration_per_task_ms": 1125092.9698996656,
      "avg_output_tokens_per_task": 36573.67558528428,
      "last_run_timestamp": "2026-10-24T13:26:35.887Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.8-opus-20260528",
      "model_slug": "anthropic/claude-opus-4.8",
      "model_name": "Anthropic: Claude Opus 4.8",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.7753333,
      "accuracy_std_dev": 0.025307041621058755,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.48832906800000003,
      "avg_duration_per_task_ms": 125426.254,
      "avg_output_tokens_per_task": 7994.034,
      "last_run_timestamp": "2026-10-24T12:05:00.275Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-fable-5.1-20260831",
      "model_slug": "anthropic/claude-fable-5.1",
      "model_name": "Anthropic: Claude Fable 5.1",
      "provider_name": "Google",
      "accuracy": 0.7746668,
      "accuracy_std_dev": 0.012927338819726197,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.758212416,
      "avg_duration_per_task_ms": 112131.876,
      "avg_output_tokens_per_task": 6866.944,
      "last_run_timestamp": "2026-10-01T08:05:27.141Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-sonnet-5.5-20260928",
      "model_slug": "anthropic/claude-sonnet-5.5",
      "model_name": "Anthropic: Claude Sonnet 5.5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.7746666,
      "accuracy_std_dev": 0.02508632562652409,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.16812989199999998,
      "avg_duration_per_task_ms": 82939.64,
      "avg_output_tokens_per_task": 6304.564,
      "last_run_timestamp": "2026-10-24T11:50:50.507Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Cloudflare",
      "accuracy": 0.7745533125,
      "accuracy_std_dev": 0.016606913050345267,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.08764765178571429,
      "avg_duration_per_task_ms": 134355.6294642857,
      "avg_output_tokens_per_task": 4417.004464285715,
      "last_run_timestamp": "2026-10-24T12:55:13.106Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Alibaba",
      "accuracy": 0.7743589230769231,
      "accuracy_std_dev": 0.016086511681250408,
      "run_count": 13,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.12578370615384615,
      "avg_duration_per_task_ms": 93781.64923076922,
      "avg_output_tokens_per_task": 4504.9,
      "last_run_timestamp": "2026-10-24T13:03:36.417Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "DeepInfra",
      "accuracy": 0.7742857142857142,
      "accuracy_std_dev": 0.026945723560596418,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.21623159142857146,
      "avg_duration_per_task_ms": 562306.5571428571,
      "avg_output_tokens_per_task": 12602.945714285714,
      "last_run_timestamp": "2026-10-10T11:06:43.832Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "DekaLLM",
      "accuracy": 0.7740741111111111,
      "accuracy_std_dev": 0.022098307604602994,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.05655496222222222,
      "avg_duration_per_task_ms": 155063.69555555555,
      "avg_output_tokens_per_task": 16071.293333333333,
      "last_run_timestamp": "2026-10-24T11:33:51.404Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.8-opus-20260528",
      "model_slug": "anthropic/claude-opus-4.8",
      "model_name": "Anthropic: Claude Opus 4.8",
      "provider_name": "Azure",
      "accuracy": 0.7739130434782608,
      "accuracy_std_dev": 0.029285348993887062,
      "run_count": 23,
      "total_questions": 1150,
      "avg_cost_per_task_usd": 0.5129370408695653,
      "avg_duration_per_task_ms": 128405.80086956522,
      "avg_output_tokens_per_task": 7970.952173913043,
      "last_run_timestamp": "2026-10-24T11:35:03.985Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.7-opus-20260416",
      "model_slug": "anthropic/claude-opus-4.7",
      "model_name": "Anthropic: Claude Opus 4.7",
      "provider_name": "Azure",
      "accuracy": 0.7738888333333334,
      "accuracy_std_dev": 0.018749437853054562,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.4345813583333333,
      "avg_duration_per_task_ms": 98445.79833333334,
      "avg_output_tokens_per_task": 5637.391666666666,
      "last_run_timestamp": "2026-10-24T11:36:40.498Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.7-opus-20260416",
      "model_slug": "anthropic/claude-opus-4.7",
      "model_name": "Anthropic: Claude Opus 4.7",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7738887916666667,
      "accuracy_std_dev": 0.031145842469981935,
      "run_count": 24,
      "total_questions": 1200,
      "avg_cost_per_task_usd": 0.4443571025000001,
      "avg_duration_per_task_ms": 79517.7575,
      "avg_output_tokens_per_task": 5656.515,
      "last_run_timestamp": "2026-10-24T12:17:57.222Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "Parasail",
      "accuracy": 0.7738695201005025,
      "accuracy_std_dev": 0.034154093534853214,
      "run_count": 8,
      "total_questions": 398,
      "avg_cost_per_task_usd": 0.18507628140703516,
      "avg_duration_per_task_ms": 590522.5326633166,
      "avg_output_tokens_per_task": 28427.434673366835,
      "last_run_timestamp": "2026-10-24T19:18:27.046Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Relace",
      "accuracy": 0.7733335,
      "accuracy_std_dev": 0.02708037419885479,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.13855224,
      "avg_duration_per_task_ms": 310185.72,
      "avg_output_tokens_per_task": 21233.035,
      "last_run_timestamp": "2026-10-24T12:39:23.138Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Reka",
      "accuracy": 0.7733335,
      "accuracy_std_dev": 0.00666650000000002,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.11407636,
      "avg_duration_per_task_ms": 360681.75,
      "avg_output_tokens_per_task": 19997.86,
      "last_run_timestamp": "2026-10-01T22:40:52.402Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7733335,
      "accuracy_std_dev": 0.013333499999999998,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01881542,
      "avg_duration_per_task_ms": 96553.18,
      "avg_output_tokens_per_task": 11186.19,
      "last_run_timestamp": "2026-10-24T11:34:09.882Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Io Net",
      "accuracy": 0.7733335,
      "accuracy_std_dev": 0.013333499999999998,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.10867763,
      "avg_duration_per_task_ms": 521628.03,
      "avg_output_tokens_per_task": 12092.47,
      "last_run_timestamp": "2026-09-19T08:11:59.798Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "DeepInfra",
      "accuracy": 0.7733335,
      "accuracy_std_dev": 0.03126950946609175,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.068859715,
      "avg_duration_per_task_ms": 704699.005,
      "avg_output_tokens_per_task": 13178.825,
      "last_run_timestamp": "2026-10-24T12:26:26.393Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-sonnet-5.5-20260928",
      "model_slug": "anthropic/claude-sonnet-5.5",
      "model_name": "Anthropic: Claude Sonnet 5.5",
      "provider_name": "Anthropic",
      "accuracy": 0.7733334,
      "accuracy_std_dev": 0.013984054413509715,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.17623406799999997,
      "avg_duration_per_task_ms": 74817.192,
      "avg_output_tokens_per_task": 6393.728,
      "last_run_timestamp": "2026-10-24T11:36:43.533Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Baidu",
      "accuracy": 0.7733332857142857,
      "accuracy_std_dev": 0.022537305934283946,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.08668051142857143,
      "avg_duration_per_task_ms": 364495.4428571428,
      "avg_output_tokens_per_task": 20483.94,
      "last_run_timestamp": "2026-10-24T12:45:56.375Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Reka",
      "accuracy": 0.773333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.012345539999999999,
      "avg_duration_per_task_ms": 657478.84,
      "avg_output_tokens_per_task": 14196.68,
      "last_run_timestamp": "2026-10-03T00:00:06.435Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "DeepInfra",
      "accuracy": 0.773333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.0051899,
      "avg_duration_per_task_ms": 123477.22,
      "avg_output_tokens_per_task": 11624.46,
      "last_run_timestamp": "2026-10-20T20:17:11.291Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "GMICloud",
      "accuracy": 0.773333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.26629643999999997,
      "avg_duration_per_task_ms": 438741.58,
      "avg_output_tokens_per_task": 27876.64,
      "last_run_timestamp": "2026-08-29T16:58:30.455Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "Makora",
      "accuracy": 0.773333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.11485508,
      "avg_duration_per_task_ms": 381375.48,
      "avg_output_tokens_per_task": 14199.1,
      "last_run_timestamp": "2026-10-13T03:50:19.636Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "StreamLake",
      "accuracy": 0.773333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.057549260000000005,
      "avg_duration_per_task_ms": 140710.34,
      "avg_output_tokens_per_task": 4743.94,
      "last_run_timestamp": "2026-09-11T04:09:34.110Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-opus-5-20260723",
      "model_slug": "anthropic/claude-opus-5",
      "model_name": "Anthropic: Claude Opus 5",
      "provider_name": "Anthropic",
      "accuracy": 0.77277775,
      "accuracy_std_dev": 0.02530680323919834,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.47471332,
      "avg_duration_per_task_ms": 112995.27166666667,
      "avg_output_tokens_per_task": 7220.64,
      "last_run_timestamp": "2026-10-24T11:37:27.338Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Parasail",
      "accuracy": 0.7726668,
      "accuracy_std_dev": 0.0325508757356849,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.065897038,
      "avg_duration_per_task_ms": 245761.102,
      "avg_output_tokens_per_task": 11799.874,
      "last_run_timestamp": "2026-10-24T12:58:13.946Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "GMICloud",
      "accuracy": 0.7726668,
      "accuracy_std_dev": 0.027030792318391252,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.058849759999999994,
      "avg_duration_per_task_ms": 890114.912,
      "avg_output_tokens_per_task": 40932.588,
      "last_run_timestamp": "2026-10-24T13:04:00.434Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Parasail",
      "accuracy": 0.7725925555555556,
      "accuracy_std_dev": 0.027115371049200147,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.02128270888888889,
      "avg_duration_per_task_ms": 231570.07111111112,
      "avg_output_tokens_per_task": 12386.437777777777,
      "last_run_timestamp": "2026-10-24T19:11:05.620Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "DeepInfra",
      "accuracy": 0.7723808571428572,
      "accuracy_std_dev": 0.03553283429420726,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.0410159,
      "avg_duration_per_task_ms": 732886.1685714286,
      "avg_output_tokens_per_task": 12629.145714285714,
      "last_run_timestamp": "2026-10-24T12:41:08.943Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Cerebras",
      "accuracy": 0.7722223333333333,
      "accuracy_std_dev": 0.051949695516164705,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.2672586,
      "avg_duration_per_task_ms": 26529.913333333334,
      "avg_output_tokens_per_task": 12210.046666666667,
      "last_run_timestamp": "2026-10-24T12:34:49.317Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta/muse-glimmer-30b-20260810",
      "model_slug": "meta/muse-glimmer-30b",
      "model_name": "Meta: Muse Glimmer 30B",
      "provider_name": "DeepInfra",
      "accuracy": 0.7722222222222223,
      "accuracy_std_dev": 0.03095202181433487,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.03593436666666667,
      "avg_duration_per_task_ms": 446735.1666666667,
      "avg_output_tokens_per_task": 13815.786666666667,
      "last_run_timestamp": "2026-10-08T22:42:47.098Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-fable-5.1-20260831",
      "model_slug": "anthropic/claude-fable-5.1",
      "model_name": "Anthropic: Claude Fable 5.1",
      "provider_name": "Azure",
      "accuracy": 0.772,
      "accuracy_std_dev": 0.02679938639596063,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.808827636,
      "avg_duration_per_task_ms": 183544.5,
      "avg_output_tokens_per_task": 6940.7,
      "last_run_timestamp": "2026-10-01T08:03:00.737Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "DeepInfra",
      "accuracy": 0.771929874493927,
      "accuracy_std_dev": 0.030837182396580943,
      "run_count": 5,
      "total_questions": 247,
      "avg_cost_per_task_usd": 0.011004716599190284,
      "avg_duration_per_task_ms": 1072772.7165991904,
      "avg_output_tokens_per_task": 18798.943319838058,
      "last_run_timestamp": "2026-10-17T12:03:02.755Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "DeepInfra",
      "accuracy": 0.771867598108747,
      "accuracy_std_dev": 0.030418188929615536,
      "run_count": 17,
      "total_questions": 846,
      "avg_cost_per_task_usd": 0.03435285106382978,
      "avg_duration_per_task_ms": 501754.59338061465,
      "avg_output_tokens_per_task": 12458.040189125295,
      "last_run_timestamp": "2026-10-24T19:21:21.689Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Baidu",
      "accuracy": 0.7718518888888889,
      "accuracy_std_dev": 0.020313150614233704,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.04094091999999999,
      "avg_duration_per_task_ms": 200471.00222222222,
      "avg_output_tokens_per_task": 13143.968888888889,
      "last_run_timestamp": "2026-10-24T19:48:57.247Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Novita",
      "accuracy": 0.7718518888888889,
      "accuracy_std_dev": 0.02102986445480936,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.05397537555555555,
      "avg_duration_per_task_ms": 110633.31777777777,
      "avg_output_tokens_per_task": 4708.497777777778,
      "last_run_timestamp": "2026-10-24T13:04:30.805Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Venice",
      "accuracy": 0.7714285714285715,
      "accuracy_std_dev": 0.02435571983191243,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.21782835142857143,
      "avg_duration_per_task_ms": 377656.6857142857,
      "avg_output_tokens_per_task": 17802.317142857144,
      "last_run_timestamp": "2026-10-24T12:37:29.521Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "DeepInfra",
      "accuracy": 0.7714285714285715,
      "accuracy_std_dev": 0.023289574128825135,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.044844714285714286,
      "avg_duration_per_task_ms": 100224.04857142858,
      "avg_output_tokens_per_task": 4956.551428571429,
      "last_run_timestamp": "2026-10-24T12:38:46.757Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Chutes",
      "accuracy": 0.7713333,
      "accuracy_std_dev": 0.0351567708046402,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.060528674000000005,
      "avg_duration_per_task_ms": 302911.6,
      "avg_output_tokens_per_task": 12614.202,
      "last_run_timestamp": "2026-10-24T12:18:13.146Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "AkashML",
      "accuracy": 0.7711946416184972,
      "accuracy_std_dev": 0.02510296875371517,
      "run_count": 14,
      "total_questions": 692,
      "avg_cost_per_task_usd": 0.054176160404624286,
      "avg_duration_per_task_ms": 260129.0578034682,
      "avg_output_tokens_per_task": 11031.143063583815,
      "last_run_timestamp": "2026-10-24T12:08:01.603Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.1-20251113",
      "model_slug": "openai/gpt-5.1",
      "model_name": "OpenAI: GPT-5.1",
      "provider_name": "auto-routing",
      "accuracy": 0.771111,
      "accuracy_std_dev": 0.04085521560339636,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.4149127733333333,
      "avg_duration_per_task_ms": 472514.2266666667,
      "avg_output_tokens_per_task": 36470.113333333335,
      "last_run_timestamp": "2026-10-01T16:00:45.071Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "SiliconFlow",
      "accuracy": 0.77083325,
      "accuracy_std_dev": 0.040679972209153495,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.11823473500000002,
      "avg_duration_per_task_ms": 271446.195,
      "avg_output_tokens_per_task": 11914.5925,
      "last_run_timestamp": "2026-10-17T21:03:47.447Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Io Net",
      "accuracy": 0.770833,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 48,
      "avg_cost_per_task_usd": 0.019129833333333335,
      "avg_duration_per_task_ms": 643112.25,
      "avg_output_tokens_per_task": 13416,
      "last_run_timestamp": "2026-08-01T08:05:58.651Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Parasail",
      "accuracy": 0.7707407777777777,
      "accuracy_std_dev": 0.023188894510460925,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.11877632888888891,
      "avg_duration_per_task_ms": 486710.77777777775,
      "avg_output_tokens_per_task": 21345.64,
      "last_run_timestamp": "2026-10-17T11:37:24.257Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta/muse-glimmer-30b-20260810",
      "model_slug": "meta/muse-glimmer-30b",
      "model_name": "Meta: Muse Glimmer 30B",
      "provider_name": "Parasail",
      "accuracy": 0.7706935637583894,
      "accuracy_std_dev": 0.013644816256244217,
      "run_count": 3,
      "total_questions": 149,
      "avg_cost_per_task_usd": 0.09587820805369127,
      "avg_duration_per_task_ms": 288021.6644295302,
      "avg_output_tokens_per_task": 13114.979865771813,
      "last_run_timestamp": "2026-08-29T08:48:12.957Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "NextBit",
      "accuracy": 0.7706666,
      "accuracy_std_dev": 0.0204828947622156,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.10916058,
      "avg_duration_per_task_ms": 258128.42,
      "avg_output_tokens_per_task": 11359.708,
      "last_run_timestamp": "2026-10-03T16:07:05.245Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7706664000000001,
      "accuracy_std_dev": 0.02686579651229421,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.079545256,
      "avg_duration_per_task_ms": 210434.584,
      "avg_output_tokens_per_task": 9339.868,
      "last_run_timestamp": "2026-08-29T16:11:58.076Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.8-opus-20260528",
      "model_slug": "anthropic/claude-opus-4.8",
      "model_name": "Anthropic: Claude Opus 4.8",
      "provider_name": "Google",
      "accuracy": 0.7704762000000002,
      "accuracy_std_dev": 0.027384900751848115,
      "run_count": 35,
      "total_questions": 1750,
      "avg_cost_per_task_usd": 0.5769547302857142,
      "avg_duration_per_task_ms": 116495.80342857142,
      "avg_output_tokens_per_task": 8134.803428571428,
      "last_run_timestamp": "2026-10-24T12:27:09.155Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Mistral",
      "accuracy": 0.7704761904761904,
      "accuracy_std_dev": 0.01947142566164718,
      "run_count": 21,
      "total_questions": 1050,
      "avg_cost_per_task_usd": 0.06930062190476191,
      "avg_duration_per_task_ms": 74426.91428571429,
      "avg_output_tokens_per_task": 4672.551428571429,
      "last_run_timestamp": "2026-10-24T12:23:14.304Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Parasail",
      "accuracy": 0.770416625,
      "accuracy_std_dev": 0.0211762125800242,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.079187595,
      "avg_duration_per_task_ms": 95610.7625,
      "avg_output_tokens_per_task": 4413.6475,
      "last_run_timestamp": "2026-10-24T13:02:59.051Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7703703333333334,
      "accuracy_std_dev": 0.030325214851304485,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.10216244444444443,
      "avg_duration_per_task_ms": 234870.98,
      "avg_output_tokens_per_task": 11319.771111111111,
      "last_run_timestamp": "2026-10-24T19:49:41.079Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.5-opus-20251124",
      "model_slug": "anthropic/claude-opus-4.5",
      "model_name": "Anthropic: Claude Opus 4.5",
      "provider_name": "auto-routing",
      "accuracy": 0.7700000833333334,
      "accuracy_std_dev": 0.020276053200998522,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.5360124333333335,
      "avg_duration_per_task_ms": 174546.98666666666,
      "avg_output_tokens_per_task": 10626.863333333333,
      "last_run_timestamp": "2026-10-24T11:55:03.241Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Relace",
      "accuracy": 0.77,
      "accuracy_std_dev": 0.018855956915521412,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.065556992,
      "avg_duration_per_task_ms": 170454.924,
      "avg_output_tokens_per_task": 10274.532,
      "last_run_timestamp": "2026-10-24T19:27:30.834Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "StreamLake",
      "accuracy": 0.77,
      "accuracy_std_dev": 0.03454471498507404,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.017919986000000002,
      "avg_duration_per_task_ms": 291502.854,
      "avg_output_tokens_per_task": 11512.416,
      "last_run_timestamp": "2026-10-24T19:22:55.840Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta/muse-spark-open-20261015",
      "model_slug": "meta/muse-spark-open",
      "model_name": "Meta: Muse Spark Open",
      "provider_name": "Fireworks",
      "accuracy": 0.77,
      "accuracy_std_dev": 0.010000000000000009,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.10142027,
      "avg_duration_per_task_ms": 175426.11,
      "avg_output_tokens_per_task": 10178.13,
      "last_run_timestamp": "2026-10-24T11:42:19.491Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "StreamLake",
      "accuracy": 0.77,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.05546486,
      "avg_duration_per_task_ms": 677432.76,
      "avg_output_tokens_per_task": 33432.56,
      "last_run_timestamp": "2026-10-24T13:08:41.848Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Inceptron",
      "accuracy": 0.77,
      "accuracy_std_dev": 0.022852121225391757,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.047772247500000004,
      "avg_duration_per_task_ms": 117312.3025,
      "avg_output_tokens_per_task": 4701.9325,
      "last_run_timestamp": "2026-10-24T12:40:31.665Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.5-opus-20251124",
      "model_slug": "anthropic/claude-opus-4.5",
      "model_name": "Anthropic: Claude Opus 4.5",
      "provider_name": "Anthropic",
      "accuracy": 0.7697778666666667,
      "accuracy_std_dev": 0.027741171258298072,
      "run_count": 15,
      "total_questions": 750,
      "avg_cost_per_task_usd": 0.5468098066666667,
      "avg_duration_per_task_ms": 184159.76266666668,
      "avg_output_tokens_per_task": 10508.508,
      "last_run_timestamp": "2026-10-24T12:18:32.884Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Phala",
      "accuracy": 0.7696297777777777,
      "accuracy_std_dev": 0.028477632067516422,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.1409867,
      "avg_duration_per_task_ms": 323727.4288888889,
      "avg_output_tokens_per_task": 20180.546666666665,
      "last_run_timestamp": "2026-10-24T12:56:30.925Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Venice",
      "accuracy": 0.7695238571428571,
      "accuracy_std_dev": 0.029354336946199627,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.12993965428571427,
      "avg_duration_per_task_ms": 178407.48,
      "avg_output_tokens_per_task": 8114.634285714285,
      "last_run_timestamp": "2026-10-17T12:47:59.459Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Reka",
      "accuracy": 0.7695237142857143,
      "accuracy_std_dev": 0.02386662296246901,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.11461242285714283,
      "avg_duration_per_task_ms": 472929.0742857143,
      "avg_output_tokens_per_task": 11699.722857142857,
      "last_run_timestamp": "2026-10-23T07:26:58.509Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-122b-a10b-20260224",
      "model_slug": "qwen/qwen3.5-122b-a10b",
      "model_name": "Qwen: Qwen3.5-122B-A10B",
      "provider_name": "auto-routing",
      "accuracy": 0.7694444166666666,
      "accuracy_std_dev": 0.029496861625316274,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.16601085666666665,
      "avg_duration_per_task_ms": 573194.7333333333,
      "avg_output_tokens_per_task": 34801.78666666667,
      "last_run_timestamp": "2026-10-24T18:49:47.645Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "DekaLLM",
      "accuracy": 0.769144,
      "accuracy_std_dev": 0.05227625153611027,
      "run_count": 3,
      "total_questions": 148,
      "avg_cost_per_task_usd": 0.03692847297297297,
      "avg_duration_per_task_ms": 119447.77702702703,
      "avg_output_tokens_per_task": 4304.4121621621625,
      "last_run_timestamp": "2026-10-24T12:53:05.664Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.8-opus-20260528",
      "model_slug": "anthropic/claude-opus-4.8",
      "model_name": "Anthropic: Claude Opus 4.8",
      "provider_name": "auto-routing",
      "accuracy": 0.7688890833333333,
      "accuracy_std_dev": 0.019876007624178182,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.4929804266666666,
      "avg_duration_per_task_ms": 127454.61666666667,
      "avg_output_tokens_per_task": 8112.49,
      "last_run_timestamp": "2026-10-24T11:39:26.193Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Modal",
      "accuracy": 0.7688888888888888,
      "accuracy_std_dev": 0.02914421358022683,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.08110119333333334,
      "avg_duration_per_task_ms": 48308.12,
      "avg_output_tokens_per_task": 4858.82,
      "last_run_timestamp": "2026-10-24T12:49:25.219Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.5-flash-lite-20260721",
      "model_slug": "google/gemini-3.5-flash-lite",
      "model_name": "Google: Gemini 3.5 Flash Lite",
      "provider_name": "auto-routing",
      "accuracy": 0.7688886666666668,
      "accuracy_std_dev": 0.021998731726069008,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.09783254,
      "avg_duration_per_task_ms": 102292.53333333334,
      "avg_output_tokens_per_task": 23933.02,
      "last_run_timestamp": "2026-10-01T19:04:42.463Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Sail Research",
      "accuracy": 0.7688235294117647,
      "accuracy_std_dev": 0.026394292226378947,
      "run_count": 17,
      "total_questions": 850,
      "avg_cost_per_task_usd": 0.04698485764705882,
      "avg_duration_per_task_ms": 91736.21058823529,
      "avg_output_tokens_per_task": 4343.287058823529,
      "last_run_timestamp": "2026-10-24T12:48:21.011Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Friendli",
      "accuracy": 0.7685714285714286,
      "accuracy_std_dev": 0.017715488952222757,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.029891637142857145,
      "avg_duration_per_task_ms": 131231.82571428572,
      "avg_output_tokens_per_task": 12135.551428571429,
      "last_run_timestamp": "2026-10-24T18:43:12.442Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "CoreWeave",
      "accuracy": 0.7684563154362417,
      "accuracy_std_dev": 0.006017520216464235,
      "run_count": 3,
      "total_questions": 149,
      "avg_cost_per_task_usd": 0.1581116510067114,
      "avg_duration_per_task_ms": 537705.5637583892,
      "avg_output_tokens_per_task": 10979.71812080537,
      "last_run_timestamp": "2026-08-08T08:27:49.545Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "auto-routing",
      "accuracy": 0.768333375,
      "accuracy_std_dev": 0.032787198976649036,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.05900729999999999,
      "avg_duration_per_task_ms": 114247.045,
      "avg_output_tokens_per_task": 6145.8525,
      "last_run_timestamp": "2026-10-24T12:32:52.472Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Venice",
      "accuracy": 0.7681483333333333,
      "accuracy_std_dev": 0.02753713940602159,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.1370029888888889,
      "avg_duration_per_task_ms": 279741.1088888889,
      "avg_output_tokens_per_task": 11838.22,
      "last_run_timestamp": "2026-10-24T19:21:41.094Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "CoreWeave",
      "accuracy": 0.7680176486486487,
      "accuracy_std_dev": 0.018065309426264085,
      "run_count": 3,
      "total_questions": 148,
      "avg_cost_per_task_usd": 0.06668277702702702,
      "avg_duration_per_task_ms": 92879.05405405405,
      "avg_output_tokens_per_task": 4281.108108108108,
      "last_run_timestamp": "2026-10-24T13:05:47.020Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Cloudflare",
      "accuracy": 0.768,
      "accuracy_std_dev": 0.028875333573830795,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.15511289399999997,
      "avg_duration_per_task_ms": 603929.03,
      "avg_output_tokens_per_task": 12401.186,
      "last_run_timestamp": "2026-10-24T12:59:12.979Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Relace",
      "accuracy": 0.768,
      "accuracy_std_dev": 0.011469922022402766,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.030947572,
      "avg_duration_per_task_ms": 58222.324,
      "avg_output_tokens_per_task": 4803.824,
      "last_run_timestamp": "2026-10-24T12:19:36.645Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.8-opus-20260528",
      "model_slug": "anthropic/claude-opus-4.8",
      "model_name": "Anthropic: Claude Opus 4.8",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7679999714285717,
      "accuracy_std_dev": 0.020082359314277674,
      "run_count": 35,
      "total_questions": 1750,
      "avg_cost_per_task_usd": 0.5321429325714288,
      "avg_duration_per_task_ms": 119418.44171428571,
      "avg_output_tokens_per_task": 8085.556571428571,
      "last_run_timestamp": "2026-10-24T12:05:52.446Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "auto-routing",
      "accuracy": 0.7679998,
      "accuracy_std_dev": 0.04721585044622197,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.07315105200000001,
      "avg_duration_per_task_ms": 1015313.192,
      "avg_output_tokens_per_task": 46954.436,
      "last_run_timestamp": "2026-10-24T13:12:06.662Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7679165625000001,
      "accuracy_std_dev": 0.021885852467886505,
      "run_count": 16,
      "total_questions": 800,
      "avg_cost_per_task_usd": 0.019775938749999996,
      "avg_duration_per_task_ms": 242328.7,
      "avg_output_tokens_per_task": 16787.80875,
      "last_run_timestamp": "2026-10-24T12:30:46.104Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Io Net",
      "accuracy": 0.7675584849498328,
      "accuracy_std_dev": 0.01898112793159446,
      "run_count": 6,
      "total_questions": 299,
      "avg_cost_per_task_usd": 0.01286576254180602,
      "avg_duration_per_task_ms": 339011.96321070235,
      "avg_output_tokens_per_task": 16426.635451505015,
      "last_run_timestamp": "2026-10-24T15:07:38.934Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Phala",
      "accuracy": 0.767500125,
      "accuracy_std_dev": 0.01839017891727471,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.05273846000000001,
      "avg_duration_per_task_ms": 114111.54,
      "avg_output_tokens_per_task": 4459.3875,
      "last_run_timestamp": "2026-10-24T13:02:56.928Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Decart",
      "accuracy": 0.7675,
      "accuracy_std_dev": 0.019561171859068173,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0578507225,
      "avg_duration_per_task_ms": 50974.42,
      "avg_output_tokens_per_task": 4677.1725,
      "last_run_timestamp": "2026-10-24T12:36:03.010Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "StreamLake",
      "accuracy": 0.7673333999999999,
      "accuracy_std_dev": 0.027235571513004833,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.044897582,
      "avg_duration_per_task_ms": 196958.324,
      "avg_output_tokens_per_task": 9659.966,
      "last_run_timestamp": "2026-10-24T19:32:27.920Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Alibaba",
      "accuracy": 0.767037,
      "accuracy_std_dev": 0.029247427051128982,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.11265956888888888,
      "avg_duration_per_task_ms": 168759.15333333332,
      "avg_output_tokens_per_task": 9396.86,
      "last_run_timestamp": "2026-10-24T20:38:25.417Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "Novita",
      "accuracy": 0.7666669999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.0337152,
      "avg_duration_per_task_ms": 192245.9,
      "avg_output_tokens_per_task": 17702.16,
      "last_run_timestamp": "2026-09-01T12:41:31.535Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta/muse-spark-open-20261015",
      "model_slug": "meta/muse-spark-open",
      "model_name": "Meta: Muse Spark Open",
      "provider_name": "Modal",
      "accuracy": 0.7666669999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.30783086,
      "avg_duration_per_task_ms": 74384.36,
      "avg_output_tokens_per_task": 8511.9,
      "last_run_timestamp": "2026-10-24T11:55:43.315Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Parasail",
      "accuracy": 0.7666669999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.0912162,
      "avg_duration_per_task_ms": 191018.42,
      "avg_output_tokens_per_task": 9260.76,
      "last_run_timestamp": "2026-08-04T03:12:10.013Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "NextBit",
      "accuracy": 0.76666675,
      "accuracy_std_dev": 0.024267078752653787,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.036940722499999995,
      "avg_duration_per_task_ms": 209470.17,
      "avg_output_tokens_per_task": 17358.2575,
      "last_run_timestamp": "2026-10-03T08:39:48.671Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Cloudflare",
      "accuracy": 0.76666675,
      "accuracy_std_dev": 0.01666670000142501,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.15229175250000002,
      "avg_duration_per_task_ms": 351001.0475,
      "avg_output_tokens_per_task": 16911.185,
      "last_run_timestamp": "2026-10-24T20:11:33.520Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Wafer",
      "accuracy": 0.7666666666666667,
      "accuracy_std_dev": 0.05192598721470999,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.02354878666666667,
      "avg_duration_per_task_ms": 257749.54666666666,
      "avg_output_tokens_per_task": 13740.606666666667,
      "last_run_timestamp": "2026-10-24T19:12:17.948Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Novita",
      "accuracy": 0.7666666666666667,
      "accuracy_std_dev": 0.01721334538471048,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.1022698822222222,
      "avg_duration_per_task_ms": 247139.55333333334,
      "avg_output_tokens_per_task": 17955.155555555557,
      "last_run_timestamp": "2026-10-24T19:31:34.515Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.1-pro-preview-20260219",
      "model_slug": "google/gemini-3.1-pro-preview",
      "model_name": "Google: Gemini 3.1 Pro Preview",
      "provider_name": "Google",
      "accuracy": 0.7666666666666667,
      "accuracy_std_dev": 0.014401440259764175,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.3460653,
      "avg_duration_per_task_ms": 170899.55333333334,
      "avg_output_tokens_per_task": 12611.153333333334,
      "last_run_timestamp": "2026-10-01T19:04:40.361Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.7-opus-20260416",
      "model_slug": "anthropic/claude-opus-4.7",
      "model_name": "Anthropic: Claude Opus 4.7",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.7666666,
      "accuracy_std_dev": 0.03169302061400902,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.435956584,
      "avg_duration_per_task_ms": 91965.526,
      "avg_output_tokens_per_task": 5586.83,
      "last_run_timestamp": "2026-10-24T11:44:03.067Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.8-flash-20260902",
      "model_slug": "google/gemini-3.8-flash",
      "model_name": "Google: Gemini 3.8 Flash",
      "provider_name": "Google",
      "accuracy": 0.7666665,
      "accuracy_std_dev": 0.013333499999999998,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.27949224,
      "avg_duration_per_task_ms": 363191.33,
      "avg_output_tokens_per_task": 35460.86,
      "last_run_timestamp": "2026-10-15T21:31:19.497Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "DekaLLM",
      "accuracy": 0.7666665,
      "accuracy_std_dev": 0.026666499999999982,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01352679,
      "avg_duration_per_task_ms": 90848.6,
      "avg_output_tokens_per_task": 4507.38,
      "last_run_timestamp": "2026-10-10T11:18:50.119Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "DeepInfra",
      "accuracy": 0.7666663333333334,
      "accuracy_std_dev": 0.024944382578492925,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.06548607333333334,
      "avg_duration_per_task_ms": 180351.34,
      "avg_output_tokens_per_task": 8330.653333333334,
      "last_run_timestamp": "2026-08-15T08:17:17.868Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.7663690982142857,
      "accuracy_std_dev": 0.02313623805704157,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.09936721205357145,
      "avg_duration_per_task_ms": 46610.31919642857,
      "avg_output_tokens_per_task": 4301.232142857143,
      "last_run_timestamp": "2026-10-24T12:21:21.947Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "GMICloud",
      "accuracy": 0.7662500624999999,
      "accuracy_std_dev": 0.022418780493340704,
      "run_count": 16,
      "total_questions": 800,
      "avg_cost_per_task_usd": 0.01870876375,
      "avg_duration_per_task_ms": 186867.03125,
      "avg_output_tokens_per_task": 18628.71375,
      "last_run_timestamp": "2026-10-24T11:38:35.907Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Alibaba",
      "accuracy": 0.7661538461538462,
      "accuracy_std_dev": 0.023745191285965687,
      "run_count": 13,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.09027926615384615,
      "avg_duration_per_task_ms": 183761.18615384615,
      "avg_output_tokens_per_task": 11217.321538461538,
      "last_run_timestamp": "2026-10-24T20:01:36.231Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.7-opus-20260416",
      "model_slug": "anthropic/claude-opus-4.7",
      "model_name": "Anthropic: Claude Opus 4.7",
      "provider_name": "auto-routing",
      "accuracy": 0.7661111666666667,
      "accuracy_std_dev": 0.02300705876766714,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.41899941,
      "avg_duration_per_task_ms": 94556.285,
      "avg_output_tokens_per_task": 5609.2,
      "last_run_timestamp": "2026-10-24T12:23:35.338Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "auto-routing",
      "accuracy": 0.7659258888888889,
      "accuracy_std_dev": 0.023400799584840606,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.049908673333333334,
      "avg_duration_per_task_ms": 223450.7577777778,
      "avg_output_tokens_per_task": 12477.913333333334,
      "last_run_timestamp": "2026-10-24T19:41:33.487Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "DekaLLM",
      "accuracy": 0.7658335,
      "accuracy_std_dev": 0.029095183832552104,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.11472753,
      "avg_duration_per_task_ms": 363943.795,
      "avg_output_tokens_per_task": 19034.395,
      "last_run_timestamp": "2026-10-24T12:25:43.493Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.5-opus-20251124",
      "model_slug": "anthropic/claude-opus-4.5",
      "model_name": "Anthropic: Claude Opus 4.5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.765833375,
      "accuracy_std_dev": 0.0314650846333579,
      "run_count": 24,
      "total_questions": 1200,
      "avg_cost_per_task_usd": 0.56580197,
      "avg_duration_per_task_ms": 182718.72916666666,
      "avg_output_tokens_per_task": 10626.925,
      "last_run_timestamp": "2026-10-24T12:21:43.024Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7657141428571429,
      "accuracy_std_dev": 0.02265784142050462,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.01868964,
      "avg_duration_per_task_ms": 296461.9485714286,
      "avg_output_tokens_per_task": 14208.894285714287,
      "last_run_timestamp": "2026-10-24T13:09:04.430Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7656249117647059,
      "accuracy_std_dev": 0.03944849186508596,
      "run_count": 11,
      "total_questions": 544,
      "avg_cost_per_task_usd": 0.09746370772058824,
      "avg_duration_per_task_ms": 725535.5073529412,
      "avg_output_tokens_per_task": 10412.310661764706,
      "last_run_timestamp": "2026-10-24T19:52:35.108Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Parasail",
      "accuracy": 0.7655824146341464,
      "accuracy_std_dev": 0.030599644599243303,
      "run_count": 5,
      "total_questions": 246,
      "avg_cost_per_task_usd": 0.07519595528455283,
      "avg_duration_per_task_ms": 226711.57317073172,
      "avg_output_tokens_per_task": 10093.731707317073,
      "last_run_timestamp": "2026-10-10T19:17:08.161Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Z.AI",
      "accuracy": 0.7651851111111111,
      "accuracy_std_dev": 0.030107298743897986,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.009121904444444445,
      "avg_duration_per_task_ms": 193621.67333333334,
      "avg_output_tokens_per_task": 4487.197777777777,
      "last_run_timestamp": "2026-10-24T13:10:16.181Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Makora",
      "accuracy": 0.7651515252525253,
      "accuracy_std_dev": 0.02642279772753065,
      "run_count": 8,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.07224725252525253,
      "avg_duration_per_task_ms": 56760.18939393939,
      "avg_output_tokens_per_task": 4516.429292929293,
      "last_run_timestamp": "2026-10-24T12:38:39.898Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.7-opus-20260416",
      "model_slug": "anthropic/claude-opus-4.7",
      "model_name": "Anthropic: Claude Opus 4.7",
      "provider_name": "Google",
      "accuracy": 0.7650000277777779,
      "accuracy_std_dev": 0.02201993695556385,
      "run_count": 36,
      "total_questions": 1800,
      "avg_cost_per_task_usd": 0.6106352950000001,
      "avg_duration_per_task_ms": 72113.83055555556,
      "avg_output_tokens_per_task": 5697.888888888889,
      "last_run_timestamp": "2026-10-24T12:22:02.797Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Ionstream",
      "accuracy": 0.765,
      "accuracy_std_dev": 0.03840572873934304,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.08522504,
      "avg_duration_per_task_ms": 558480.23,
      "avg_output_tokens_per_task": 11022.365,
      "last_run_timestamp": "2026-08-29T17:19:42.622Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "Tencent",
      "accuracy": 0.765,
      "accuracy_std_dev": 0.02682235800596212,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.093536965,
      "avg_duration_per_task_ms": 395836.625,
      "avg_output_tokens_per_task": 14656.925,
      "last_run_timestamp": "2026-10-24T12:32:30.905Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Wafer",
      "accuracy": 0.7648484545454546,
      "accuracy_std_dev": 0.029417433910596247,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.012073312727272727,
      "avg_duration_per_task_ms": 225241.07636363636,
      "avg_output_tokens_per_task": 5963.167272727273,
      "last_run_timestamp": "2026-10-24T13:10:13.421Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Morph",
      "accuracy": 0.7647713411371238,
      "accuracy_std_dev": 0.037669588541078686,
      "run_count": 12,
      "total_questions": 598,
      "avg_cost_per_task_usd": 0.09814772575250837,
      "avg_duration_per_task_ms": 324868.6220735786,
      "avg_output_tokens_per_task": 16071.638795986622,
      "last_run_timestamp": "2026-10-24T19:50:25.118Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "OpenInference",
      "accuracy": 0.7646666,
      "accuracy_std_dev": 0.021868245879356662,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.019245832,
      "avg_duration_per_task_ms": 612553.372,
      "avg_output_tokens_per_task": 12922.088,
      "last_run_timestamp": "2026-08-29T16:47:05.637Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Ionstream",
      "accuracy": 0.7645742242152466,
      "accuracy_std_dev": 0.03220926464716497,
      "run_count": 9,
      "total_questions": 446,
      "avg_cost_per_task_usd": 0.05124673542600897,
      "avg_duration_per_task_ms": 396898.4529147982,
      "avg_output_tokens_per_task": 12964.300448430493,
      "last_run_timestamp": "2026-10-24T12:42:05.792Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Cohere",
      "accuracy": 0.7636566923076923,
      "accuracy_std_dev": 0.016849090853317095,
      "run_count": 6,
      "total_questions": 299,
      "avg_cost_per_task_usd": 0.028887086956521736,
      "avg_duration_per_task_ms": 106220.35785953177,
      "avg_output_tokens_per_task": 18019.585284280936,
      "last_run_timestamp": "2026-10-24T11:41:59.518Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.6-opus-20260205",
      "model_slug": "anthropic/claude-opus-4.6",
      "model_name": "Anthropic: Claude Opus 4.6",
      "provider_name": "Anthropic",
      "accuracy": 0.7635555333333334,
      "accuracy_std_dev": 0.026203346503494974,
      "run_count": 15,
      "total_questions": 750,
      "avg_cost_per_task_usd": 0.5199028973333334,
      "avg_duration_per_task_ms": 219812.216,
      "avg_output_tokens_per_task": 10143.609333333334,
      "last_run_timestamp": "2026-10-24T12:29:32.505Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "GMICloud",
      "accuracy": 0.7633335,
      "accuracy_std_dev": 0.0033334999999999892,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01887093,
      "avg_duration_per_task_ms": 108015.59,
      "avg_output_tokens_per_task": 13223.41,
      "last_run_timestamp": "2026-10-24T12:18:28.261Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "Darkbloom",
      "accuracy": 0.7633333333333333,
      "accuracy_std_dev": 0.026317266547944456,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.06636648333333332,
      "avg_duration_per_task_ms": 1144475.0366666666,
      "avg_output_tokens_per_task": 64035,
      "last_run_timestamp": "2026-10-24T19:42:45.616Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Sail Research",
      "accuracy": 0.7633332777777778,
      "accuracy_std_dev": 0.01808766376046748,
      "run_count": 18,
      "total_questions": 900,
      "avg_cost_per_task_usd": 0.010060852222222218,
      "avg_duration_per_task_ms": 327666.52,
      "avg_output_tokens_per_task": 14799.732222222223,
      "last_run_timestamp": "2026-10-24T13:02:06.972Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "CoreWeave",
      "accuracy": 0.7633332500000001,
      "accuracy_std_dev": 0.03605539719220273,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.157135485,
      "avg_duration_per_task_ms": 411704.8,
      "avg_output_tokens_per_task": 26150.13,
      "last_run_timestamp": "2026-08-29T08:26:04.467Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Xiaomi",
      "accuracy": 0.7633332500000001,
      "accuracy_std_dev": 0.013743584364622665,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.01189864,
      "avg_duration_per_task_ms": 642869.1,
      "avg_output_tokens_per_task": 23140.98,
      "last_run_timestamp": "2026-10-24T13:28:08.470Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Venice",
      "accuracy": 0.7631776302895322,
      "accuracy_std_dev": 0.02017755686675451,
      "run_count": 9,
      "total_questions": 449,
      "avg_cost_per_task_usd": 0.011049703786191535,
      "avg_duration_per_task_ms": 296930.06904231623,
      "avg_output_tokens_per_task": 12882.661469933184,
      "last_run_timestamp": "2026-10-24T19:59:46.522Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-122b-a10b-20260224",
      "model_slug": "qwen/qwen3.5-122b-a10b",
      "model_name": "Qwen: Qwen3.5-122B-A10B",
      "provider_name": "Novita",
      "accuracy": 0.763025131092437,
      "accuracy_std_dev": 0.03297689657885083,
      "run_count": 12,
      "total_questions": 595,
      "avg_cost_per_task_usd": 0.19578441512605044,
      "avg_duration_per_task_ms": 529598.2705882353,
      "avg_output_tokens_per_task": 27119.653781512607,
      "last_run_timestamp": "2026-10-24T19:14:04.604Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Novita",
      "accuracy": 0.762963,
      "accuracy_std_dev": 0.025745351040959277,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.08904649555555556,
      "avg_duration_per_task_ms": 300656.5133333333,
      "avg_output_tokens_per_task": 12058.504444444445,
      "last_run_timestamp": "2026-10-24T12:17:31.834Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7629628888888889,
      "accuracy_std_dev": 0.02354127566374736,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.01074868,
      "avg_duration_per_task_ms": 1059485.7177777777,
      "avg_output_tokens_per_task": 13128.448888888888,
      "last_run_timestamp": "2026-10-24T19:49:45.575Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Novita",
      "accuracy": 0.7629628888888889,
      "accuracy_std_dev": 0.031128845034949126,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.008403426666666667,
      "avg_duration_per_task_ms": 205005.57333333333,
      "avg_output_tokens_per_task": 4237.573333333334,
      "last_run_timestamp": "2026-10-24T13:15:23.369Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Together",
      "accuracy": 0.7629022732240437,
      "accuracy_std_dev": 0.028478290118661355,
      "run_count": 11,
      "total_questions": 549,
      "avg_cost_per_task_usd": 0.01388206375227687,
      "avg_duration_per_task_ms": 226746.868852459,
      "avg_output_tokens_per_task": 15915.701275045538,
      "last_run_timestamp": "2026-10-17T11:00:48.267Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "NextBit",
      "accuracy": 0.7628571428571429,
      "accuracy_std_dev": 0.023328310442945685,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.11986269428571429,
      "avg_duration_per_task_ms": 415507.14,
      "avg_output_tokens_per_task": 18962.734285714287,
      "last_run_timestamp": "2026-10-03T08:33:31.578Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Alibaba",
      "accuracy": 0.762857,
      "accuracy_std_dev": 0.03062451702709541,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.03217967714285715,
      "avg_duration_per_task_ms": 195766.24,
      "avg_output_tokens_per_task": 14823.671428571428,
      "last_run_timestamp": "2026-10-24T11:41:00.069Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Novita",
      "accuracy": 0.7625627587939698,
      "accuracy_std_dev": 0.017640643984489356,
      "run_count": 8,
      "total_questions": 398,
      "avg_cost_per_task_usd": 0.01927761809045226,
      "avg_duration_per_task_ms": 121120.84924623116,
      "avg_output_tokens_per_task": 14540.939698492462,
      "last_run_timestamp": "2026-10-24T11:49:30.492Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "GMICloud",
      "accuracy": 0.762499875,
      "accuracy_std_dev": 0.0406800464676403,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0091677075,
      "avg_duration_per_task_ms": 191141.5625,
      "avg_output_tokens_per_task": 15674.26,
      "last_run_timestamp": "2026-10-24T19:03:39.853Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Alibaba",
      "accuracy": 0.7624696569343066,
      "accuracy_std_dev": 0.03657436469296787,
      "run_count": 11,
      "total_questions": 548,
      "avg_cost_per_task_usd": 0.19588743978102188,
      "avg_duration_per_task_ms": 477892.8375912409,
      "avg_output_tokens_per_task": 27946.843065693432,
      "last_run_timestamp": "2026-10-24T19:19:34.860Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Alibaba",
      "accuracy": 0.7624473392405063,
      "accuracy_std_dev": 0.03272282893772328,
      "run_count": 8,
      "total_questions": 395,
      "avg_cost_per_task_usd": 0.01104873670886076,
      "avg_duration_per_task_ms": 137884.06329113923,
      "avg_output_tokens_per_task": 12786.113924050633,
      "last_run_timestamp": "2026-10-24T19:21:18.651Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "DeepInfra",
      "accuracy": 0.7622223333333333,
      "accuracy_std_dev": 0.030469494967188978,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.029655940000000006,
      "avg_duration_per_task_ms": 256385.88,
      "avg_output_tokens_per_task": 18856.086666666666,
      "last_run_timestamp": "2026-10-08T22:27:05.549Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "GMICloud",
      "accuracy": 0.7622223333333333,
      "accuracy_std_dev": 0.013698571612479267,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.011449053333333332,
      "avg_duration_per_task_ms": 599408.32,
      "avg_output_tokens_per_task": 23110.746666666666,
      "last_run_timestamp": "2026-10-24T14:24:42.966Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.5-opus-20251124",
      "model_slug": "anthropic/claude-opus-4.5",
      "model_name": "Anthropic: Claude Opus 4.5",
      "provider_name": "Google",
      "accuracy": 0.7622221666666666,
      "accuracy_std_dev": 0.029855561759448156,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.54613815,
      "avg_duration_per_task_ms": 190695.96833333332,
      "avg_output_tokens_per_task": 10651.453333333333,
      "last_run_timestamp": "2026-10-24T12:29:33.066Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Morph",
      "accuracy": 0.762143986599665,
      "accuracy_std_dev": 0.04074106053036531,
      "run_count": 12,
      "total_questions": 597,
      "avg_cost_per_task_usd": 0.06808563986599665,
      "avg_duration_per_task_ms": 207102.986599665,
      "avg_output_tokens_per_task": 11782.765494137353,
      "last_run_timestamp": "2026-10-24T12:36:04.401Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.5-opus-20251124",
      "model_slug": "anthropic/claude-opus-4.5",
      "model_name": "Anthropic: Claude Opus 4.5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.762,
      "accuracy_std_dev": 0.02386079071196091,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.5457165,
      "avg_duration_per_task_ms": 180407.204,
      "avg_output_tokens_per_task": 10745.142,
      "last_run_timestamp": "2026-10-24T11:51:51.137Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Chutes",
      "accuracy": 0.7619049795918367,
      "accuracy_std_dev": 0.01847199999999999,
      "run_count": 2,
      "total_questions": 98,
      "avg_cost_per_task_usd": 0.1408940612244898,
      "avg_duration_per_task_ms": 345711.98979591834,
      "avg_output_tokens_per_task": 8276.57142857143,
      "last_run_timestamp": "2026-08-01T08:06:00.902Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Baidu",
      "accuracy": 0.7617847878787878,
      "accuracy_std_dev": 0.03876959831491683,
      "run_count": 4,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.02413183333333333,
      "avg_duration_per_task_ms": 113963.81313131313,
      "avg_output_tokens_per_task": 15997.81313131313,
      "last_run_timestamp": "2026-10-24T12:54:33.303Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Together",
      "accuracy": 0.76166675,
      "accuracy_std_dev": 0.011902275693643652,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.080519975,
      "avg_duration_per_task_ms": 148683.485,
      "avg_output_tokens_per_task": 8604.13,
      "last_run_timestamp": "2026-08-22T08:39:22.990Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "SiliconFlow",
      "accuracy": 0.76166675,
      "accuracy_std_dev": 0.02229839929653024,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.065100265,
      "avg_duration_per_task_ms": 116150.4225,
      "avg_output_tokens_per_task": 4604.0675,
      "last_run_timestamp": "2026-10-24T12:22:07.097Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "DeepInfra",
      "accuracy": 0.7615039124579126,
      "accuracy_std_dev": 0.01612860952882591,
      "run_count": 6,
      "total_questions": 297,
      "avg_cost_per_task_usd": 0.1581539494949495,
      "avg_duration_per_task_ms": 807649.4747474748,
      "avg_output_tokens_per_task": 30684.649831649833,
      "last_run_timestamp": "2026-10-24T18:54:54.259Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "DeepInfra",
      "accuracy": 0.7614941091954023,
      "accuracy_std_dev": 0.008329841446907167,
      "run_count": 7,
      "total_questions": 348,
      "avg_cost_per_task_usd": 0.007317272988505747,
      "avg_duration_per_task_ms": 290620.7356321839,
      "avg_output_tokens_per_task": 12448.798850574713,
      "last_run_timestamp": "2026-10-24T20:14:29.971Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Baidu",
      "accuracy": 0.761111,
      "accuracy_std_dev": 0.016063062451890448,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.05401205,
      "avg_duration_per_task_ms": 73585.80666666667,
      "avg_output_tokens_per_task": 4477.26,
      "last_run_timestamp": "2026-10-24T12:38:41.606Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "BaseTen",
      "accuracy": 0.7610546224489795,
      "accuracy_std_dev": 0.016475935184003962,
      "run_count": 4,
      "total_questions": 196,
      "avg_cost_per_task_usd": 0.2764416326530612,
      "avg_duration_per_task_ms": 159672.02551020408,
      "avg_output_tokens_per_task": 7623.877551020408,
      "last_run_timestamp": "2026-10-24T12:38:26.300Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Azure",
      "accuracy": 0.760833375,
      "accuracy_std_dev": 0.017460477198071506,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.17206684999999997,
      "avg_duration_per_task_ms": 195271.6725,
      "avg_output_tokens_per_task": 11186.875,
      "last_run_timestamp": "2026-10-24T19:43:42.996Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7607406666666667,
      "accuracy_std_dev": 0.02887947805006949,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.18594795111111112,
      "avg_duration_per_task_ms": 175274.95555555556,
      "avg_output_tokens_per_task": 9597.25111111111,
      "last_run_timestamp": "2026-10-24T19:21:02.236Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Relace",
      "accuracy": 0.7607406666666667,
      "accuracy_std_dev": 0.034633054297881376,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.006436806666666667,
      "avg_duration_per_task_ms": 125547.04444444444,
      "avg_output_tokens_per_task": 4262.8533333333335,
      "last_run_timestamp": "2026-10-24T14:25:23.734Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Together",
      "accuracy": 0.7606666,
      "accuracy_std_dev": 0.031048465341140474,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.027007422000000003,
      "avg_duration_per_task_ms": 150781.166,
      "avg_output_tokens_per_task": 8984.718,
      "last_run_timestamp": "2026-10-24T19:16:17.608Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "InferenceNet",
      "accuracy": 0.7606059090909091,
      "accuracy_std_dev": 0.02402931176124294,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.03610860181818182,
      "avg_duration_per_task_ms": 72225.87454545454,
      "avg_output_tokens_per_task": 4745.927272727273,
      "last_run_timestamp": "2026-10-24T12:22:54.608Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Crusoe",
      "accuracy": 0.7603748473895582,
      "accuracy_std_dev": 0.02026019621425223,
      "run_count": 5,
      "total_questions": 249,
      "avg_cost_per_task_usd": 0.03634712449799197,
      "avg_duration_per_task_ms": 152002.0281124498,
      "avg_output_tokens_per_task": 23794.738955823294,
      "last_run_timestamp": "2026-10-24T12:53:48.922Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "DekaLLM",
      "accuracy": 0.7603120367892977,
      "accuracy_std_dev": 0.027860534151775026,
      "run_count": 6,
      "total_questions": 299,
      "avg_cost_per_task_usd": 0.06456417725752508,
      "avg_duration_per_task_ms": 376665.0802675585,
      "avg_output_tokens_per_task": 52393.28428093645,
      "last_run_timestamp": "2026-10-24T19:21:23.089Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Ionstream",
      "accuracy": 0.76,
      "accuracy_std_dev": 0.0066669999999999785,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01747962,
      "avg_duration_per_task_ms": 589736.28,
      "avg_output_tokens_per_task": 13675.16,
      "last_run_timestamp": "2026-08-01T08:06:01.204Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "SiliconFlow",
      "accuracy": 0.76,
      "accuracy_std_dev": 0.01333300000000004,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.054461335,
      "avg_duration_per_task_ms": 91417,
      "avg_output_tokens_per_task": 15101.33,
      "last_run_timestamp": "2026-10-03T08:13:12.597Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "DekaLLM",
      "accuracy": 0.76,
      "accuracy_std_dev": 0.014401594564491825,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.019761673333333334,
      "avg_duration_per_task_ms": 291193.25333333336,
      "avg_output_tokens_per_task": 17811.393333333333,
      "last_run_timestamp": "2026-10-10T11:23:23.352Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "GLM 5.3",
      "provider_name": "digitalocean",
      "accuracy": 0.76,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.01597332,
      "avg_duration_per_task_ms": 42256.58,
      "avg_output_tokens_per_task": 1764.66,
      "last_run_timestamp": "2026-10-01T23:50:45.025Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "ParetoAlpha",
      "accuracy": 0.76,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.32747625999999996,
      "avg_duration_per_task_ms": 2629117.52,
      "avg_output_tokens_per_task": 11787.92,
      "last_run_timestamp": "2026-10-24T12:30:30.307Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3-ultra-550b-a55b-20260604",
      "model_slug": "nvidia/nemotron-3-ultra-550b-a55b",
      "model_name": "NVIDIA: Nemotron 3 Ultra",
      "provider_name": "Venice",
      "accuracy": 0.76,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.19103682,
      "avg_duration_per_task_ms": 233230.24,
      "avg_output_tokens_per_task": 15236.66,
      "last_run_timestamp": "2026-08-04T03:12:00.236Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "Novita",
      "accuracy": 0.76,
      "accuracy_std_dev": 0.02108195647941624,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.08341662,
      "avg_duration_per_task_ms": 533566.115,
      "avg_output_tokens_per_task": 14923.395,
      "last_run_timestamp": "2026-10-24T12:36:09.902Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "Cohere",
      "accuracy": 0.76,
      "accuracy_std_dev": 0.0066669999999999785,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.10151575999999998,
      "avg_duration_per_task_ms": 434936.03,
      "avg_output_tokens_per_task": 13962.46,
      "last_run_timestamp": "2026-10-24T12:32:07.608Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Makora",
      "accuracy": 0.76,
      "accuracy_std_dev": 0.0066669999999999785,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.07038507,
      "avg_duration_per_task_ms": 256302.49,
      "avg_output_tokens_per_task": 10045.27,
      "last_run_timestamp": "2026-08-26T00:00:05.638Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Reka",
      "accuracy": 0.7595238571428571,
      "accuracy_std_dev": 0.02174373003832582,
      "run_count": 14,
      "total_questions": 700,
      "avg_cost_per_task_usd": 0.04527854285714286,
      "avg_duration_per_task_ms": 136160.14428571428,
      "avg_output_tokens_per_task": 4500.364285714286,
      "last_run_timestamp": "2026-10-24T12:17:41.928Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Venice",
      "accuracy": 0.7594445,
      "accuracy_std_dev": 0.03225570769776,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.01983395166666667,
      "avg_duration_per_task_ms": 192524.21833333332,
      "avg_output_tokens_per_task": 14719.435,
      "last_run_timestamp": "2026-10-24T12:16:56.429Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "DeepInfra",
      "accuracy": 0.759408620967742,
      "accuracy_std_dev": 0.023253272624729635,
      "run_count": 5,
      "total_questions": 248,
      "avg_cost_per_task_usd": 0.012160826612903226,
      "avg_duration_per_task_ms": 309190.9919354839,
      "avg_output_tokens_per_task": 4125.725806451613,
      "last_run_timestamp": "2026-10-17T11:47:18.976Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "ModelRun",
      "accuracy": 0.7593733097072419,
      "accuracy_std_dev": 0.029570836436633496,
      "run_count": 13,
      "total_questions": 649,
      "avg_cost_per_task_usd": 0.08832456086286594,
      "avg_duration_per_task_ms": 143648.46687211093,
      "avg_output_tokens_per_task": 12389.725731895223,
      "last_run_timestamp": "2026-10-24T19:07:01.850Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.5-flash-lite-20260721",
      "model_slug": "google/gemini-3.5-flash-lite",
      "model_name": "Google: Gemini 3.5 Flash Lite",
      "provider_name": "Google",
      "accuracy": 0.7593333,
      "accuracy_std_dev": 0.016452043441773433,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.083344808,
      "avg_duration_per_task_ms": 106115.21,
      "avg_output_tokens_per_task": 23741.31,
      "last_run_timestamp": "2026-10-01T19:05:47.886Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "CoreWeave",
      "accuracy": 0.7592594444444444,
      "accuracy_std_dev": 0.017340275597650393,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.15033249555555556,
      "avg_duration_per_task_ms": 165755.92888888888,
      "avg_output_tokens_per_task": 20125.08,
      "last_run_timestamp": "2026-10-24T11:39:34.603Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Reka",
      "accuracy": 0.7592592436412315,
      "accuracy_std_dev": 0.02502658689065603,
      "run_count": 15,
      "total_questions": 747,
      "avg_cost_per_task_usd": 0.05607525167336011,
      "avg_duration_per_task_ms": 275516.45113788487,
      "avg_output_tokens_per_task": 12015.718875502009,
      "last_run_timestamp": "2026-10-24T11:39:58.119Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Azure",
      "accuracy": 0.759166625,
      "accuracy_std_dev": 0.02777639760631993,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0257473675,
      "avg_duration_per_task_ms": 253688.2675,
      "avg_output_tokens_per_task": 13099.1275,
      "last_run_timestamp": "2026-10-24T19:09:33.707Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.6-flash-20260721",
      "model_slug": "google/gemini-3.6-flash",
      "model_name": "Google: Gemini 3.6 Flash",
      "provider_name": "Google",
      "accuracy": 0.7590477142857143,
      "accuracy_std_dev": 0.013997181448055748,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.24529245999999996,
      "avg_duration_per_task_ms": 151053.55142857143,
      "avg_output_tokens_per_task": 23224.54,
      "last_run_timestamp": "2026-10-01T19:04:58.090Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Makora",
      "accuracy": 0.7588886666666668,
      "accuracy_std_dev": 0.011811189506377245,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.010121153333333334,
      "avg_duration_per_task_ms": 244060.09,
      "avg_output_tokens_per_task": 16283.283333333333,
      "last_run_timestamp": "2026-10-24T13:01:16.778Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "Venice",
      "accuracy": 0.7585186666666667,
      "accuracy_std_dev": 0.029444037268463494,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.33338275111111115,
      "avg_duration_per_task_ms": 628430.3377777778,
      "avg_output_tokens_per_task": 27569.568888888887,
      "last_run_timestamp": "2026-10-24T18:52:57.046Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Cloudflare",
      "accuracy": 0.7585185555555556,
      "accuracy_std_dev": 0.020793455498792097,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.02593214222222222,
      "avg_duration_per_task_ms": 205207.60444444444,
      "avg_output_tokens_per_task": 13619.955555555556,
      "last_run_timestamp": "2026-10-24T20:44:23.337Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "CoreWeave",
      "accuracy": 0.7584615384615384,
      "accuracy_std_dev": 0.02110672856848985,
      "run_count": 13,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.021854490769230767,
      "avg_duration_per_task_ms": 124647.87076923077,
      "avg_output_tokens_per_task": 17096.62,
      "last_run_timestamp": "2026-10-24T12:20:00.386Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "auto-routing",
      "accuracy": 0.7583335,
      "accuracy_std_dev": 0.034760921546328406,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.12354291499999999,
      "avg_duration_per_task_ms": 1021866.15,
      "avg_output_tokens_per_task": 26151.615,
      "last_run_timestamp": "2026-10-24T12:47:02.906Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta/muse-glimmer-30b-20260810",
      "model_slug": "meta/muse-glimmer-30b",
      "model_name": "Meta: Muse Glimmer 30B",
      "provider_name": "Phala",
      "accuracy": 0.758333375,
      "accuracy_std_dev": 0.012360224066916245,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.034803532500000005,
      "avg_duration_per_task_ms": 275650.13,
      "avg_output_tokens_per_task": 13717.48,
      "last_run_timestamp": "2026-10-03T08:20:09.018Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.1-20251113",
      "model_slug": "openai/gpt-5.1",
      "model_name": "OpenAI: GPT-5.1",
      "provider_name": "OpenAI",
      "accuracy": 0.7583332500000001,
      "accuracy_std_dev": 0.005527984369324845,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.40697005500000005,
      "avg_duration_per_task_ms": 402356.405,
      "avg_output_tokens_per_task": 35595.7,
      "last_run_timestamp": "2026-10-02T23:12:57.355Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Mancer 2",
      "accuracy": 0.7583332500000001,
      "accuracy_std_dev": 0.03460084697211181,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0886916525,
      "avg_duration_per_task_ms": 581833.97,
      "avg_output_tokens_per_task": 12683.3375,
      "last_run_timestamp": "2026-10-24T12:19:32.894Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-sonnet-5-20260630",
      "model_slug": "anthropic/claude-sonnet-5",
      "model_name": "Anthropic: Claude Sonnet 5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7582857142857146,
      "accuracy_std_dev": 0.02995624352529768,
      "run_count": 35,
      "total_questions": 1750,
      "avg_cost_per_task_usd": 0.20326376057142856,
      "avg_duration_per_task_ms": 118129.27428571429,
      "avg_output_tokens_per_task": 7790.011428571429,
      "last_run_timestamp": "2026-10-24T12:18:37.751Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "CoreWeave",
      "accuracy": 0.7581818181818182,
      "accuracy_std_dev": 0.034942042529720004,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.0916818309090909,
      "avg_duration_per_task_ms": 244976.56,
      "avg_output_tokens_per_task": 12409.656363636364,
      "last_run_timestamp": "2026-10-24T12:33:35.006Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Alibaba",
      "accuracy": 0.758,
      "accuracy_std_dev": 0.02952224773285396,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.10296491399999999,
      "avg_duration_per_task_ms": 278724.296,
      "avg_output_tokens_per_task": 12256.98,
      "last_run_timestamp": "2026-10-24T12:28:42.014Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-122b-a10b-20260224",
      "model_slug": "qwen/qwen3.5-122b-a10b",
      "model_name": "Qwen: Qwen3.5-122B-A10B",
      "provider_name": "DeepInfra",
      "accuracy": 0.7579964141414142,
      "accuracy_std_dev": 0.023533015514113357,
      "run_count": 8,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.20713992929292932,
      "avg_duration_per_task_ms": 664979.7727272727,
      "avg_output_tokens_per_task": 53099.67929292929,
      "last_run_timestamp": "2026-09-26T08:05:12.736Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "SambaNova",
      "accuracy": 0.7579907534246576,
      "accuracy_std_dev": 0.01133630408123485,
      "run_count": 3,
      "total_questions": 146,
      "avg_cost_per_task_usd": 0.06852380136986301,
      "avg_duration_per_task_ms": 156936.7876712329,
      "avg_output_tokens_per_task": 10408.410958904109,
      "last_run_timestamp": "2026-10-24T18:33:39.875Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "StreamLake",
      "accuracy": 0.7579486923076924,
      "accuracy_std_dev": 0.021623878857143158,
      "run_count": 13,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.0068238523076923074,
      "avg_duration_per_task_ms": 261267.2046153846,
      "avg_output_tokens_per_task": 15563.432307692308,
      "last_run_timestamp": "2026-10-24T13:09:56.475Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.757778,
      "accuracy_std_dev": 0.025724014817546128,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.03362461666666666,
      "avg_duration_per_task_ms": 226709.77666666667,
      "avg_output_tokens_per_task": 35231.33666666667,
      "last_run_timestamp": "2026-10-24T11:55:02.754Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "auto-routing",
      "accuracy": 0.7577777777777778,
      "accuracy_std_dev": 0.03142696805391397,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.04451809777777778,
      "avg_duration_per_task_ms": 185433.6288888889,
      "avg_output_tokens_per_task": 9376.346666666666,
      "last_run_timestamp": "2026-10-24T19:48:57.855Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "GMICloud",
      "accuracy": 0.7577777777777778,
      "accuracy_std_dev": 0.022662374322886335,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.06675582666666667,
      "avg_duration_per_task_ms": 168314.9311111111,
      "avg_output_tokens_per_task": 4664.111111111111,
      "last_run_timestamp": "2026-10-24T14:02:36.188Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3-flash-preview-20251217",
      "model_slug": "google/gemini-3-flash-preview",
      "model_name": "Google: Gemini 3 Flash Preview",
      "provider_name": "Google AI Studio",
      "accuracy": 0.7577776666666667,
      "accuracy_std_dev": 0.04629468145358481,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.12420428666666666,
      "avg_duration_per_task_ms": 116279.86,
      "avg_output_tokens_per_task": 19029.046666666665,
      "last_run_timestamp": "2026-10-01T19:05:00.104Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.1-pro-preview-20260219",
      "model_slug": "google/gemini-3.1-pro-preview",
      "model_name": "Google: Gemini 3.1 Pro Preview",
      "provider_name": "auto-routing",
      "accuracy": 0.7577776666666667,
      "accuracy_std_dev": 0.017497908167041637,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.35916456,
      "avg_duration_per_task_ms": 165675.65333333332,
      "avg_output_tokens_per_task": 12758.866666666667,
      "last_run_timestamp": "2026-10-01T19:06:01.702Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "AkashML",
      "accuracy": 0.7575758811188811,
      "accuracy_std_dev": 0.010732768960317529,
      "run_count": 3,
      "total_questions": 143,
      "avg_cost_per_task_usd": 0.27724311888111886,
      "avg_duration_per_task_ms": 338699.5594405594,
      "avg_output_tokens_per_task": 6540.4965034965035,
      "last_run_timestamp": "2026-10-24T11:53:36.261Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Cloudflare",
      "accuracy": 0.7575504530201342,
      "accuracy_std_dev": 0.028065656837905205,
      "run_count": 12,
      "total_questions": 596,
      "avg_cost_per_task_usd": 0.19407245637583895,
      "avg_duration_per_task_ms": 482897.4211409396,
      "avg_output_tokens_per_task": 20396.08389261745,
      "last_run_timestamp": "2026-10-24T14:01:08.267Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Parasail",
      "accuracy": 0.7575376909547739,
      "accuracy_std_dev": 0.025888825973329584,
      "run_count": 16,
      "total_questions": 796,
      "avg_cost_per_task_usd": 0.019801728643216082,
      "avg_duration_per_task_ms": 423933.55778894474,
      "avg_output_tokens_per_task": 16756.603015075376,
      "last_run_timestamp": "2026-10-24T12:17:05.499Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7574406116071428,
      "accuracy_std_dev": 0.06727408355010252,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.25612264285714287,
      "avg_duration_per_task_ms": 471727.34375,
      "avg_output_tokens_per_task": 29191.647321428572,
      "last_run_timestamp": "2026-10-24T19:02:15.249Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Modal",
      "accuracy": 0.7574072222222222,
      "accuracy_std_dev": 0.030298005023351983,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.015441244444444443,
      "avg_duration_per_task_ms": 72124.47777777778,
      "avg_output_tokens_per_task": 4310.673333333333,
      "last_run_timestamp": "2026-10-17T11:14:15.607Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Alibaba",
      "accuracy": 0.7573529926470588,
      "accuracy_std_dev": 0.018618180288085897,
      "run_count": 11,
      "total_questions": 544,
      "avg_cost_per_task_usd": 0.03312612867647059,
      "avg_duration_per_task_ms": 167023.56985294117,
      "avg_output_tokens_per_task": 16331.753676470587,
      "last_run_timestamp": "2026-10-24T12:47:41.904Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7573333999999999,
      "accuracy_std_dev": 0.032000033334982636,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.028732296,
      "avg_duration_per_task_ms": 171140.192,
      "avg_output_tokens_per_task": 15777.572,
      "last_run_timestamp": "2026-10-24T12:23:36.194Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "NextBit",
      "accuracy": 0.7573333999999999,
      "accuracy_std_dev": 0.03991088294989224,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.023026752,
      "avg_duration_per_task_ms": 161149.652,
      "avg_output_tokens_per_task": 18473.992,
      "last_run_timestamp": "2026-10-03T08:55:31.322Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Cloudflare",
      "accuracy": 0.7573332666666667,
      "accuracy_std_dev": 0.01944421688991929,
      "run_count": 15,
      "total_questions": 750,
      "avg_cost_per_task_usd": 0.02663897066666666,
      "avg_duration_per_task_ms": 283461.33733333333,
      "avg_output_tokens_per_task": 16864.825333333334,
      "last_run_timestamp": "2026-10-24T13:28:03.080Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.5-opus-20251124",
      "model_slug": "anthropic/claude-opus-4.5",
      "model_name": "Anthropic: Claude Opus 4.5",
      "provider_name": "Azure",
      "accuracy": 0.75722225,
      "accuracy_std_dev": 0.026311355692061323,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.5355160966666667,
      "avg_duration_per_task_ms": 190831.025,
      "avg_output_tokens_per_task": 10368.328333333333,
      "last_run_timestamp": "2026-10-24T11:54:04.367Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7571428571428571,
      "accuracy_std_dev": 0.024651550802730272,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.012830634285714286,
      "avg_duration_per_task_ms": 221282.58285714284,
      "avg_output_tokens_per_task": 4997.4742857142855,
      "last_run_timestamp": "2026-10-24T12:58:11.932Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "SiliconFlow",
      "accuracy": 0.757037111111111,
      "accuracy_std_dev": 0.02311772460181651,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.014514133333333333,
      "avg_duration_per_task_ms": 244582.29555555555,
      "avg_output_tokens_per_task": 15046.255555555555,
      "last_run_timestamp": "2026-10-24T19:19:42.648Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "Alibaba",
      "accuracy": 0.757037,
      "accuracy_std_dev": 0.01666255916718679,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.17754851555555554,
      "avg_duration_per_task_ms": 320852.2622222222,
      "avg_output_tokens_per_task": 10861.113333333333,
      "last_run_timestamp": "2026-10-24T11:53:33.499Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Sail Research",
      "accuracy": 0.7568475720930232,
      "accuracy_std_dev": 0.02459803490467476,
      "run_count": 13,
      "total_questions": 645,
      "avg_cost_per_task_usd": 0.09061050697674419,
      "avg_duration_per_task_ms": 702536.334883721,
      "avg_output_tokens_per_task": 21508.96434108527,
      "last_run_timestamp": "2026-10-24T12:57:12.879Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7566669999999999,
      "accuracy_std_dev": 0.010000000000000009,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.023368849999999997,
      "avg_duration_per_task_ms": 302412.09,
      "avg_output_tokens_per_task": 11606.21,
      "last_run_timestamp": "2026-10-24T12:08:01.600Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-sonnet-5-20260630",
      "model_slug": "anthropic/claude-sonnet-5",
      "model_name": "Anthropic: Claude Sonnet 5",
      "provider_name": "Anthropic",
      "accuracy": 0.7566666666666667,
      "accuracy_std_dev": 0.03853813887854761,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.19762862999999997,
      "avg_duration_per_task_ms": 102033.36166666666,
      "avg_output_tokens_per_task": 7661.711666666667,
      "last_run_timestamp": "2026-10-24T12:18:01.499Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7566666666666667,
      "accuracy_std_dev": 0.04895203245581898,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.02191566,
      "avg_duration_per_task_ms": 332956.19333333336,
      "avg_output_tokens_per_task": 12348.723333333333,
      "last_run_timestamp": "2026-10-24T12:55:32.102Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Parasail",
      "accuracy": 0.7566665,
      "accuracy_std_dev": 0.0033334999999999892,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.018845109999999998,
      "avg_duration_per_task_ms": 92258.91,
      "avg_output_tokens_per_task": 11470.03,
      "last_run_timestamp": "2026-10-24T12:23:52.812Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Wafer",
      "accuracy": 0.7566665,
      "accuracy_std_dev": 0.01666650000000003,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.010595590000000002,
      "avg_duration_per_task_ms": 149870.75,
      "avg_output_tokens_per_task": 13511.14,
      "last_run_timestamp": "2026-10-24T12:17:44.341Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Wafer",
      "accuracy": 0.7564445999999999,
      "accuracy_std_dev": 0.020921957750331767,
      "run_count": 15,
      "total_questions": 750,
      "avg_cost_per_task_usd": 0.25126205733333334,
      "avg_duration_per_task_ms": 263807.58666666667,
      "avg_output_tokens_per_task": 10194.865333333333,
      "last_run_timestamp": "2026-10-24T12:05:03.668Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "DeepInfra",
      "accuracy": 0.7563351081871345,
      "accuracy_std_dev": 0.013338594901982123,
      "run_count": 7,
      "total_questions": 342,
      "avg_cost_per_task_usd": 0.033182684210526316,
      "avg_duration_per_task_ms": 2146298.769005848,
      "avg_output_tokens_per_task": 50213.23391812865,
      "last_run_timestamp": "2026-10-09T04:05:56.641Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "Alibaba",
      "accuracy": 0.7562962222222223,
      "accuracy_std_dev": 0.08255839119176704,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.07683437111111112,
      "avg_duration_per_task_ms": 192163.8777777778,
      "avg_output_tokens_per_task": 23858.006666666668,
      "last_run_timestamp": "2026-10-24T18:45:10.073Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7562961111111112,
      "accuracy_std_dev": 0.02650155778836174,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.013197128888888889,
      "avg_duration_per_task_ms": 222309.89333333334,
      "avg_output_tokens_per_task": 14339.34888888889,
      "last_run_timestamp": "2026-10-24T19:54:36.988Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-sonnet-5-20260630",
      "model_slug": "anthropic/claude-sonnet-5",
      "model_name": "Anthropic: Claude Sonnet 5",
      "provider_name": "auto-routing",
      "accuracy": 0.7561111666666667,
      "accuracy_std_dev": 0.02924456159366313,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.20050754833333334,
      "avg_duration_per_task_ms": 123808.41166666667,
      "avg_output_tokens_per_task": 7847.445,
      "last_run_timestamp": "2026-10-24T11:58:02.887Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "SiliconFlow",
      "accuracy": 0.756,
      "accuracy_std_dev": 0.024073867807230344,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.07011482399999999,
      "avg_duration_per_task_ms": 334056.692,
      "avg_output_tokens_per_task": 14633.816,
      "last_run_timestamp": "2026-10-24T12:59:11.337Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "InferenceNet",
      "accuracy": 0.7559648947368421,
      "accuracy_std_dev": 0.020191449941144954,
      "run_count": 19,
      "total_questions": 950,
      "avg_cost_per_task_usd": 0.29131978947368414,
      "avg_duration_per_task_ms": 264712.04,
      "avg_output_tokens_per_task": 11238.59157894737,
      "last_run_timestamp": "2026-10-24T12:24:12.062Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "auto-routing",
      "accuracy": 0.7559523303571429,
      "accuracy_std_dev": 0.02997543982811411,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.02883498660714286,
      "avg_duration_per_task_ms": 493727.48883928574,
      "avg_output_tokens_per_task": 12075.370535714286,
      "last_run_timestamp": "2026-10-24T19:23:19.868Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Modal",
      "accuracy": 0.755892303030303,
      "accuracy_std_dev": 0.022789500000000018,
      "run_count": 2,
      "total_questions": 99,
      "avg_cost_per_task_usd": 0.027549767676767676,
      "avg_duration_per_task_ms": 104945.30303030302,
      "avg_output_tokens_per_task": 10081.373737373737,
      "last_run_timestamp": "2026-10-24T12:19:36.298Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.75583325,
      "accuracy_std_dev": 0.019982613972839007,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.008674565,
      "avg_duration_per_task_ms": 119249.875,
      "avg_output_tokens_per_task": 4714.72,
      "last_run_timestamp": "2026-10-24T12:46:25.479Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-sonnet-5-20260630",
      "model_slug": "anthropic/claude-sonnet-5",
      "model_name": "Anthropic: Claude Sonnet 5",
      "provider_name": "Azure",
      "accuracy": 0.7556522173913043,
      "accuracy_std_dev": 0.03227948562565843,
      "run_count": 23,
      "total_questions": 1150,
      "avg_cost_per_task_usd": 0.20032831565217393,
      "avg_duration_per_task_ms": 100773.6104347826,
      "avg_output_tokens_per_task": 7631.826086956522,
      "last_run_timestamp": "2026-10-24T12:48:21.016Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Venice",
      "accuracy": 0.7556052242152467,
      "accuracy_std_dev": 0.025010430481878735,
      "run_count": 9,
      "total_questions": 446,
      "avg_cost_per_task_usd": 0.020433513452914796,
      "avg_duration_per_task_ms": 233457.43946188342,
      "avg_output_tokens_per_task": 4244.6569506726455,
      "last_run_timestamp": "2026-10-24T12:41:47.992Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Novita",
      "accuracy": 0.7555556,
      "accuracy_std_dev": 0.02330687812585232,
      "run_count": 15,
      "total_questions": 750,
      "avg_cost_per_task_usd": 0.028939747999999994,
      "avg_duration_per_task_ms": 174269.36933333334,
      "avg_output_tokens_per_task": 17673.329333333335,
      "last_run_timestamp": "2026-10-24T12:19:39.208Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "StreamLake",
      "accuracy": 0.7555555555555555,
      "accuracy_std_dev": 0.02330683151215116,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.007330602222222223,
      "avg_duration_per_task_ms": 193346.8088888889,
      "avg_output_tokens_per_task": 4415.6,
      "last_run_timestamp": "2026-10-24T14:44:52.720Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "CoreWeave",
      "accuracy": 0.7554201788617886,
      "accuracy_std_dev": 0.023120046817426663,
      "run_count": 10,
      "total_questions": 492,
      "avg_cost_per_task_usd": 0.025219461382113822,
      "avg_duration_per_task_ms": 380343.8069105691,
      "avg_output_tokens_per_task": 12305.247967479674,
      "last_run_timestamp": "2026-10-24T19:18:15.431Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Wafer",
      "accuracy": 0.7553846153846154,
      "accuracy_std_dev": 0.022400361015423448,
      "run_count": 13,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.023595866153846155,
      "avg_duration_per_task_ms": 114714.57846153846,
      "avg_output_tokens_per_task": 14825.998461538462,
      "last_run_timestamp": "2026-10-24T12:22:58.748Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "CoreWeave",
      "accuracy": 0.7553335,
      "accuracy_std_dev": 0.031134064945811373,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.085864622,
      "avg_duration_per_task_ms": 71299.624,
      "avg_output_tokens_per_task": 12122.088,
      "last_run_timestamp": "2026-10-24T19:37:32.789Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "Darkbloom",
      "accuracy": 0.755319,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 47,
      "avg_cost_per_task_usd": 0.014918489361702128,
      "avg_duration_per_task_ms": 876881.4042553192,
      "avg_output_tokens_per_task": 9452.702127659575,
      "last_run_timestamp": "2026-10-24T14:20:18.925Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-sonnet-5-20260630",
      "model_slug": "anthropic/claude-sonnet-5",
      "model_name": "Anthropic: Claude Sonnet 5",
      "provider_name": "Google",
      "accuracy": 0.7551851666666668,
      "accuracy_std_dev": 0.03354053891064762,
      "run_count": 36,
      "total_questions": 1800,
      "avg_cost_per_task_usd": 0.22096761222222225,
      "avg_duration_per_task_ms": 95197.85388888889,
      "avg_output_tokens_per_task": 7539.6577777777775,
      "last_run_timestamp": "2026-10-24T12:46:28.470Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Together",
      "accuracy": 0.755,
      "accuracy_std_dev": 0.04014562657874455,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.161091306,
      "avg_duration_per_task_ms": 311751.64,
      "avg_output_tokens_per_task": 23196.802,
      "last_run_timestamp": "2026-10-24T12:52:46.316Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Parasail",
      "accuracy": 0.7548484545454546,
      "accuracy_std_dev": 0.021805424301655014,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.01682924181818182,
      "avg_duration_per_task_ms": 119778.08181818182,
      "avg_output_tokens_per_task": 4470.658181818182,
      "last_run_timestamp": "2026-10-24T13:29:07.045Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Venice",
      "accuracy": 0.7548147777777778,
      "accuracy_std_dev": 0.027716071710502704,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.021845168888888888,
      "avg_duration_per_task_ms": 476468.12444444443,
      "avg_output_tokens_per_task": 14231.342222222222,
      "last_run_timestamp": "2026-10-24T18:47:38.219Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "InferenceNet",
      "accuracy": 0.7545454545454545,
      "accuracy_std_dev": 0.03163137195128692,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.008343416363636365,
      "avg_duration_per_task_ms": 139678.11272727273,
      "avg_output_tokens_per_task": 4440.807272727273,
      "last_run_timestamp": "2026-10-24T13:04:30.804Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Ionstream",
      "accuracy": 0.7542953608247424,
      "accuracy_std_dev": 0.02652500000000002,
      "run_count": 2,
      "total_questions": 97,
      "avg_cost_per_task_usd": 0.021299783505154636,
      "avg_duration_per_task_ms": 633495.5257731959,
      "avg_output_tokens_per_task": 26267.113402061856,
      "last_run_timestamp": "2026-08-08T08:30:53.738Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Wafer",
      "accuracy": 0.7542857142857143,
      "accuracy_std_dev": 0.014876559353697425,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.05917716571428572,
      "avg_duration_per_task_ms": 372934.26285714284,
      "avg_output_tokens_per_task": 14097.657142857142,
      "last_run_timestamp": "2026-10-24T13:33:32.917Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "DeepInfra",
      "accuracy": 0.7541274103773585,
      "accuracy_std_dev": 0.021560862869117733,
      "run_count": 17,
      "total_questions": 848,
      "avg_cost_per_task_usd": 0.007241307783018868,
      "avg_duration_per_task_ms": 204331.1474056604,
      "avg_output_tokens_per_task": 13534.349056603774,
      "last_run_timestamp": "2026-10-24T12:31:25.936Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Crusoe",
      "accuracy": 0.7539998,
      "accuracy_std_dev": 0.010832125062055002,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.09518565200000001,
      "avg_duration_per_task_ms": 207718.192,
      "avg_output_tokens_per_task": 9678.708,
      "last_run_timestamp": "2026-08-29T16:31:41.554Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "NextBit",
      "accuracy": 0.7537880151515151,
      "accuracy_std_dev": 0.02397695833731574,
      "run_count": 8,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.01387228787878788,
      "avg_duration_per_task_ms": 200605.87121212122,
      "avg_output_tokens_per_task": 4857.164141414141,
      "last_run_timestamp": "2026-10-17T11:28:07.798Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Moonshot AI",
      "accuracy": 0.7537201651785714,
      "accuracy_std_dev": 0.044602290458818805,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.11035437723214285,
      "avg_duration_per_task_ms": 289472.76116071426,
      "avg_output_tokens_per_task": 11809.0625,
      "last_run_timestamp": "2026-10-24T19:31:21.574Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "Darkbloom",
      "accuracy": 0.7534912732240436,
      "accuracy_std_dev": 0.02055401911633615,
      "run_count": 11,
      "total_questions": 549,
      "avg_cost_per_task_usd": 0.027412590163934426,
      "avg_duration_per_task_ms": 556770.8961748633,
      "avg_output_tokens_per_task": 22246.586520947178,
      "last_run_timestamp": "2026-10-24T20:27:36.167Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Darkbloom",
      "accuracy": 0.7534013163265306,
      "accuracy_std_dev": 0.008336108804472242,
      "run_count": 4,
      "total_questions": 196,
      "avg_cost_per_task_usd": 0.03488917857142857,
      "avg_duration_per_task_ms": 592188.2653061225,
      "avg_output_tokens_per_task": 12444.612244897959,
      "last_run_timestamp": "2026-10-24T12:44:02.644Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "x-ai/grok-4.6-20260810",
      "model_slug": "x-ai/grok-4.6",
      "model_name": "SpaceXAI: Grok 4.6",
      "provider_name": "auto-routing",
      "accuracy": 0.7533335,
      "accuracy_std_dev": 0.00666650000000002,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.33087047999999997,
      "avg_duration_per_task_ms": 241177.17,
      "avg_output_tokens_per_task": 11982.77,
      "last_run_timestamp": "2026-10-01T16:01:02.676Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "auto-routing",
      "accuracy": 0.753333375,
      "accuracy_std_dev": 0.01452975873799614,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.1921822125,
      "avg_duration_per_task_ms": 294611.6675,
      "avg_output_tokens_per_task": 11751.955,
      "last_run_timestamp": "2026-10-24T11:55:23.806Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3-flash-preview-20251217",
      "model_slug": "google/gemini-3-flash-preview",
      "model_name": "Google: Gemini 3 Flash Preview",
      "provider_name": "Google",
      "accuracy": 0.7533333333333333,
      "accuracy_std_dev": 0.028803034820811656,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.11451421333333334,
      "avg_duration_per_task_ms": 146387.56666666668,
      "avg_output_tokens_per_task": 17546.113333333335,
      "last_run_timestamp": "2026-10-01T19:05:47.885Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Morph",
      "accuracy": 0.7533333333333333,
      "accuracy_std_dev": 0.014401903168987377,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.021082753333333332,
      "avg_duration_per_task_ms": 915986.2133333334,
      "avg_output_tokens_per_task": 13268.026666666667,
      "last_run_timestamp": "2026-08-15T08:15:52.453Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.4-20260305",
      "model_slug": "openai/gpt-5.4",
      "model_name": "OpenAI: GPT-5.4",
      "provider_name": "auto-routing",
      "accuracy": 0.7533333333333333,
      "accuracy_std_dev": 0.005443174461857933,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.29158895333333334,
      "avg_duration_per_task_ms": 171961.5,
      "avg_output_tokens_per_task": 10696.2,
      "last_run_timestamp": "2026-10-01T19:04:36.169Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Phala",
      "accuracy": 0.7533333333333333,
      "accuracy_std_dev": 0.012765636746968635,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.17238889666666665,
      "avg_duration_per_task_ms": 505286.37666666665,
      "avg_output_tokens_per_task": 9803.186666666666,
      "last_run_timestamp": "2026-10-03T16:16:10.656Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Phala",
      "accuracy": 0.7533331428571429,
      "accuracy_std_dev": 0.0230940520074423,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.011956362857142858,
      "avg_duration_per_task_ms": 437970.4885714286,
      "avg_output_tokens_per_task": 6008.582857142857,
      "last_run_timestamp": "2026-10-24T12:34:15.321Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "BaseTen",
      "accuracy": 0.753333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.024162740000000002,
      "avg_duration_per_task_ms": 65962.82,
      "avg_output_tokens_per_task": 8135.38,
      "last_run_timestamp": "2026-10-24T12:37:26.070Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Decart",
      "accuracy": 0.753333,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01197713,
      "avg_duration_per_task_ms": 212560.45,
      "avg_output_tokens_per_task": 13957.42,
      "last_run_timestamp": "2026-10-24T11:49:49.163Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Baidu",
      "accuracy": 0.753333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.0185477,
      "avg_duration_per_task_ms": 97588.58,
      "avg_output_tokens_per_task": 10717.4,
      "last_run_timestamp": "2026-10-24T11:41:40.972Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3-flash-preview-20251217",
      "model_slug": "google/gemini-3-flash-preview",
      "model_name": "Google: Gemini 3 Flash Preview",
      "provider_name": "auto-routing",
      "accuracy": 0.753333,
      "accuracy_std_dev": 0.032659863237109024,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.12209446666666665,
      "avg_duration_per_task_ms": 147223.81333333332,
      "avg_output_tokens_per_task": 19831.306666666667,
      "last_run_timestamp": "2026-10-01T19:04:26.702Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.3-codex-20260224",
      "model_slug": "openai/gpt-5.3-codex",
      "model_name": "OpenAI: GPT-5.3-Codex",
      "provider_name": "OpenAI",
      "accuracy": 0.753333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.29971074000000003,
      "avg_duration_per_task_ms": 180697.5,
      "avg_output_tokens_per_task": 13569.24,
      "last_run_timestamp": "2026-10-02T23:12:05.500Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "CoreWeave",
      "accuracy": 0.753333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.10139794,
      "avg_duration_per_task_ms": 119629.34,
      "avg_output_tokens_per_task": 9506.02,
      "last_run_timestamp": "2026-08-01T17:46:04.356Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Ionstream",
      "accuracy": 0.753333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.09723594,
      "avg_duration_per_task_ms": 178772.1,
      "avg_output_tokens_per_task": 8310.58,
      "last_run_timestamp": "2026-08-01T08:05:58.203Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Morph",
      "accuracy": 0.7533329999999999,
      "accuracy_std_dev": 0.03999999999999998,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.0194427,
      "avg_duration_per_task_ms": 127110.83,
      "avg_output_tokens_per_task": 12815.84,
      "last_run_timestamp": "2026-10-24T12:03:02.865Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.6-opus-20260205",
      "model_slug": "anthropic/claude-opus-4.6",
      "model_name": "Anthropic: Claude Opus 4.6",
      "provider_name": "auto-routing",
      "accuracy": 0.7527778333333334,
      "accuracy_std_dev": 0.026311368008376572,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.501247295,
      "avg_duration_per_task_ms": 196372.42666666667,
      "avg_output_tokens_per_task": 10155.583333333334,
      "last_run_timestamp": "2026-10-24T12:42:05.323Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Baidu",
      "accuracy": 0.7527272727272727,
      "accuracy_std_dev": 0.029875735292863623,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.013207607272727274,
      "avg_duration_per_task_ms": 136477.7381818182,
      "avg_output_tokens_per_task": 16262.556363636364,
      "last_run_timestamp": "2026-10-24T11:51:53.770Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.5-20260423",
      "model_slug": "openai/gpt-5.5",
      "model_name": "OpenAI: GPT-5.5",
      "provider_name": "Azure",
      "accuracy": 0.7525925555555556,
      "accuracy_std_dev": 0.03021638651133634,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.5157013133333334,
      "avg_duration_per_task_ms": 178548.1511111111,
      "avg_output_tokens_per_task": 7818.084444444445,
      "last_run_timestamp": "2026-10-01T19:05:47.886Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Fireworks",
      "accuracy": 0.7525337229729728,
      "accuracy_std_dev": 0.02409323915008227,
      "run_count": 12,
      "total_questions": 592,
      "avg_cost_per_task_usd": 0.014006842905405404,
      "avg_duration_per_task_ms": 57600.26858108108,
      "avg_output_tokens_per_task": 4072.2314189189187,
      "last_run_timestamp": "2026-10-24T12:20:04.669Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.5-sonnet-20250929",
      "model_slug": "anthropic/claude-sonnet-4.5",
      "model_name": "Anthropic: Claude Sonnet 4.5",
      "provider_name": "Anthropic",
      "accuracy": 0.7524445999999999,
      "accuracy_std_dev": 0.02858304219824523,
      "run_count": 15,
      "total_questions": 750,
      "avg_cost_per_task_usd": 0.35143002133333334,
      "avg_duration_per_task_ms": 230047.736,
      "avg_output_tokens_per_task": 10847.656,
      "last_run_timestamp": "2026-10-24T11:48:43.120Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "GMICloud",
      "accuracy": 0.7522223333333334,
      "accuracy_std_dev": 0.033866116211077726,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.024190810000000004,
      "avg_duration_per_task_ms": 128051.35,
      "avg_output_tokens_per_task": 10623.603333333333,
      "last_run_timestamp": "2026-10-24T18:55:24.450Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Novita",
      "accuracy": 0.7522223333333333,
      "accuracy_std_dev": 0.03297965328636568,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.014283683333333335,
      "avg_duration_per_task_ms": 164802.21666666667,
      "avg_output_tokens_per_task": 14931.153333333334,
      "last_run_timestamp": "2026-10-03T16:15:22.012Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "auto-routing",
      "accuracy": 0.75222225,
      "accuracy_std_dev": 0.04696988331211997,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.13807676666666668,
      "avg_duration_per_task_ms": 829053.1233333333,
      "avg_output_tokens_per_task": 31392.836666666666,
      "last_run_timestamp": "2026-10-24T19:43:40.189Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.1-20251113",
      "model_slug": "openai/gpt-5.1",
      "model_name": "OpenAI: GPT-5.1",
      "provider_name": "Azure",
      "accuracy": 0.7522221666666666,
      "accuracy_std_dev": 0.01511278994115765,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.44058496666666663,
      "avg_duration_per_task_ms": 458389.25,
      "avg_output_tokens_per_task": 35681.92333333333,
      "last_run_timestamp": "2026-10-01T16:01:13.647Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.6-sonnet-20260217",
      "model_slug": "anthropic/claude-sonnet-4.6",
      "model_name": "Anthropic: Claude Sonnet 4.6",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7521905714285715,
      "accuracy_std_dev": 0.023501964796022273,
      "run_count": 35,
      "total_questions": 1750,
      "avg_cost_per_task_usd": 0.37098750228571425,
      "avg_duration_per_task_ms": 159626.864,
      "avg_output_tokens_per_task": 11846.640571428572,
      "last_run_timestamp": "2026-10-24T12:39:24.516Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-opus-5.5-20260921",
      "model_slug": "anthropic/claude-opus-5.5",
      "model_name": "Anthropic: Claude Opus 5.5",
      "provider_name": "auto-routing",
      "accuracy": 0.752,
      "accuracy_std_dev": 0.01857135189478677,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.33952648,
      "avg_duration_per_task_ms": 120001.212,
      "avg_output_tokens_per_task": 7107.392,
      "last_run_timestamp": "2026-10-24T12:22:52.909Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-opus-5.5-20260921",
      "model_slug": "anthropic/claude-opus-5.5",
      "model_name": "Anthropic: Claude Opus 5.5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7518518333333334,
      "accuracy_std_dev": 0.017715982107345284,
      "run_count": 18,
      "total_questions": 900,
      "avg_cost_per_task_usd": 0.36871217222222225,
      "avg_duration_per_task_ms": 149259.27444444445,
      "avg_output_tokens_per_task": 7150.745555555555,
      "last_run_timestamp": "2026-10-24T12:24:28.843Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.6-sonnet-20260217",
      "model_slug": "anthropic/claude-sonnet-4.6",
      "model_name": "Anthropic: Claude Sonnet 4.6",
      "provider_name": "Azure",
      "accuracy": 0.7516666666666667,
      "accuracy_std_dev": 0.021963791754208158,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.34539720999999995,
      "avg_duration_per_task_ms": 219756.98166666666,
      "avg_output_tokens_per_task": 11912.11,
      "last_run_timestamp": "2026-10-24T12:39:07.663Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Sail Research",
      "accuracy": 0.7514285714285714,
      "accuracy_std_dev": 0.021296004758486267,
      "run_count": 14,
      "total_questions": 700,
      "avg_cost_per_task_usd": 0.009337331428571431,
      "avg_duration_per_task_ms": 230829.7457142857,
      "avg_output_tokens_per_task": 4265.0371428571425,
      "last_run_timestamp": "2026-10-24T13:34:36.032Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.6-opus-20260205",
      "model_slug": "anthropic/claude-opus-4.6",
      "model_name": "Anthropic: Claude Opus 4.6",
      "provider_name": "Google",
      "accuracy": 0.7513888333333334,
      "accuracy_std_dev": 0.027452241920692434,
      "run_count": 24,
      "total_questions": 1200,
      "avg_cost_per_task_usd": 0.5585198816666668,
      "avg_duration_per_task_ms": 202142.5925,
      "avg_output_tokens_per_task": 9870.238333333333,
      "last_run_timestamp": "2026-10-24T12:58:30.427Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.5-20260423",
      "model_slug": "openai/gpt-5.5",
      "model_name": "OpenAI: GPT-5.5",
      "provider_name": "auto-routing",
      "accuracy": 0.7511113333333332,
      "accuracy_std_dev": 0.02199873172606905,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.48754586,
      "avg_duration_per_task_ms": 141651.29333333333,
      "avg_output_tokens_per_task": 7913.286666666667,
      "last_run_timestamp": "2026-10-01T19:04:40.362Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-opus-5.5-20260921",
      "model_slug": "anthropic/claude-opus-5.5",
      "model_name": "Anthropic: Claude Opus 5.5",
      "provider_name": "Anthropic",
      "accuracy": 0.7511111666666667,
      "accuracy_std_dev": 0.02454511593248011,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.3494615833333333,
      "avg_duration_per_task_ms": 100555.84,
      "avg_output_tokens_per_task": 7111.15,
      "last_run_timestamp": "2026-10-24T11:44:08.041Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-opus-5.5-20260921",
      "model_slug": "anthropic/claude-opus-5.5",
      "model_name": "Anthropic: Claude Opus 5.5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.751111,
      "accuracy_std_dev": 0.01523486904002351,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.3475124233333333,
      "avg_duration_per_task_ms": 124013.22666666667,
      "avg_output_tokens_per_task": 7125.08,
      "last_run_timestamp": "2026-10-24T11:31:04.549Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "Parasail",
      "accuracy": 0.751111,
      "accuracy_std_dev": 0.018921497998308716,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.03688384,
      "avg_duration_per_task_ms": 282057.35833333334,
      "avg_output_tokens_per_task": 21913.646666666667,
      "last_run_timestamp": "2026-10-24T19:02:49.020Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "Cerebras",
      "accuracy": 0.751111,
      "accuracy_std_dev": 0.03861798340151904,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.4379569333333333,
      "avg_duration_per_task_ms": 39099.41333333333,
      "avg_output_tokens_per_task": 9339.68,
      "last_run_timestamp": "2026-08-15T14:16:11.865Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Decart",
      "accuracy": 0.7506668,
      "accuracy_std_dev": 0.025854254755455656,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.264866352,
      "avg_duration_per_task_ms": 264537.464,
      "avg_output_tokens_per_task": 10440.212,
      "last_run_timestamp": "2026-10-24T11:49:28.605Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Relace",
      "accuracy": 0.7506666,
      "accuracy_std_dev": 0.024073823498563778,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.022084748,
      "avg_duration_per_task_ms": 137222.84,
      "avg_output_tokens_per_task": 13584.26,
      "last_run_timestamp": "2026-10-24T19:40:37.108Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Phala",
      "accuracy": 0.7505555,
      "accuracy_std_dev": 0.03574174773170259,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.03120006,
      "avg_duration_per_task_ms": 258878.755,
      "avg_output_tokens_per_task": 17037.54,
      "last_run_timestamp": "2026-10-24T12:34:23.300Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Fireworks",
      "accuracy": 0.7505555,
      "accuracy_std_dev": 0.03274012621004996,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.02124632666666667,
      "avg_duration_per_task_ms": 354797.86666666664,
      "avg_output_tokens_per_task": 18494.39,
      "last_run_timestamp": "2026-09-19T08:13:24.062Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Novita",
      "accuracy": 0.7503703333333334,
      "accuracy_std_dev": 0.021798491033402586,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.09827831777777778,
      "avg_duration_per_task_ms": 290553.8022222222,
      "avg_output_tokens_per_task": 12820.055555555555,
      "last_run_timestamp": "2026-10-24T19:44:46.340Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "StreamLake",
      "accuracy": 0.7503347710843373,
      "accuracy_std_dev": 0.02780906636854249,
      "run_count": 10,
      "total_questions": 498,
      "avg_cost_per_task_usd": 0.004999224899598394,
      "avg_duration_per_task_ms": 263913.3353413655,
      "avg_output_tokens_per_task": 12415.361445783132,
      "last_run_timestamp": "2026-10-24T20:11:39.941Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Inceptron",
      "accuracy": 0.7502788193979932,
      "accuracy_std_dev": 0.03287689055865162,
      "run_count": 6,
      "total_questions": 299,
      "avg_cost_per_task_usd": 0.01826405016722408,
      "avg_duration_per_task_ms": 135784.91304347827,
      "avg_output_tokens_per_task": 4013.7859531772574,
      "last_run_timestamp": "2026-10-24T12:16:33.907Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-opus-5.5-20260921",
      "model_slug": "anthropic/claude-opus-5.5",
      "model_name": "Anthropic: Claude Opus 5.5",
      "provider_name": "Azure",
      "accuracy": 0.7500001666666666,
      "accuracy_std_dev": 0.023333261905247805,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.3534387216666667,
      "avg_duration_per_task_ms": 87888.735,
      "avg_output_tokens_per_task": 7005.235,
      "last_run_timestamp": "2026-10-24T12:23:35.337Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-sonnet-5-20260630",
      "model_slug": "anthropic/claude-sonnet-5",
      "model_name": "Anthropic: Claude Sonnet 5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.7500001,
      "accuracy_std_dev": 0.02333341428702625,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.196243942,
      "avg_duration_per_task_ms": 117146.822,
      "avg_output_tokens_per_task": 7664.984,
      "last_run_timestamp": "2026-10-24T12:32:11.296Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Morph",
      "accuracy": 0.75,
      "accuracy_std_dev": 0.0033330000000000304,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.02619736,
      "avg_duration_per_task_ms": 832580.72,
      "avg_output_tokens_per_task": 18187.39,
      "last_run_timestamp": "2026-08-08T08:29:47.435Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "BaseTen",
      "accuracy": 0.75,
      "accuracy_std_dev": 0.0033330000000000304,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.28014248,
      "avg_duration_per_task_ms": 223020.56,
      "avg_output_tokens_per_task": 18715.38,
      "last_run_timestamp": "2026-09-12T08:12:13.731Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.75,
      "accuracy_std_dev": 0.010000000000000009,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.300415095,
      "avg_duration_per_task_ms": 278099.495,
      "avg_output_tokens_per_task": 10648.38,
      "last_run_timestamp": "2026-10-24T12:46:10.016Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.5-20260423",
      "model_slug": "openai/gpt-5.5",
      "model_name": "OpenAI: GPT-5.5",
      "provider_name": "OpenAI",
      "accuracy": 0.75,
      "accuracy_std_dev": 0.04384306609373936,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.50077502,
      "avg_duration_per_task_ms": 139265.845,
      "avg_output_tokens_per_task": 7986.47,
      "last_run_timestamp": "2026-10-02T23:08:08.058Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Venice",
      "accuracy": 0.75,
      "accuracy_std_dev": 0.016666999999999987,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01952235,
      "avg_duration_per_task_ms": 2267441.92,
      "avg_output_tokens_per_task": 19667.19,
      "last_run_timestamp": "2026-10-03T08:13:52.062Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.75,
      "accuracy_std_dev": 0.010000000000000009,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.09412079000000001,
      "avg_duration_per_task_ms": 469968.05,
      "avg_output_tokens_per_task": 4703.42,
      "last_run_timestamp": "2026-10-24T12:22:38.011Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "Baidu",
      "accuracy": 0.7496296666666666,
      "accuracy_std_dev": 0.022246872749020503,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.06072559777777777,
      "avg_duration_per_task_ms": 205446.72222222222,
      "avg_output_tokens_per_task": 10600.893333333333,
      "last_run_timestamp": "2026-10-24T20:08:17.474Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "Phala",
      "accuracy": 0.7494445,
      "accuracy_std_dev": 0.024412988600674583,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.08046244666666666,
      "avg_duration_per_task_ms": 794170.4033333333,
      "avg_output_tokens_per_task": 35486.02166666667,
      "last_run_timestamp": "2026-10-24T19:16:16.979Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "Google",
      "accuracy": 0.7491623969849246,
      "accuracy_std_dev": 0.02175219497401357,
      "run_count": 8,
      "total_questions": 398,
      "avg_cost_per_task_usd": 0.11730947487437185,
      "avg_duration_per_task_ms": 117420.94974874372,
      "avg_output_tokens_per_task": 10306.74623115578,
      "last_run_timestamp": "2026-10-17T19:14:08.120Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "auto-routing",
      "accuracy": 0.7488889166666667,
      "accuracy_std_dev": 0.027397460908687426,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.02489698,
      "avg_duration_per_task_ms": 157509.05666666667,
      "avg_output_tokens_per_task": 8743.431666666667,
      "last_run_timestamp": "2026-10-24T19:12:17.342Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Baidu",
      "accuracy": 0.7488888888888889,
      "accuracy_std_dev": 0.02330689507632377,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.007562466666666667,
      "avg_duration_per_task_ms": 152556.50444444444,
      "avg_output_tokens_per_task": 12996.975555555555,
      "last_run_timestamp": "2026-10-24T19:02:53.466Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "DeepInfra",
      "accuracy": 0.7488888888888889,
      "accuracy_std_dev": 0.016329946740366617,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.01262068,
      "avg_duration_per_task_ms": 336899.4511111111,
      "avg_output_tokens_per_task": 16364.997777777779,
      "last_run_timestamp": "2026-10-24T12:24:10.218Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.6-sonnet-20260217",
      "model_slug": "anthropic/claude-sonnet-4.6",
      "model_name": "Anthropic: Claude Sonnet 4.6",
      "provider_name": "Anthropic",
      "accuracy": 0.7488888666666667,
      "accuracy_std_dev": 0.02560864244707678,
      "run_count": 15,
      "total_questions": 750,
      "avg_cost_per_task_usd": 0.34009223066666666,
      "avg_duration_per_task_ms": 209111.22666666665,
      "avg_output_tokens_per_task": 11827.101333333334,
      "last_run_timestamp": "2026-10-24T13:09:08.696Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.5-flash-20260519",
      "model_slug": "google/gemini-3.5-flash",
      "model_name": "Google: Gemini 3.5 Flash",
      "provider_name": "Google",
      "accuracy": 0.7488737702702704,
      "accuracy_std_dev": 0.035681240826082414,
      "run_count": 6,
      "total_questions": 296,
      "avg_cost_per_task_usd": 0.3075426081081082,
      "avg_duration_per_task_ms": 149494.7972972973,
      "avg_output_tokens_per_task": 20582.41554054054,
      "last_run_timestamp": "2026-10-01T19:05:00.071Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "DeepInfra",
      "accuracy": 0.748479288321168,
      "accuracy_std_dev": 0.026337582738596148,
      "run_count": 11,
      "total_questions": 548,
      "avg_cost_per_task_usd": 0.04487602919708029,
      "avg_duration_per_task_ms": 472772.2098540146,
      "avg_output_tokens_per_task": 23140.4598540146,
      "last_run_timestamp": "2026-10-24T19:38:23.863Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.6-opus-20260205",
      "model_slug": "anthropic/claude-opus-4.6",
      "model_name": "Anthropic: Claude Opus 4.6",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7483334166666666,
      "accuracy_std_dev": 0.03210214167429313,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.5444264083333333,
      "avg_duration_per_task_ms": 182768.61333333334,
      "avg_output_tokens_per_task": 10413.385,
      "last_run_timestamp": "2026-10-24T12:39:07.404Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "GMICloud",
      "accuracy": 0.748333375,
      "accuracy_std_dev": 0.028431152223122702,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.021531867499999996,
      "avg_duration_per_task_ms": 145411.285,
      "avg_output_tokens_per_task": 16569.425,
      "last_run_timestamp": "2026-10-24T11:58:02.888Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta/muse-glimmer-30b-20260810",
      "model_slug": "meta/muse-glimmer-30b",
      "model_name": "Meta: Muse Glimmer 30B",
      "provider_name": "Together",
      "accuracy": 0.748333375,
      "accuracy_std_dev": 0.018782375754530515,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0418515775,
      "avg_duration_per_task_ms": 191146.0675,
      "avg_output_tokens_per_task": 12827.9725,
      "last_run_timestamp": "2026-10-03T08:33:09.729Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "AkashML",
      "accuracy": 0.748275724137931,
      "accuracy_std_dev": 0.011196981765934397,
      "run_count": 3,
      "total_questions": 145,
      "avg_cost_per_task_usd": 0.07340827586206897,
      "avg_duration_per_task_ms": 233348.6,
      "avg_output_tokens_per_task": 7130.5931034482755,
      "last_run_timestamp": "2026-08-15T08:10:05.152Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Nebius",
      "accuracy": 0.748275551724138,
      "accuracy_std_dev": 0.011091913731282908,
      "run_count": 3,
      "total_questions": 145,
      "avg_cost_per_task_usd": 0.00999303448275862,
      "avg_duration_per_task_ms": 128162.52413793103,
      "avg_output_tokens_per_task": 3554.88275862069,
      "last_run_timestamp": "2026-10-24T12:44:22.556Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "auto-routing",
      "accuracy": 0.7480002,
      "accuracy_std_dev": 0.020396130352593862,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.030171100000000003,
      "avg_duration_per_task_ms": 197857.312,
      "avg_output_tokens_per_task": 18663.832,
      "last_run_timestamp": "2026-10-03T08:18:33.163Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Fireworks",
      "accuracy": 0.747906015075377,
      "accuracy_std_dev": 0.009429275581931,
      "run_count": 4,
      "total_questions": 199,
      "avg_cost_per_task_usd": 0.0214910351758794,
      "avg_duration_per_task_ms": 74311.23618090453,
      "avg_output_tokens_per_task": 12217.904522613066,
      "last_run_timestamp": "2026-10-24T13:10:59.376Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "CoreWeave",
      "accuracy": 0.7477778333333334,
      "accuracy_std_dev": 0.022580213671978926,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.018034246666666667,
      "avg_duration_per_task_ms": 75225.08333333333,
      "avg_output_tokens_per_task": 4268.986666666667,
      "last_run_timestamp": "2026-10-24T12:38:26.301Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.6-opus-20260205",
      "model_slug": "anthropic/claude-opus-4.6",
      "model_name": "Anthropic: Claude Opus 4.6",
      "provider_name": "Azure",
      "accuracy": 0.74777775,
      "accuracy_std_dev": 0.02650409355843297,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.52115986,
      "avg_duration_per_task_ms": 227711.20333333334,
      "avg_output_tokens_per_task": 10054.68,
      "last_run_timestamp": "2026-10-24T12:47:43.608Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Ionstream",
      "accuracy": 0.747691088161209,
      "accuracy_std_dev": 0.03008101142513954,
      "run_count": 8,
      "total_questions": 397,
      "avg_cost_per_task_usd": 0.09758945340050378,
      "avg_duration_per_task_ms": 397184.00755667506,
      "avg_output_tokens_per_task": 25302.57934508816,
      "last_run_timestamp": "2026-10-24T13:02:31.690Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "x-ai/grok-4.6-20260810",
      "model_slug": "x-ai/grok-4.6",
      "model_name": "SpaceXAI: Grok 4.6",
      "provider_name": "xAI",
      "accuracy": 0.747619,
      "accuracy_std_dev": 0.025305279849978234,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.33171734571428574,
      "avg_duration_per_task_ms": 235739.5942857143,
      "avg_output_tokens_per_task": 12164.34,
      "last_run_timestamp": "2026-10-01T16:01:04.865Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Reka",
      "accuracy": 0.7475554666666667,
      "accuracy_std_dev": 0.02281875035978576,
      "run_count": 15,
      "total_questions": 750,
      "avg_cost_per_task_usd": 0.013525546666666666,
      "avg_duration_per_task_ms": 139030.82666666666,
      "avg_output_tokens_per_task": 15549.14,
      "last_run_timestamp": "2026-10-24T12:22:53.795Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Mara",
      "accuracy": 0.7473276647230321,
      "accuracy_std_dev": 0.024479710940170173,
      "run_count": 7,
      "total_questions": 343,
      "avg_cost_per_task_usd": 0.13171293294460643,
      "avg_duration_per_task_ms": 209506.55976676385,
      "avg_output_tokens_per_task": 11363.093294460641,
      "last_run_timestamp": "2026-10-24T19:49:56.630Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "Xiaomi",
      "accuracy": 0.7472728181818181,
      "accuracy_std_dev": 0.03871218656330843,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.023786492727272724,
      "avg_duration_per_task_ms": 413071.56363636366,
      "avg_output_tokens_per_task": 15329.412727272727,
      "last_run_timestamp": "2026-10-17T11:15:55.168Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "GMICloud",
      "accuracy": 0.7472727272727273,
      "accuracy_std_dev": 0.0336894846720913,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.10508834363636363,
      "avg_duration_per_task_ms": 310050.1672727273,
      "avg_output_tokens_per_task": 18603.976363636364,
      "last_run_timestamp": "2026-10-24T12:22:55.595Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7472726363636364,
      "accuracy_std_dev": 0.03984348764688863,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.04194175272727273,
      "avg_duration_per_task_ms": 190838.81636363638,
      "avg_output_tokens_per_task": 19126.147272727274,
      "last_run_timestamp": "2026-10-24T12:19:21.144Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "Modal",
      "accuracy": 0.7472221666666666,
      "accuracy_std_dev": 0.011613686586906374,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.17439429333333334,
      "avg_duration_per_task_ms": 161420.59666666668,
      "avg_output_tokens_per_task": 12499.9,
      "last_run_timestamp": "2026-10-10T11:19:48.441Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-122b-a10b-20260224",
      "model_slug": "qwen/qwen3.5-122b-a10b",
      "model_name": "Qwen: Qwen3.5-122B-A10B",
      "provider_name": "Alibaba",
      "accuracy": 0.7471783002257337,
      "accuracy_std_dev": 0.040296469031969095,
      "run_count": 9,
      "total_questions": 443,
      "avg_cost_per_task_usd": 0.11695101128668173,
      "avg_duration_per_task_ms": 466244.8329571106,
      "avg_output_tokens_per_task": 21735.31828442438,
      "last_run_timestamp": "2026-10-03T08:18:33.162Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "BaseTen",
      "accuracy": 0.747143,
      "accuracy_std_dev": 0.033305956785965055,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.014612914285714287,
      "avg_duration_per_task_ms": 50786.11142857143,
      "avg_output_tokens_per_task": 4096.908571428571,
      "last_run_timestamp": "2026-10-24T13:27:44.702Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "AkashML",
      "accuracy": 0.746766898623279,
      "accuracy_std_dev": 0.01597020524668026,
      "run_count": 16,
      "total_questions": 799,
      "avg_cost_per_task_usd": 0.045530197747183976,
      "avg_duration_per_task_ms": 308691.72590738424,
      "avg_output_tokens_per_task": 26256.58448060075,
      "last_run_timestamp": "2026-10-24T19:41:04.558Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Cloudflare",
      "accuracy": 0.746667,
      "accuracy_std_dev": 0.020000000000000018,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.00955731,
      "avg_duration_per_task_ms": 207347.51,
      "avg_output_tokens_per_task": 13127.37,
      "last_run_timestamp": "2026-10-24T12:22:57.425Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta/muse-spark-open-20261015",
      "model_slug": "meta/muse-spark-open",
      "model_name": "Meta: Muse Spark Open",
      "provider_name": "auto-routing",
      "accuracy": 0.746667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.08652918,
      "avg_duration_per_task_ms": 60257.98,
      "avg_output_tokens_per_task": 8644.22,
      "last_run_timestamp": "2026-10-24T12:40:48.538Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Relace",
      "accuracy": 0.746667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.00930136,
      "avg_duration_per_task_ms": 89811.58,
      "avg_output_tokens_per_task": 12020.36,
      "last_run_timestamp": "2026-10-24T15:18:33.444Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2-thinking-20251106",
      "model_slug": "moonshotai/kimi-k2-thinking",
      "model_name": "MoonshotAI: Kimi K2 Thinking",
      "provider_name": "Novita",
      "accuracy": 0.746667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.0524172,
      "avg_duration_per_task_ms": 255934.74,
      "avg_output_tokens_per_task": 8431.46,
      "last_run_timestamp": "2026-08-01T17:38:23.802Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "InferenceNet",
      "accuracy": 0.746667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.07074242,
      "avg_duration_per_task_ms": 800280.44,
      "avg_output_tokens_per_task": 15183.42,
      "last_run_timestamp": "2026-10-02T06:00:07.943Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "StreamLake",
      "accuracy": 0.7466667,
      "accuracy_std_dev": 0.02875178835150955,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.069247084,
      "avg_duration_per_task_ms": 254648.06,
      "avg_output_tokens_per_task": 9736.448,
      "last_run_timestamp": "2026-10-24T19:33:48.037Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.6-flash-20260721",
      "model_slug": "google/gemini-3.6-flash",
      "model_name": "Google: Gemini 3.6 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.7466666666666667,
      "accuracy_std_dev": 0.023726809342082992,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.21412848666666662,
      "avg_duration_per_task_ms": 142371.99333333335,
      "avg_output_tokens_per_task": 22756.126666666667,
      "last_run_timestamp": "2026-10-01T19:04:26.676Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-sol-20260709",
      "model_slug": "openai/gpt-5.6-sol",
      "model_name": "OpenAI: GPT-5.6 Sol",
      "provider_name": "auto-routing",
      "accuracy": 0.7466666666666667,
      "accuracy_std_dev": 0.028803034820811656,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.18399340666666666,
      "avg_duration_per_task_ms": 116870.40666666666,
      "avg_output_tokens_per_task": 5117.2266666666665,
      "last_run_timestamp": "2026-10-01T19:04:55.696Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "GMICloud",
      "accuracy": 0.746666625,
      "accuracy_std_dev": 0.025165982331996805,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0084872125,
      "avg_duration_per_task_ms": 222581.33,
      "avg_output_tokens_per_task": 5098.45,
      "last_run_timestamp": "2026-10-17T11:55:03.480Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Crusoe",
      "accuracy": 0.7466666000000001,
      "accuracy_std_dev": 0.030983694638309366,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.022173396,
      "avg_duration_per_task_ms": 122667.268,
      "avg_output_tokens_per_task": 7404.18,
      "last_run_timestamp": "2026-10-24T12:20:04.753Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta/muse-glimmer-30b-20260810",
      "model_slug": "meta/muse-glimmer-30b",
      "model_name": "Meta: Muse Glimmer 30B",
      "provider_name": "auto-routing",
      "accuracy": 0.7466665714285714,
      "accuracy_std_dev": 0.03380611384628165,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.03743281428571429,
      "avg_duration_per_task_ms": 197093.65714285715,
      "avg_output_tokens_per_task": 13601.3,
      "last_run_timestamp": "2026-10-03T08:15:57.213Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "CoreWeave",
      "accuracy": 0.7466665714285714,
      "accuracy_std_dev": 0.029814069333871523,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.02163002857142857,
      "avg_duration_per_task_ms": 164450.89714285714,
      "avg_output_tokens_per_task": 8529.557142857142,
      "last_run_timestamp": "2026-10-24T18:59:48.400Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7466665555555556,
      "accuracy_std_dev": 0.03471233459324771,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.07409725333333333,
      "avg_duration_per_task_ms": 215903.96666666667,
      "avg_output_tokens_per_task": 9487.915555555555,
      "last_run_timestamp": "2026-10-24T19:32:20.110Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Together",
      "accuracy": 0.7466665,
      "accuracy_std_dev": 0.026666500000000037,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.019887960000000003,
      "avg_duration_per_task_ms": 51072.42,
      "avg_output_tokens_per_task": 13340.71,
      "last_run_timestamp": "2026-10-24T12:26:21.165Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Relace",
      "accuracy": 0.7466665,
      "accuracy_std_dev": 0.0466665,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.011416649999999999,
      "avg_duration_per_task_ms": 95996.46,
      "avg_output_tokens_per_task": 13669.03,
      "last_run_timestamp": "2026-10-24T12:07:39.217Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta/muse-glimmer-30b-20260810",
      "model_slug": "meta/muse-glimmer-30b",
      "model_name": "Meta: Muse Glimmer 30B",
      "provider_name": "Fireworks",
      "accuracy": 0.7466665,
      "accuracy_std_dev": 0.011547101609639251,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.04422623666666667,
      "avg_duration_per_task_ms": 266653.04,
      "avg_output_tokens_per_task": 13360.296666666667,
      "last_run_timestamp": "2026-09-19T08:01:36.427Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.5-20260423",
      "model_slug": "openai/gpt-5.5",
      "model_name": "OpenAI: GPT-5.5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7466665,
      "accuracy_std_dev": 0.00666650000000002,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.58250779,
      "avg_duration_per_task_ms": 102747.13,
      "avg_output_tokens_per_task": 7546.85,
      "last_run_timestamp": "2026-10-01T19:04:42.459Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Sail Research",
      "accuracy": 0.7466665,
      "accuracy_std_dev": 0.019999888889274717,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.06283585333333332,
      "avg_duration_per_task_ms": 361191.69,
      "avg_output_tokens_per_task": 8869.28,
      "last_run_timestamp": "2026-08-29T08:25:32.049Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Wafer",
      "accuracy": 0.7463769999999998,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 46,
      "avg_cost_per_task_usd": 0.05924082608695652,
      "avg_duration_per_task_ms": 289784.32608695654,
      "avg_output_tokens_per_task": 9531.695652173914,
      "last_run_timestamp": "2026-08-15T14:23:29.125Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "Novita",
      "accuracy": 0.7459258888888889,
      "accuracy_std_dev": 0.025809336792221565,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.17245649777777777,
      "avg_duration_per_task_ms": 324708.56444444443,
      "avg_output_tokens_per_task": 11055.957777777778,
      "last_run_timestamp": "2026-10-24T12:33:16.196Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Venice",
      "accuracy": 0.7459258888888889,
      "accuracy_std_dev": 0.032423682687006024,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.10307451777777778,
      "avg_duration_per_task_ms": 332680.06888888887,
      "avg_output_tokens_per_task": 7684.475555555556,
      "last_run_timestamp": "2026-10-24T19:37:38.705Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Decart",
      "accuracy": 0.74583325,
      "accuracy_std_dev": 0.03402396010442498,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.02499106,
      "avg_duration_per_task_ms": 929627.78,
      "avg_output_tokens_per_task": 52531.49,
      "last_run_timestamp": "2026-10-24T12:31:49.408Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.7455556666666666,
      "accuracy_std_dev": 0.022166719578884227,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.03517121333333333,
      "avg_duration_per_task_ms": 100449.04,
      "avg_output_tokens_per_task": 17842.213333333333,
      "last_run_timestamp": "2026-10-24T12:20:51.421Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "InferenceNet",
      "accuracy": 0.7455556666666666,
      "accuracy_std_dev": 0.020846357411627908,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.018734011111111108,
      "avg_duration_per_task_ms": 376517.3044444444,
      "avg_output_tokens_per_task": 24556.922222222223,
      "last_run_timestamp": "2026-10-24T11:55:52.478Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.6-sonnet-20260217",
      "model_slug": "anthropic/claude-sonnet-4.6",
      "model_name": "Anthropic: Claude Sonnet 4.6",
      "provider_name": "auto-routing",
      "accuracy": 0.7455555833333333,
      "accuracy_std_dev": 0.027329632768170454,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.3244978516666666,
      "avg_duration_per_task_ms": 197778.63333333333,
      "avg_output_tokens_per_task": 11507.296666666667,
      "last_run_timestamp": "2026-10-24T12:37:04.719Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-122b-a10b-20260224",
      "model_slug": "qwen/qwen3.5-122b-a10b",
      "model_name": "Qwen: Qwen3.5-122B-A10B",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7455555833333333,
      "accuracy_std_dev": 0.10487495876237436,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.15016765666666665,
      "avg_duration_per_task_ms": 591828.39,
      "avg_output_tokens_per_task": 27540.295,
      "last_run_timestamp": "2026-10-24T19:26:44.780Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Morph",
      "accuracy": 0.7454957162162162,
      "accuracy_std_dev": 0.015757653442057918,
      "run_count": 3,
      "total_questions": 148,
      "avg_cost_per_task_usd": 0.12553497972972974,
      "avg_duration_per_task_ms": 598677.7297297297,
      "avg_output_tokens_per_task": 24585.641891891893,
      "last_run_timestamp": "2026-08-15T08:15:35.382Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.6-sonnet-20260217",
      "model_slug": "anthropic/claude-sonnet-4.6",
      "model_name": "Anthropic: Claude Sonnet 4.6",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.7453334,
      "accuracy_std_dev": 0.02543827737170895,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.346131274,
      "avg_duration_per_task_ms": 200655.086,
      "avg_output_tokens_per_task": 11839.45,
      "last_run_timestamp": "2026-10-24T13:00:12.600Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "auto-routing",
      "accuracy": 0.7451852222222222,
      "accuracy_std_dev": 0.02368052570633507,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.03867792222222222,
      "avg_duration_per_task_ms": 612636.5933333334,
      "avg_output_tokens_per_task": 15988.397777777778,
      "last_run_timestamp": "2026-10-24T19:37:04.226Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Phala",
      "accuracy": 0.7451852222222222,
      "accuracy_std_dev": 0.03170500430103102,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.13362894444444445,
      "avg_duration_per_task_ms": 336443.09555555554,
      "avg_output_tokens_per_task": 11874.128888888888,
      "last_run_timestamp": "2026-10-24T19:38:37.255Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7451852222222222,
      "accuracy_std_dev": 0.04034135688727106,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.11528159555555555,
      "avg_duration_per_task_ms": 255335.0888888889,
      "avg_output_tokens_per_task": 9618.166666666666,
      "last_run_timestamp": "2026-10-24T19:48:41.460Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "DeepInfra",
      "accuracy": 0.7451851111111111,
      "accuracy_std_dev": 0.030270966502370805,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.4099032177777778,
      "avg_duration_per_task_ms": 619397.1977777778,
      "avg_output_tokens_per_task": 9848.733333333334,
      "last_run_timestamp": "2026-10-24T12:05:22.704Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "StreamLake",
      "accuracy": 0.745000125,
      "accuracy_std_dev": 0.03312433437609538,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.015775572499999998,
      "avg_duration_per_task_ms": 148938.23,
      "avg_output_tokens_per_task": 15126.975,
      "last_run_timestamp": "2026-10-24T12:05:54.269Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "SambaNova",
      "accuracy": 0.745,
      "accuracy_std_dev": 0.026822730826670128,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.15284815,
      "avg_duration_per_task_ms": 494527.49,
      "avg_output_tokens_per_task": 28246.935,
      "last_run_timestamp": "2026-10-17T19:47:21.934Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Relace",
      "accuracy": 0.745,
      "accuracy_std_dev": 0.028332999999999997,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01250761,
      "avg_duration_per_task_ms": 355978.98,
      "avg_output_tokens_per_task": 23604.2,
      "last_run_timestamp": "2026-10-24T12:29:47.596Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Near AI",
      "accuracy": 0.745,
      "accuracy_std_dev": 0.040380248197602775,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.007515549999999999,
      "avg_duration_per_task_ms": 328112.605,
      "avg_output_tokens_per_task": 3998.705,
      "last_run_timestamp": "2026-10-24T13:13:17.945Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Morph",
      "accuracy": 0.7448485454545455,
      "accuracy_std_dev": 0.021944100032763564,
      "run_count": 22,
      "total_questions": 1100,
      "avg_cost_per_task_usd": 0.4515615263636363,
      "avg_duration_per_task_ms": 587567.8727272727,
      "avg_output_tokens_per_task": 10197.047272727274,
      "last_run_timestamp": "2026-10-24T11:44:25.226Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Io Net",
      "accuracy": 0.7448276551724138,
      "accuracy_std_dev": 0.0035438860340341496,
      "run_count": 3,
      "total_questions": 145,
      "avg_cost_per_task_usd": 0.022715406896551726,
      "avg_duration_per_task_ms": 289516.0620689655,
      "avg_output_tokens_per_task": 10103.63448275862,
      "last_run_timestamp": "2026-08-22T08:32:23.660Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Together",
      "accuracy": 0.744681,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 47,
      "avg_cost_per_task_usd": 0.05595836170212766,
      "avg_duration_per_task_ms": 34218.255319148935,
      "avg_output_tokens_per_task": 6745.085106382979,
      "last_run_timestamp": "2026-08-04T03:12:18.736Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Cloudflare",
      "accuracy": 0.7446668,
      "accuracy_std_dev": 0.025130744341543096,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.079246066,
      "avg_duration_per_task_ms": 290947.32,
      "avg_output_tokens_per_task": 11272.336,
      "last_run_timestamp": "2026-10-24T23:01:57.882Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "DeepInfra",
      "accuracy": 0.7445560603015074,
      "accuracy_std_dev": 0.018800290223823686,
      "run_count": 8,
      "total_questions": 398,
      "avg_cost_per_task_usd": 0.0472956608040201,
      "avg_duration_per_task_ms": 744919.2989949749,
      "avg_output_tokens_per_task": 14283.391959798995,
      "last_run_timestamp": "2026-10-24T19:19:10.662Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "amazon/nova-micro-v1",
      "model_slug": "amazon/nova-micro-v1",
      "model_name": "Amazon: Nova Micro 1.0",
      "provider_name": "Anthropic",
      "accuracy": 0.7444443333333334,
      "accuracy_std_dev": 0.00831485358994506,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 1.2319787333333332,
      "avg_duration_per_task_ms": 98531.56,
      "avg_output_tokens_per_task": 5524.42,
      "last_run_timestamp": "2026-10-01T19:04:26.700Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "Reka",
      "accuracy": 0.7444443333333334,
      "accuracy_std_dev": 0.006285236475706844,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.04041968666666667,
      "avg_duration_per_task_ms": 291588.68666666665,
      "avg_output_tokens_per_task": 24330.48,
      "last_run_timestamp": "2026-09-30T11:25:05.946Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Venice",
      "accuracy": 0.7444070973154362,
      "accuracy_std_dev": 0.02348016052421823,
      "run_count": 12,
      "total_questions": 596,
      "avg_cost_per_task_usd": 0.1433310234899329,
      "avg_duration_per_task_ms": 614878.5134228187,
      "avg_output_tokens_per_task": 26835.629194630874,
      "last_run_timestamp": "2026-10-24T20:02:49.919Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Inceptron",
      "accuracy": 0.7442960934891485,
      "accuracy_std_dev": 0.02614118042622386,
      "run_count": 12,
      "total_questions": 599,
      "avg_cost_per_task_usd": 0.011909569282136896,
      "avg_duration_per_task_ms": 437548.9916527546,
      "avg_output_tokens_per_task": 14218.791318864774,
      "last_run_timestamp": "2026-10-24T12:42:25.209Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Relace",
      "accuracy": 0.7442423636363636,
      "accuracy_std_dev": 0.04102458399827359,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.011285496363636365,
      "avg_duration_per_task_ms": 301862.60363636364,
      "avg_output_tokens_per_task": 16012.072727272727,
      "last_run_timestamp": "2026-10-24T12:46:15.681Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "Novita",
      "accuracy": 0.7442423636363636,
      "accuracy_std_dev": 0.026858821460209414,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.02083015818181818,
      "avg_duration_per_task_ms": 406483.71090909094,
      "avg_output_tokens_per_task": 12085.40909090909,
      "last_run_timestamp": "2026-10-17T11:29:07.053Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "DigitalOcean",
      "accuracy": 0.74416675,
      "accuracy_std_dev": 0.03012731088543251,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.064155875,
      "avg_duration_per_task_ms": 1722921.32,
      "avg_output_tokens_per_task": 8915.945,
      "last_run_timestamp": "2026-08-29T16:20:58.762Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "Together",
      "accuracy": 0.744131569416499,
      "accuracy_std_dev": 0.0294471423722235,
      "run_count": 10,
      "total_questions": 497,
      "avg_cost_per_task_usd": 0.06635609456740442,
      "avg_duration_per_task_ms": 1016068.6076458753,
      "avg_output_tokens_per_task": 74326.28571428571,
      "last_run_timestamp": "2026-10-24T19:38:23.433Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Wafer",
      "accuracy": 0.744036743119266,
      "accuracy_std_dev": 0.02171918727532931,
      "run_count": 11,
      "total_questions": 545,
      "avg_cost_per_task_usd": 0.025592662385321107,
      "avg_duration_per_task_ms": 271579.7009174312,
      "avg_output_tokens_per_task": 19610.990825688074,
      "last_run_timestamp": "2026-10-24T12:32:51.108Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Ambient",
      "accuracy": 0.7439998000000001,
      "accuracy_std_dev": 0.025508095447524134,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.010447856000000002,
      "avg_duration_per_task_ms": 427694.796,
      "avg_output_tokens_per_task": 12591.384,
      "last_run_timestamp": "2026-08-29T09:31:13.940Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "StreamLake",
      "accuracy": 0.7439515846774194,
      "accuracy_std_dev": 0.044830576271223645,
      "run_count": 10,
      "total_questions": 496,
      "avg_cost_per_task_usd": 0.07215105040322582,
      "avg_duration_per_task_ms": 329377.40524193546,
      "avg_output_tokens_per_task": 12649.491935483871,
      "last_run_timestamp": "2026-10-24T19:51:25.882Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "DeepInfra",
      "accuracy": 0.7439125188916877,
      "accuracy_std_dev": 0.0319877187270267,
      "run_count": 8,
      "total_questions": 397,
      "avg_cost_per_task_usd": 0.06879249370277078,
      "avg_duration_per_task_ms": 463688.63979848864,
      "avg_output_tokens_per_task": 10615.753148614609,
      "last_run_timestamp": "2026-10-24T19:42:30.512Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Parasail",
      "accuracy": 0.7438111658031088,
      "accuracy_std_dev": 0.030675189182609607,
      "run_count": 12,
      "total_questions": 579,
      "avg_cost_per_task_usd": 0.021574207253886007,
      "avg_duration_per_task_ms": 94130.40587219344,
      "avg_output_tokens_per_task": 7255.53195164076,
      "last_run_timestamp": "2026-10-24T18:57:18.438Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "ModelRun",
      "accuracy": 0.7438096428571428,
      "accuracy_std_dev": 0.027105232069017705,
      "run_count": 14,
      "total_questions": 700,
      "avg_cost_per_task_usd": 0.06181787428571429,
      "avg_duration_per_task_ms": 140380.01,
      "avg_output_tokens_per_task": 9906.655714285715,
      "last_run_timestamp": "2026-10-24T19:18:50.638Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Io Net",
      "accuracy": 0.743738,
      "accuracy_std_dev": 0.01688456971344439,
      "run_count": 7,
      "total_questions": 346,
      "avg_cost_per_task_usd": 0.00938970520231214,
      "avg_duration_per_task_ms": 347914.32658959535,
      "avg_output_tokens_per_task": 3877.916184971098,
      "last_run_timestamp": "2026-10-03T08:46:08.334Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7436412530120482,
      "accuracy_std_dev": 0.036080178615965866,
      "run_count": 10,
      "total_questions": 498,
      "avg_cost_per_task_usd": 0.07742782530120482,
      "avg_duration_per_task_ms": 824019.6144578314,
      "avg_output_tokens_per_task": 14546.923694779116,
      "last_run_timestamp": "2026-10-24T20:35:30.856Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Mancer 2",
      "accuracy": 0.7433335,
      "accuracy_std_dev": 0.0033334999999999892,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.06141422999999999,
      "avg_duration_per_task_ms": 1060280.1,
      "avg_output_tokens_per_task": 12094.68,
      "last_run_timestamp": "2026-10-24T12:42:07.189Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "InferenceNet",
      "accuracy": 0.7433335,
      "accuracy_std_dev": 0.023333285714832386,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.006871515,
      "avg_duration_per_task_ms": 97438.585,
      "avg_output_tokens_per_task": 9380.645,
      "last_run_timestamp": "2026-10-24T12:26:07.751Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "StreamLake",
      "accuracy": 0.7433335,
      "accuracy_std_dev": 0.03666649999999999,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.03822515,
      "avg_duration_per_task_ms": 176603.12,
      "avg_output_tokens_per_task": 16434.51,
      "last_run_timestamp": "2026-10-24T18:56:11.637Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "DeepInfra",
      "accuracy": 0.7433335,
      "accuracy_std_dev": 0.01666650000000003,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.03870923,
      "avg_duration_per_task_ms": 112805.69,
      "avg_output_tokens_per_task": 14592.13,
      "last_run_timestamp": "2026-10-24T12:30:47.912Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "x-ai/grok-4.6-20260810",
      "model_slug": "x-ai/grok-4.6",
      "model_name": "SpaceXAI: Grok 4.6",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7433335,
      "accuracy_std_dev": 0.043333500000000025,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.64823893,
      "avg_duration_per_task_ms": 264133.83,
      "avg_output_tokens_per_task": 10511,
      "last_run_timestamp": "2026-10-01T16:00:37.003Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Cloudflare",
      "accuracy": 0.7433333333333333,
      "accuracy_std_dev": 0.017532868606008445,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.012350623333333333,
      "avg_duration_per_task_ms": 125234.99333333333,
      "avg_output_tokens_per_task": 4198.8,
      "last_run_timestamp": "2026-10-03T08:23:09.816Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "Google",
      "accuracy": 0.7433332500000001,
      "accuracy_std_dev": 0.01774311029726281,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.013128406666666667,
      "avg_duration_per_task_ms": 72447.74833333334,
      "avg_output_tokens_per_task": 10872.788333333334,
      "last_run_timestamp": "2026-10-24T12:05:47.786Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "auto-routing",
      "accuracy": 0.7433332500000001,
      "accuracy_std_dev": 0.022110882189263745,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.010621789999999999,
      "avg_duration_per_task_ms": 218070.5775,
      "avg_output_tokens_per_task": 14077.9775,
      "last_run_timestamp": "2026-10-24T19:37:39.139Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.6-sonnet-20260217",
      "model_slug": "anthropic/claude-sonnet-4.6",
      "model_name": "Anthropic: Claude Sonnet 4.6",
      "provider_name": "Google",
      "accuracy": 0.7431372352941178,
      "accuracy_std_dev": 0.03152894888087569,
      "run_count": 34,
      "total_questions": 1700,
      "avg_cost_per_task_usd": 0.37377429588235295,
      "avg_duration_per_task_ms": 200222.53705882354,
      "avg_output_tokens_per_task": 11930.03411764706,
      "last_run_timestamp": "2026-10-24T13:14:09.404Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "auto-routing",
      "accuracy": 0.7430303636363637,
      "accuracy_std_dev": 0.0482721820820095,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.009174501818181817,
      "avg_duration_per_task_ms": 294400.0709090909,
      "avg_output_tokens_per_task": 15379.543636363636,
      "last_run_timestamp": "2026-10-24T11:53:54.891Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-opus-5.5-20260921",
      "model_slug": "anthropic/claude-opus-5.5",
      "model_name": "Anthropic: Claude Opus 5.5",
      "provider_name": "Google",
      "accuracy": 0.7429629444444444,
      "accuracy_std_dev": 0.018622075637599298,
      "run_count": 18,
      "total_questions": 900,
      "avg_cost_per_task_usd": 0.35968263222222224,
      "avg_duration_per_task_ms": 87204.23444444445,
      "avg_output_tokens_per_task": 7123.5233333333335,
      "last_run_timestamp": "2026-10-24T12:18:54.175Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Alibaba",
      "accuracy": 0.742962888888889,
      "accuracy_std_dev": 0.03570948009343192,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.35507732222222216,
      "avg_duration_per_task_ms": 386491.71777777775,
      "avg_output_tokens_per_task": 9636.25111111111,
      "last_run_timestamp": "2026-10-24T12:16:41.815Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7429378258859785,
      "accuracy_std_dev": 0.02605274813906652,
      "run_count": 13,
      "total_questions": 649,
      "avg_cost_per_task_usd": 0.051288061633281964,
      "avg_duration_per_task_ms": 216591.9815100154,
      "avg_output_tokens_per_task": 9230.906009244993,
      "last_run_timestamp": "2026-10-24T19:33:32.803Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "AkashML",
      "accuracy": 0.7426668,
      "accuracy_std_dev": 0.03690245640821219,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.011645404,
      "avg_duration_per_task_ms": 585840.916,
      "avg_output_tokens_per_task": 18794.172,
      "last_run_timestamp": "2026-08-29T08:43:07.210Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.6-opus-20260205",
      "model_slug": "anthropic/claude-opus-4.6",
      "model_name": "Anthropic: Claude Opus 4.6",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.7426666000000001,
      "accuracy_std_dev": 0.02908997383704564,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.504919826,
      "avg_duration_per_task_ms": 191444.46,
      "avg_output_tokens_per_task": 9912.906,
      "last_run_timestamp": "2026-10-24T11:43:06.361Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7426666000000001,
      "accuracy_std_dev": 0.01768882635564045,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.056281936000000005,
      "avg_duration_per_task_ms": 224029.02,
      "avg_output_tokens_per_task": 9140.692,
      "last_run_timestamp": "2026-10-04T00:00:34.896Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Sail Research",
      "accuracy": 0.7425926666666667,
      "accuracy_std_dev": 0.03165311687654157,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.013586744444444443,
      "avg_duration_per_task_ms": 236886.98,
      "avg_output_tokens_per_task": 22693.746666666666,
      "last_run_timestamp": "2026-10-24T12:29:34.092Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "DeepInfra",
      "accuracy": 0.7425595803571429,
      "accuracy_std_dev": 0.02504151285854947,
      "run_count": 7,
      "total_questions": 336,
      "avg_cost_per_task_usd": 0.15002946726190475,
      "avg_duration_per_task_ms": 711525.6666666666,
      "avg_output_tokens_per_task": 29928.91369047619,
      "last_run_timestamp": "2026-10-24T19:30:34.911Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "Together",
      "accuracy": 0.7425,
      "accuracy_std_dev": 0.01697637247028941,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.22007109249999998,
      "avg_duration_per_task_ms": 261814.77,
      "avg_output_tokens_per_task": 11695.5675,
      "last_run_timestamp": "2026-10-24T11:58:56.378Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "Novita",
      "accuracy": 0.7424244545454545,
      "accuracy_std_dev": 0.04131889777277492,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.009867483636363635,
      "avg_duration_per_task_ms": 412569.9109090909,
      "avg_output_tokens_per_task": 14715.354545454546,
      "last_run_timestamp": "2026-10-17T11:26:35.790Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.1-pro-preview-20260219",
      "model_slug": "google/gemini-3.1-pro-preview",
      "model_name": "Google: Gemini 3.1 Pro Preview",
      "provider_name": "Google AI Studio",
      "accuracy": 0.7422223333333333,
      "accuracy_std_dev": 0.006285707880227596,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.38184122000000004,
      "avg_duration_per_task_ms": 149071.25333333333,
      "avg_output_tokens_per_task": 12794.173333333334,
      "last_run_timestamp": "2026-10-01T19:04:55.722Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "inclusionai/ling-3.0-flash-20260723",
      "model_slug": "inclusionai/ling-3.0-flash",
      "model_name": "inclusionAI: Ling 3.0 Flash",
      "provider_name": "Novita",
      "accuracy": 0.7422223333333333,
      "accuracy_std_dev": 0.003142853940113798,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.0028365266666666666,
      "avg_duration_per_task_ms": 68565.14666666667,
      "avg_output_tokens_per_task": 11675.026666666667,
      "last_run_timestamp": "2026-10-01T16:01:02.671Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7422223333333333,
      "accuracy_std_dev": 0.028284218870442763,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.0722646911111111,
      "avg_duration_per_task_ms": 320607.4111111111,
      "avg_output_tokens_per_task": 11515.595555555556,
      "last_run_timestamp": "2026-10-24T22:27:21.158Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7421122716049383,
      "accuracy_std_dev": 0.04949233809146621,
      "run_count": 5,
      "total_questions": 243,
      "avg_cost_per_task_usd": 0.07618280658436215,
      "avg_duration_per_task_ms": 605963.6255144033,
      "avg_output_tokens_per_task": 7686.909465020576,
      "last_run_timestamp": "2026-08-29T17:13:58.312Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Reka",
      "accuracy": 0.7417778,
      "accuracy_std_dev": 0.0192398742345162,
      "run_count": 15,
      "total_questions": 750,
      "avg_cost_per_task_usd": 0.016580628,
      "avg_duration_per_task_ms": 168506.08666666667,
      "avg_output_tokens_per_task": 4247.114666666666,
      "last_run_timestamp": "2026-10-24T13:06:52.407Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "Anthropic",
      "accuracy": 0.7416665,
      "accuracy_std_dev": 0.03140250343921645,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.012469425000000001,
      "avg_duration_per_task_ms": 90319.205,
      "avg_output_tokens_per_task": 10893.135,
      "last_run_timestamp": "2026-10-24T12:16:35.471Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Z.AI",
      "accuracy": 0.7414814444444444,
      "accuracy_std_dev": 0.014665955839677083,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.11785513555555556,
      "avg_duration_per_task_ms": 351306.1688888889,
      "avg_output_tokens_per_task": 9806.686666666666,
      "last_run_timestamp": "2026-10-24T20:00:28.580Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Ambient",
      "accuracy": 0.7413334,
      "accuracy_std_dev": 0.02578533261061415,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.064516224,
      "avg_duration_per_task_ms": 309719.048,
      "avg_output_tokens_per_task": 8304.048,
      "last_run_timestamp": "2026-08-29T08:12:20.454Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "auto-routing",
      "accuracy": 0.7412120909090909,
      "accuracy_std_dev": 0.033492690704511933,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.027686032727272727,
      "avg_duration_per_task_ms": 442749.78545454546,
      "avg_output_tokens_per_task": 14610.476363636364,
      "last_run_timestamp": "2026-10-17T13:55:17.194Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7411110000000001,
      "accuracy_std_dev": 0.02078704487575536,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.06692624,
      "avg_duration_per_task_ms": 156822.04,
      "avg_output_tokens_per_task": 7280.366666666667,
      "last_run_timestamp": "2026-08-15T08:13:33.738Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Relace",
      "accuracy": 0.7410191879699248,
      "accuracy_std_dev": 0.023857300276526983,
      "run_count": 8,
      "total_questions": 399,
      "avg_cost_per_task_usd": 0.22161520300751877,
      "avg_duration_per_task_ms": 375984.74686716794,
      "avg_output_tokens_per_task": 9942.047619047618,
      "last_run_timestamp": "2026-10-24T12:46:10.015Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "SiliconFlow",
      "accuracy": 0.74083325,
      "accuracy_std_dev": 0.030080976362769567,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.05869578,
      "avg_duration_per_task_ms": 714580.6575,
      "avg_output_tokens_per_task": 15701.38,
      "last_run_timestamp": "2026-10-24T19:50:11.361Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Morph",
      "accuracy": 0.7406779661016949,
      "accuracy_std_dev": 0.0192449897289312,
      "run_count": 6,
      "total_questions": 295,
      "avg_cost_per_task_usd": 0.02004586440677966,
      "avg_duration_per_task_ms": 1117850.5423728814,
      "avg_output_tokens_per_task": 22657.922033898307,
      "last_run_timestamp": "2026-09-12T08:11:08.573Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7405555,
      "accuracy_std_dev": 0.01855080944101005,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.0133585,
      "avg_duration_per_task_ms": 101370.275,
      "avg_output_tokens_per_task": 10830.071666666667,
      "last_run_timestamp": "2026-10-24T11:56:06.063Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-sol-20260709",
      "model_slug": "openai/gpt-5.6-sol",
      "model_name": "OpenAI: GPT-5.6 Sol",
      "provider_name": "Azure",
      "accuracy": 0.7403703333333334,
      "accuracy_std_dev": 0.030325141571529955,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.3409674066666667,
      "avg_duration_per_task_ms": 128710.22,
      "avg_output_tokens_per_task": 5246.528888888889,
      "last_run_timestamp": "2026-10-01T19:05:16.562Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Fireworks",
      "accuracy": 0.7403685728643216,
      "accuracy_std_dev": 0.02451522166373828,
      "run_count": 12,
      "total_questions": 597,
      "avg_cost_per_task_usd": 0.05593858793969849,
      "avg_duration_per_task_ms": 169768.54773869348,
      "avg_output_tokens_per_task": 30975.941373534337,
      "last_run_timestamp": "2026-10-24T12:06:17.420Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "GMICloud",
      "accuracy": 0.7400002,
      "accuracy_std_dev": 0.01264900523203309,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.034869184,
      "avg_duration_per_task_ms": 129959.044,
      "avg_output_tokens_per_task": 16349.368,
      "last_run_timestamp": "2026-10-03T08:21:01.017Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Modal",
      "accuracy": 0.7400000999999999,
      "accuracy_std_dev": 0.02022089128821974,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.37108522400000005,
      "avg_duration_per_task_ms": 184691.406,
      "avg_output_tokens_per_task": 10675.186,
      "last_run_timestamp": "2026-10-24T12:36:29.628Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.2-exp",
      "model_slug": "deepseek/deepseek-v3.2-exp",
      "model_name": "DeepSeek: DeepSeek V3.2 Exp",
      "provider_name": "AtlasCloud",
      "accuracy": 0.74,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.08086026,
      "avg_duration_per_task_ms": 522771.74,
      "avg_output_tokens_per_task": 13545.48,
      "last_run_timestamp": "2026-08-01T17:38:10.889Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "ModelRun",
      "accuracy": 0.74,
      "accuracy_std_dev": 0.033333,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.024783569999999998,
      "avg_duration_per_task_ms": 140465.05,
      "avg_output_tokens_per_task": 20671.25,
      "last_run_timestamp": "2026-10-24T12:32:09.787Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "OpenInference",
      "accuracy": 0.74,
      "accuracy_std_dev": 0.020000000000000018,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.02650346,
      "avg_duration_per_task_ms": 245657.46,
      "avg_output_tokens_per_task": 33745.71,
      "last_run_timestamp": "2026-10-24T12:08:32.744Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.74,
      "accuracy_std_dev": 0.033333,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.02533324,
      "avg_duration_per_task_ms": 117353.99,
      "avg_output_tokens_per_task": 11491.8,
      "last_run_timestamp": "2026-10-24T11:51:14.108Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta/muse-spark-open-20261015",
      "model_slug": "meta/muse-spark-open",
      "model_name": "Meta: Muse Spark Open",
      "provider_name": "Together",
      "accuracy": 0.74,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.08715152,
      "avg_duration_per_task_ms": 69751.7,
      "avg_output_tokens_per_task": 8640,
      "last_run_timestamp": "2026-10-24T11:33:58.313Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-sol-pro-20260709",
      "model_slug": "openai/gpt-5.6-sol-pro",
      "model_name": "OpenAI: GPT-5.6 Sol Pro",
      "provider_name": "OpenAI",
      "accuracy": 0.74,
      "accuracy_std_dev": 0.044555500139339316,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 1.0935073933333332,
      "avg_duration_per_task_ms": 295552.94,
      "avg_output_tokens_per_task": 22860.126666666667,
      "last_run_timestamp": "2026-10-02T23:08:29.026Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Cohere",
      "accuracy": 0.74,
      "accuracy_std_dev": 0.01333300000000004,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01884005,
      "avg_duration_per_task_ms": 67349.35,
      "avg_output_tokens_per_task": 22014.7,
      "last_run_timestamp": "2026-10-24T11:48:47.797Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.74,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.026460249999999998,
      "avg_duration_per_task_ms": 73471.9,
      "avg_output_tokens_per_task": 4407.31,
      "last_run_timestamp": "2026-10-24T13:04:21.529Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.5-sonnet-20250929",
      "model_slug": "anthropic/claude-sonnet-4.5",
      "model_name": "Anthropic: Claude Sonnet 4.5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.7399999,
      "accuracy_std_dev": 0.028906131607152153,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.34949705000000003,
      "avg_duration_per_task_ms": 220413.082,
      "avg_output_tokens_per_task": 10856.716,
      "last_run_timestamp": "2026-10-24T12:39:24.524Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Morph",
      "accuracy": 0.7399998571428572,
      "accuracy_std_dev": 0.023094052007442304,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.01886730285714286,
      "avg_duration_per_task_ms": 206247.68,
      "avg_output_tokens_per_task": 4550.928571428572,
      "last_run_timestamp": "2026-10-10T11:50:34.352Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Makora",
      "accuracy": 0.7395835,
      "accuracy_std_dev": 0.022012461748973016,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.042535722500000005,
      "avg_duration_per_task_ms": 134957.37,
      "avg_output_tokens_per_task": 25971.215,
      "last_run_timestamp": "2026-10-23T03:54:04.118Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Friendli",
      "accuracy": 0.7395367713004484,
      "accuracy_std_dev": 0.016058910754384415,
      "run_count": 9,
      "total_questions": 446,
      "avg_cost_per_task_usd": 0.017859735426008966,
      "avg_duration_per_task_ms": 83110.88116591929,
      "avg_output_tokens_per_task": 4085.5941704035877,
      "last_run_timestamp": "2026-10-24T12:42:47.974Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Mancer 2",
      "accuracy": 0.7393333,
      "accuracy_std_dev": 0.04912106993144592,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.04842688399999999,
      "avg_duration_per_task_ms": 284804.346,
      "avg_output_tokens_per_task": 14992.81,
      "last_run_timestamp": "2026-10-24T12:17:04.178Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "Venice",
      "accuracy": 0.7392593333333333,
      "accuracy_std_dev": 0.036167601020188705,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.04373026666666667,
      "avg_duration_per_task_ms": 581663.7222222222,
      "avg_output_tokens_per_task": 9145.893333333333,
      "last_run_timestamp": "2026-10-24T19:13:45.181Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7392592222222222,
      "accuracy_std_dev": 0.04548160272700695,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.16905808444444445,
      "avg_duration_per_task_ms": 289473.89555555553,
      "avg_output_tokens_per_task": 9765.786666666667,
      "last_run_timestamp": "2026-10-24T11:42:23.888Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Phala",
      "accuracy": 0.7392592222222222,
      "accuracy_std_dev": 0.03955854406875342,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.12468993111111111,
      "avg_duration_per_task_ms": 675089.5488888889,
      "avg_output_tokens_per_task": 35068.42222222222,
      "last_run_timestamp": "2026-10-24T11:39:42.358Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7392592222222222,
      "accuracy_std_dev": 0.02938791032145241,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.1616571777777778,
      "avg_duration_per_task_ms": 223287.44222222222,
      "avg_output_tokens_per_task": 9621.846666666666,
      "last_run_timestamp": "2026-10-24T19:37:14.525Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Venice",
      "accuracy": 0.73916675,
      "accuracy_std_dev": 0.029521604613189638,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.079499585,
      "avg_duration_per_task_ms": 501652.2525,
      "avg_output_tokens_per_task": 11995.7875,
      "last_run_timestamp": "2026-10-24T23:08:37.435Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "CoreWeave",
      "accuracy": 0.738889,
      "accuracy_std_dev": 0.017812306747489716,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.022993023333333338,
      "avg_duration_per_task_ms": 282964.01666666666,
      "avg_output_tokens_per_task": 13071.093333333334,
      "last_run_timestamp": "2026-08-29T16:33:29.042Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "StreamLake",
      "accuracy": 0.738667,
      "accuracy_std_dev": 0.020396078054371124,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.019569312,
      "avg_duration_per_task_ms": 507361.484,
      "avg_output_tokens_per_task": 14742.804,
      "last_run_timestamp": "2026-08-29T16:31:42.782Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Decart",
      "accuracy": 0.7386668,
      "accuracy_std_dev": 0.03166503013357166,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.009092168000000001,
      "avg_duration_per_task_ms": 101935.708,
      "avg_output_tokens_per_task": 5027.116,
      "last_run_timestamp": "2026-10-17T11:17:02.460Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.4-20260305",
      "model_slug": "openai/gpt-5.4",
      "model_name": "OpenAI: GPT-5.4",
      "provider_name": "Azure",
      "accuracy": 0.7385184444444445,
      "accuracy_std_dev": 0.017992647486688756,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.3103866155555556,
      "avg_duration_per_task_ms": 191198.25333333333,
      "avg_output_tokens_per_task": 10879.702222222222,
      "last_run_timestamp": "2026-10-01T19:05:18.005Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Makora",
      "accuracy": 0.7384824959349594,
      "accuracy_std_dev": 0.030529621622286764,
      "run_count": 5,
      "total_questions": 246,
      "avg_cost_per_task_usd": 0.012032963414634147,
      "avg_duration_per_task_ms": 61127.26829268293,
      "avg_output_tokens_per_task": 4200.292682926829,
      "last_run_timestamp": "2026-10-24T12:38:23.545Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "DeepInfra",
      "accuracy": 0.7383335,
      "accuracy_std_dev": 0.0016665000000000152,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.08170183999999998,
      "avg_duration_per_task_ms": 327224.56,
      "avg_output_tokens_per_task": 8775.96,
      "last_run_timestamp": "2026-08-22T16:12:20.461Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Parasail",
      "accuracy": 0.73833325,
      "accuracy_std_dev": 0.02153795745161318,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.34754801999999996,
      "avg_duration_per_task_ms": 197375.055,
      "avg_output_tokens_per_task": 8811.4275,
      "last_run_timestamp": "2026-10-24T12:17:34.198Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "StreamLake",
      "accuracy": 0.7382021797752809,
      "accuracy_std_dev": 0.021880590311850887,
      "run_count": 9,
      "total_questions": 445,
      "avg_cost_per_task_usd": 0.020498339325842697,
      "avg_duration_per_task_ms": 307841.11235955055,
      "avg_output_tokens_per_task": 10994.101123595505,
      "last_run_timestamp": "2026-10-17T11:09:59.070Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Crusoe",
      "accuracy": 0.73809525,
      "accuracy_std_dev": 0.03175464476067312,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.09706365848214285,
      "avg_duration_per_task_ms": 342421.30803571426,
      "avg_output_tokens_per_task": 11142.66517857143,
      "last_run_timestamp": "2026-10-24T20:22:47.990Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "Venice",
      "accuracy": 0.7380950401002507,
      "accuracy_std_dev": 0.022649397541656617,
      "run_count": 8,
      "total_questions": 399,
      "avg_cost_per_task_usd": 0.053091303258145366,
      "avg_duration_per_task_ms": 777194.3383458647,
      "avg_output_tokens_per_task": 14178.676691729323,
      "last_run_timestamp": "2026-10-24T19:40:41.765Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Together",
      "accuracy": 0.7377776666666667,
      "accuracy_std_dev": 0.017777750001367207,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.014543953333333333,
      "avg_duration_per_task_ms": 73473.39777777778,
      "avg_output_tokens_per_task": 4452.611111111111,
      "last_run_timestamp": "2026-10-24T12:18:30.719Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "Xiaomi",
      "accuracy": 0.7375756363636364,
      "accuracy_std_dev": 0.024498683461000814,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.008059856363636365,
      "avg_duration_per_task_ms": 393755.9490909091,
      "avg_output_tokens_per_task": 14461.965454545454,
      "last_run_timestamp": "2026-10-17T12:09:02.862Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Moonshot AI",
      "accuracy": 0.7374528407310704,
      "accuracy_std_dev": 0.027468652356798965,
      "run_count": 23,
      "total_questions": 1149,
      "avg_cost_per_task_usd": 0.12826692602262835,
      "avg_duration_per_task_ms": 160716.89469103568,
      "avg_output_tokens_per_task": 9299.472584856398,
      "last_run_timestamp": "2026-10-24T19:48:18.775Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Ionstream",
      "accuracy": 0.7373334,
      "accuracy_std_dev": 0.020482933819157834,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.039714615999999994,
      "avg_duration_per_task_ms": 174773.632,
      "avg_output_tokens_per_task": 27822.08,
      "last_run_timestamp": "2026-10-21T12:36:51.217Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Phala",
      "accuracy": 0.7371428571428571,
      "accuracy_std_dev": 0.023866691370602513,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.022481400000000002,
      "avg_duration_per_task_ms": 144173.6,
      "avg_output_tokens_per_task": 16498.17142857143,
      "last_run_timestamp": "2026-10-24T12:33:37.198Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "Novita",
      "accuracy": 0.737037,
      "accuracy_std_dev": 0.03543182483775474,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.09546396444444445,
      "avg_duration_per_task_ms": 226782.94222222222,
      "avg_output_tokens_per_task": 12230.953333333333,
      "last_run_timestamp": "2026-10-24T19:51:30.269Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "auto-routing",
      "accuracy": 0.737036888888889,
      "accuracy_std_dev": 0.020872504739459674,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.06686444222222222,
      "avg_duration_per_task_ms": 272840.34222222224,
      "avg_output_tokens_per_task": 12781.422222222222,
      "last_run_timestamp": "2026-10-24T20:26:13.288Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-sol-pro-20260709",
      "model_slug": "openai/gpt-5.6-sol-pro",
      "model_name": "OpenAI: GPT-5.6 Sol Pro",
      "provider_name": "auto-routing",
      "accuracy": 0.736667,
      "accuracy_std_dev": 0.02999999999999997,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 1.2879983000000002,
      "avg_duration_per_task_ms": 310547.73,
      "avg_output_tokens_per_task": 22913.4,
      "last_run_timestamp": "2026-10-01T16:00:34.660Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "Venice",
      "accuracy": 0.7366668,
      "accuracy_std_dev": 0.022900761755015906,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.063241744,
      "avg_duration_per_task_ms": 1530537.112,
      "avg_output_tokens_per_task": 25222.716,
      "last_run_timestamp": "2026-10-03T08:06:43.536Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Fireworks",
      "accuracy": 0.7366665,
      "accuracy_std_dev": 0.01666650000000003,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.17216302,
      "avg_duration_per_task_ms": 234612.69,
      "avg_output_tokens_per_task": 17080.8,
      "last_run_timestamp": "2026-09-19T08:12:02.180Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Phala",
      "accuracy": 0.7366665,
      "accuracy_std_dev": 0.0033334999999999892,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.020005850000000002,
      "avg_duration_per_task_ms": 329585.74,
      "avg_output_tokens_per_task": 11689.47,
      "last_run_timestamp": "2026-08-29T16:43:42.382Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "reflection/beam-1.0-501b-a23b-20261016",
      "model_slug": "reflection/beam-1.0-501b-a23b",
      "model_name": "Reflection: Beam 1.0 501B A23B",
      "provider_name": "Io Net",
      "accuracy": 0.7366665,
      "accuracy_std_dev": 0.0033334999999999892,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.02539905,
      "avg_duration_per_task_ms": 216631.23,
      "avg_output_tokens_per_task": 14336.27,
      "last_run_timestamp": "2026-10-24T12:02:36.303Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7365772046979866,
      "accuracy_std_dev": 0.061530610928569984,
      "run_count": 12,
      "total_questions": 596,
      "avg_cost_per_task_usd": 0.15594882046979863,
      "avg_duration_per_task_ms": 1382559.5033557047,
      "avg_output_tokens_per_task": 28131.454697986577,
      "last_run_timestamp": "2026-10-24T20:45:53.033Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Makora",
      "accuracy": 0.7365591209677419,
      "accuracy_std_dev": 0.016931426040354632,
      "run_count": 5,
      "total_questions": 248,
      "avg_cost_per_task_usd": 0.23607216935483874,
      "avg_duration_per_task_ms": 378114.625,
      "avg_output_tokens_per_task": 11505.427419354839,
      "last_run_timestamp": "2026-10-24T11:40:39.669Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Baidu",
      "accuracy": 0.7362963333333333,
      "accuracy_std_dev": 0.01605880825106412,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.07133713555555556,
      "avg_duration_per_task_ms": 247650.13333333333,
      "avg_output_tokens_per_task": 13143.48,
      "last_run_timestamp": "2026-10-24T19:41:01.989Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7361904285714286,
      "accuracy_std_dev": 0.035964376429450086,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.024463299999999997,
      "avg_duration_per_task_ms": 153502.6,
      "avg_output_tokens_per_task": 16394.091428571428,
      "last_run_timestamp": "2026-10-24T11:47:20.359Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Minimax",
      "accuracy": 0.7361111666666666,
      "accuracy_std_dev": 0.029994961512542214,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.02623104,
      "avg_duration_per_task_ms": 96319.25333333333,
      "avg_output_tokens_per_task": 9025.273333333333,
      "last_run_timestamp": "2026-10-24T19:08:17.063Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Novita",
      "accuracy": 0.7361111666666666,
      "accuracy_std_dev": 0.03656122127754426,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.025397581666666662,
      "avg_duration_per_task_ms": 117939.91666666667,
      "avg_output_tokens_per_task": 8683.921666666667,
      "last_run_timestamp": "2026-10-24T19:07:29.704Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "DeepInfra",
      "accuracy": 0.736,
      "accuracy_std_dev": 0.02908991425219403,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.035272232,
      "avg_duration_per_task_ms": 341115.712,
      "avg_output_tokens_per_task": 8366.916,
      "last_run_timestamp": "2026-08-29T16:11:10.137Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Baidu",
      "accuracy": 0.735926,
      "accuracy_std_dev": 0.03280219513318514,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.08689671777777778,
      "avg_duration_per_task_ms": 176906.36444444445,
      "avg_output_tokens_per_task": 10416.388888888889,
      "last_run_timestamp": "2026-10-24T19:53:44.150Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.5-sonnet-20250929",
      "model_slug": "anthropic/claude-sonnet-4.5",
      "model_name": "Anthropic: Claude Sonnet 4.5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7358025185185185,
      "accuracy_std_dev": 0.032868308079781576,
      "run_count": 27,
      "total_questions": 1350,
      "avg_cost_per_task_usd": 0.35590564518518514,
      "avg_duration_per_task_ms": 206939.88074074074,
      "avg_output_tokens_per_task": 10319.86,
      "last_run_timestamp": "2026-10-24T12:18:44.336Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7355555833333334,
      "accuracy_std_dev": 0.043489812913406,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.08088585999999999,
      "avg_duration_per_task_ms": 169768.16333333333,
      "avg_output_tokens_per_task": 8239.883333333333,
      "last_run_timestamp": "2026-10-24T19:35:51.798Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "auto-routing",
      "accuracy": 0.7355555,
      "accuracy_std_dev": 0.026293678535280442,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.07820470500000001,
      "avg_duration_per_task_ms": 167526.81,
      "avg_output_tokens_per_task": 8785.875,
      "last_run_timestamp": "2026-10-24T19:32:01.303Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "GMICloud",
      "accuracy": 0.7354165,
      "accuracy_std_dev": 0.0429125354005097,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.006874592499999999,
      "avg_duration_per_task_ms": 421319.53,
      "avg_output_tokens_per_task": 14336.97,
      "last_run_timestamp": "2026-10-10T15:31:23.439Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Phala",
      "accuracy": 0.7351191205357143,
      "accuracy_std_dev": 0.049992352433912764,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.30342583258928574,
      "avg_duration_per_task_ms": 387074.1026785714,
      "avg_output_tokens_per_task": 11534.020089285714,
      "last_run_timestamp": "2026-10-24T13:33:47.625Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.735,
      "accuracy_std_dev": 0.026822730826670128,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.012519140000000002,
      "avg_duration_per_task_ms": 95304.945,
      "avg_output_tokens_per_task": 10824.09,
      "last_run_timestamp": "2026-10-24T11:50:10.636Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.5-flash-lite-20260721",
      "model_slug": "google/gemini-3.5-flash-lite",
      "model_name": "Google: Gemini 3.5 Flash Lite",
      "provider_name": "Google AI Studio",
      "accuracy": 0.7349997500000001,
      "accuracy_std_dev": 0.002886895683515417,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.12295462999999998,
      "avg_duration_per_task_ms": 89986.35,
      "avg_output_tokens_per_task": 23350.75,
      "last_run_timestamp": "2026-10-01T19:05:46.177Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "AkashML",
      "accuracy": 0.7349570573065901,
      "accuracy_std_dev": 0.02405249324595975,
      "run_count": 7,
      "total_questions": 349,
      "avg_cost_per_task_usd": 0.067489111747851,
      "avg_duration_per_task_ms": 124543.75071633238,
      "avg_output_tokens_per_task": 4309.329512893983,
      "last_run_timestamp": "2026-10-24T12:40:32.142Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "Groq",
      "accuracy": 0.7349569369627508,
      "accuracy_std_dev": 0.02758952341386278,
      "run_count": 7,
      "total_questions": 349,
      "avg_cost_per_task_usd": 0.13143638108882524,
      "avg_duration_per_task_ms": 67067.49856733525,
      "avg_output_tokens_per_task": 9819,
      "last_run_timestamp": "2026-10-24T18:49:16.415Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "Novita",
      "accuracy": 0.7348148888888889,
      "accuracy_std_dev": 0.026439290919406052,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.05043111333333333,
      "avg_duration_per_task_ms": 363684.43555555557,
      "avg_output_tokens_per_task": 10090.468888888889,
      "last_run_timestamp": "2026-10-24T20:19:38.537Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Inceptron",
      "accuracy": 0.7348008584905661,
      "accuracy_std_dev": 0.031872568589016145,
      "run_count": 13,
      "total_questions": 636,
      "avg_cost_per_task_usd": 0.06298322641509434,
      "avg_duration_per_task_ms": 126873.33333333333,
      "avg_output_tokens_per_task": 8176.9606918239,
      "last_run_timestamp": "2026-10-24T19:13:30.742Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.8-flash-20260902",
      "model_slug": "google/gemini-3.8-flash",
      "model_name": "Google: Gemini 3.8 Flash",
      "provider_name": "Google AI Studio",
      "accuracy": 0.734694,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 49,
      "avg_cost_per_task_usd": 0.25339910204081634,
      "avg_duration_per_task_ms": 146531.16326530612,
      "avg_output_tokens_per_task": 29831.102040816328,
      "last_run_timestamp": "2026-10-01T08:03:06.861Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Relace",
      "accuracy": 0.73458325,
      "accuracy_std_dev": 0.027785801480927293,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0151816525,
      "avg_duration_per_task_ms": 268037.2675,
      "avg_output_tokens_per_task": 21770.385,
      "last_run_timestamp": "2026-10-24T12:56:13.097Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.5-sonnet-20250929",
      "model_slug": "anthropic/claude-sonnet-4.5",
      "model_name": "Anthropic: Claude Sonnet 4.5",
      "provider_name": "Google",
      "accuracy": 0.7344444166666667,
      "accuracy_std_dev": 0.025361547399420575,
      "run_count": 24,
      "total_questions": 1200,
      "avg_cost_per_task_usd": 0.3672292141666667,
      "avg_duration_per_task_ms": 228755.93166666667,
      "avg_output_tokens_per_task": 10889.619166666667,
      "last_run_timestamp": "2026-10-24T12:16:43.052Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "Azure",
      "accuracy": 0.7341665,
      "accuracy_std_dev": 0.02066601477668108,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.012598340000000001,
      "avg_duration_per_task_ms": 74617.86,
      "avg_output_tokens_per_task": 10550.06,
      "last_run_timestamp": "2026-10-24T11:45:42.778Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Morph",
      "accuracy": 0.7340742222222222,
      "accuracy_std_dev": 0.027296881245453555,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.017404215555555554,
      "avg_duration_per_task_ms": 575071.7644444444,
      "avg_output_tokens_per_task": 18088.03111111111,
      "last_run_timestamp": "2026-10-24T11:35:52.428Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "auto-routing",
      "accuracy": 0.7340738888888889,
      "accuracy_std_dev": 0.03993117810668319,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.09019846222222223,
      "avg_duration_per_task_ms": 227703.9688888889,
      "avg_output_tokens_per_task": 10092.924444444445,
      "last_run_timestamp": "2026-10-24T20:31:27.967Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "DeepInfra",
      "accuracy": 0.7338888333333333,
      "accuracy_std_dev": 0.030937109644226465,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.03372703333333333,
      "avg_duration_per_task_ms": 429862.66333333333,
      "avg_output_tokens_per_task": 9026.426666666666,
      "last_run_timestamp": "2026-10-24T20:40:57.027Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "Novita",
      "accuracy": 0.733796,
      "accuracy_std_dev": 0.0059017228557995535,
      "run_count": 3,
      "total_questions": 144,
      "avg_cost_per_task_usd": 0.20467558333333333,
      "avg_duration_per_task_ms": 370954.7361111111,
      "avg_output_tokens_per_task": 23212.9375,
      "last_run_timestamp": "2026-10-24T19:18:46.938Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "DeepInfra",
      "accuracy": 0.7334173324937028,
      "accuracy_std_dev": 0.036811483722601776,
      "run_count": 8,
      "total_questions": 397,
      "avg_cost_per_task_usd": 0.08725528211586901,
      "avg_duration_per_task_ms": 292129.2695214106,
      "avg_output_tokens_per_task": 14240.488664987406,
      "last_run_timestamp": "2026-09-26T08:12:00.292Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.5-flash-20260519",
      "model_slug": "google/gemini-3.5-flash",
      "model_name": "Google: Gemini 3.5 Flash",
      "provider_name": "Google AI Studio",
      "accuracy": 0.7333336666666668,
      "accuracy_std_dev": 0.03771236166328251,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.44470873333333333,
      "avg_duration_per_task_ms": 130557.96,
      "avg_output_tokens_per_task": 20380.893333333333,
      "last_run_timestamp": "2026-10-01T19:04:52.658Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Ionstream",
      "accuracy": 0.7333335,
      "accuracy_std_dev": 0.033333500000000016,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.018280730000000002,
      "avg_duration_per_task_ms": 88642.31,
      "avg_output_tokens_per_task": 11899.17,
      "last_run_timestamp": "2026-10-24T11:54:38.633Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Phala",
      "accuracy": 0.7333335,
      "accuracy_std_dev": 0.013333499999999998,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.08394800999999999,
      "avg_duration_per_task_ms": 134148.76,
      "avg_output_tokens_per_task": 8723.94,
      "last_run_timestamp": "2026-10-17T19:55:39.408Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "Together",
      "accuracy": 0.7333335,
      "accuracy_std_dev": 0.013333499999999998,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.047618709999999995,
      "avg_duration_per_task_ms": 115029.78,
      "avg_output_tokens_per_task": 15522.32,
      "last_run_timestamp": "2026-10-24T12:31:48.669Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Parasail",
      "accuracy": 0.733333375,
      "accuracy_std_dev": 0.03507936636378106,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.07472340000000001,
      "avg_duration_per_task_ms": 205272.835,
      "avg_output_tokens_per_task": 11169.0925,
      "last_run_timestamp": "2026-10-24T19:35:24.250Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Decart",
      "accuracy": 0.7333333333333333,
      "accuracy_std_dev": 0.005443174461857933,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.008480786666666665,
      "avg_duration_per_task_ms": 397846.0266666667,
      "avg_output_tokens_per_task": 19127.326666666668,
      "last_run_timestamp": "2026-08-22T08:38:00.859Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "InferenceNet",
      "accuracy": 0.7333333333333333,
      "accuracy_std_dev": 0.05760637824832321,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.04898614,
      "avg_duration_per_task_ms": 106829.63333333333,
      "avg_output_tokens_per_task": 9897.326666666666,
      "last_run_timestamp": "2026-10-10T11:51:48.927Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "CoreWeave",
      "accuracy": 0.73333325,
      "accuracy_std_dev": 0.02357013176007932,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.04045966999999999,
      "avg_duration_per_task_ms": 165902.775,
      "avg_output_tokens_per_task": 35276.36,
      "last_run_timestamp": "2026-10-24T12:26:22.907Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Together",
      "accuracy": 0.73333325,
      "accuracy_std_dev": 0.03377488109064932,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.30121887833333333,
      "avg_duration_per_task_ms": 325660.4733333333,
      "avg_output_tokens_per_task": 9348.415,
      "last_run_timestamp": "2026-10-24T13:17:19.762Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-sol-20260709",
      "model_slug": "openai/gpt-5.6-sol",
      "model_name": "OpenAI: GPT-5.6 Sol",
      "provider_name": "OpenAI",
      "accuracy": 0.73333325,
      "accuracy_std_dev": 0.012472414008021857,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.17039939499999998,
      "avg_duration_per_task_ms": 114141.445,
      "avg_output_tokens_per_task": 5083.04,
      "last_run_timestamp": "2026-10-02T23:08:37.533Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "NextBit",
      "accuracy": 0.7333330000000001,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01186947,
      "avg_duration_per_task_ms": 176514.74,
      "avg_output_tokens_per_task": 14201.61,
      "last_run_timestamp": "2026-08-29T16:50:07.898Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7333330000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.058005,
      "avg_duration_per_task_ms": 191243.62,
      "avg_output_tokens_per_task": 30812.32,
      "last_run_timestamp": "2026-09-12T08:09:12.993Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "auto-routing",
      "accuracy": 0.7333330000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.01490002,
      "avg_duration_per_task_ms": 87918.52,
      "avg_output_tokens_per_task": 12302.3,
      "last_run_timestamp": "2026-10-24T11:33:56.850Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "auto-routing",
      "accuracy": 0.7333330000000001,
      "accuracy_std_dev": 0.020000000000000018,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.038872460000000005,
      "avg_duration_per_task_ms": 158714.4,
      "avg_output_tokens_per_task": 14542.76,
      "last_run_timestamp": "2026-10-24T13:34:38.188Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "CoreWeave",
      "accuracy": 0.7333330000000001,
      "accuracy_std_dev": 0.020000000000000018,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.02595887,
      "avg_duration_per_task_ms": 77664,
      "avg_output_tokens_per_task": 16596.59,
      "last_run_timestamp": "2026-10-24T12:17:54.586Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.2-20251211",
      "model_slug": "openai/gpt-5.2",
      "model_name": "OpenAI: GPT-5.2",
      "provider_name": "auto-routing",
      "accuracy": 0.7333330000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.2412942,
      "avg_duration_per_task_ms": 207134.84,
      "avg_output_tokens_per_task": 12538.34,
      "last_run_timestamp": "2026-08-01T08:06:12.346Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "BaseTen",
      "accuracy": 0.7332197098976109,
      "accuracy_std_dev": 0.022906695901829432,
      "run_count": 6,
      "total_questions": 293,
      "avg_cost_per_task_usd": 0.04657865870307167,
      "avg_duration_per_task_ms": 71132.01023890785,
      "avg_output_tokens_per_task": 19231.02389078498,
      "last_run_timestamp": "2026-10-24T12:19:39.574Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Inceptron",
      "accuracy": 0.7331107060133629,
      "accuracy_std_dev": 0.030637914267565063,
      "run_count": 9,
      "total_questions": 449,
      "avg_cost_per_task_usd": 0.06578883518930957,
      "avg_duration_per_task_ms": 155774.63919821827,
      "avg_output_tokens_per_task": 11358.256124721604,
      "last_run_timestamp": "2026-10-24T19:41:50.588Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "ModelRun",
      "accuracy": 0.7328571428571429,
      "accuracy_std_dev": 0.02503509736024079,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.06745117142857142,
      "avg_duration_per_task_ms": 76562.92857142857,
      "avg_output_tokens_per_task": 8496.594285714285,
      "last_run_timestamp": "2026-09-26T08:08:19.642Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Moonshot AI",
      "accuracy": 0.7327847422303474,
      "accuracy_std_dev": 0.025507937405721676,
      "run_count": 11,
      "total_questions": 547,
      "avg_cost_per_task_usd": 0.35576682998171844,
      "avg_duration_per_task_ms": 402866.0219378428,
      "avg_output_tokens_per_task": 9582.784277879342,
      "last_run_timestamp": "2026-10-24T12:29:09.762Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "OpenInference",
      "accuracy": 0.7325,
      "accuracy_std_dev": 0.04596937990662913,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.012367795,
      "avg_duration_per_task_ms": 953069.875,
      "avg_output_tokens_per_task": 22737,
      "last_run_timestamp": "2026-10-17T10:32:07.729Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Nebius",
      "accuracy": 0.7325,
      "accuracy_std_dev": 0.029474565306379014,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.08144697000000001,
      "avg_duration_per_task_ms": 228561.56,
      "avg_output_tokens_per_task": 29757.14,
      "last_run_timestamp": "2026-10-24T12:23:33.838Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "Venice",
      "accuracy": 0.7318182727272727,
      "accuracy_std_dev": 0.040982052557610026,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.08459119272727272,
      "avg_duration_per_task_ms": 333975.78909090906,
      "avg_output_tokens_per_task": 9334.785454545454,
      "last_run_timestamp": "2026-10-24T19:37:38.817Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.4-20260305",
      "model_slug": "openai/gpt-5.4",
      "model_name": "OpenAI: GPT-5.4",
      "provider_name": "OpenAI",
      "accuracy": 0.73166675,
      "accuracy_std_dev": 0.015184321229067165,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.29879598,
      "avg_duration_per_task_ms": 177541.565,
      "avg_output_tokens_per_task": 10870.785,
      "last_run_timestamp": "2026-10-02T23:13:07.333Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Sail Research",
      "accuracy": 0.7315151818181818,
      "accuracy_std_dev": 0.012091042761393707,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.23347585818181818,
      "avg_duration_per_task_ms": 257016.06181818183,
      "avg_output_tokens_per_task": 10120.247272727273,
      "last_run_timestamp": "2026-10-24T12:33:16.806Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Fireworks",
      "accuracy": 0.7312639731543625,
      "accuracy_std_dev": 0.031328778264981705,
      "run_count": 12,
      "total_questions": 596,
      "avg_cost_per_task_usd": 0.10314290604026845,
      "avg_duration_per_task_ms": 101786.30704697987,
      "avg_output_tokens_per_task": 7660.4093959731545,
      "last_run_timestamp": "2026-09-26T08:26:47.303Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Reka",
      "accuracy": 0.7311113333333332,
      "accuracy_std_dev": 0.0332591470759883,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.03472362,
      "avg_duration_per_task_ms": 267399.06,
      "avg_output_tokens_per_task": 33925.37333333334,
      "last_run_timestamp": "2026-10-23T07:22:37.672Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-terra-20260709",
      "model_slug": "openai/gpt-5.6-terra",
      "model_name": "OpenAI: GPT-5.6 Terra",
      "provider_name": "auto-routing",
      "accuracy": 0.7311113333333332,
      "accuracy_std_dev": 0.011331263428624758,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.10536965333333334,
      "avg_duration_per_task_ms": 88766.42666666667,
      "avg_output_tokens_per_task": 4017.0533333333333,
      "last_run_timestamp": "2026-10-01T19:04:40.360Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.5-flash-20260519",
      "model_slug": "google/gemini-3.5-flash",
      "model_name": "Google: Gemini 3.5 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.7311110000000001,
      "accuracy_std_dev": 0.022662199996175744,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.3988358066666667,
      "avg_duration_per_task_ms": 172929.82666666666,
      "avg_output_tokens_per_task": 27148.5,
      "last_run_timestamp": "2026-10-01T19:04:40.362Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Modal",
      "accuracy": 0.7309522857142857,
      "accuracy_std_dev": 0.01941313112092864,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.04753026285714287,
      "avg_duration_per_task_ms": 161306.4342857143,
      "avg_output_tokens_per_task": 28780.68857142857,
      "last_run_timestamp": "2026-10-24T12:41:08.949Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7306985367647059,
      "accuracy_std_dev": 0.04536902430892472,
      "run_count": 11,
      "total_questions": 544,
      "avg_cost_per_task_usd": 0.025314847426470583,
      "avg_duration_per_task_ms": 110991.27205882352,
      "avg_output_tokens_per_task": 8956.389705882353,
      "last_run_timestamp": "2026-10-24T19:18:47.147Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta/muse-spark-open-20261015",
      "model_slug": "meta/muse-spark-open",
      "model_name": "Meta: Muse Spark Open",
      "provider_name": "Nebius",
      "accuracy": 0.7306399292929292,
      "accuracy_std_dev": 0.010748499999999994,
      "run_count": 2,
      "total_questions": 99,
      "avg_cost_per_task_usd": 0.08618277777777777,
      "avg_duration_per_task_ms": 204214.55555555556,
      "avg_output_tokens_per_task": 8369.424242424242,
      "last_run_timestamp": "2026-10-24T12:28:29.312Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "Z.AI",
      "accuracy": 0.7304762857142857,
      "accuracy_std_dev": 0.028026480838736798,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.04970431714285714,
      "avg_duration_per_task_ms": 453038.12,
      "avg_output_tokens_per_task": 9603.891428571429,
      "last_run_timestamp": "2026-10-24T20:04:11.026Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "Z.AI",
      "accuracy": 0.7303704444444444,
      "accuracy_std_dev": 0.039013983635817075,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.09884042444444445,
      "avg_duration_per_task_ms": 220383.57555555555,
      "avg_output_tokens_per_task": 11667.057777777778,
      "last_run_timestamp": "2026-10-24T20:41:03.023Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "GMICloud",
      "accuracy": 0.7303704444444444,
      "accuracy_std_dev": 0.02882252301050966,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.05139691333333334,
      "avg_duration_per_task_ms": 265789.9711111111,
      "avg_output_tokens_per_task": 10843.32,
      "last_run_timestamp": "2026-10-24T19:37:52.115Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Alibaba",
      "accuracy": 0.7303473382084096,
      "accuracy_std_dev": 0.033471576963840254,
      "run_count": 11,
      "total_questions": 547,
      "avg_cost_per_task_usd": 0.07826806032906765,
      "avg_duration_per_task_ms": 206680.5575868373,
      "avg_output_tokens_per_task": 8422.92870201097,
      "last_run_timestamp": "2026-10-24T19:29:28.512Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Crusoe",
      "accuracy": 0.7300001666666667,
      "accuracy_std_dev": 0.03605552816336059,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.11662036,
      "avg_duration_per_task_ms": 82058.93666666666,
      "avg_output_tokens_per_task": 9160.97,
      "last_run_timestamp": "2026-09-12T08:12:28.966Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Crusoe",
      "accuracy": 0.73,
      "accuracy_std_dev": 0.04333300000000001,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.023320379999999998,
      "avg_duration_per_task_ms": 139023.54,
      "avg_output_tokens_per_task": 12073.4,
      "last_run_timestamp": "2026-10-24T12:04:17.803Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Novita",
      "accuracy": 0.73,
      "accuracy_std_dev": 0.023332999999999993,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.0186985,
      "avg_duration_per_task_ms": 114277.39,
      "avg_output_tokens_per_task": 11475.52,
      "last_run_timestamp": "2026-10-24T11:36:27.069Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "GMICloud",
      "accuracy": 0.73,
      "accuracy_std_dev": 0.010000000000000009,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.014710339999999999,
      "avg_duration_per_task_ms": 195605.73,
      "avg_output_tokens_per_task": 13552.09,
      "last_run_timestamp": "2026-10-24T11:46:43.066Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "SiliconFlow",
      "accuracy": 0.73,
      "accuracy_std_dev": 0.02860989210046065,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.08111631999999999,
      "avg_duration_per_task_ms": 225903.98,
      "avg_output_tokens_per_task": 8790.808333333332,
      "last_run_timestamp": "2026-10-24T19:45:17.304Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Mistral",
      "accuracy": 0.73,
      "accuracy_std_dev": 0.04333300000000001,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01638479,
      "avg_duration_per_task_ms": 69570.4,
      "avg_output_tokens_per_task": 4309.49,
      "last_run_timestamp": "2026-10-24T12:32:51.108Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Cohere",
      "accuracy": 0.73,
      "accuracy_std_dev": 0.010000000000000009,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.02344918,
      "avg_duration_per_task_ms": 35583.1,
      "avg_output_tokens_per_task": 4177.41,
      "last_run_timestamp": "2026-10-24T11:37:33.416Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.5-sonnet-20250929",
      "model_slug": "anthropic/claude-sonnet-4.5",
      "model_name": "Anthropic: Claude Sonnet 4.5",
      "provider_name": "auto-routing",
      "accuracy": 0.7294444166666667,
      "accuracy_std_dev": 0.03271169279849418,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.33465930833333335,
      "avg_duration_per_task_ms": 207392.885,
      "avg_output_tokens_per_task": 10389.821666666667,
      "last_run_timestamp": "2026-10-24T12:22:38.802Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "auto-routing",
      "accuracy": 0.7290908181818182,
      "accuracy_std_dev": 0.028606911904192237,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.2888466727272727,
      "avg_duration_per_task_ms": 345375.0236363636,
      "avg_output_tokens_per_task": 10075.607272727273,
      "last_run_timestamp": "2026-10-24T12:45:08.706Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Venice",
      "accuracy": 0.7285714285714285,
      "accuracy_std_dev": 0.026598516660344282,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.04723099714285714,
      "avg_duration_per_task_ms": 142250.88571428572,
      "avg_output_tokens_per_task": 24794.32857142857,
      "last_run_timestamp": "2026-10-24T13:07:38.963Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Together",
      "accuracy": 0.7285712857142858,
      "accuracy_std_dev": 0.024871537523352083,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.03253488857142857,
      "avg_duration_per_task_ms": 92457.76285714285,
      "avg_output_tokens_per_task": 23099.80285714286,
      "last_run_timestamp": "2026-10-24T11:48:04.395Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "Novita",
      "accuracy": 0.7284945564516129,
      "accuracy_std_dev": 0.025519129645032956,
      "run_count": 5,
      "total_questions": 248,
      "avg_cost_per_task_usd": 0.06162772983870967,
      "avg_duration_per_task_ms": 509436.8830645161,
      "avg_output_tokens_per_task": 14294.16935483871,
      "last_run_timestamp": "2026-08-30T00:01:09.483Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.5-sonnet-20250929",
      "model_slug": "anthropic/claude-sonnet-4.5",
      "model_name": "Anthropic: Claude Sonnet 4.5",
      "provider_name": "Azure",
      "accuracy": 0.72833325,
      "accuracy_std_dev": 0.029829774773216668,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.35398706500000005,
      "avg_duration_per_task_ms": 238402.66166666665,
      "avg_output_tokens_per_task": 10886.34,
      "last_run_timestamp": "2026-10-24T11:45:41.975Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Chutes",
      "accuracy": 0.72833325,
      "accuracy_std_dev": 0.03443989530961877,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.09767092499999999,
      "avg_duration_per_task_ms": 823173.835,
      "avg_output_tokens_per_task": 9664.385,
      "last_run_timestamp": "2026-10-24T20:25:57.765Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta/muse-spark-open-20261015",
      "model_slug": "meta/muse-spark-open",
      "model_name": "Meta: Muse Spark Open",
      "provider_name": "Makora",
      "accuracy": 0.7276631288659794,
      "accuracy_std_dev": 0.019224116824642454,
      "run_count": 4,
      "total_questions": 194,
      "avg_cost_per_task_usd": 0.07521926804123712,
      "avg_duration_per_task_ms": 73859.24742268042,
      "avg_output_tokens_per_task": 8316.994845360825,
      "last_run_timestamp": "2026-10-24T11:42:23.362Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "StreamLake",
      "accuracy": 0.7275,
      "accuracy_std_dev": 0.022220636056377873,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.06640188,
      "avg_duration_per_task_ms": 199646.51,
      "avg_output_tokens_per_task": 9529.36,
      "last_run_timestamp": "2026-10-24T19:34:44.643Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7272047264573992,
      "accuracy_std_dev": 0.08169417811313268,
      "run_count": 9,
      "total_questions": 446,
      "avg_cost_per_task_usd": 0.009946286995515695,
      "avg_duration_per_task_ms": 180161.19058295965,
      "avg_output_tokens_per_task": 4234.017937219731,
      "last_run_timestamp": "2026-10-24T12:25:45.296Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Fireworks",
      "accuracy": 0.726667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.21341358,
      "avg_duration_per_task_ms": 101630.84,
      "avg_output_tokens_per_task": 8827.48,
      "last_run_timestamp": "2026-08-26T21:00:03.913Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.2-20251211",
      "model_slug": "openai/gpt-5.2",
      "model_name": "OpenAI: GPT-5.2",
      "provider_name": "OpenAI",
      "accuracy": 0.726667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.27023132,
      "avg_duration_per_task_ms": 228964.68,
      "avg_output_tokens_per_task": 14058.24,
      "last_run_timestamp": "2026-08-01T08:05:59.194Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-astra-pro-20260903",
      "model_slug": "openai/gpt-6-astra-pro",
      "model_name": "OpenAI: GPT-6 Astra Pro",
      "provider_name": "auto-routing",
      "accuracy": 0.726667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 2.91364046,
      "avg_duration_per_task_ms": 331183.04,
      "avg_output_tokens_per_task": 12422.4,
      "last_run_timestamp": "2026-10-01T08:02:52.271Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6.1-sol-20260929",
      "model_slug": "openai/gpt-6.1-sol",
      "model_name": "OpenAI: GPT-6.1 Sol",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.726667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.11300302000000001,
      "avg_duration_per_task_ms": 353321.78,
      "avg_output_tokens_per_task": 3658.14,
      "last_run_timestamp": "2026-10-03T00:00:06.432Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.7-plus-20260602",
      "model_slug": "qwen/qwen3.7-plus",
      "model_name": "Qwen: Qwen3.7 Plus",
      "provider_name": "Alibaba",
      "accuracy": 0.726667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.05448868,
      "avg_duration_per_task_ms": 350679.6,
      "avg_output_tokens_per_task": 17629,
      "last_run_timestamp": "2026-10-01T08:04:38.753Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "auto-routing",
      "accuracy": 0.7266666666666667,
      "accuracy_std_dev": 0.047453712340436424,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.011438460000000001,
      "avg_duration_per_task_ms": 86937.99333333333,
      "avg_output_tokens_per_task": 10682.4,
      "last_run_timestamp": "2026-10-24T11:51:31.389Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Nebius",
      "accuracy": 0.7266666666666667,
      "accuracy_std_dev": 0.010886757164351354,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.02419578,
      "avg_duration_per_task_ms": 133541.80666666667,
      "avg_output_tokens_per_task": 11460.326666666666,
      "last_run_timestamp": "2026-10-24T12:24:12.065Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Phala",
      "accuracy": 0.7266665,
      "accuracy_std_dev": 0.00666650000000002,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01900955,
      "avg_duration_per_task_ms": 214869.47,
      "avg_output_tokens_per_task": 11727.86,
      "last_run_timestamp": "2026-10-24T12:08:01.610Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Modal",
      "accuracy": 0.7266665,
      "accuracy_std_dev": 0.026666500000000037,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.018476000000000003,
      "avg_duration_per_task_ms": 104934.72,
      "avg_output_tokens_per_task": 10917.83,
      "last_run_timestamp": "2026-10-24T11:48:43.775Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.4-20260305",
      "model_slug": "openai/gpt-5.4",
      "model_name": "OpenAI: GPT-5.4",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7266665,
      "accuracy_std_dev": 0.03162256578378802,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.325251515,
      "avg_duration_per_task_ms": 113433.705,
      "avg_output_tokens_per_task": 10925.445,
      "last_run_timestamp": "2026-10-01T19:04:58.091Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Nebius",
      "accuracy": 0.7263158421052631,
      "accuracy_std_dev": 0.007406999999999997,
      "run_count": 2,
      "total_questions": 95,
      "avg_cost_per_task_usd": 0.013109768421052631,
      "avg_duration_per_task_ms": 155953.90526315788,
      "avg_output_tokens_per_task": 8164.357894736842,
      "last_run_timestamp": "2026-10-24T13:07:13.980Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "auto-routing",
      "accuracy": 0.7261505656565657,
      "accuracy_std_dev": 0.07912140749203615,
      "run_count": 12,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.04286990067340068,
      "avg_duration_per_task_ms": 271792.98484848486,
      "avg_output_tokens_per_task": 24134.688552188552,
      "last_run_timestamp": "2026-10-24T19:21:10.232Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Venice",
      "accuracy": 0.7261110833333333,
      "accuracy_std_dev": 0.02574195007977683,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.08434359333333334,
      "avg_duration_per_task_ms": 650589.2116666667,
      "avg_output_tokens_per_task": 9213.296666666667,
      "last_run_timestamp": "2026-10-24T19:18:30.102Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "auto-routing",
      "accuracy": 0.7260606363636364,
      "accuracy_std_dev": 0.037546404958694955,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.009626858181818184,
      "avg_duration_per_task_ms": 366334.8836363636,
      "avg_output_tokens_per_task": 14262.687272727273,
      "last_run_timestamp": "2026-10-17T13:13:33.362Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "DeepInfra",
      "accuracy": 0.7259886915254237,
      "accuracy_std_dev": 0.03421740126220705,
      "run_count": 6,
      "total_questions": 295,
      "avg_cost_per_task_usd": 0.01908770508474576,
      "avg_duration_per_task_ms": 310174.90847457625,
      "avg_output_tokens_per_task": 6632.271186440678,
      "last_run_timestamp": "2026-10-09T02:01:31.527Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Novita",
      "accuracy": 0.7259258888888889,
      "accuracy_std_dev": 0.02618905414457831,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.12695806222222222,
      "avg_duration_per_task_ms": 251909.17111111112,
      "avg_output_tokens_per_task": 10238.07111111111,
      "last_run_timestamp": "2026-10-24T22:56:39.524Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Cloudflare",
      "accuracy": 0.7256409230769231,
      "accuracy_std_dev": 0.024647466274717048,
      "run_count": 13,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.07219294769230769,
      "avg_duration_per_task_ms": 219921.15384615384,
      "avg_output_tokens_per_task": 8518.82,
      "last_run_timestamp": "2026-10-24T19:52:00.596Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "Novita",
      "accuracy": 0.7251852222222223,
      "accuracy_std_dev": 0.024500620077313125,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.05734841777777778,
      "avg_duration_per_task_ms": 328949.4666666667,
      "avg_output_tokens_per_task": 9955.037777777778,
      "last_run_timestamp": "2026-10-24T20:09:59.536Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Chutes",
      "accuracy": 0.7251850000000001,
      "accuracy_std_dev": 0.03496421058739923,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.05816458666666668,
      "avg_duration_per_task_ms": 362209.84222222224,
      "avg_output_tokens_per_task": 11818.435555555556,
      "last_run_timestamp": "2026-10-24T20:11:31.655Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Fireworks",
      "accuracy": 0.7250074958979034,
      "accuracy_std_dev": 0.02896808645605747,
      "run_count": 22,
      "total_questions": 1097,
      "avg_cost_per_task_usd": 0.4289184184138559,
      "avg_duration_per_task_ms": 233970.86873290793,
      "avg_output_tokens_per_task": 10583.862351868733,
      "last_run_timestamp": "2026-10-24T12:16:44.774Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7249999166666666,
      "accuracy_std_dev": 0.035603221088965,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.08727709499999999,
      "avg_duration_per_task_ms": 139180.04,
      "avg_output_tokens_per_task": 9126.01,
      "last_run_timestamp": "2026-10-24T18:58:26.138Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.1-terminus",
      "model_slug": "deepseek/deepseek-v3.1-terminus",
      "model_name": "DeepSeek: DeepSeek V3.1 Terminus",
      "provider_name": "Novita",
      "accuracy": 0.7244622983870967,
      "accuracy_std_dev": 0.027938338852551706,
      "run_count": 5,
      "total_questions": 248,
      "avg_cost_per_task_usd": 0.061894955645161294,
      "avg_duration_per_task_ms": 511647.25,
      "avg_output_tokens_per_task": 14459.54435483871,
      "last_run_timestamp": "2026-08-29T16:33:37.924Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Parasail",
      "accuracy": 0.7242855714285714,
      "accuracy_std_dev": 0.024734242792748124,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.04387056285714286,
      "avg_duration_per_task_ms": 165104.78285714285,
      "avg_output_tokens_per_task": 29616.194285714286,
      "last_run_timestamp": "2026-10-24T12:35:19.896Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "Friendli",
      "accuracy": 0.7241666250000001,
      "accuracy_std_dev": 0.018540625440755097,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.1069103825,
      "avg_duration_per_task_ms": 322528.1575,
      "avg_output_tokens_per_task": 14528.2375,
      "last_run_timestamp": "2026-10-17T19:16:53.251Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "Novita",
      "accuracy": 0.724,
      "accuracy_std_dev": 0.031721020021430596,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.053489452,
      "avg_duration_per_task_ms": 274950.848,
      "avg_output_tokens_per_task": 8530.916,
      "last_run_timestamp": "2026-08-29T17:21:26.058Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Phala",
      "accuracy": 0.7233335,
      "accuracy_std_dev": 0.023333500000000007,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.02063872,
      "avg_duration_per_task_ms": 638594.61,
      "avg_output_tokens_per_task": 13867.61,
      "last_run_timestamp": "2026-08-29T16:33:36.025Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-luna-20260709",
      "model_slug": "openai/gpt-5.6-luna",
      "model_name": "OpenAI: GPT-5.6 Luna",
      "provider_name": "OpenAI",
      "accuracy": 0.72333325,
      "accuracy_std_dev": 0.01374358436462263,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.011429815,
      "avg_duration_per_task_ms": 101515.345,
      "avg_output_tokens_per_task": 5183.985,
      "last_run_timestamp": "2026-10-02T23:07:05.152Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Baidu",
      "accuracy": 0.7231578947368421,
      "accuracy_std_dev": 0.03181579445709768,
      "run_count": 19,
      "total_questions": 950,
      "avg_cost_per_task_usd": 0.06943197789473686,
      "avg_duration_per_task_ms": 118880.57368421053,
      "avg_output_tokens_per_task": 8790.421052631578,
      "last_run_timestamp": "2026-10-24T19:51:31.088Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "Venice",
      "accuracy": 0.722963,
      "accuracy_std_dev": 0.024567656927124495,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.2753419711111111,
      "avg_duration_per_task_ms": 183403.79555555555,
      "avg_output_tokens_per_task": 11487.52,
      "last_run_timestamp": "2026-10-24T12:33:54.727Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "auto-routing",
      "accuracy": 0.722963,
      "accuracy_std_dev": 0.015748325844002865,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.05190752666666666,
      "avg_duration_per_task_ms": 384896.1533333333,
      "avg_output_tokens_per_task": 9409.375555555556,
      "last_run_timestamp": "2026-10-24T20:09:28.564Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "Moonshot AI",
      "accuracy": 0.7226663999999999,
      "accuracy_std_dev": 0.028782797331739687,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.04632364,
      "avg_duration_per_task_ms": 161759.348,
      "avg_output_tokens_per_task": 8208.956,
      "last_run_timestamp": "2026-08-29T17:36:31.073Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Novita",
      "accuracy": 0.7224079799331103,
      "accuracy_std_dev": 0.027594435979393265,
      "run_count": 12,
      "total_questions": 598,
      "avg_cost_per_task_usd": 0.046122013377926416,
      "avg_duration_per_task_ms": 197207.33444816055,
      "avg_output_tokens_per_task": 9716.61872909699,
      "last_run_timestamp": "2026-10-24T19:18:28.440Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Alibaba",
      "accuracy": 0.7223809999999999,
      "accuracy_std_dev": 0.03270676301623258,
      "run_count": 28,
      "total_questions": 1400,
      "avg_cost_per_task_usd": 0.1190975942857143,
      "avg_duration_per_task_ms": 133109.795,
      "avg_output_tokens_per_task": 8591.442857142858,
      "last_run_timestamp": "2026-10-24T19:56:04.170Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "OpenInference",
      "accuracy": 0.7223148931552588,
      "accuracy_std_dev": 0.03960298431691171,
      "run_count": 12,
      "total_questions": 599,
      "avg_cost_per_task_usd": 0.025822385642737894,
      "avg_duration_per_task_ms": 283083.48080133554,
      "avg_output_tokens_per_task": 7842.495826377295,
      "last_run_timestamp": "2026-10-24T12:07:12.594Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Io Net",
      "accuracy": 0.7222223333333333,
      "accuracy_std_dev": 0.011330936570096713,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.018792733333333332,
      "avg_duration_per_task_ms": 301192.17333333334,
      "avg_output_tokens_per_task": 19692.3,
      "last_run_timestamp": "2026-10-01T16:00:06.404Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Reka",
      "accuracy": 0.72166675,
      "accuracy_std_dev": 0.012801898733683979,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.01389351,
      "avg_duration_per_task_ms": 148770.74,
      "avg_output_tokens_per_task": 13265.9,
      "last_run_timestamp": "2026-10-24T12:22:53.798Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Makora",
      "accuracy": 0.7210882448979592,
      "accuracy_std_dev": 0.0076780598851532815,
      "run_count": 5,
      "total_questions": 245,
      "avg_cost_per_task_usd": 0.009489710204081632,
      "avg_duration_per_task_ms": 145499.9755102041,
      "avg_output_tokens_per_task": 16260.81224489796,
      "last_run_timestamp": "2026-09-26T08:06:43.933Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Nebius",
      "accuracy": 0.7206667999999999,
      "accuracy_std_dev": 0.037882852175621616,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.29679829,
      "avg_duration_per_task_ms": 498320.864,
      "avg_output_tokens_per_task": 9522.336,
      "last_run_timestamp": "2026-10-24T22:09:08.519Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Novita",
      "accuracy": 0.7205555,
      "accuracy_std_dev": 0.020268346362000044,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.08116760333333334,
      "avg_duration_per_task_ms": 238268.15166666667,
      "avg_output_tokens_per_task": 9240.011666666667,
      "last_run_timestamp": "2026-10-24T19:37:32.748Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-sol-pro-20260709",
      "model_slug": "openai/gpt-5.6-sol-pro",
      "model_name": "OpenAI: GPT-5.6 Sol Pro",
      "provider_name": "Azure",
      "accuracy": 0.7205384141414142,
      "accuracy_std_dev": 0.051112796809483045,
      "run_count": 4,
      "total_questions": 198,
      "avg_cost_per_task_usd": 1.6761988181818182,
      "avg_duration_per_task_ms": 313695.66161616164,
      "avg_output_tokens_per_task": 23404.97474747475,
      "last_run_timestamp": "2026-10-01T16:01:13.647Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Sail Research",
      "accuracy": 0.7200002499999999,
      "accuracy_std_dev": 0.016329931619192377,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.005028675,
      "avg_duration_per_task_ms": 140463.585,
      "avg_output_tokens_per_task": 11883.17,
      "last_run_timestamp": "2026-10-24T12:43:45.098Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Io Net",
      "accuracy": 0.72,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.12671924,
      "avg_duration_per_task_ms": 566714.78,
      "avg_output_tokens_per_task": 16714.02,
      "last_run_timestamp": "2026-09-28T18:00:05.892Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.6-flash-20260721",
      "model_slug": "google/gemini-3.6-flash",
      "model_name": "Google: Gemini 3.6 Flash",
      "provider_name": "Google AI Studio",
      "accuracy": 0.72,
      "accuracy_std_dev": 0.03231762505197435,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.36211610499999997,
      "avg_duration_per_task_ms": 150104.23,
      "avg_output_tokens_per_task": 23123.4,
      "last_run_timestamp": "2026-10-01T19:04:55.721Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "inclusionai/ling-3.0-flash-vl-20260910",
      "model_slug": "inclusionai/ling-3.0-flash-vl",
      "model_name": "inclusionAI: Ling 3.0 Flash VL",
      "provider_name": "Novita",
      "accuracy": 0.72,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.00317077,
      "avg_duration_per_task_ms": 128719.23,
      "avg_output_tokens_per_task": 15118.03,
      "last_run_timestamp": "2026-10-01T08:00:27.045Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Venice",
      "accuracy": 0.72,
      "accuracy_std_dev": 0.03165460057329021,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.028238483636363637,
      "avg_duration_per_task_ms": 217753.1509090909,
      "avg_output_tokens_per_task": 9627.723636363637,
      "last_run_timestamp": "2026-10-24T19:14:23.110Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2-thinking-20251106",
      "model_slug": "moonshotai/kimi-k2-thinking",
      "model_name": "MoonshotAI: Kimi K2 Thinking",
      "provider_name": "auto-routing",
      "accuracy": 0.72,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.053717240000000006,
      "avg_duration_per_task_ms": 263357.62,
      "avg_output_tokens_per_task": 8791.84,
      "last_run_timestamp": "2026-08-01T17:38:27.943Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3-ultra-550b-a55b-20260604",
      "model_slug": "nvidia/nemotron-3-ultra-550b-a55b",
      "model_name": "NVIDIA: Nemotron 3 Ultra",
      "provider_name": "BaseTen",
      "accuracy": 0.72,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.15154258,
      "avg_duration_per_task_ms": 105755.46,
      "avg_output_tokens_per_task": 13133.32,
      "last_run_timestamp": "2026-08-04T03:12:01.476Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.7-max-20260520",
      "model_slug": "qwen/qwen3.7-max",
      "model_name": "Qwen: Qwen3.7 Max",
      "provider_name": "AtlasCloud",
      "accuracy": 0.72,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.402901,
      "avg_duration_per_task_ms": 196016.4,
      "avg_output_tokens_per_task": 16256.52,
      "last_run_timestamp": "2026-09-27T00:00:03.103Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "Parasail",
      "accuracy": 0.72,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.028463379999999996,
      "avg_duration_per_task_ms": 583244.64,
      "avg_output_tokens_per_task": 43866.54,
      "last_run_timestamp": "2026-08-08T08:27:52.152Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Venice",
      "accuracy": 0.72,
      "accuracy_std_dev": 0.05409184233782393,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.11944148833333335,
      "avg_duration_per_task_ms": 138004.31,
      "avg_output_tokens_per_task": 9585.921666666667,
      "last_run_timestamp": "2026-10-24T19:35:31.077Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7199998461538462,
      "accuracy_std_dev": 0.060595320121208315,
      "run_count": 13,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.28468824153846156,
      "avg_duration_per_task_ms": 424519.81384615385,
      "avg_output_tokens_per_task": 10050.18923076923,
      "last_run_timestamp": "2026-10-24T12:30:07.991Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.1-flash-lite-20260507",
      "model_slug": "google/gemini-3.1-flash-lite",
      "model_name": "Google: Gemini 3.1 Flash Lite",
      "provider_name": "auto-routing",
      "accuracy": 0.7199996666666667,
      "accuracy_std_dev": 0.009428090415820642,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.09322211333333333,
      "avg_duration_per_task_ms": 149908.11333333334,
      "avg_output_tokens_per_task": 49346.81333333333,
      "last_run_timestamp": "2026-10-01T19:04:52.649Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta/muse-spark-open-20261015",
      "model_slug": "meta/muse-spark-open",
      "model_name": "Meta: Muse Spark Open",
      "provider_name": "BaseTen",
      "accuracy": 0.719858,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 47,
      "avg_cost_per_task_usd": 0.0657194255319149,
      "avg_duration_per_task_ms": 49852.255319148935,
      "avg_output_tokens_per_task": 5715.191489361702,
      "last_run_timestamp": "2026-10-24T12:16:29.775Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "Alibaba",
      "accuracy": 0.7193332,
      "accuracy_std_dev": 0.03143937799893632,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.045152332,
      "avg_duration_per_task_ms": 326196.928,
      "avg_output_tokens_per_task": 13427.836,
      "last_run_timestamp": "2026-10-04T00:00:34.898Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-sol-20260709",
      "model_slug": "openai/gpt-5.6-sol",
      "model_name": "OpenAI: GPT-5.6 Sol",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7188551919191919,
      "accuracy_std_dev": 0.027525589315898408,
      "run_count": 4,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.29904142424242425,
      "avg_duration_per_task_ms": 154139.7323232323,
      "avg_output_tokens_per_task": 5036.570707070707,
      "last_run_timestamp": "2026-10-01T19:05:12.160Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Ambient",
      "accuracy": 0.7185185555555555,
      "accuracy_std_dev": 0.0497675821798155,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.07125336888888889,
      "avg_duration_per_task_ms": 245562.12,
      "avg_output_tokens_per_task": 8258.706666666667,
      "last_run_timestamp": "2026-10-24T19:39:33.038Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Venice",
      "accuracy": 0.7185183333333334,
      "accuracy_std_dev": 0.033815486586933986,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.24051653333333337,
      "avg_duration_per_task_ms": 300854.7,
      "avg_output_tokens_per_task": 21246.093333333334,
      "last_run_timestamp": "2026-10-24T12:24:27.575Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "Minimax",
      "accuracy": 0.7177777777777777,
      "accuracy_std_dev": 0.018986661986643713,
      "run_count": 18,
      "total_questions": 900,
      "avg_cost_per_task_usd": 0.03833726666666666,
      "avg_duration_per_task_ms": 195653.57888888888,
      "avg_output_tokens_per_task": 9865.436666666666,
      "last_run_timestamp": "2026-10-24T19:52:24.184Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "Venice",
      "accuracy": 0.7177776666666666,
      "accuracy_std_dev": 0.013698571612479267,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.10180220666666666,
      "avg_duration_per_task_ms": 128107.32666666666,
      "avg_output_tokens_per_task": 16602.8,
      "last_run_timestamp": "2026-10-24T12:39:45.200Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Phala",
      "accuracy": 0.7173648633879781,
      "accuracy_std_dev": 0.031638898989055704,
      "run_count": 11,
      "total_questions": 549,
      "avg_cost_per_task_usd": 0.0967067795992714,
      "avg_duration_per_task_ms": 190644.16393442624,
      "avg_output_tokens_per_task": 10605.225865209472,
      "last_run_timestamp": "2026-10-17T19:53:39.136Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Wafer",
      "accuracy": 0.7171491336302895,
      "accuracy_std_dev": 0.026956007306143996,
      "run_count": 18,
      "total_questions": 898,
      "avg_cost_per_task_usd": 0.10596277951002227,
      "avg_duration_per_task_ms": 105231.55345211581,
      "avg_output_tokens_per_task": 9573.741648106905,
      "last_run_timestamp": "2026-10-24T19:32:20.737Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-terra-20260709",
      "model_slug": "openai/gpt-5.6-terra",
      "model_name": "OpenAI: GPT-5.6 Terra",
      "provider_name": "Azure",
      "accuracy": 0.7170371111111111,
      "accuracy_std_dev": 0.016956299841680886,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.12728777111111111,
      "avg_duration_per_task_ms": 101323.44222222222,
      "avg_output_tokens_per_task": 3885.56,
      "last_run_timestamp": "2026-10-01T19:04:52.651Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "SambaNova",
      "accuracy": 0.7166669999999999,
      "accuracy_std_dev": 0.009999999999999953,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.04331858,
      "avg_duration_per_task_ms": 119875.46,
      "avg_output_tokens_per_task": 9185.81,
      "last_run_timestamp": "2026-10-24T12:22:21.794Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.1-terminus",
      "model_slug": "deepseek/deepseek-v3.1-terminus",
      "model_name": "DeepSeek: DeepSeek V3.1 Terminus",
      "provider_name": "auto-routing",
      "accuracy": 0.7166665000000001,
      "accuracy_std_dev": 0.015275361427802623,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.058710445,
      "avg_duration_per_task_ms": 517833.175,
      "avg_output_tokens_per_task": 13935.145,
      "last_run_timestamp": "2026-08-29T16:22:37.842Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "Phala",
      "accuracy": 0.716250125,
      "accuracy_std_dev": 0.028550118578727003,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.2090713325,
      "avg_duration_per_task_ms": 1406671.1525,
      "avg_output_tokens_per_task": 13817.695,
      "last_run_timestamp": "2026-10-24T19:32:17.023Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "Chutes",
      "accuracy": 0.7162160675675675,
      "accuracy_std_dev": 0.02984900053044771,
      "run_count": 3,
      "total_questions": 148,
      "avg_cost_per_task_usd": 0.1296834662162162,
      "avg_duration_per_task_ms": 769826.7162162162,
      "avg_output_tokens_per_task": 28396.02027027027,
      "last_run_timestamp": "2026-08-15T14:15:18.888Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "auto-routing",
      "accuracy": 0.7161110833333333,
      "accuracy_std_dev": 0.032025355021342926,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.04424442,
      "avg_duration_per_task_ms": 122982.75,
      "avg_output_tokens_per_task": 8577.505,
      "last_run_timestamp": "2026-10-24T19:43:58.683Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "CoreWeave",
      "accuracy": 0.7155646528925619,
      "accuracy_std_dev": 0.07922982153103719,
      "run_count": 5,
      "total_questions": 242,
      "avg_cost_per_task_usd": 0.11734726446280992,
      "avg_duration_per_task_ms": 254637.0785123967,
      "avg_output_tokens_per_task": 42006.561983471074,
      "last_run_timestamp": "2026-08-29T17:18:52.408Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "GMICloud",
      "accuracy": 0.7155554444444445,
      "accuracy_std_dev": 0.018324967789034047,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.02354926,
      "avg_duration_per_task_ms": 203685.72222222222,
      "avg_output_tokens_per_task": 9821.226666666667,
      "last_run_timestamp": "2026-10-24T19:46:12.078Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Relace",
      "accuracy": 0.7155553333333334,
      "accuracy_std_dev": 0.029979285029648195,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.13937678,
      "avg_duration_per_task_ms": 263819.3466666667,
      "avg_output_tokens_per_task": 10311.9,
      "last_run_timestamp": "2026-10-03T16:18:50.927Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.2-exp",
      "model_slug": "deepseek/deepseek-v3.2-exp",
      "model_name": "DeepSeek: DeepSeek V3.2 Exp",
      "provider_name": "auto-routing",
      "accuracy": 0.715278,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 48,
      "avg_cost_per_task_usd": 0.07260889583333334,
      "avg_duration_per_task_ms": 556749.9791666666,
      "avg_output_tokens_per_task": 12329.4375,
      "last_run_timestamp": "2026-08-01T17:38:07.305Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "Wafer",
      "accuracy": 0.715278,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 48,
      "avg_cost_per_task_usd": 0.009582875,
      "avg_duration_per_task_ms": 164124.5,
      "avg_output_tokens_per_task": 18445.333333333332,
      "last_run_timestamp": "2026-10-10T11:44:02.317Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-terra-20260709",
      "model_slug": "openai/gpt-5.6-terra",
      "model_name": "OpenAI: GPT-5.6 Terra",
      "provider_name": "OpenAI",
      "accuracy": 0.715,
      "accuracy_std_dev": 0.029202378139459823,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.09407665500000001,
      "avg_duration_per_task_ms": 83159.84,
      "avg_output_tokens_per_task": 3660.855,
      "last_run_timestamp": "2026-10-02T23:10:36.500Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "StreamLake",
      "accuracy": 0.7148716923076923,
      "accuracy_std_dev": 0.0339082212572173,
      "run_count": 13,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.03673959538461538,
      "avg_duration_per_task_ms": 186566.58,
      "avg_output_tokens_per_task": 9052.430769230768,
      "last_run_timestamp": "2026-10-24T19:49:36.340Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.1-flash-lite-20260507",
      "model_slug": "google/gemini-3.1-flash-lite",
      "model_name": "Google: Gemini 3.1 Flash Lite",
      "provider_name": "Google",
      "accuracy": 0.7148146666666667,
      "accuracy_std_dev": 0.031075936024161484,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.09082762666666666,
      "avg_duration_per_task_ms": 156907.88888888888,
      "avg_output_tokens_per_task": 48272.70222222222,
      "last_run_timestamp": "2026-10-01T19:06:49.017Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Z.AI",
      "accuracy": 0.7144445,
      "accuracy_std_dev": 0.03331469664302328,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.122428735,
      "avg_duration_per_task_ms": 164804.72666666665,
      "avg_output_tokens_per_task": 11653.74,
      "last_run_timestamp": "2026-10-24T19:52:24.144Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Phala",
      "accuracy": 0.7133782062780268,
      "accuracy_std_dev": 0.10832537981105755,
      "run_count": 9,
      "total_questions": 446,
      "avg_cost_per_task_usd": 0.20968350448430492,
      "avg_duration_per_task_ms": 1204177.4775784754,
      "avg_output_tokens_per_task": 56724.81614349776,
      "last_run_timestamp": "2026-10-10T19:16:51.412Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "CoreWeave",
      "accuracy": 0.7133333846153846,
      "accuracy_std_dev": 0.03807331209762081,
      "run_count": 13,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.08142609230769231,
      "avg_duration_per_task_ms": 97247.61076923077,
      "avg_output_tokens_per_task": 8604.487692307692,
      "last_run_timestamp": "2026-10-24T23:37:48.880Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7133333749999999,
      "accuracy_std_dev": 0.059814328458859874,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.10449861,
      "avg_duration_per_task_ms": 230117.0725,
      "avg_output_tokens_per_task": 11136.0225,
      "last_run_timestamp": "2026-10-17T19:18:15.219Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7133333333333334,
      "accuracy_std_dev": 0.14919525114109886,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.06109567666666668,
      "avg_duration_per_task_ms": 151262.36,
      "avg_output_tokens_per_task": 17284.825,
      "last_run_timestamp": "2026-10-24T18:32:59.224Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Together",
      "accuracy": 0.7133333333333334,
      "accuracy_std_dev": 0.03953422333736796,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.10266909833333333,
      "avg_duration_per_task_ms": 104721.82833333334,
      "avg_output_tokens_per_task": 9163.678333333333,
      "last_run_timestamp": "2026-10-24T19:12:36.271Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "InferenceNet",
      "accuracy": 0.71333325,
      "accuracy_std_dev": 0.016996797086142445,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.01124519,
      "avg_duration_per_task_ms": 291598.365,
      "avg_output_tokens_per_task": 13318.715,
      "last_run_timestamp": "2026-10-24T13:02:30.014Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "auto-routing",
      "accuracy": 0.7133332222222223,
      "accuracy_std_dev": 0.02222223333499722,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.06405054444444444,
      "avg_duration_per_task_ms": 308107.3911111111,
      "avg_output_tokens_per_task": 9732.942222222222,
      "last_run_timestamp": "2026-10-24T20:27:01.220Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Cloudflare",
      "accuracy": 0.713333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.05452906,
      "avg_duration_per_task_ms": 272872.18,
      "avg_output_tokens_per_task": 39719.44,
      "last_run_timestamp": "2026-10-13T20:01:29.821Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "inclusionai/ling-3.0-flash-vl-20260910",
      "model_slug": "inclusionai/ling-3.0-flash-vl",
      "model_name": "inclusionAI: Ling 3.0 Flash VL",
      "provider_name": "DeepInfra",
      "accuracy": 0.713333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.006318460000000001,
      "avg_duration_per_task_ms": 234030.92,
      "avg_output_tokens_per_task": 14773.36,
      "last_run_timestamp": "2026-10-01T08:05:57.971Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "GMICloud",
      "accuracy": 0.713333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.09868048,
      "avg_duration_per_task_ms": 318461,
      "avg_output_tokens_per_task": 11794.24,
      "last_run_timestamp": "2026-08-29T17:18:09.782Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-flash-20260826",
      "model_slug": "qwen/qwen3.8-flash",
      "model_name": "Qwen: Qwen3.8 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.713333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.01665106,
      "avg_duration_per_task_ms": 327560.6,
      "avg_output_tokens_per_task": 17227.24,
      "last_run_timestamp": "2026-10-01T08:00:41.459Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.8-flash-20260902",
      "model_slug": "google/gemini-3.8-flash",
      "model_name": "Google: Gemini 3.8 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.712766,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 47,
      "avg_cost_per_task_usd": 0.21022357446808512,
      "avg_duration_per_task_ms": 261220.25531914894,
      "avg_output_tokens_per_task": 27609.17021276596,
      "last_run_timestamp": "2026-10-01T08:02:55.088Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5-mini-2025-08-07",
      "model_slug": "openai/gpt-5-mini",
      "model_name": "OpenAI: GPT-5 Mini",
      "provider_name": "OpenAI",
      "accuracy": 0.712121101010101,
      "accuracy_std_dev": 0.029111223892169177,
      "run_count": 4,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.12218062626262625,
      "avg_duration_per_task_ms": 537818.2474747475,
      "avg_output_tokens_per_task": 55422.06060606061,
      "last_run_timestamp": "2026-10-02T23:27:09.304Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "Venice",
      "accuracy": 0.7118518888888888,
      "accuracy_std_dev": 0.032626193960083476,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.07140289111111112,
      "avg_duration_per_task_ms": 302781.4844444444,
      "avg_output_tokens_per_task": 10769.26888888889,
      "last_run_timestamp": "2026-10-24T22:01:31.925Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Decart",
      "accuracy": 0.711711777027027,
      "accuracy_std_dev": 0.11123612518334841,
      "run_count": 6,
      "total_questions": 296,
      "avg_cost_per_task_usd": 0.07281375000000001,
      "avg_duration_per_task_ms": 155901.73648648648,
      "avg_output_tokens_per_task": 15030.263513513513,
      "last_run_timestamp": "2026-10-03T16:16:30.145Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-luna-pro-20260709",
      "model_slug": "openai/gpt-5.6-luna-pro",
      "model_name": "OpenAI: GPT-5.6 Luna Pro",
      "provider_name": "OpenAI",
      "accuracy": 0.71166675,
      "accuracy_std_dev": 0.019075733568790994,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.07509325,
      "avg_duration_per_task_ms": 322731.32,
      "avg_output_tokens_per_task": 33483.435,
      "last_run_timestamp": "2026-10-02T23:12:53.530Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "auto-routing",
      "accuracy": 0.7113821300813009,
      "accuracy_std_dev": 0.04920965638205573,
      "run_count": 10,
      "total_questions": 492,
      "avg_cost_per_task_usd": 0.03321126422764228,
      "avg_duration_per_task_ms": 1038484.2195121951,
      "avg_output_tokens_per_task": 47360.833333333336,
      "last_run_timestamp": "2026-10-17T19:19:46.010Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Nebius",
      "accuracy": 0.7113332,
      "accuracy_std_dev": 0.017587783639788142,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.28712958000000005,
      "avg_duration_per_task_ms": 155529.044,
      "avg_output_tokens_per_task": 10850.46,
      "last_run_timestamp": "2026-10-24T19:05:28.397Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "DeepInfra",
      "accuracy": 0.7112820769230769,
      "accuracy_std_dev": 0.0340322401972047,
      "run_count": 13,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.04383589384615384,
      "avg_duration_per_task_ms": 144055.18,
      "avg_output_tokens_per_task": 8486.646153846154,
      "last_run_timestamp": "2026-10-24T19:47:13.308Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-luna-pro-20260709",
      "model_slug": "openai/gpt-5.6-luna-pro",
      "model_name": "OpenAI: GPT-5.6 Luna Pro",
      "provider_name": "auto-routing",
      "accuracy": 0.7111113333333333,
      "accuracy_std_dev": 0.011331263428624743,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.07186812,
      "avg_duration_per_task_ms": 348295.26,
      "avg_output_tokens_per_task": 33945.17333333333,
      "last_run_timestamp": "2026-10-01T16:00:47.532Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "Mara",
      "accuracy": 0.7108843469387754,
      "accuracy_std_dev": 0.017917000000000016,
      "run_count": 2,
      "total_questions": 98,
      "avg_cost_per_task_usd": 0.04385853061224489,
      "avg_duration_per_task_ms": 61398.17346938775,
      "avg_output_tokens_per_task": 8040.4795918367345,
      "last_run_timestamp": "2026-08-08T08:31:07.742Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "StreamLake",
      "accuracy": 0.7106667999999999,
      "accuracy_std_dev": 0.037381050458220114,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.144933328,
      "avg_duration_per_task_ms": 336674.688,
      "avg_output_tokens_per_task": 10779.192,
      "last_run_timestamp": "2026-08-30T00:01:09.490Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-luna-20260709",
      "model_slug": "openai/gpt-5.6-luna",
      "model_name": "OpenAI: GPT-5.6 Luna",
      "provider_name": "Azure",
      "accuracy": 0.7103705555555555,
      "accuracy_std_dev": 0.030809822424065814,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.03036840888888889,
      "avg_duration_per_task_ms": 105409.86666666667,
      "avg_output_tokens_per_task": 5076.06,
      "last_run_timestamp": "2026-10-01T19:05:16.550Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "CoreWeave",
      "accuracy": 0.7103703333333333,
      "accuracy_std_dev": 0.03759572984617624,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.06715483777777777,
      "avg_duration_per_task_ms": 69489.32,
      "avg_output_tokens_per_task": 9660.7,
      "last_run_timestamp": "2026-10-24T19:30:31.496Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Novita",
      "accuracy": 0.71,
      "accuracy_std_dev": 0.003332999999999975,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01784487,
      "avg_duration_per_task_ms": 239614.14,
      "avg_output_tokens_per_task": 16662.97,
      "last_run_timestamp": "2026-10-24T11:51:42.268Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-terra-20260709",
      "model_slug": "openai/gpt-5.6-terra",
      "model_name": "OpenAI: GPT-5.6 Terra",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.71,
      "accuracy_std_dev": 0.01795033271279393,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.12391315,
      "avg_duration_per_task_ms": 76197.77,
      "avg_output_tokens_per_task": 4134.025,
      "last_run_timestamp": "2026-10-01T19:04:38.304Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "OpenInference",
      "accuracy": 0.70990758186398,
      "accuracy_std_dev": 0.05927908384168378,
      "run_count": 8,
      "total_questions": 397,
      "avg_cost_per_task_usd": 0.04869609571788413,
      "avg_duration_per_task_ms": 986154.4534005037,
      "avg_output_tokens_per_task": 53874.76574307305,
      "last_run_timestamp": "2026-10-24T12:28:11.805Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7093335999999999,
      "accuracy_std_dev": 0.027840743923968677,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.049385444,
      "avg_duration_per_task_ms": 869963.78,
      "avg_output_tokens_per_task": 8710.828,
      "last_run_timestamp": "2026-08-29T16:47:05.512Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Io Net",
      "accuracy": 0.7088886666666667,
      "accuracy_std_dev": 0.01749773883169545,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.11639628666666667,
      "avg_duration_per_task_ms": 755791.72,
      "avg_output_tokens_per_task": 27174.273333333334,
      "last_run_timestamp": "2026-08-15T08:15:24.293Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "DeepInfra",
      "accuracy": 0.7079150969899666,
      "accuracy_std_dev": 0.016686286941344667,
      "run_count": 6,
      "total_questions": 299,
      "avg_cost_per_task_usd": 0.04951336454849498,
      "avg_duration_per_task_ms": 411808.29431438126,
      "avg_output_tokens_per_task": 8694.879598662208,
      "last_run_timestamp": "2026-10-24T21:22:48.118Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "inclusionai/ling-3.0-flash-fin-20260827",
      "model_slug": "inclusionai/ling-3.0-flash-fin",
      "model_name": "inclusionAI: Ling 3.0 Flash Fin",
      "provider_name": "auto-routing",
      "accuracy": 0.707483,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 49,
      "avg_cost_per_task_usd": 0.004842469387755102,
      "avg_duration_per_task_ms": 161558.55102040817,
      "avg_output_tokens_per_task": 17024.020408163266,
      "last_run_timestamp": "2026-10-01T08:01:49.153Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7074075555555556,
      "accuracy_std_dev": 0.05206310721745101,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.03110885111111111,
      "avg_duration_per_task_ms": 310847.1533333333,
      "avg_output_tokens_per_task": 10019.133333333333,
      "last_run_timestamp": "2026-10-24T19:49:45.253Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "auto-routing",
      "accuracy": 0.7074073333333334,
      "accuracy_std_dev": 0.012350780677080023,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.02628152222222222,
      "avg_duration_per_task_ms": 205434.89555555556,
      "avg_output_tokens_per_task": 10011.104444444445,
      "last_run_timestamp": "2026-10-24T19:34:31.773Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "Makora",
      "accuracy": 0.7073334,
      "accuracy_std_dev": 0.025681872553223187,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.024487520000000002,
      "avg_duration_per_task_ms": 117939.624,
      "avg_output_tokens_per_task": 30855.352,
      "last_run_timestamp": "2026-10-24T19:30:58.294Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Parasail",
      "accuracy": 0.7072222499999999,
      "accuracy_std_dev": 0.02940247900581683,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.11202355833333333,
      "avg_duration_per_task_ms": 94561.55833333333,
      "avg_output_tokens_per_task": 9737.35,
      "last_run_timestamp": "2026-10-24T19:14:21.866Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Nebius",
      "accuracy": 0.706667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.3541182,
      "avg_duration_per_task_ms": 1646927.84,
      "avg_output_tokens_per_task": 24534.74,
      "last_run_timestamp": "2026-09-10T00:00:02.948Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Cohere",
      "accuracy": 0.706667,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.07705153,
      "avg_duration_per_task_ms": 159916.36,
      "avg_output_tokens_per_task": 34851.38,
      "last_run_timestamp": "2026-10-24T12:16:31.422Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "ModelRun",
      "accuracy": 0.706667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.1838298,
      "avg_duration_per_task_ms": 329811.9,
      "avg_output_tokens_per_task": 9540.84,
      "last_run_timestamp": "2026-08-15T00:16:18.405Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.1",
      "model_slug": "minimax/minimax-m2.1",
      "model_name": "MiniMax: MiniMax M2.1",
      "provider_name": "Minimax",
      "accuracy": 0.7066669999999999,
      "accuracy_std_dev": 0.06,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.02912445,
      "avg_duration_per_task_ms": 175317.22,
      "avg_output_tokens_per_task": 10432.57,
      "last_run_timestamp": "2026-08-01T17:38:12.859Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-astra-pro-20260903",
      "model_slug": "openai/gpt-6-astra-pro",
      "model_name": "OpenAI: GPT-6 Astra Pro",
      "provider_name": "OpenAI",
      "accuracy": 0.7066669999999999,
      "accuracy_std_dev": 0.019999999999999962,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 3.04321209,
      "avg_duration_per_task_ms": 317897.29,
      "avg_output_tokens_per_task": 13272.53,
      "last_run_timestamp": "2026-10-02T23:13:16.494Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5-2025-08-07",
      "model_slug": "openai/gpt-5",
      "model_name": "OpenAI: GPT-5",
      "provider_name": "OpenAI",
      "accuracy": 0.7066666666666667,
      "accuracy_std_dev": 0.005443174461857933,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.5994524266666668,
      "avg_duration_per_task_ms": 636541.5466666666,
      "avg_output_tokens_per_task": 54354.86,
      "last_run_timestamp": "2026-10-02T23:22:41.519Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "Z.AI",
      "accuracy": 0.7066666666666667,
      "accuracy_std_dev": 0.04388527130041595,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.06779559555555556,
      "avg_duration_per_task_ms": 346332.98,
      "avg_output_tokens_per_task": 9979.706666666667,
      "last_run_timestamp": "2026-10-24T20:13:56.072Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Makora",
      "accuracy": 0.7066665000000001,
      "accuracy_std_dev": 0.027888567949071897,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.021449605000000004,
      "avg_duration_per_task_ms": 108412.625,
      "avg_output_tokens_per_task": 11247.685,
      "last_run_timestamp": "2026-10-24T11:28:47.983Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "auto-routing",
      "accuracy": 0.7060302060301508,
      "accuracy_std_dev": 0.10430171386655158,
      "run_count": 8,
      "total_questions": 398,
      "avg_cost_per_task_usd": 0.07453727386934673,
      "avg_duration_per_task_ms": 742520.2713567839,
      "avg_output_tokens_per_task": 45595.6608040201,
      "last_run_timestamp": "2026-10-24T18:56:22.719Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5-2025-08-07",
      "model_slug": "openai/gpt-5",
      "model_name": "OpenAI: GPT-5",
      "provider_name": "Azure",
      "accuracy": 0.7056452258064517,
      "accuracy_std_dev": 0.012405393094940622,
      "run_count": 5,
      "total_questions": 248,
      "avg_cost_per_task_usd": 0.6654760282258064,
      "avg_duration_per_task_ms": 686538.5927419355,
      "avg_output_tokens_per_task": 55521.342741935485,
      "last_run_timestamp": "2026-10-01T16:01:02.669Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "BaseTen",
      "accuracy": 0.7051358812949641,
      "accuracy_std_dev": 0.03208361099796414,
      "run_count": 34,
      "total_questions": 1668,
      "avg_cost_per_task_usd": 0.09995068285371703,
      "avg_duration_per_task_ms": 72045.10491606714,
      "avg_output_tokens_per_task": 8049.988009592326,
      "last_run_timestamp": "2026-10-24T19:37:14.677Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Nebius",
      "accuracy": 0.70512827090301,
      "accuracy_std_dev": 0.04983484902405144,
      "run_count": 6,
      "total_questions": 299,
      "avg_cost_per_task_usd": 0.1901660066889632,
      "avg_duration_per_task_ms": 613546.7023411371,
      "avg_output_tokens_per_task": 9664.80602006689,
      "last_run_timestamp": "2026-10-24T21:19:54.882Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Inceptron",
      "accuracy": 0.7046738967642526,
      "accuracy_std_dev": 0.03234510691217085,
      "run_count": 13,
      "total_questions": 649,
      "avg_cost_per_task_usd": 0.07402583821263482,
      "avg_duration_per_task_ms": 210999.13251155624,
      "avg_output_tokens_per_task": 9489.39907550077,
      "last_run_timestamp": "2026-10-24T19:09:54.424Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Decart",
      "accuracy": 0.70395825,
      "accuracy_std_dev": 0.024642497310286955,
      "run_count": 16,
      "total_questions": 800,
      "avg_cost_per_task_usd": 0.13601082625,
      "avg_duration_per_task_ms": 82905.08,
      "avg_output_tokens_per_task": 10495.6875,
      "last_run_timestamp": "2026-10-24T19:06:47.526Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7033335,
      "accuracy_std_dev": 0.016666499999999973,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01719149,
      "avg_duration_per_task_ms": 73722.59,
      "avg_output_tokens_per_task": 12133.38,
      "last_run_timestamp": "2026-10-24T12:37:04.711Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Parasail",
      "accuracy": 0.703333,
      "accuracy_std_dev": 0.010000000000000009,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.012750089999999999,
      "avg_duration_per_task_ms": 89612.54,
      "avg_output_tokens_per_task": 11420.52,
      "last_run_timestamp": "2026-10-24T12:33:10.739Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.7-plus-20260602",
      "model_slug": "qwen/qwen3.7-plus",
      "model_name": "Qwen: Qwen3.7 Plus",
      "provider_name": "auto-routing",
      "accuracy": 0.703333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.05425128,
      "avg_duration_per_task_ms": 396980.46,
      "avg_output_tokens_per_task": 20673.9,
      "last_run_timestamp": "2026-10-01T08:05:20.277Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-sol-20260922",
      "model_slug": "openai/gpt-6-sol",
      "model_name": "OpenAI: GPT-6 Sol",
      "provider_name": "OpenAI",
      "accuracy": 0.7022223333333333,
      "accuracy_std_dev": 0.011330936570096727,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.13357544,
      "avg_duration_per_task_ms": 105586.74666666667,
      "avg_output_tokens_per_task": 5260.826666666667,
      "last_run_timestamp": "2026-10-02T23:07:17.250Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "Venice",
      "accuracy": 0.7022222222222222,
      "accuracy_std_dev": 0.06584673187668091,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.07167732666666665,
      "avg_duration_per_task_ms": 274025.66444444447,
      "avg_output_tokens_per_task": 26720.986666666668,
      "last_run_timestamp": "2026-10-24T18:56:04.577Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Cloudflare",
      "accuracy": 0.701764705882353,
      "accuracy_std_dev": 0.03595790471184932,
      "run_count": 17,
      "total_questions": 850,
      "avg_cost_per_task_usd": 0.1260328188235294,
      "avg_duration_per_task_ms": 127118.39411764705,
      "avg_output_tokens_per_task": 8942.11294117647,
      "last_run_timestamp": "2026-10-24T19:19:29.660Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Parasail",
      "accuracy": 0.7016665000000001,
      "accuracy_std_dev": 0.005000166672222055,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.052099805000000006,
      "avg_duration_per_task_ms": 224630.845,
      "avg_output_tokens_per_task": 6764.525,
      "last_run_timestamp": "2026-08-22T08:43:02.347Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "DeepInfra",
      "accuracy": 0.7012261270903011,
      "accuracy_std_dev": 0.02520740151666215,
      "run_count": 6,
      "total_questions": 299,
      "avg_cost_per_task_usd": 0.025383806020066887,
      "avg_duration_per_task_ms": 238193.83946488294,
      "avg_output_tokens_per_task": 9583.702341137123,
      "last_run_timestamp": "2026-08-29T16:45:30.973Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5-mini-2025-08-07",
      "model_slug": "openai/gpt-5-mini",
      "model_name": "OpenAI: GPT-5 Mini",
      "provider_name": "Azure",
      "accuracy": 0.701111,
      "accuracy_std_dev": 0.011811220865487748,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.13736028333333333,
      "avg_duration_per_task_ms": 494538.36333333334,
      "avg_output_tokens_per_task": 57774.223333333335,
      "last_run_timestamp": "2026-10-01T19:05:16.547Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Novita",
      "accuracy": 0.7000836992481203,
      "accuracy_std_dev": 0.028456037794892915,
      "run_count": 8,
      "total_questions": 399,
      "avg_cost_per_task_usd": 0.020681045112781956,
      "avg_duration_per_task_ms": 573578.8195488722,
      "avg_output_tokens_per_task": 9328.899749373433,
      "last_run_timestamp": "2026-10-24T19:39:11.501Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-luna-20260709",
      "model_slug": "openai/gpt-5.6-luna",
      "model_name": "OpenAI: GPT-5.6 Luna",
      "provider_name": "auto-routing",
      "accuracy": 0.7000003333333334,
      "accuracy_std_dev": 0.009428090415820642,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.010915493333333333,
      "avg_duration_per_task_ms": 115952.82,
      "avg_output_tokens_per_task": 5345.16,
      "last_run_timestamp": "2026-10-01T19:04:36.191Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5-mini-2025-08-07",
      "model_slug": "openai/gpt-5-mini",
      "model_name": "OpenAI: GPT-5 Mini",
      "provider_name": "auto-routing",
      "accuracy": 0.7000003333333332,
      "accuracy_std_dev": 0.024944382578492925,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.12734470666666667,
      "avg_duration_per_task_ms": 591806.9066666666,
      "avg_output_tokens_per_task": 57532.753333333334,
      "last_run_timestamp": "2026-10-01T19:04:42.411Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "GLM 5.3 Flash",
      "provider_name": "digitalocean",
      "accuracy": 0.7,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.00357748,
      "avg_duration_per_task_ms": 48793.08,
      "avg_output_tokens_per_task": 1567.76,
      "last_run_timestamp": "2026-10-01T23:50:45.022Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "inclusionai/ling-3.0-flash-vl-20260910",
      "model_slug": "inclusionai/ling-3.0-flash-vl",
      "model_name": "inclusionAI: Ling 3.0 Flash VL",
      "provider_name": "auto-routing",
      "accuracy": 0.7,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.0030046400000000003,
      "avg_duration_per_task_ms": 98848.12,
      "avg_output_tokens_per_task": 17113.74,
      "last_run_timestamp": "2026-10-01T08:00:19.901Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.7,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.12726106,
      "avg_duration_per_task_ms": 181226.44,
      "avg_output_tokens_per_task": 9838.56,
      "last_run_timestamp": "2026-10-04T03:00:16.405Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "Phala",
      "accuracy": 0.7,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.07376152,
      "avg_duration_per_task_ms": 750952.84,
      "avg_output_tokens_per_task": 11282.28,
      "last_run_timestamp": "2026-08-29T16:44:16.336Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "Io Net",
      "accuracy": 0.7,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.01919496,
      "avg_duration_per_task_ms": 285286.98,
      "avg_output_tokens_per_task": 14572.04,
      "last_run_timestamp": "2026-10-24T12:28:14.169Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.2-20251211",
      "model_slug": "openai/gpt-5.2",
      "model_name": "OpenAI: GPT-5.2",
      "provider_name": "Azure",
      "accuracy": 0.7,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.31292506,
      "avg_duration_per_task_ms": 261751.42,
      "avg_output_tokens_per_task": 14948.8,
      "last_run_timestamp": "2026-08-01T08:06:08.193Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6.1-sol-20260929",
      "model_slug": "openai/gpt-6.1-sol",
      "model_name": "OpenAI: GPT-6.1 Sol",
      "provider_name": "auto-routing",
      "accuracy": 0.7,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.1012818,
      "avg_duration_per_task_ms": 144103.72,
      "avg_output_tokens_per_task": 3638.14,
      "last_run_timestamp": "2026-10-01T08:03:00.822Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-flash-20260826",
      "model_slug": "qwen/qwen3.8-flash",
      "model_name": "Qwen: Qwen3.8 Flash",
      "provider_name": "Alibaba",
      "accuracy": 0.7,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.019795359999999998,
      "avg_duration_per_task_ms": 396615.94,
      "avg_output_tokens_per_task": 23218.96,
      "last_run_timestamp": "2026-10-01T08:03:06.120Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "StreamLake",
      "accuracy": 0.6986667999999999,
      "accuracy_std_dev": 0.029028574428655626,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.04290838,
      "avg_duration_per_task_ms": 173950.452,
      "avg_output_tokens_per_task": 8370.488,
      "last_run_timestamp": "2026-08-29T17:03:30.001Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "SiliconFlow",
      "accuracy": 0.698333375,
      "accuracy_std_dev": 0.13007476749540772,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0452811775,
      "avg_duration_per_task_ms": 317194.5925,
      "avg_output_tokens_per_task": 10517.9925,
      "last_run_timestamp": "2026-10-24T21:05:35.420Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Morph",
      "accuracy": 0.6977779999999999,
      "accuracy_std_dev": 0.03624346343825322,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.10053378,
      "avg_duration_per_task_ms": 541324.4933333333,
      "avg_output_tokens_per_task": 9996.04,
      "last_run_timestamp": "2026-08-22T08:53:18.310Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-luna-pro-20260709",
      "model_slug": "openai/gpt-5.6-luna-pro",
      "model_name": "OpenAI: GPT-5.6 Luna Pro",
      "provider_name": "Azure",
      "accuracy": 0.6977778333333333,
      "accuracy_std_dev": 0.017915990515148434,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.21024033333333336,
      "avg_duration_per_task_ms": 335029.48333333334,
      "avg_output_tokens_per_task": 34166.46666666667,
      "last_run_timestamp": "2026-10-01T16:01:04.829Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "GMICloud",
      "accuracy": 0.6972220833333334,
      "accuracy_std_dev": 0.04077554452008854,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.09417427166666667,
      "avg_duration_per_task_ms": 206387.00666666665,
      "avg_output_tokens_per_task": 7840.663333333333,
      "last_run_timestamp": "2026-10-24T19:36:22.203Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "Novita",
      "accuracy": 0.6970371111111111,
      "accuracy_std_dev": 0.052409631153580175,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.0251458,
      "avg_duration_per_task_ms": 210354.41555555555,
      "avg_output_tokens_per_task": 9795.00888888889,
      "last_run_timestamp": "2026-10-24T19:49:21.066Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "DeepInfra",
      "accuracy": 0.6969084516129033,
      "accuracy_std_dev": 0.011597141847886495,
      "run_count": 10,
      "total_questions": 496,
      "avg_cost_per_task_usd": 0.02167151008064516,
      "avg_duration_per_task_ms": 544748.2903225806,
      "avg_output_tokens_per_task": 18890.834677419356,
      "last_run_timestamp": "2026-10-24T19:03:53.319Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2-thinking-20251106",
      "model_slug": "moonshotai/kimi-k2-thinking",
      "model_name": "MoonshotAI: Kimi K2 Thinking",
      "provider_name": "Google",
      "accuracy": 0.696667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.11266726,
      "avg_duration_per_task_ms": 191409.04,
      "avg_output_tokens_per_task": 8829.08,
      "last_run_timestamp": "2026-08-01T17:38:46.618Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-astra-20260903",
      "model_slug": "openai/gpt-6-astra",
      "model_name": "OpenAI: GPT-6 Astra",
      "provider_name": "Azure",
      "accuracy": 0.6966665000000001,
      "accuracy_std_dev": 0.0033334999999999892,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.62488186,
      "avg_duration_per_task_ms": 152081.62,
      "avg_output_tokens_per_task": 3927.45,
      "last_run_timestamp": "2026-10-01T08:05:41.499Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-sol-20260922",
      "model_slug": "openai/gpt-6-sol",
      "model_name": "OpenAI: GPT-6 Sol",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.6966665000000001,
      "accuracy_std_dev": 0.023333500000000007,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.14710045000000002,
      "avg_duration_per_task_ms": 163276.98,
      "avg_output_tokens_per_task": 5053.91,
      "last_run_timestamp": "2026-10-01T08:02:59.984Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Reka",
      "accuracy": 0.6958335,
      "accuracy_std_dev": 0.03781341986980283,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.132227115,
      "avg_duration_per_task_ms": 205681.995,
      "avg_output_tokens_per_task": 9407.695,
      "last_run_timestamp": "2026-10-01T22:50:26.018Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "Venice",
      "accuracy": 0.6956845625,
      "accuracy_std_dev": 0.03279928505657131,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.045980859375,
      "avg_duration_per_task_ms": 547809.8258928572,
      "avg_output_tokens_per_task": 9085.082589285714,
      "last_run_timestamp": "2026-10-24T20:03:06.027Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Venice",
      "accuracy": 0.6952381877551019,
      "accuracy_std_dev": 0.03106996917024538,
      "run_count": 5,
      "total_questions": 245,
      "avg_cost_per_task_usd": 0.016887012244897963,
      "avg_duration_per_task_ms": 257280.0285714286,
      "avg_output_tokens_per_task": 8914.297959183674,
      "last_run_timestamp": "2026-08-29T17:28:21.827Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "auto-routing",
      "accuracy": 0.6948148888888889,
      "accuracy_std_dev": 0.030756431516909117,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.04968136888888889,
      "avg_duration_per_task_ms": 351365.8222222222,
      "avg_output_tokens_per_task": 9136.788888888888,
      "last_run_timestamp": "2026-10-24T19:50:05.039Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.7-flash-20260119",
      "model_slug": "z-ai/glm-4.7-flash",
      "model_name": "Z.ai: GLM 4.7 Flash",
      "provider_name": "Venice",
      "accuracy": 0.6937392385786801,
      "accuracy_std_dev": 0.0037967340817470817,
      "run_count": 4,
      "total_questions": 197,
      "avg_cost_per_task_usd": 0.005706243654822335,
      "avg_duration_per_task_ms": 284564.64467005077,
      "avg_output_tokens_per_task": 8902.248730964468,
      "last_run_timestamp": "2026-08-23T00:00:19.944Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Together",
      "accuracy": 0.6933334999999999,
      "accuracy_std_dev": 0.006666499999999964,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01276862,
      "avg_duration_per_task_ms": 66785.21,
      "avg_output_tokens_per_task": 8630.75,
      "last_run_timestamp": "2026-10-24T12:58:47.706Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "ModelRun",
      "accuracy": 0.6933334999999999,
      "accuracy_std_dev": 0.006666499999999964,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01353958,
      "avg_duration_per_task_ms": 117341.27,
      "avg_output_tokens_per_task": 12550.54,
      "last_run_timestamp": "2026-10-24T12:06:17.107Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-luna-pro-20260922",
      "model_slug": "openai/gpt-6-luna-pro",
      "model_name": "OpenAI: GPT-6 Luna Pro",
      "provider_name": "OpenAI",
      "accuracy": 0.6933333333333334,
      "accuracy_std_dev": 0.014401903168987411,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.03953364,
      "avg_duration_per_task_ms": 309829.13333333336,
      "avg_output_tokens_per_task": 32261.04,
      "last_run_timestamp": "2026-10-02T23:10:33.891Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.2-exp",
      "model_slug": "deepseek/deepseek-v3.2-exp",
      "model_name": "DeepSeek: DeepSeek V3.2 Exp",
      "provider_name": "SiliconFlow",
      "accuracy": 0.693333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.07614706,
      "avg_duration_per_task_ms": 539148.38,
      "avg_output_tokens_per_task": 13067.92,
      "last_run_timestamp": "2026-08-01T17:38:16.661Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "Ionstream",
      "accuracy": 0.693333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.015592,
      "avg_duration_per_task_ms": 1084747.56,
      "avg_output_tokens_per_task": 17745.48,
      "last_run_timestamp": "2026-07-31T03:44:20.825Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "BaseTen",
      "accuracy": 0.693333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.03237594,
      "avg_duration_per_task_ms": 69982.76,
      "avg_output_tokens_per_task": 15576.66,
      "last_run_timestamp": "2026-10-20T06:20:53.664Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-sol-20260922",
      "model_slug": "openai/gpt-6-sol",
      "model_name": "OpenAI: GPT-6 Sol",
      "provider_name": "auto-routing",
      "accuracy": 0.693333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.12999802000000002,
      "avg_duration_per_task_ms": 96250.54,
      "avg_output_tokens_per_task": 4982.06,
      "last_run_timestamp": "2026-10-01T08:05:35.867Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "reflection/beam-1.0-501b-a23b-20261016",
      "model_slug": "reflection/beam-1.0-501b-a23b",
      "model_name": "Reflection: Beam 1.0 501B A23B",
      "provider_name": "auto-routing",
      "accuracy": 0.693333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.02754302,
      "avg_duration_per_task_ms": 209408.8,
      "avg_output_tokens_per_task": 14914.54,
      "last_run_timestamp": "2026-10-24T12:03:24.771Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.5-air",
      "model_slug": "z-ai/glm-4.5-air",
      "model_name": "Z.ai: GLM 4.5 Air",
      "provider_name": "auto-routing",
      "accuracy": 0.693333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.01748764,
      "avg_duration_per_task_ms": 289873.3,
      "avg_output_tokens_per_task": 9630.44,
      "last_run_timestamp": "2026-08-01T17:39:06.362Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Mistral",
      "accuracy": 0.6930301363636364,
      "accuracy_std_dev": 0.023993062780151963,
      "run_count": 22,
      "total_questions": 1100,
      "avg_cost_per_task_usd": 0.10643902272727274,
      "avg_duration_per_task_ms": 76241.01909090909,
      "avg_output_tokens_per_task": 9720.23090909091,
      "last_run_timestamp": "2026-10-24T19:25:31.791Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "Novita",
      "accuracy": 0.6917516734693877,
      "accuracy_std_dev": 0.023775708138770536,
      "run_count": 8,
      "total_questions": 392,
      "avg_cost_per_task_usd": 0.03340652295918367,
      "avg_duration_per_task_ms": 558953.0178571428,
      "avg_output_tokens_per_task": 18907.030612244896,
      "last_run_timestamp": "2026-10-17T19:21:01.980Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Friendli",
      "accuracy": 0.6905556666666667,
      "accuracy_std_dev": 0.0689850851662074,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.11194472833333334,
      "avg_duration_per_task_ms": 74478.39833333333,
      "avg_output_tokens_per_task": 9151.01,
      "last_run_timestamp": "2026-10-24T19:20:41.265Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.7-flash-20260119",
      "model_slug": "z-ai/glm-4.7-flash",
      "model_name": "Z.ai: GLM 4.7 Flash",
      "provider_name": "DeepInfra",
      "accuracy": 0.690093718875502,
      "accuracy_std_dev": 0.022147823156238193,
      "run_count": 5,
      "total_questions": 249,
      "avg_cost_per_task_usd": 0.006209787148594378,
      "avg_duration_per_task_ms": 270789.8915662651,
      "avg_output_tokens_per_task": 9428.562248995984,
      "last_run_timestamp": "2026-08-30T00:01:09.439Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-astra-pro-20260903",
      "model_slug": "openai/gpt-6-astra-pro",
      "model_name": "OpenAI: GPT-6 Astra Pro",
      "provider_name": "Azure",
      "accuracy": 0.69,
      "accuracy_std_dev": 0.009999999999999953,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 2.99530482,
      "avg_duration_per_task_ms": 320001.13,
      "avg_output_tokens_per_task": 12475.35,
      "last_run_timestamp": "2026-10-01T08:05:20.373Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5-2025-08-07",
      "model_slug": "openai/gpt-5",
      "model_name": "OpenAI: GPT-5",
      "provider_name": "auto-routing",
      "accuracy": 0.6891892297297297,
      "accuracy_std_dev": 0.03170486152346707,
      "run_count": 3,
      "total_questions": 148,
      "avg_cost_per_task_usd": 0.6622049662162162,
      "avg_duration_per_task_ms": 777791.8783783783,
      "avg_output_tokens_per_task": 59355.4527027027,
      "last_run_timestamp": "2026-10-01T16:01:02.610Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "auto-routing",
      "accuracy": 0.6888886666666667,
      "accuracy_std_dev": 0.04262959037257258,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.01831292444444444,
      "avg_duration_per_task_ms": 377813.63333333336,
      "avg_output_tokens_per_task": 20424.906666666666,
      "last_run_timestamp": "2026-10-24T21:02:58.473Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "Parasail",
      "accuracy": 0.6886632385786803,
      "accuracy_std_dev": 0.0900995905873051,
      "run_count": 8,
      "total_questions": 394,
      "avg_cost_per_task_usd": 0.1329306802030457,
      "avg_duration_per_task_ms": 1780497.6497461929,
      "avg_output_tokens_per_task": 98848.87817258883,
      "last_run_timestamp": "2026-10-24T19:18:22.830Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "inclusionai/ling-3.0-flash-20260723",
      "model_slug": "inclusionai/ling-3.0-flash",
      "model_name": "inclusionAI: Ling 3.0 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.686667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.00219582,
      "avg_duration_per_task_ms": 78121.92,
      "avg_output_tokens_per_task": 11659.46,
      "last_run_timestamp": "2026-10-01T16:00:49.880Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.5-air",
      "model_slug": "z-ai/glm-4.5-air",
      "model_name": "Z.ai: GLM 4.5 Air",
      "provider_name": "Novita",
      "accuracy": 0.686667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.01622654,
      "avg_duration_per_task_ms": 255665.46,
      "avg_output_tokens_per_task": 10102.66,
      "last_run_timestamp": "2026-08-01T17:38:31.690Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Minimax",
      "accuracy": 0.6866668,
      "accuracy_std_dev": 0.03373094853335731,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.035621146,
      "avg_duration_per_task_ms": 161374.6,
      "avg_output_tokens_per_task": 10041.704,
      "last_run_timestamp": "2026-08-29T17:02:26.936Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "OpenInference",
      "accuracy": 0.68666675,
      "accuracy_std_dev": 0.14596428025783398,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.093884975,
      "avg_duration_per_task_ms": 1131069.095,
      "avg_output_tokens_per_task": 24772.465,
      "last_run_timestamp": "2026-10-17T19:19:25.280Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Wafer",
      "accuracy": 0.6866665000000001,
      "accuracy_std_dev": 0.026666499999999982,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01561806,
      "avg_duration_per_task_ms": 117250.14,
      "avg_output_tokens_per_task": 14519.81,
      "last_run_timestamp": "2026-10-24T12:06:50.408Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "Io Net",
      "accuracy": 0.6863144308943089,
      "accuracy_std_dev": 0.02557346986546797,
      "run_count": 5,
      "total_questions": 246,
      "avg_cost_per_task_usd": 0.023131280487804877,
      "avg_duration_per_task_ms": 301545.0325203252,
      "avg_output_tokens_per_task": 16644.90243902439,
      "last_run_timestamp": "2026-10-24T20:27:13.776Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-luna-pro-20260922",
      "model_slug": "openai/gpt-6-luna-pro",
      "model_name": "OpenAI: GPT-6 Luna Pro",
      "provider_name": "Azure",
      "accuracy": 0.6850001666666666,
      "accuracy_std_dev": 0.0143694957278334,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.04138883,
      "avg_duration_per_task_ms": 242662.93333333332,
      "avg_output_tokens_per_task": 30866.096666666668,
      "last_run_timestamp": "2026-10-01T08:05:57.699Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "DigitalOcean",
      "accuracy": 0.6837050185873607,
      "accuracy_std_dev": 0.11487260701179351,
      "run_count": 11,
      "total_questions": 538,
      "avg_cost_per_task_usd": 0.32785563568773235,
      "avg_duration_per_task_ms": 1690067.0799256505,
      "avg_output_tokens_per_task": 54783.62267657993,
      "last_run_timestamp": "2026-10-24T12:39:42.207Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Mancer 2",
      "accuracy": 0.6833335555555555,
      "accuracy_std_dev": 0.09394000318064849,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.05554264666666667,
      "avg_duration_per_task_ms": 504546.8688888889,
      "avg_output_tokens_per_task": 15954.128888888888,
      "last_run_timestamp": "2026-10-24T19:51:41.428Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "inclusionai/ling-3.0-flash-20260723",
      "model_slug": "inclusionai/ling-3.0-flash",
      "model_name": "inclusionAI: Ling 3.0 Flash",
      "provider_name": "DeepInfra",
      "accuracy": 0.6833333333333333,
      "accuracy_std_dev": 0.07779388246014092,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.00909372,
      "avg_duration_per_task_ms": 253755.2,
      "avg_output_tokens_per_task": 15941.146666666667,
      "last_run_timestamp": "2026-10-08T22:38:42.670Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6.1-sol-20260929",
      "model_slug": "openai/gpt-6.1-sol",
      "model_name": "OpenAI: GPT-6.1 Sol",
      "provider_name": "Azure",
      "accuracy": 0.6833333333333333,
      "accuracy_std_dev": 0.012619766977070334,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.10710260999999997,
      "avg_duration_per_task_ms": 135398.91,
      "avg_output_tokens_per_task": 4027.3166666666666,
      "last_run_timestamp": "2026-10-01T08:06:09.665Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "Darkbloom",
      "accuracy": 0.683333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.00634196,
      "avg_duration_per_task_ms": 404260.22,
      "avg_output_tokens_per_task": 17374,
      "last_run_timestamp": "2026-10-17T11:31:13.423Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6.1-sol-20260929",
      "model_slug": "openai/gpt-6.1-sol",
      "model_name": "OpenAI: GPT-6.1 Sol",
      "provider_name": "OpenAI",
      "accuracy": 0.6822223333333333,
      "accuracy_std_dev": 0.011330936570096727,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.09746202,
      "avg_duration_per_task_ms": 132162.47333333333,
      "avg_output_tokens_per_task": 3507.6066666666666,
      "last_run_timestamp": "2026-10-02T23:10:34.290Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "DekaLLM",
      "accuracy": 0.6822220000000001,
      "accuracy_std_dev": 0.020607975203789406,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.015206260000000001,
      "avg_duration_per_task_ms": 252877.44,
      "avg_output_tokens_per_task": 17489.156666666666,
      "last_run_timestamp": "2026-10-24T20:33:46.929Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "NextBit",
      "accuracy": 0.682,
      "accuracy_std_dev": 0.03438995977607416,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.017993061999999997,
      "avg_duration_per_task_ms": 315808.552,
      "avg_output_tokens_per_task": 20586.058,
      "last_run_timestamp": "2026-10-10T21:10:48.057Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-luna-20260709",
      "model_slug": "openai/gpt-5.6-luna",
      "model_name": "OpenAI: GPT-5.6 Luna",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.6816665000000001,
      "accuracy_std_dev": 0.02327362599059285,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.013877145,
      "avg_duration_per_task_ms": 70751.9,
      "avg_output_tokens_per_task": 5304.49,
      "last_run_timestamp": "2026-10-01T19:04:52.652Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "thinkingmachines/inkling-20260715",
      "model_slug": "thinkingmachines/inkling",
      "model_name": "Thinking Machines: Inkling",
      "provider_name": "Together",
      "accuracy": 0.6813334,
      "accuracy_std_dev": 0.0026667999999999913,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.08704178,
      "avg_duration_per_task_ms": 82752.556,
      "avg_output_tokens_per_task": 3576.016,
      "last_run_timestamp": "2026-10-01T16:01:04.859Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "AtlasCloud",
      "accuracy": 0.6811112222222222,
      "accuracy_std_dev": 0.11108899779983991,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.09013350666666665,
      "avg_duration_per_task_ms": 216735.6,
      "avg_output_tokens_per_task": 20823.96222222222,
      "last_run_timestamp": "2026-10-24T19:19:49.639Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-luna-20260922",
      "model_slug": "openai/gpt-6-luna",
      "model_name": "OpenAI: GPT-6 Luna",
      "provider_name": "Azure",
      "accuracy": 0.6811111666666666,
      "accuracy_std_dev": 0.008088865998326882,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.00903582,
      "avg_duration_per_task_ms": 105469.92666666667,
      "avg_output_tokens_per_task": 8459.866666666667,
      "last_run_timestamp": "2026-10-01T08:05:22.516Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4-sonnet-20250522",
      "model_slug": "anthropic/claude-sonnet-4",
      "model_name": "Anthropic: Claude Sonnet 4",
      "provider_name": "auto-routing",
      "accuracy": 0.68,
      "accuracy_std_dev": 0.019999999999999962,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.33301777,
      "avg_duration_per_task_ms": 171524.99,
      "avg_output_tokens_per_task": 8948.71,
      "last_run_timestamp": "2026-09-01T12:41:24.616Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Minimax",
      "accuracy": 0.68,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.01807364,
      "avg_duration_per_task_ms": 208774.72,
      "avg_output_tokens_per_task": 16251.38,
      "last_run_timestamp": "2026-10-24T12:17:54.200Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-astra-20260903",
      "model_slug": "openai/gpt-6-astra",
      "model_name": "OpenAI: GPT-6 Astra",
      "provider_name": "auto-routing",
      "accuracy": 0.68,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.58024224,
      "avg_duration_per_task_ms": 141487.98,
      "avg_output_tokens_per_task": 3731.58,
      "last_run_timestamp": "2026-10-01T08:05:19.702Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-luna-20260922",
      "model_slug": "openai/gpt-6-luna",
      "model_name": "OpenAI: GPT-6 Luna",
      "provider_name": "OpenAI",
      "accuracy": 0.68,
      "accuracy_std_dev": 0.03810292105162893,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.00883898,
      "avg_duration_per_task_ms": 114396.12,
      "avg_output_tokens_per_task": 8841.28,
      "last_run_timestamp": "2026-10-02T23:08:44.329Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.5-air",
      "model_slug": "z-ai/glm-4.5-air",
      "model_name": "Z.ai: GLM 4.5 Air",
      "provider_name": "Z.AI",
      "accuracy": 0.68,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.01975412,
      "avg_duration_per_task_ms": 454730.8,
      "avg_output_tokens_per_task": 9931.28,
      "last_run_timestamp": "2026-08-01T17:38:51.399Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "CoreWeave",
      "accuracy": 0.6785714285714286,
      "accuracy_std_dev": 0.020303791411297286,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.017469171428571427,
      "avg_duration_per_task_ms": 119261.06857142856,
      "avg_output_tokens_per_task": 21365.42,
      "last_run_timestamp": "2026-10-24T11:42:22.586Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "Fireworks",
      "accuracy": 0.6777776666666667,
      "accuracy_std_dev": 0.013698571612479227,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.0340762,
      "avg_duration_per_task_ms": 244093.44,
      "avg_output_tokens_per_task": 27086.806666666667,
      "last_run_timestamp": "2026-09-19T08:01:53.393Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.7-flash-20260119",
      "model_slug": "z-ai/glm-4.7-flash",
      "model_name": "Z.ai: GLM 4.7 Flash",
      "provider_name": "Novita",
      "accuracy": 0.6773334,
      "accuracy_std_dev": 0.025854265068649684,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.00809352,
      "avg_duration_per_task_ms": 287886.284,
      "avg_output_tokens_per_task": 12626.116,
      "last_run_timestamp": "2026-08-30T00:01:09.470Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "DigitalOcean",
      "accuracy": 0.676667,
      "accuracy_std_dev": 0.010000000000000009,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01545023,
      "avg_duration_per_task_ms": 128578.81,
      "avg_output_tokens_per_task": 15653.52,
      "last_run_timestamp": "2026-10-24T12:17:01.920Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-astra-20260903",
      "model_slug": "openai/gpt-6-astra",
      "model_name": "OpenAI: GPT-6 Astra",
      "provider_name": "OpenAI",
      "accuracy": 0.6766665000000001,
      "accuracy_std_dev": 0.0033335000000000448,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.5967202500000001,
      "avg_duration_per_task_ms": 137724.15,
      "avg_output_tokens_per_task": 4015.19,
      "last_run_timestamp": "2026-10-02T23:06:01.512Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Tenstorrent",
      "accuracy": 0.676,
      "accuracy_std_dev": 0.19009238689542513,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.011841847999999999,
      "avg_duration_per_task_ms": 391925.292,
      "avg_output_tokens_per_task": 10879.944,
      "last_run_timestamp": "2026-10-24T13:02:10.008Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Nebius",
      "accuracy": 0.675675972972973,
      "accuracy_std_dev": 0.02284471405428887,
      "run_count": 3,
      "total_questions": 148,
      "avg_cost_per_task_usd": 0.04042408108108108,
      "avg_duration_per_task_ms": 518660.3783783784,
      "avg_output_tokens_per_task": 20990.527027027027,
      "last_run_timestamp": "2026-09-26T08:11:26.490Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "thinkingmachines/inkling-20260715",
      "model_slug": "thinkingmachines/inkling",
      "model_name": "Thinking Machines: Inkling",
      "provider_name": "DeepInfra",
      "accuracy": 0.6755556666666667,
      "accuracy_std_dev": 0.006285236475706844,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.07120372000000001,
      "avg_duration_per_task_ms": 144319.38666666666,
      "avg_output_tokens_per_task": 2561.846666666667,
      "last_run_timestamp": "2026-10-08T22:25:52.178Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.7-flash-20260119",
      "model_slug": "z-ai/glm-4.7-flash",
      "model_name": "Z.ai: GLM 4.7 Flash",
      "provider_name": "Cloudflare",
      "accuracy": 0.6755553333333334,
      "accuracy_std_dev": 0.01422909556351193,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.015154949999999999,
      "avg_duration_per_task_ms": 230024.84,
      "avg_output_tokens_per_task": 10278.443333333333,
      "last_run_timestamp": "2026-10-13T22:01:15.953Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Tenstorrent",
      "accuracy": 0.675,
      "accuracy_std_dev": 0.011667000000000038,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.1079227,
      "avg_duration_per_task_ms": 145751.42,
      "avg_output_tokens_per_task": 18113.26,
      "last_run_timestamp": "2026-10-04T05:20:00.959Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.7-flash-20260119",
      "model_slug": "z-ai/glm-4.7-flash",
      "model_name": "Z.ai: GLM 4.7 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.6746668,
      "accuracy_std_dev": 0.022070706580442783,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.00831342,
      "avg_duration_per_task_ms": 264349.076,
      "avg_output_tokens_per_task": 9566.64,
      "last_run_timestamp": "2026-08-30T00:01:02.161Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "AtlasCloud",
      "accuracy": 0.6738888333333334,
      "accuracy_std_dev": 0.07846715716233696,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.051633876666666675,
      "avg_duration_per_task_ms": 318522.69,
      "avg_output_tokens_per_task": 7956.61,
      "last_run_timestamp": "2026-10-03T16:13:53.555Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-astra-20260903",
      "model_slug": "openai/gpt-6-astra",
      "model_name": "OpenAI: GPT-6 Astra",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.6733334999999999,
      "accuracy_std_dev": 0.00666650000000002,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.64995528,
      "avg_duration_per_task_ms": 85231.23,
      "avg_output_tokens_per_task": 3880.06,
      "last_run_timestamp": "2026-10-01T08:05:26.145Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.1-flash-lite-20260507",
      "model_slug": "google/gemini-3.1-flash-lite",
      "model_name": "Google: Gemini 3.1 Flash Lite",
      "provider_name": "Google AI Studio",
      "accuracy": 0.6733333333333333,
      "accuracy_std_dev": 0.02372680934208296,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.11501549333333332,
      "avg_duration_per_task_ms": 145439.36666666667,
      "avg_output_tokens_per_task": 56059.84,
      "last_run_timestamp": "2026-10-01T19:05:16.548Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2-0905",
      "model_slug": "moonshotai/kimi-k2-0905",
      "model_name": "MoonshotAI: Kimi K2 0905",
      "provider_name": "Novita",
      "accuracy": 0.673333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.09564197999999999,
      "avg_duration_per_task_ms": 166878.82,
      "avg_output_tokens_per_task": 4054.64,
      "last_run_timestamp": "2026-08-01T17:38:24.442Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.2-chat-20251211",
      "model_slug": "openai/gpt-5.2-chat",
      "model_name": "OpenAI: GPT-5.2 Chat",
      "provider_name": "auto-routing",
      "accuracy": 0.673333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.1116277,
      "avg_duration_per_task_ms": 79582.44,
      "avg_output_tokens_per_task": 4280.42,
      "last_run_timestamp": "2026-08-01T08:05:57.504Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-luna-20260922",
      "model_slug": "openai/gpt-6-luna",
      "model_name": "OpenAI: GPT-6 Luna",
      "provider_name": "auto-routing",
      "accuracy": 0.673333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.00928508,
      "avg_duration_per_task_ms": 121838.98,
      "avg_output_tokens_per_task": 8898.44,
      "last_run_timestamp": "2026-10-01T08:00:22.469Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.5-air",
      "model_slug": "z-ai/glm-4.5-air",
      "model_name": "Z.ai: GLM 4.5 Air",
      "provider_name": "SiliconFlow",
      "accuracy": 0.673333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.04504326,
      "avg_duration_per_task_ms": 481727.7,
      "avg_output_tokens_per_task": 10202.42,
      "last_run_timestamp": "2026-08-01T17:38:21.896Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Friendli",
      "accuracy": 0.6725926666666666,
      "accuracy_std_dev": 0.08293649247064486,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.09095060666666666,
      "avg_duration_per_task_ms": 106719.20666666667,
      "avg_output_tokens_per_task": 8550.382222222222,
      "last_run_timestamp": "2026-10-24T19:23:31.089Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "thinkingmachines/inkling-20260715",
      "model_slug": "thinkingmachines/inkling",
      "model_name": "Thinking Machines: Inkling",
      "provider_name": "auto-routing",
      "accuracy": 0.672,
      "accuracy_std_dev": 0.028409779224766934,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.07962264,
      "avg_duration_per_task_ms": 80557.024,
      "avg_output_tokens_per_task": 3531.236,
      "last_run_timestamp": "2026-10-01T16:00:45.064Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-sol-20260922",
      "model_slug": "openai/gpt-6-sol",
      "model_name": "OpenAI: GPT-6 Sol",
      "provider_name": "Azure",
      "accuracy": 0.6706668,
      "accuracy_std_dev": 0.015548970672041286,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.142799016,
      "avg_duration_per_task_ms": 115124.684,
      "avg_output_tokens_per_task": 5068.812,
      "last_run_timestamp": "2026-10-01T08:04:38.837Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "DigitalOcean",
      "accuracy": 0.6706667,
      "accuracy_std_dev": 0.0663460470232402,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.061452046,
      "avg_duration_per_task_ms": 791003.158,
      "avg_output_tokens_per_task": 18308.462,
      "last_run_timestamp": "2026-10-05T20:43:11.175Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.5-haiku-20251001",
      "model_slug": "anthropic/claude-haiku-4.5",
      "model_name": "Anthropic: Claude Haiku 4.5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.6689523428571428,
      "accuracy_std_dev": 0.030318660008971234,
      "run_count": 35,
      "total_questions": 1750,
      "avg_cost_per_task_usd": 0.12536032800000002,
      "avg_duration_per_task_ms": 121567.65485714286,
      "avg_output_tokens_per_task": 12712.501714285714,
      "last_run_timestamp": "2026-10-24T13:04:43.619Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "Cloudflare",
      "accuracy": 0.668691016194332,
      "accuracy_std_dev": 0.033329650538972036,
      "run_count": 10,
      "total_questions": 494,
      "avg_cost_per_task_usd": 0.021791346153846154,
      "avg_duration_per_task_ms": 276569.14979757083,
      "avg_output_tokens_per_task": 16852.0971659919,
      "last_run_timestamp": "2026-10-24T19:52:46.621Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-luna-pro-20260922",
      "model_slug": "openai/gpt-6-luna-pro",
      "model_name": "OpenAI: GPT-6 Luna Pro",
      "provider_name": "auto-routing",
      "accuracy": 0.666667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.038762819999999996,
      "avg_duration_per_task_ms": 314594.08,
      "avg_output_tokens_per_task": 31952.48,
      "last_run_timestamp": "2026-10-01T08:00:24.890Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.5-haiku-20251001",
      "model_slug": "anthropic/claude-haiku-4.5",
      "model_name": "Anthropic: Claude Haiku 4.5",
      "provider_name": "auto-routing",
      "accuracy": 0.666111,
      "accuracy_std_dev": 0.03259831473404721,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.117344875,
      "avg_duration_per_task_ms": 125188.30166666667,
      "avg_output_tokens_per_task": 12661.381666666666,
      "last_run_timestamp": "2026-10-24T12:09:26.665Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "thinkingmachines/inkling-20260715",
      "model_slug": "thinkingmachines/inkling",
      "model_name": "Thinking Machines: Inkling",
      "provider_name": "BaseTen",
      "accuracy": 0.6655251643835617,
      "accuracy_std_dev": 0.050919510401766024,
      "run_count": 3,
      "total_questions": 146,
      "avg_cost_per_task_usd": 0.0655772808219178,
      "avg_duration_per_task_ms": 50910.90410958904,
      "avg_output_tokens_per_task": 3110.958904109589,
      "last_run_timestamp": "2026-09-12T08:11:26.919Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "DekaLLM",
      "accuracy": 0.6649998333333333,
      "accuracy_std_dev": 0.024626931209664655,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.008708743333333335,
      "avg_duration_per_task_ms": 613076.7966666666,
      "avg_output_tokens_per_task": 17440.79,
      "last_run_timestamp": "2026-10-24T19:39:37.861Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "auto-routing",
      "accuracy": 0.6640001999999999,
      "accuracy_std_dev": 0.027519806201352526,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.023235323999999998,
      "avg_duration_per_task_ms": 326459.196,
      "avg_output_tokens_per_task": 10364.68,
      "last_run_timestamp": "2026-08-29T16:11:34.887Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "SambaNova",
      "accuracy": 0.6632998585858586,
      "accuracy_std_dev": 0.0034015000000000017,
      "run_count": 2,
      "total_questions": 99,
      "avg_cost_per_task_usd": 0.07516567676767677,
      "avg_duration_per_task_ms": 78894.82828282828,
      "avg_output_tokens_per_task": 13736.282828282829,
      "last_run_timestamp": "2026-08-15T14:12:46.929Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.5-haiku-20251001",
      "model_slug": "anthropic/claude-haiku-4.5",
      "model_name": "Anthropic: Claude Haiku 4.5",
      "provider_name": "Anthropic",
      "accuracy": 0.6625000833333333,
      "accuracy_std_dev": 0.026566641399376315,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.11466395333333333,
      "avg_duration_per_task_ms": 126856.15333333334,
      "avg_output_tokens_per_task": 12309.436666666666,
      "last_run_timestamp": "2026-10-24T12:22:57.423Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4-sonnet-20250522",
      "model_slug": "anthropic/claude-sonnet-4",
      "model_name": "Anthropic: Claude Sonnet 4",
      "provider_name": "Google",
      "accuracy": 0.6622221666666667,
      "accuracy_std_dev": 0.0198761101108564,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.31710583,
      "avg_duration_per_task_ms": 192864.10333333333,
      "avg_output_tokens_per_task": 8814.376666666667,
      "last_run_timestamp": "2026-09-01T13:04:20.137Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "DigitalOcean",
      "accuracy": 0.66,
      "accuracy_std_dev": 0.017384335224563518,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.026960704,
      "avg_duration_per_task_ms": 420608.976,
      "avg_output_tokens_per_task": 10298.344,
      "last_run_timestamp": "2026-08-29T16:36:42.164Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.66,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.013329200000000001,
      "avg_duration_per_task_ms": 278453.68,
      "avg_output_tokens_per_task": 15070.16,
      "last_run_timestamp": "2026-10-24T12:29:49.995Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "GMICloud",
      "accuracy": 0.66,
      "accuracy_std_dev": 0.1609540552497306,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.1285084711111111,
      "avg_duration_per_task_ms": 355132.86666666664,
      "avg_output_tokens_per_task": 8634.402222222223,
      "last_run_timestamp": "2026-10-24T19:43:33.860Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.5-haiku-20251001",
      "model_slug": "anthropic/claude-haiku-4.5",
      "model_name": "Anthropic: Claude Haiku 4.5",
      "provider_name": "Google",
      "accuracy": 0.659259222222222,
      "accuracy_std_dev": 0.028664879599443933,
      "run_count": 36,
      "total_questions": 1800,
      "avg_cost_per_task_usd": 0.12198797388888888,
      "avg_duration_per_task_ms": 121356.62555555555,
      "avg_output_tokens_per_task": 12615.958333333334,
      "last_run_timestamp": "2026-10-24T12:56:50.982Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Near AI",
      "accuracy": 0.6588886666666667,
      "accuracy_std_dev": 0.03965918385723817,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.006265106666666667,
      "avg_duration_per_task_ms": 362004.0733333333,
      "avg_output_tokens_per_task": 19036.246666666666,
      "last_run_timestamp": "2026-10-24T19:34:04.376Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Parasail",
      "accuracy": 0.6585712857142858,
      "accuracy_std_dev": 0.014125931834894325,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.023240128571428573,
      "avg_duration_per_task_ms": 158768.29142857142,
      "avg_output_tokens_per_task": 17172.022857142856,
      "last_run_timestamp": "2026-10-24T19:17:27.524Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Friendli",
      "accuracy": 0.6573334,
      "accuracy_std_dev": 0.0067987188381341416,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.029082692,
      "avg_duration_per_task_ms": 234875.74,
      "avg_output_tokens_per_task": 10738.14,
      "last_run_timestamp": "2026-08-29T16:14:58.314Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "auto-routing",
      "accuracy": 0.6566381031518624,
      "accuracy_std_dev": 0.020222279406836508,
      "run_count": 7,
      "total_questions": 349,
      "avg_cost_per_task_usd": 0.01698823782234957,
      "avg_duration_per_task_ms": 325791.0286532951,
      "avg_output_tokens_per_task": 22916.045845272205,
      "last_run_timestamp": "2026-10-24T12:47:29.776Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "SiliconFlow",
      "accuracy": 0.656,
      "accuracy_std_dev": 0.02480152325160696,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.022391364000000004,
      "avg_duration_per_task_ms": 255228.56,
      "avg_output_tokens_per_task": 9636.224,
      "last_run_timestamp": "2026-08-29T16:50:17.446Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "Phala",
      "accuracy": 0.6557823102040816,
      "accuracy_std_dev": 0.009318408855593308,
      "run_count": 5,
      "total_questions": 245,
      "avg_cost_per_task_usd": 0.014797057142857142,
      "avg_duration_per_task_ms": 277907.2775510204,
      "avg_output_tokens_per_task": 20245.579591836733,
      "last_run_timestamp": "2026-10-10T11:18:54.565Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "DigitalOcean",
      "accuracy": 0.654618485943775,
      "accuracy_std_dev": 0.11865847802618236,
      "run_count": 10,
      "total_questions": 498,
      "avg_cost_per_task_usd": 0.04366364257028112,
      "avg_duration_per_task_ms": 1933703.5220883535,
      "avg_output_tokens_per_task": 71251.46586345382,
      "last_run_timestamp": "2026-10-24T12:24:29.901Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "auto-routing",
      "accuracy": 0.65416675,
      "accuracy_std_dev": 0.03373725891262509,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.013133230000000001,
      "avg_duration_per_task_ms": 415006.955,
      "avg_output_tokens_per_task": 18075.7325,
      "last_run_timestamp": "2026-10-24T19:51:37.271Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "inclusionai/ling-3.0-flash-fin-20260827",
      "model_slug": "inclusionai/ling-3.0-flash-fin",
      "model_name": "inclusionAI: Ling 3.0 Flash Fin",
      "provider_name": "Novita",
      "accuracy": 0.6533335,
      "accuracy_std_dev": 0.00666650000000002,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.00488981,
      "avg_duration_per_task_ms": 111736.2,
      "avg_output_tokens_per_task": 17377.29,
      "last_run_timestamp": "2026-10-01T08:04:38.584Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.4-nano-20260317",
      "model_slug": "openai/gpt-5.4-nano",
      "model_name": "OpenAI: GPT-5.4 Nano",
      "provider_name": "auto-routing",
      "accuracy": 0.6533333333333333,
      "accuracy_std_dev": 0.023726809342082957,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.03210012,
      "avg_duration_per_task_ms": 143919.89333333334,
      "avg_output_tokens_per_task": 17353.273333333334,
      "last_run_timestamp": "2026-10-01T19:04:28.503Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "GMICloud",
      "accuracy": 0.653333142857143,
      "accuracy_std_dev": 0.10498677770139653,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.017404814285714287,
      "avg_duration_per_task_ms": 257125.88,
      "avg_output_tokens_per_task": 9629.577142857142,
      "last_run_timestamp": "2026-10-24T19:35:00.192Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.1",
      "model_slug": "minimax/minimax-m2.1",
      "model_name": "MiniMax: MiniMax M2.1",
      "provider_name": "Novita",
      "accuracy": 0.653333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.023979720000000003,
      "avg_duration_per_task_ms": 166174,
      "avg_output_tokens_per_task": 10916.26,
      "last_run_timestamp": "2026-08-01T17:38:10.202Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "CoreWeave",
      "accuracy": 0.653333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.07713578,
      "avg_duration_per_task_ms": 238048.66,
      "avg_output_tokens_per_task": 10413.86,
      "last_run_timestamp": "2026-08-01T17:40:48.165Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-luna-20260922",
      "model_slug": "openai/gpt-6-luna",
      "model_name": "OpenAI: GPT-6 Luna",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.653333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.00923216,
      "avg_duration_per_task_ms": 145753.66,
      "avg_output_tokens_per_task": 8192.38,
      "last_run_timestamp": "2026-09-23T06:00:08.549Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "AkashML",
      "accuracy": 0.653243932885906,
      "accuracy_std_dev": 0.02323319256351424,
      "run_count": 15,
      "total_questions": 745,
      "avg_cost_per_task_usd": 0.010499983892617451,
      "avg_duration_per_task_ms": 309969.21610738256,
      "avg_output_tokens_per_task": 17108.903355704697,
      "last_run_timestamp": "2026-10-24T18:50:11.994Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "DeepInfra",
      "accuracy": 0.65166675,
      "accuracy_std_dev": 0.038657908491504044,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.021260025000000002,
      "avg_duration_per_task_ms": 249515.925,
      "avg_output_tokens_per_task": 26907.185,
      "last_run_timestamp": "2026-10-24T12:38:08.218Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.4-nano-20260317",
      "model_slug": "openai/gpt-5.4-nano",
      "model_name": "OpenAI: GPT-5.4 Nano",
      "provider_name": "OpenAI",
      "accuracy": 0.65166675,
      "accuracy_std_dev": 0.012801703448662586,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.031589305,
      "avg_duration_per_task_ms": 139715.76,
      "avg_output_tokens_per_task": 17049.6,
      "last_run_timestamp": "2026-10-02T23:09:54.584Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "thinkingmachines/inkling-small-20260730",
      "model_slug": "thinkingmachines/inkling-small",
      "model_name": "Thinking Machines: Inkling Small",
      "provider_name": "Together",
      "accuracy": 0.65166675,
      "accuracy_std_dev": 0.04506177549972369,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.04474220500000001,
      "avg_duration_per_task_ms": 212693.005,
      "avg_output_tokens_per_task": 3236.785,
      "last_run_timestamp": "2026-09-12T08:12:16.067Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "SiliconFlow",
      "accuracy": 0.650067849593496,
      "accuracy_std_dev": 0.11323840378051966,
      "run_count": 10,
      "total_questions": 492,
      "avg_cost_per_task_usd": 0.025260294715447155,
      "avg_duration_per_task_ms": 965913.6768292683,
      "avg_output_tokens_per_task": 32745.57113821138,
      "last_run_timestamp": "2026-10-17T21:28:32.796Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.5-haiku-20251001",
      "model_slug": "anthropic/claude-haiku-4.5",
      "model_name": "Anthropic: Claude Haiku 4.5",
      "provider_name": "Azure",
      "accuracy": 0.6500000833333334,
      "accuracy_std_dev": 0.02441931916010468,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.11641354333333334,
      "avg_duration_per_task_ms": 142724.34,
      "avg_output_tokens_per_task": 12635.95,
      "last_run_timestamp": "2026-10-24T11:52:54.886Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Inceptron",
      "accuracy": 0.65,
      "accuracy_std_dev": 0.025963164791681315,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.022032366666666667,
      "avg_duration_per_task_ms": 277950.80666666664,
      "avg_output_tokens_per_task": 10197.613333333333,
      "last_run_timestamp": "2026-08-15T14:15:19.430Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Crusoe",
      "accuracy": 0.6499997499999999,
      "accuracy_std_dev": 0.0378593669609979,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.015033419999999999,
      "avg_duration_per_task_ms": 89205.15,
      "avg_output_tokens_per_task": 19468.825,
      "last_run_timestamp": "2026-10-24T19:10:30.177Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "Venice",
      "accuracy": 0.6490788090452262,
      "accuracy_std_dev": 0.06101351971284725,
      "run_count": 8,
      "total_questions": 398,
      "avg_cost_per_task_usd": 0.02417512814070352,
      "avg_duration_per_task_ms": 427158.12060301506,
      "avg_output_tokens_per_task": 15670.065326633166,
      "last_run_timestamp": "2026-10-24T19:06:03.144Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Groq",
      "accuracy": 0.6477778888888889,
      "accuracy_std_dev": 0.032735479299800314,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.033562786666666664,
      "avg_duration_per_task_ms": 67754.29111111112,
      "avg_output_tokens_per_task": 16126.108888888888,
      "last_run_timestamp": "2026-10-24T18:54:16.139Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.4-mini-20260317",
      "model_slug": "openai/gpt-5.4-mini",
      "model_name": "OpenAI: GPT-5.4 Mini",
      "provider_name": "Azure",
      "accuracy": 0.6477778333333334,
      "accuracy_std_dev": 0.03430091421724631,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.09916395333333333,
      "avg_duration_per_task_ms": 153090.92333333334,
      "avg_output_tokens_per_task": 12859.136666666667,
      "last_run_timestamp": "2026-10-01T19:04:42.436Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "Darkbloom",
      "accuracy": 0.6475644699140402,
      "accuracy_std_dev": 0.027096557835463376,
      "run_count": 7,
      "total_questions": 349,
      "avg_cost_per_task_usd": 0.014083598853868194,
      "avg_duration_per_task_ms": 819841.329512894,
      "avg_output_tokens_per_task": 25954.131805157594,
      "last_run_timestamp": "2026-10-24T20:01:47.516Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "Io Net",
      "accuracy": 0.6460437828282828,
      "accuracy_std_dev": 0.038169636351874946,
      "run_count": 8,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.012887035353535355,
      "avg_duration_per_task_ms": 941376.398989899,
      "avg_output_tokens_per_task": 22837.7297979798,
      "last_run_timestamp": "2026-10-24T14:58:16.664Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Krea",
      "accuracy": 0.6455556666666666,
      "accuracy_std_dev": 0.09964144484545016,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.11177372666666668,
      "avg_duration_per_task_ms": 710817.72,
      "avg_output_tokens_per_task": 87845.5,
      "last_run_timestamp": "2026-09-29T19:45:03.655Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "AtlasCloud",
      "accuracy": 0.6437036666666667,
      "accuracy_std_dev": 0.12512151876564725,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.05068651111111112,
      "avg_duration_per_task_ms": 788989.3488888888,
      "avg_output_tokens_per_task": 25827.417777777777,
      "last_run_timestamp": "2026-10-24T19:37:12.172Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.4-nano-20260317",
      "model_slug": "openai/gpt-5.4-nano",
      "model_name": "OpenAI: GPT-5.4 Nano",
      "provider_name": "Azure",
      "accuracy": 0.6433335,
      "accuracy_std_dev": 0.014272415770172448,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.031060726666666667,
      "avg_duration_per_task_ms": 229987.72666666665,
      "avg_output_tokens_per_task": 16358.803333333333,
      "last_run_timestamp": "2026-10-01T19:05:16.543Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "Google",
      "accuracy": 0.6415203403508771,
      "accuracy_std_dev": 0.08180204086577275,
      "run_count": 6,
      "total_questions": 285,
      "avg_cost_per_task_usd": 0.12277472280701755,
      "avg_duration_per_task_ms": 351597.06315789477,
      "avg_output_tokens_per_task": 9359.947368421053,
      "last_run_timestamp": "2026-10-17T19:51:53.622Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "Venice",
      "accuracy": 0.6403186580882353,
      "accuracy_std_dev": 0.0713480377599717,
      "run_count": 11,
      "total_questions": 544,
      "avg_cost_per_task_usd": 0.024952294117647062,
      "avg_duration_per_task_ms": 269151.14705882355,
      "avg_output_tokens_per_task": 25044.069852941175,
      "last_run_timestamp": "2026-10-17T19:25:29.604Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4-sonnet-20250522",
      "model_slug": "anthropic/claude-sonnet-4",
      "model_name": "Anthropic: Claude Sonnet 4",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.64,
      "accuracy_std_dev": 0.020000000000000018,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.29234208,
      "avg_duration_per_task_ms": 227202.07,
      "avg_output_tokens_per_task": 8702.01,
      "last_run_timestamp": "2026-09-01T13:00:51.629Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "ibm-granite/granite-4.2-8b-20260831",
      "model_slug": "ibm-granite/granite-4.2-8b",
      "model_name": "IBM: Granite 4.2 8B",
      "provider_name": "DeepInfra",
      "accuracy": 0.64,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.01684376,
      "avg_duration_per_task_ms": 1178898.4,
      "avg_output_tokens_per_task": 43776.78,
      "last_run_timestamp": "2026-10-09T00:39:20.903Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Novita",
      "accuracy": 0.64,
      "accuracy_std_dev": 0.02022095722758942,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.024095932,
      "avg_duration_per_task_ms": 145745.204,
      "avg_output_tokens_per_task": 10183.88,
      "last_run_timestamp": "2026-08-29T16:49:08.574Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "StreamLake",
      "accuracy": 0.64,
      "accuracy_std_dev": 0.03238665020035263,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.021093272000000003,
      "avg_duration_per_task_ms": 219291.42,
      "avg_output_tokens_per_task": 9989.872,
      "last_run_timestamp": "2026-08-29T16:17:15.093Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "DeepInfra",
      "accuracy": 0.6386381963882618,
      "accuracy_std_dev": 0.04009666605876078,
      "run_count": 18,
      "total_questions": 886,
      "avg_cost_per_task_usd": 0.0201472381489842,
      "avg_duration_per_task_ms": 321295.43115124153,
      "avg_output_tokens_per_task": 17837.213318284423,
      "last_run_timestamp": "2026-10-24T19:19:43.152Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "inclusionai/ling-3.0-flash-fin-20260827",
      "model_slug": "inclusionai/ling-3.0-flash-fin",
      "model_name": "inclusionAI: Ling 3.0 Flash Fin",
      "provider_name": "DeepInfra",
      "accuracy": 0.638298,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 47,
      "avg_cost_per_task_usd": 0.020538510638297872,
      "avg_duration_per_task_ms": 686659.4680851063,
      "avg_output_tokens_per_task": 47094.31914893617,
      "last_run_timestamp": "2026-10-09T00:23:16.829Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.4-mini-20260317",
      "model_slug": "openai/gpt-5.4-mini",
      "model_name": "OpenAI: GPT-5.4 Mini",
      "provider_name": "auto-routing",
      "accuracy": 0.6377776666666667,
      "accuracy_std_dev": 0.04228062043326973,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.08944130666666666,
      "avg_duration_per_task_ms": 118157.82666666666,
      "avg_output_tokens_per_task": 12612.513333333334,
      "last_run_timestamp": "2026-10-01T19:04:42.468Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-235b-a22b-thinking-2507",
      "model_slug": "qwen/qwen3-235b-a22b-thinking-2507",
      "model_name": "Qwen: Qwen3 235B A22B Thinking 2507",
      "provider_name": "Venice",
      "accuracy": 0.6377776666666667,
      "accuracy_std_dev": 0.032810732962790655,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.20948366666666668,
      "avg_duration_per_task_ms": 978205.8133333334,
      "avg_output_tokens_per_task": 27155.76,
      "last_run_timestamp": "2026-08-22T16:17:16.779Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Mara",
      "accuracy": 0.637681,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 46,
      "avg_cost_per_task_usd": 0.04568002173913044,
      "avg_duration_per_task_ms": 70889.17391304347,
      "avg_output_tokens_per_task": 8168.934782608696,
      "last_run_timestamp": "2026-08-01T17:41:39.392Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Venice",
      "accuracy": 0.6366665,
      "accuracy_std_dev": 0.0033334999999999892,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.008583809999999999,
      "avg_duration_per_task_ms": 1454771.53,
      "avg_output_tokens_per_task": 17935.13,
      "last_run_timestamp": "2026-10-24T20:43:26.554Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Phala",
      "accuracy": 0.6353466979865772,
      "accuracy_std_dev": 0.036572065936844325,
      "run_count": 3,
      "total_questions": 149,
      "avg_cost_per_task_usd": 0.033807637583892616,
      "avg_duration_per_task_ms": 262695.20805369125,
      "avg_output_tokens_per_task": 17180.040268456374,
      "last_run_timestamp": "2026-10-24T18:33:47.845Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Mara",
      "accuracy": 0.6307017947368421,
      "accuracy_std_dev": 0.04099517109977711,
      "run_count": 4,
      "total_questions": 190,
      "avg_cost_per_task_usd": 0.02501245263157895,
      "avg_duration_per_task_ms": 66232.84736842105,
      "avg_output_tokens_per_task": 14146.842105263158,
      "last_run_timestamp": "2026-10-24T18:40:18.531Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "AtlasCloud",
      "accuracy": 0.6293332,
      "accuracy_std_dev": 0.02133341250151978,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.028055644,
      "avg_duration_per_task_ms": 202913.916,
      "avg_output_tokens_per_task": 9696.648,
      "last_run_timestamp": "2026-08-29T16:13:54.380Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "SambaNova",
      "accuracy": 0.6284643303370787,
      "accuracy_std_dev": 0.02768998520123195,
      "run_count": 9,
      "total_questions": 445,
      "avg_cost_per_task_usd": 0.041866957303370776,
      "avg_duration_per_task_ms": 82960.69662921349,
      "avg_output_tokens_per_task": 22261.50786516854,
      "last_run_timestamp": "2026-10-24T19:23:06.220Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-235b-a22b-thinking-2507",
      "model_slug": "qwen/qwen3-235b-a22b-thinking-2507",
      "model_name": "Qwen: Qwen3 235B A22B Thinking 2507",
      "provider_name": "Alibaba",
      "accuracy": 0.6266666666666667,
      "accuracy_std_dev": 0.014401903168987377,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.12269806,
      "avg_duration_per_task_ms": 610304.8666666667,
      "avg_output_tokens_per_task": 27046.16,
      "last_run_timestamp": "2026-08-22T16:05:00.942Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Cerebras",
      "accuracy": 0.622962888888889,
      "accuracy_std_dev": 0.02416214886251472,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.06782037777777777,
      "avg_duration_per_task_ms": 31573.64888888889,
      "avg_output_tokens_per_task": 21383.52666666667,
      "last_run_timestamp": "2026-10-24T18:29:40.588Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.4-mini-20260317",
      "model_slug": "openai/gpt-5.4-mini",
      "model_name": "OpenAI: GPT-5.4 Mini",
      "provider_name": "OpenAI",
      "accuracy": 0.6216667499999999,
      "accuracy_std_dev": 0.04997217837144482,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.08547625,
      "avg_duration_per_task_ms": 106037.93,
      "avg_output_tokens_per_task": 11996.815,
      "last_run_timestamp": "2026-10-02T23:10:27.235Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.6192592222222222,
      "accuracy_std_dev": 0.03841860903196025,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.09472958,
      "avg_duration_per_task_ms": 55630.27777777778,
      "avg_output_tokens_per_task": 3606.46,
      "last_run_timestamp": "2026-10-24T19:00:38.447Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Parasail",
      "accuracy": 0.619048,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 49,
      "avg_cost_per_task_usd": 0.019175836734693877,
      "avg_duration_per_task_ms": 171955.81632653062,
      "avg_output_tokens_per_task": 9348.57142857143,
      "last_run_timestamp": "2026-08-01T17:40:52.342Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-235b-a22b-thinking-2507",
      "model_slug": "qwen/qwen3-235b-a22b-thinking-2507",
      "model_name": "Qwen: Qwen3 235B A22B Thinking 2507",
      "provider_name": "auto-routing",
      "accuracy": 0.616161808080808,
      "accuracy_std_dev": 0.009872706603940993,
      "run_count": 4,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.12580894444444446,
      "avg_duration_per_task_ms": 627518.9242424242,
      "avg_output_tokens_per_task": 29626.045454545456,
      "last_run_timestamp": "2026-08-22T16:01:40.197Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.1",
      "model_slug": "minimax/minimax-m2.1",
      "model_name": "MiniMax: MiniMax M2.1",
      "provider_name": "auto-routing",
      "accuracy": 0.613333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.025167079999999998,
      "avg_duration_per_task_ms": 171182.22,
      "avg_output_tokens_per_task": 11292.18,
      "last_run_timestamp": "2026-08-01T17:38:10.364Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "thinkingmachines/inkling-small-20260730",
      "model_slug": "thinkingmachines/inkling-small",
      "model_name": "Thinking Machines: Inkling Small",
      "provider_name": "DeepInfra",
      "accuracy": 0.6076188571428572,
      "accuracy_std_dev": 0.03274303997331158,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.046068719999999994,
      "avg_duration_per_task_ms": 191059.44,
      "avg_output_tokens_per_task": 3405.154285714286,
      "last_run_timestamp": "2026-10-08T22:20:40.622Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.6-20251208",
      "model_slug": "z-ai/glm-4.6v",
      "model_name": "Z.ai: GLM 4.6V",
      "provider_name": "auto-routing",
      "accuracy": 0.606667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.03212956,
      "avg_duration_per_task_ms": 335001.52,
      "avg_output_tokens_per_task": 8851.18,
      "last_run_timestamp": "2026-10-01T08:03:01.921Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "thinkingmachines/inkling-small-20260730",
      "model_slug": "thinkingmachines/inkling-small",
      "model_name": "Thinking Machines: Inkling Small",
      "provider_name": "auto-routing",
      "accuracy": 0.6066668000000001,
      "accuracy_std_dev": 0.026666399999999996,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.046513716,
      "avg_duration_per_task_ms": 209556.196,
      "avg_output_tokens_per_task": 3673.824,
      "last_run_timestamp": "2026-10-01T16:00:32.452Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-2.5-pro",
      "model_slug": "google/gemini-2.5-pro",
      "model_name": "Google: Gemini 2.5 Pro",
      "provider_name": "Google",
      "accuracy": 0.6060269955357143,
      "accuracy_std_dev": 0.032348691330028045,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.19772456473214287,
      "avg_duration_per_task_ms": 188670.52455357142,
      "avg_output_tokens_per_task": 13047.183035714286,
      "last_run_timestamp": "2026-10-01T19:04:55.725Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Mancer 2",
      "accuracy": 0.6024903390804598,
      "accuracy_std_dev": 0.08896202865084331,
      "run_count": 7,
      "total_questions": 348,
      "avg_cost_per_task_usd": 0.031245040229885054,
      "avg_duration_per_task_ms": 1940985.2241379311,
      "avg_output_tokens_per_task": 61478.612068965514,
      "last_run_timestamp": "2026-10-24T19:11:22.177Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2-0905",
      "model_slug": "moonshotai/kimi-k2-0905",
      "model_name": "MoonshotAI: Kimi K2 0905",
      "provider_name": "auto-routing",
      "accuracy": 0.597222,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 48,
      "avg_cost_per_task_usd": 0.09316766666666666,
      "avg_duration_per_task_ms": 117445.22916666667,
      "avg_output_tokens_per_task": 3558.4166666666665,
      "last_run_timestamp": "2026-08-01T17:38:23.537Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-next-2025-02-03",
      "model_slug": "qwen/qwen3-coder-next",
      "model_name": "Qwen: Qwen3 Coder Next",
      "provider_name": "StreamLake",
      "accuracy": 0.5966665714285715,
      "accuracy_std_dev": 0.02742942459089792,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.04750767714285715,
      "avg_duration_per_task_ms": 92834.19142857143,
      "avg_output_tokens_per_task": 4625.282857142857,
      "last_run_timestamp": "2026-10-03T16:14:40.781Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-next-2025-02-03",
      "model_slug": "qwen/qwen3-coder-next",
      "model_name": "Qwen: Qwen3 Coder Next",
      "provider_name": "Parasail",
      "accuracy": 0.5950784093959732,
      "accuracy_std_dev": 0.022941326296058443,
      "run_count": 6,
      "total_questions": 298,
      "avg_cost_per_task_usd": 0.019630268456375836,
      "avg_duration_per_task_ms": 99814.84899328859,
      "avg_output_tokens_per_task": 4258.728187919463,
      "last_run_timestamp": "2026-10-03T16:14:27.398Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-235b-a22b-thinking-2507",
      "model_slug": "qwen/qwen3-235b-a22b-thinking-2507",
      "model_name": "Qwen: Qwen3 235B A22B Thinking 2507",
      "provider_name": "Novita",
      "accuracy": 0.5942761515151515,
      "accuracy_std_dev": 0.012018411802313968,
      "run_count": 4,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.15294240909090911,
      "avg_duration_per_task_ms": 508469.1313131313,
      "avg_output_tokens_per_task": 26213.378787878788,
      "last_run_timestamp": "2026-08-22T16:20:42.106Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-2.5-pro",
      "model_slug": "google/gemini-2.5-pro",
      "model_name": "Google: Gemini 2.5 Pro",
      "provider_name": "auto-routing",
      "accuracy": 0.5933333333333334,
      "accuracy_std_dev": 0.02880349773127486,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.21649743333333332,
      "avg_duration_per_task_ms": 191120.76666666666,
      "avg_output_tokens_per_task": 14122.313333333334,
      "last_run_timestamp": "2026-10-01T19:04:42.461Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "BaseTen",
      "accuracy": 0.593333,
      "accuracy_std_dev": 0.020000000000000018,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.04552234,
      "avg_duration_per_task_ms": 32685.39,
      "avg_output_tokens_per_task": 3125.98,
      "last_run_timestamp": "2026-09-26T08:22:32.737Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.6-20251208",
      "model_slug": "z-ai/glm-4.6v",
      "model_name": "Z.ai: GLM 4.6V",
      "provider_name": "Z.AI",
      "accuracy": 0.593333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.01896666,
      "avg_duration_per_task_ms": 306972.98,
      "avg_output_tokens_per_task": 7436.64,
      "last_run_timestamp": "2026-10-01T08:02:59.889Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-2.5-flash",
      "model_slug": "google/gemini-2.5-flash",
      "model_name": "Google: Gemini 2.5 Flash",
      "provider_name": "Google",
      "accuracy": 0.5881481111111111,
      "accuracy_std_dev": 0.007554077566224961,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.029889999999999996,
      "avg_duration_per_task_ms": 78610.16,
      "avg_output_tokens_per_task": 8108.122222222222,
      "last_run_timestamp": "2026-10-01T19:05:46.198Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "SiliconFlow",
      "accuracy": 0.584762,
      "accuracy_std_dev": 0.13245977235587955,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.03989074857142857,
      "avg_duration_per_task_ms": 436437.8257142857,
      "avg_output_tokens_per_task": 8373.411428571428,
      "last_run_timestamp": "2026-10-10T19:10:33.654Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "ibm-granite/granite-4.2-8b-20260831",
      "model_slug": "ibm-granite/granite-4.2-8b",
      "model_name": "IBM: Granite 4.2 8B",
      "provider_name": "auto-routing",
      "accuracy": 0.583333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.015297799999999999,
      "avg_duration_per_task_ms": 833284.24,
      "avg_output_tokens_per_task": 38596.74,
      "last_run_timestamp": "2026-10-01T08:04:13.086Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "AtlasCloud",
      "accuracy": 0.5801346121212121,
      "accuracy_std_dev": 0.2505881880506142,
      "run_count": 10,
      "total_questions": 495,
      "avg_cost_per_task_usd": 0.015714664646464645,
      "avg_duration_per_task_ms": 289180.65252525255,
      "avg_output_tokens_per_task": 8241.527272727273,
      "last_run_timestamp": "2026-10-17T11:22:12.284Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.2-chat-20251211",
      "model_slug": "openai/gpt-5.2-chat",
      "model_name": "OpenAI: GPT-5.2 Chat",
      "provider_name": "OpenAI",
      "accuracy": 0.58,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.10846708,
      "avg_duration_per_task_ms": 79571.2,
      "avg_output_tokens_per_task": 4199.82,
      "last_run_timestamp": "2026-08-01T08:05:35.385Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-r1-0528",
      "model_slug": "deepseek/deepseek-r1-0528",
      "model_name": "DeepSeek: R1 0528",
      "provider_name": "Novita",
      "accuracy": 0.5789978786127168,
      "accuracy_std_dev": 0.05180400838836742,
      "run_count": 7,
      "total_questions": 346,
      "avg_cost_per_task_usd": 0.18764681502890174,
      "avg_duration_per_task_ms": 2044873.4393063583,
      "avg_output_tokens_per_task": 41085.294797687864,
      "last_run_timestamp": "2026-10-17T19:42:10.129Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-next-2025-02-03",
      "model_slug": "qwen/qwen3-coder-next",
      "model_name": "Qwen: Qwen3 Coder Next",
      "provider_name": "Ionstream",
      "accuracy": 0.57483,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 49,
      "avg_cost_per_task_usd": 0.017800938775510203,
      "avg_duration_per_task_ms": 142107.1224489796,
      "avg_output_tokens_per_task": 3993.1632653061224,
      "last_run_timestamp": "2026-08-08T16:03:49.463Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-2.5-flash",
      "model_slug": "google/gemini-2.5-flash",
      "model_name": "Google: Gemini 2.5 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.571111,
      "accuracy_std_dev": 0.008314764498569157,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.028662713333333336,
      "avg_duration_per_task_ms": 78035.85333333333,
      "avg_output_tokens_per_task": 8001.593333333333,
      "last_run_timestamp": "2026-10-01T19:04:40.340Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "Google",
      "accuracy": 0.5674075555555556,
      "accuracy_std_dev": 0.07922128546898255,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.03296292666666667,
      "avg_duration_per_task_ms": 389055.2444444444,
      "avg_output_tokens_per_task": 17353.69111111111,
      "last_run_timestamp": "2026-10-24T19:54:36.551Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-next-2025-02-03",
      "model_slug": "qwen/qwen3-coder-next",
      "model_name": "Qwen: Qwen3 Coder Next",
      "provider_name": "Alibaba",
      "accuracy": 0.5670103642611685,
      "accuracy_std_dev": 0.038907173466327376,
      "run_count": 6,
      "total_questions": 291,
      "avg_cost_per_task_usd": 0.07826669759450172,
      "avg_duration_per_task_ms": 96488.71821305843,
      "avg_output_tokens_per_task": 6239.993127147766,
      "last_run_timestamp": "2026-10-03T16:05:18.093Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "Crusoe",
      "accuracy": 0.5666665,
      "accuracy_std_dev": 0.1466665,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.02785182,
      "avg_duration_per_task_ms": 139069.32,
      "avg_output_tokens_per_task": 11612.72,
      "last_run_timestamp": "2026-10-24T12:08:00.201Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Cerebras",
      "accuracy": 0.5648772953020134,
      "accuracy_std_dev": 0.0424873867714486,
      "run_count": 3,
      "total_questions": 149,
      "avg_cost_per_task_usd": 0.1486826308724832,
      "avg_duration_per_task_ms": 18002.88590604027,
      "avg_output_tokens_per_task": 2546.3288590604025,
      "last_run_timestamp": "2026-08-29T17:07:32.456Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-2.5-pro",
      "model_slug": "google/gemini-2.5-pro",
      "model_name": "Google: Gemini 2.5 Pro",
      "provider_name": "Google AI Studio",
      "accuracy": 0.5622223333333333,
      "accuracy_std_dev": 0.03456997912768952,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.23936231333333335,
      "avg_duration_per_task_ms": 165980.19333333333,
      "avg_output_tokens_per_task": 13303.673333333334,
      "last_run_timestamp": "2026-10-01T19:04:52.649Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-next-2025-02-03",
      "model_slug": "qwen/qwen3-coder-next",
      "model_name": "Qwen: Qwen3 Coder Next",
      "provider_name": "auto-routing",
      "accuracy": 0.5544443333333333,
      "accuracy_std_dev": 0.10997765317352229,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.040382676666666666,
      "avg_duration_per_task_ms": 99197.56,
      "avg_output_tokens_per_task": 4360.85,
      "last_run_timestamp": "2026-10-03T16:07:50.329Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Novita",
      "accuracy": 0.5534669598997495,
      "accuracy_std_dev": 0.05099077839163053,
      "run_count": 8,
      "total_questions": 399,
      "avg_cost_per_task_usd": 0.0106168671679198,
      "avg_duration_per_task_ms": 216000.37593984962,
      "avg_output_tokens_per_task": 17010.057644110275,
      "last_run_timestamp": "2026-10-24T19:30:50.871Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-2.5-flash",
      "model_slug": "google/gemini-2.5-flash",
      "model_name": "Google: Gemini 2.5 Flash",
      "provider_name": "Google AI Studio",
      "accuracy": 0.5511113333333334,
      "accuracy_std_dev": 0.006285236475706844,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.035753513333333334,
      "avg_duration_per_task_ms": 61316.40666666667,
      "avg_output_tokens_per_task": 6659.566666666667,
      "last_run_timestamp": "2026-10-01T19:04:42.483Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Darkbloom",
      "accuracy": 0.5370368653198654,
      "accuracy_std_dev": 0.03035419429923245,
      "run_count": 6,
      "total_questions": 297,
      "avg_cost_per_task_usd": 0.0026686632996632998,
      "avg_duration_per_task_ms": 405513.9057239057,
      "avg_output_tokens_per_task": 6736.090909090909,
      "last_run_timestamp": "2026-10-24T19:31:02.973Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "auto-routing",
      "accuracy": 0.535956005076142,
      "accuracy_std_dev": 0.0166977509787995,
      "run_count": 8,
      "total_questions": 394,
      "avg_cost_per_task_usd": 0.018755781725888324,
      "avg_duration_per_task_ms": 1279767.0253807106,
      "avg_output_tokens_per_task": 84685.69796954315,
      "last_run_timestamp": "2026-10-24T19:37:33.902Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Nebius",
      "accuracy": 0.5333335,
      "accuracy_std_dev": 0.026666500000000037,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.10219375,
      "avg_duration_per_task_ms": 375443.12,
      "avg_output_tokens_per_task": 17495.06,
      "last_run_timestamp": "2026-08-15T14:16:11.776Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "AkashML",
      "accuracy": 0.5333333333333333,
      "accuracy_std_dev": 0.08960819241688912,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.05793356666666665,
      "avg_duration_per_task_ms": 732402.9933333333,
      "avg_output_tokens_per_task": 42843.026666666665,
      "last_run_timestamp": "2026-08-15T14:08:35.086Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "SiliconFlow",
      "accuracy": 0.526785875,
      "accuracy_std_dev": 0.01933119676763796,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.06628725223214285,
      "avg_duration_per_task_ms": 287109.67633928574,
      "avg_output_tokens_per_task": 4037.7633928571427,
      "last_run_timestamp": "2026-10-24T18:33:38.931Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "auto-routing",
      "accuracy": 0.5226130251256281,
      "accuracy_std_dev": 0.08810932391131698,
      "run_count": 8,
      "total_questions": 398,
      "avg_cost_per_task_usd": 0.05013379899497488,
      "avg_duration_per_task_ms": 411095.6934673367,
      "avg_output_tokens_per_task": 7609.560301507538,
      "last_run_timestamp": "2026-10-24T18:46:40.201Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4.1-2025-04-14",
      "model_slug": "openai/gpt-4.1",
      "model_name": "OpenAI: GPT-4.1",
      "provider_name": "Azure",
      "accuracy": 0.5221089591836734,
      "accuracy_std_dev": 0.023889008586586408,
      "run_count": 4,
      "total_questions": 196,
      "avg_cost_per_task_usd": 0.13644156632653062,
      "avg_duration_per_task_ms": 73125.92346938775,
      "avg_output_tokens_per_task": 2553.7397959183672,
      "last_run_timestamp": "2026-10-01T19:05:18.003Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3-nano-30b-a3b",
      "model_slug": "nvidia/nemotron-3-nano-30b-a3b",
      "model_name": "NVIDIA: Nemotron 3 Nano 30B A3B",
      "provider_name": "auto-routing",
      "accuracy": 0.517500125,
      "accuracy_std_dev": 0.024930426803594342,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0282700725,
      "avg_duration_per_task_ms": 444815,
      "avg_output_tokens_per_task": 58343.125,
      "last_run_timestamp": "2026-10-03T16:09:19.689Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-32b-04-28",
      "model_slug": "qwen/qwen3-32b",
      "model_name": "Qwen: Qwen3 32B",
      "provider_name": "SiliconFlow",
      "accuracy": 0.5173608125000001,
      "accuracy_std_dev": 0.025072499999999998,
      "run_count": 2,
      "total_questions": 96,
      "avg_cost_per_task_usd": 0.06539028125,
      "avg_duration_per_task_ms": 2881049.1770833335,
      "avg_output_tokens_per_task": 39867.583333333336,
      "last_run_timestamp": "2026-08-15T14:12:30.435Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Together",
      "accuracy": 0.5159258888888889,
      "accuracy_std_dev": 0.032271244532157746,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.030723802222222218,
      "avg_duration_per_task_ms": 255044.73555555556,
      "avg_output_tokens_per_task": 20795.11111111111,
      "last_run_timestamp": "2026-10-24T19:23:13.562Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-2.5-flash-lite",
      "model_slug": "google/gemini-2.5-flash-lite",
      "model_name": "Google: Gemini 2.5 Flash Lite",
      "provider_name": "Google AI Studio",
      "accuracy": 0.5156599127516779,
      "accuracy_std_dev": 0.03449671802746843,
      "run_count": 3,
      "total_questions": 149,
      "avg_cost_per_task_usd": 0.024214489932885906,
      "avg_duration_per_task_ms": 190286.47651006712,
      "avg_output_tokens_per_task": 48937.10738255033,
      "last_run_timestamp": "2026-10-01T19:05:00.102Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.6-20251208",
      "model_slug": "z-ai/glm-4.6v",
      "model_name": "Z.ai: GLM 4.6V",
      "provider_name": "Novita",
      "accuracy": 0.513333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.03238348,
      "avg_duration_per_task_ms": 271942.88,
      "avg_output_tokens_per_task": 7084.96,
      "last_run_timestamp": "2026-10-01T08:05:50.069Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.1-terminus",
      "model_slug": "deepseek/deepseek-v3.1-terminus",
      "model_name": "DeepSeek: DeepSeek V3.1 Terminus",
      "provider_name": "DeepInfra",
      "accuracy": 0.5104164999999999,
      "accuracy_std_dev": 0.003472500000000045,
      "run_count": 2,
      "total_questions": 96,
      "avg_cost_per_task_usd": 0.024820281250000003,
      "avg_duration_per_task_ms": 480070.4583333333,
      "avg_output_tokens_per_task": 9578.8125,
      "last_run_timestamp": "2026-08-08T16:03:43.735Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4.1-2025-04-14",
      "model_slug": "openai/gpt-4.1",
      "model_name": "OpenAI: GPT-4.1",
      "provider_name": "OpenAI",
      "accuracy": 0.5083332500000001,
      "accuracy_std_dev": 0.026404802805313633,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.12635733,
      "avg_duration_per_task_ms": 53878.245,
      "avg_output_tokens_per_task": 2800.47,
      "last_run_timestamp": "2026-10-02T23:07:04.603Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "DeepInfra",
      "accuracy": 0.507575826446281,
      "accuracy_std_dev": 0.10203663931078877,
      "run_count": 5,
      "total_questions": 242,
      "avg_cost_per_task_usd": 0.05456804958677686,
      "avg_duration_per_task_ms": 374769.78925619833,
      "avg_output_tokens_per_task": 24046.93388429752,
      "last_run_timestamp": "2026-08-29T08:10:31.240Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "DeepInfra",
      "accuracy": 0.5051021836734694,
      "accuracy_std_dev": 0.011159046957513885,
      "run_count": 4,
      "total_questions": 196,
      "avg_cost_per_task_usd": 0.019562938775510203,
      "avg_duration_per_task_ms": 1197227.9285714286,
      "avg_output_tokens_per_task": 100212.39285714286,
      "last_run_timestamp": "2026-08-29T17:16:15.676Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-2.5-flash-lite",
      "model_slug": "google/gemini-2.5-flash-lite",
      "model_name": "Google: Gemini 2.5 Flash Lite",
      "provider_name": "auto-routing",
      "accuracy": 0.5033670303030303,
      "accuracy_std_dev": 0.030340499999999992,
      "run_count": 2,
      "total_questions": 99,
      "avg_cost_per_task_usd": 0.01769851515151515,
      "avg_duration_per_task_ms": 135563.53535353535,
      "avg_output_tokens_per_task": 30988.575757575756,
      "last_run_timestamp": "2026-10-01T19:04:42.466Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3-nano-30b-a3b",
      "model_slug": "nvidia/nemotron-3-nano-30b-a3b",
      "model_name": "NVIDIA: Nemotron 3 Nano 30B A3B",
      "provider_name": "DeepInfra",
      "accuracy": 0.5033669747474748,
      "accuracy_std_dev": 0.027292446564651823,
      "run_count": 8,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.025276146464646466,
      "avg_duration_per_task_ms": 577739.3712121212,
      "avg_output_tokens_per_task": 53598.469696969696,
      "last_run_timestamp": "2026-09-26T16:00:25.689Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-r1-0528",
      "model_slug": "deepseek/deepseek-r1-0528",
      "model_name": "DeepSeek: R1 0528",
      "provider_name": "auto-routing",
      "accuracy": 0.4995211379310344,
      "accuracy_std_dev": 0.03521937973703148,
      "run_count": 7,
      "total_questions": 348,
      "avg_cost_per_task_usd": 0.16297116666666667,
      "avg_duration_per_task_ms": 2016445.1580459771,
      "avg_output_tokens_per_task": 38482.971264367814,
      "last_run_timestamp": "2026-10-17T19:35:49.791Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "Mancer 2",
      "accuracy": 0.493333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.14235776,
      "avg_duration_per_task_ms": 443941.74,
      "avg_output_tokens_per_task": 8833.66,
      "last_run_timestamp": "2026-08-06T00:00:02.908Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3-nano-30b-a3b",
      "model_slug": "nvidia/nemotron-3-nano-30b-a3b",
      "model_name": "NVIDIA: Nemotron 3 Nano 30B A3B",
      "provider_name": "Nebius",
      "accuracy": 0.48850570114942526,
      "accuracy_std_dev": 0.017439968302957548,
      "run_count": 7,
      "total_questions": 348,
      "avg_cost_per_task_usd": 0.031982068965517245,
      "avg_duration_per_task_ms": 642915.867816092,
      "avg_output_tokens_per_task": 59401.63505747126,
      "last_run_timestamp": "2026-10-23T09:26:25.528Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3-nano-30b-a3b",
      "model_slug": "nvidia/nemotron-3-nano-30b-a3b",
      "model_name": "NVIDIA: Nemotron 3 Nano 30B A3B",
      "provider_name": "Crusoe",
      "accuracy": 0.4852150080645161,
      "accuracy_std_dev": 0.023420064714684302,
      "run_count": 10,
      "total_questions": 496,
      "avg_cost_per_task_usd": 0.028619552419354842,
      "avg_duration_per_task_ms": 386610.2439516129,
      "avg_output_tokens_per_task": 66355.3125,
      "last_run_timestamp": "2026-10-03T16:14:03.053Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-32b-04-28",
      "model_slug": "qwen/qwen3-32b",
      "model_name": "Qwen: Qwen3 32B",
      "provider_name": "Nebius",
      "accuracy": 0.4850950081300813,
      "accuracy_std_dev": 0.026197876665104,
      "run_count": 5,
      "total_questions": 246,
      "avg_cost_per_task_usd": 0.02379429674796748,
      "avg_duration_per_task_ms": 820057.8048780488,
      "avg_output_tokens_per_task": 22035.906504065042,
      "last_run_timestamp": "2026-08-29T16:13:54.382Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.1-terminus",
      "model_slug": "deepseek/deepseek-v3.1-terminus",
      "model_name": "DeepSeek: DeepSeek V3.1 Terminus",
      "provider_name": "SiliconFlow",
      "accuracy": 0.48266679999999995,
      "accuracy_std_dev": 0.013727319074021705,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.067269828,
      "avg_duration_per_task_ms": 279626.944,
      "avg_output_tokens_per_task": 4093.296,
      "last_run_timestamp": "2026-08-29T16:26:59.974Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-3-12b-it",
      "model_slug": "google/gemma-3-12b-it",
      "model_name": "Google: Gemma 3 12B",
      "provider_name": "DeepInfra",
      "accuracy": 0.48,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.011001320000000002,
      "avg_duration_per_task_ms": 80205.64,
      "avg_output_tokens_per_task": 3762.88,
      "last_run_timestamp": "2026-10-08T22:08:03.352Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_slug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_name": "Qwen: Qwen3 VL 235B A22B Instruct",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.48,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.07899766,
      "avg_duration_per_task_ms": 132704.5,
      "avg_output_tokens_per_task": 3161.38,
      "last_run_timestamp": "2026-10-04T03:00:16.401Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Crusoe",
      "accuracy": 0.47916700000000007,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 96,
      "avg_cost_per_task_usd": 0.04362447916666667,
      "avg_duration_per_task_ms": 66953.94791666667,
      "avg_output_tokens_per_task": 2828.96875,
      "last_run_timestamp": "2026-08-15T14:08:14.681Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-r1-0528",
      "model_slug": "deepseek/deepseek-r1-0528",
      "model_name": "DeepSeek: R1 0528",
      "provider_name": "SiliconFlow",
      "accuracy": 0.47874697986577186,
      "accuracy_std_dev": 0.00975008705374242,
      "run_count": 6,
      "total_questions": 298,
      "avg_cost_per_task_usd": 0.16965920805369125,
      "avg_duration_per_task_ms": 2424316.9261744968,
      "avg_output_tokens_per_task": 45379.409395973154,
      "last_run_timestamp": "2026-10-17T19:38:30.603Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4.1-2025-04-14",
      "model_slug": "openai/gpt-4.1",
      "model_name": "OpenAI: GPT-4.1",
      "provider_name": "auto-routing",
      "accuracy": 0.4777776666666667,
      "accuracy_std_dev": 0.013698571612479242,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.12510532,
      "avg_duration_per_task_ms": 53067.566666666666,
      "avg_output_tokens_per_task": 2709.2933333333335,
      "last_run_timestamp": "2026-10-01T19:04:26.703Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "auto-routing",
      "accuracy": 0.4766666666666667,
      "accuracy_std_dev": 0.029501952227305586,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.07812271,
      "avg_duration_per_task_ms": 102263.82666666666,
      "avg_output_tokens_per_task": 2763.403333333333,
      "last_run_timestamp": "2026-10-03T16:07:50.332Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-next-80b-a3b-instruct-2509",
      "model_slug": "qwen/qwen3-next-80b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Next 80B A3B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.4766666666666667,
      "accuracy_std_dev": 0.02301379618595961,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.025132050000000003,
      "avg_duration_per_task_ms": 126606.66333333333,
      "avg_output_tokens_per_task": 3730.5233333333335,
      "last_run_timestamp": "2026-10-03T16:13:56.619Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-next-80b-a3b-instruct-2509",
      "model_slug": "qwen/qwen3-next-80b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Next 80B A3B Instruct",
      "provider_name": "Google",
      "accuracy": 0.4755555,
      "accuracy_std_dev": 0.023306820916847493,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.03474162,
      "avg_duration_per_task_ms": 44831.98333333333,
      "avg_output_tokens_per_task": 3658.7366666666667,
      "last_run_timestamp": "2026-10-03T16:17:11.217Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.3-chat-20260303",
      "model_slug": "openai/gpt-5.3-chat",
      "model_name": "OpenAI: GPT-5.3 Chat",
      "provider_name": "OpenAI",
      "accuracy": 0.473333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.09545477999999999,
      "avg_duration_per_task_ms": 59370.42,
      "avg_output_tokens_per_task": 2482.38,
      "last_run_timestamp": "2026-08-01T08:06:06.635Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "CoreWeave",
      "accuracy": 0.473333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.24038894,
      "avg_duration_per_task_ms": 93856.16,
      "avg_output_tokens_per_task": 2866.86,
      "last_run_timestamp": "2026-08-01T17:41:27.708Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "Google",
      "accuracy": 0.47222181249999995,
      "accuracy_std_dev": 0.0011595000000000077,
      "run_count": 2,
      "total_questions": 96,
      "avg_cost_per_task_usd": 0.07886530208333332,
      "avg_duration_per_task_ms": 55709.895833333336,
      "avg_output_tokens_per_task": 3493.7291666666665,
      "last_run_timestamp": "2026-08-15T14:15:58.674Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Parasail",
      "accuracy": 0.46820784971098267,
      "accuracy_std_dev": 0.04097451831844482,
      "run_count": 7,
      "total_questions": 346,
      "avg_cost_per_task_usd": 0.027368598265895955,
      "avg_duration_per_task_ms": 2298741.832369942,
      "avg_output_tokens_per_task": 136175.1647398844,
      "last_run_timestamp": "2026-10-24T19:05:04.624Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-next-80b-a3b-instruct-2509",
      "model_slug": "qwen/qwen3-next-80b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Next 80B A3B Instruct",
      "provider_name": "Alibaba",
      "accuracy": 0.46777799999999997,
      "accuracy_std_dev": 0.0387617763868823,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.02465040333333333,
      "avg_duration_per_task_ms": 59075.62,
      "avg_output_tokens_per_task": 3762.943333333333,
      "last_run_timestamp": "2026-10-03T16:10:49.378Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-30b-a3b-04-28",
      "model_slug": "qwen/qwen3-30b-a3b",
      "model_name": "Qwen: Qwen3 30B A3B",
      "provider_name": "auto-routing",
      "accuracy": 0.466667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.03175828,
      "avg_duration_per_task_ms": 435844.14,
      "avg_output_tokens_per_task": 22556.3,
      "last_run_timestamp": "2026-08-01T17:38:32.120Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5-nano-2025-08-07",
      "model_slug": "openai/gpt-5-nano",
      "model_name": "OpenAI: GPT-5 Nano",
      "provider_name": "auto-routing",
      "accuracy": 0.4662164729729729,
      "accuracy_std_dev": 0.03892660093903227,
      "run_count": 3,
      "total_questions": 148,
      "avg_cost_per_task_usd": 0.04219703378378378,
      "avg_duration_per_task_ms": 711257.6756756756,
      "avg_output_tokens_per_task": 98782.22297297297,
      "last_run_timestamp": "2026-10-01T19:04:24.509Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-32b-04-28",
      "model_slug": "qwen/qwen3-32b",
      "model_name": "Qwen: Qwen3 32B",
      "provider_name": "auto-routing",
      "accuracy": 0.4653334,
      "accuracy_std_dev": 0.021249757124259083,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.020888280000000002,
      "avg_duration_per_task_ms": 647002.876,
      "avg_output_tokens_per_task": 20107.14,
      "last_run_timestamp": "2026-08-29T16:16:50.456Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-next-80b-a3b-instruct-2509",
      "model_slug": "qwen/qwen3-next-80b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Next 80B A3B Instruct",
      "provider_name": "Parasail",
      "accuracy": 0.4650001666666666,
      "accuracy_std_dev": 0.03531613575226233,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.018810040000000004,
      "avg_duration_per_task_ms": 93827.55333333333,
      "avg_output_tokens_per_task": 3599.63,
      "last_run_timestamp": "2026-10-03T16:12:20.062Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Coder 30B A3B Instruct",
      "provider_name": "DigitalOcean",
      "accuracy": 0.46444433333333335,
      "accuracy_std_dev": 0.008314853589945066,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.033371,
      "avg_duration_per_task_ms": 305705.87333333335,
      "avg_output_tokens_per_task": 3594.2266666666665,
      "last_run_timestamp": "2026-08-15T08:15:42.180Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4o",
      "model_slug": "openai/gpt-4o",
      "model_name": "OpenAI: GPT-4o",
      "provider_name": "Azure",
      "accuracy": 0.463333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.26064434,
      "avg_duration_per_task_ms": 31467.4,
      "avg_output_tokens_per_task": 1819.56,
      "last_run_timestamp": "2026-08-01T08:06:06.230Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-30b-a3b-04-28",
      "model_slug": "qwen/qwen3-30b-a3b",
      "model_name": "Qwen: Qwen3 30B A3B",
      "provider_name": "DeepInfra",
      "accuracy": 0.46166650000000004,
      "accuracy_std_dev": 0.008333499999999994,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.035224570000000004,
      "avg_duration_per_task_ms": 486853.18,
      "avg_output_tokens_per_task": 23253.82,
      "last_run_timestamp": "2026-10-08T23:09:09.727Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Groq",
      "accuracy": 0.4615010526315789,
      "accuracy_std_dev": 0.036598855562569095,
      "run_count": 7,
      "total_questions": 342,
      "avg_cost_per_task_usd": 0.05408880409356725,
      "avg_duration_per_task_ms": 193973.8625730994,
      "avg_output_tokens_per_task": 87403.31578947368,
      "last_run_timestamp": "2026-10-17T18:58:49.636Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5-nano-2025-08-07",
      "model_slug": "openai/gpt-5-nano",
      "model_name": "OpenAI: GPT-5 Nano",
      "provider_name": "Azure",
      "accuracy": 0.4608501677852349,
      "accuracy_std_dev": 0.034994868451198195,
      "run_count": 6,
      "total_questions": 298,
      "avg_cost_per_task_usd": 0.04671855033557047,
      "avg_duration_per_task_ms": 767600.1241610738,
      "avg_output_tokens_per_task": 101728.0033557047,
      "last_run_timestamp": "2026-10-01T19:06:01.717Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-2.5-flash-lite",
      "model_slug": "google/gemini-2.5-flash-lite",
      "model_name": "Google: Gemini 2.5 Flash Lite",
      "provider_name": "Google",
      "accuracy": 0.46068351282051284,
      "accuracy_std_dev": 0.04225414673955088,
      "run_count": 4,
      "total_questions": 195,
      "avg_cost_per_task_usd": 0.01900871794871795,
      "avg_duration_per_task_ms": 170098.1794871795,
      "avg_output_tokens_per_task": 35854.89743589744,
      "last_run_timestamp": "2026-10-01T19:06:10.500Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-next-80b-a3b-instruct-2509",
      "model_slug": "qwen/qwen3-next-80b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Next 80B A3B Instruct",
      "provider_name": "DeepInfra",
      "accuracy": 0.4600001428571428,
      "accuracy_std_dev": 0.034226236108027545,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.02152209714285714,
      "avg_duration_per_task_ms": 223527.5657142857,
      "avg_output_tokens_per_task": 3520.9742857142855,
      "last_run_timestamp": "2026-10-08T22:47:27.753Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4.1-2025-04-14",
      "model_slug": "openai/gpt-4.1",
      "model_name": "OpenAI: GPT-4.1",
      "provider_name": "FakeProvider",
      "accuracy": 0.46,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": null,
      "avg_duration_per_task_ms": 158124.84,
      "avg_output_tokens_per_task": 4515,
      "last_run_timestamp": "2026-10-01T19:06:51.244Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.3-chat-20260303",
      "model_slug": "openai/gpt-5.3-chat",
      "model_name": "OpenAI: GPT-5.3 Chat",
      "provider_name": "auto-routing",
      "accuracy": 0.46,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.09873096,
      "avg_duration_per_task_ms": 65244.1,
      "avg_output_tokens_per_task": 2748.04,
      "last_run_timestamp": "2026-08-01T08:05:59.441Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Coder 30B A3B Instruct",
      "provider_name": "Novita",
      "accuracy": 0.45777799999999996,
      "accuracy_std_dev": 0.008314764498569207,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.00622334,
      "avg_duration_per_task_ms": 39170.17333333333,
      "avg_output_tokens_per_task": 1174.36,
      "last_run_timestamp": "2026-08-15T08:16:13.926Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Nebius",
      "accuracy": 0.454583375,
      "accuracy_std_dev": 0.040686489790646416,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.05330521999999999,
      "avg_duration_per_task_ms": 171560.5725,
      "avg_output_tokens_per_task": 3741.72,
      "last_run_timestamp": "2026-10-17T19:11:29.194Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Parasail",
      "accuracy": 0.4521073362068965,
      "accuracy_std_dev": 0.045992926219636195,
      "run_count": 7,
      "total_questions": 348,
      "avg_cost_per_task_usd": 0.015361428160919543,
      "avg_duration_per_task_ms": 128943.27873563218,
      "avg_output_tokens_per_task": 3310.485632183908,
      "last_run_timestamp": "2026-10-24T19:23:50.567Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5-nano-2025-08-07",
      "model_slug": "openai/gpt-5-nano",
      "model_name": "OpenAI: GPT-5 Nano",
      "provider_name": "OpenAI",
      "accuracy": 0.4511785656565656,
      "accuracy_std_dev": 0.03515039216708685,
      "run_count": 4,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.04521708585858586,
      "avg_duration_per_task_ms": 779859.6717171717,
      "avg_output_tokens_per_task": 105433.07575757576,
      "last_run_timestamp": "2026-10-02T23:22:06.591Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "StreamLake",
      "accuracy": 0.44952385714285714,
      "accuracy_std_dev": 0.047255747470329915,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.053731311428571425,
      "avg_duration_per_task_ms": 127204.9,
      "avg_output_tokens_per_task": 3489.102857142857,
      "last_run_timestamp": "2026-10-03T16:19:54.632Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Google",
      "accuracy": 0.44932439189189194,
      "accuracy_std_dev": 0.04806295077869809,
      "run_count": 6,
      "total_questions": 296,
      "avg_cost_per_task_usd": 0.05292330743243243,
      "avg_duration_per_task_ms": 75551.64864864865,
      "avg_output_tokens_per_task": 3306.3682432432433,
      "last_run_timestamp": "2026-10-03T16:07:05.221Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Coder 30B A3B Instruct",
      "provider_name": "Alibaba",
      "accuracy": 0.448889,
      "accuracy_std_dev": 0.0031423825355930196,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.08242190666666667,
      "avg_duration_per_task_ms": 81216.82666666666,
      "avg_output_tokens_per_task": 4095.7466666666664,
      "last_run_timestamp": "2026-08-15T08:11:06.981Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-32b-04-28",
      "model_slug": "qwen/qwen3-32b",
      "model_name": "Qwen: Qwen3 32B",
      "provider_name": "DeepInfra",
      "accuracy": 0.4446664000000001,
      "accuracy_std_dev": 0.027455050942221906,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.021282972,
      "avg_duration_per_task_ms": 631261.664,
      "avg_output_tokens_per_task": 20624.588,
      "last_run_timestamp": "2026-10-09T00:21:42.483Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Venice",
      "accuracy": 0.443749875,
      "accuracy_std_dev": 0.029881164520637004,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.040143645,
      "avg_duration_per_task_ms": 278419.015,
      "avg_output_tokens_per_task": 3569.46,
      "last_run_timestamp": "2026-10-24T19:41:30.973Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "Novita",
      "accuracy": 0.44333349999999994,
      "accuracy_std_dev": 0.07371123086466087,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.18429278333333332,
      "avg_duration_per_task_ms": 118282.36333333333,
      "avg_output_tokens_per_task": 3443.1833333333334,
      "last_run_timestamp": "2026-10-03T16:19:14.095Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4.1-mini-2025-04-14",
      "model_slug": "openai/gpt-4.1-mini",
      "model_name": "OpenAI: GPT-4.1 Mini",
      "provider_name": "OpenAI",
      "accuracy": 0.44333325,
      "accuracy_std_dev": 0.01914841159698371,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.025700495,
      "avg_duration_per_task_ms": 61023.305,
      "avg_output_tokens_per_task": 2441.375,
      "last_run_timestamp": "2026-10-02T23:07:02.872Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-chat-v3-0324",
      "model_slug": "deepseek/deepseek-chat-v3-0324",
      "model_name": "DeepSeek: DeepSeek V3 0324",
      "provider_name": "SiliconFlow",
      "accuracy": 0.4416665,
      "accuracy_std_dev": 0.017873086533948188,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.104832345,
      "avg_duration_per_task_ms": 890298.59,
      "avg_output_tokens_per_task": 21334.155,
      "last_run_timestamp": "2026-08-29T16:10:09.206Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4o-2024-05-13",
      "model_slug": "openai/gpt-4o-2024-05-13",
      "model_name": "OpenAI: GPT-4o (2024-05-13)",
      "provider_name": "OpenAI",
      "accuracy": 0.44,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.7291186999999999,
      "avg_duration_per_task_ms": 35507.54,
      "avg_output_tokens_per_task": 3272.36,
      "last_run_timestamp": "2026-08-01T08:05:57.702Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "Google",
      "accuracy": 0.43999974999999997,
      "accuracy_std_dev": 0.047081383440671115,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.050863820000000004,
      "avg_duration_per_task_ms": 37018.795,
      "avg_output_tokens_per_task": 2487.295,
      "last_run_timestamp": "2026-10-03T16:12:29.903Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta-llama/llama-4-maverick-17b-128e-instruct",
      "model_slug": "meta-llama/llama-4-maverick",
      "model_name": "Meta: Llama 4 Maverick",
      "provider_name": "DigitalOcean",
      "accuracy": 0.4391890202702703,
      "accuracy_std_dev": 0.03559610686773047,
      "run_count": 6,
      "total_questions": 296,
      "avg_cost_per_task_usd": 0.054588496621621616,
      "avg_duration_per_task_ms": 244762.44932432432,
      "avg_output_tokens_per_task": 2182.847972972973,
      "last_run_timestamp": "2026-10-05T20:09:54.588Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4o",
      "model_slug": "openai/gpt-4o",
      "model_name": "OpenAI: GPT-4o",
      "provider_name": "auto-routing",
      "accuracy": 0.43777799999999994,
      "accuracy_std_dev": 0.03861798340151904,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.19450383999999998,
      "avg_duration_per_task_ms": 50798.87333333334,
      "avg_output_tokens_per_task": 2047.5333333333333,
      "last_run_timestamp": "2026-10-01T16:01:13.623Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4.1-mini-2025-04-14",
      "model_slug": "openai/gpt-4.1-mini",
      "model_name": "OpenAI: GPT-4.1 Mini",
      "provider_name": "auto-routing",
      "accuracy": 0.4377776666666667,
      "accuracy_std_dev": 0.011330936570096718,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.02595394666666667,
      "avg_duration_per_task_ms": 66273.3,
      "avg_output_tokens_per_task": 2407.2933333333335,
      "last_run_timestamp": "2026-10-01T19:04:38.295Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4o-2024-08-06",
      "model_slug": "openai/gpt-4o-2024-08-06",
      "model_name": "OpenAI: GPT-4o (2024-08-06)",
      "provider_name": "auto-routing",
      "accuracy": 0.4377776666666667,
      "accuracy_std_dev": 0.016629618082071397,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.21387673999999998,
      "avg_duration_per_task_ms": 47267.2,
      "avg_output_tokens_per_task": 2224.56,
      "last_run_timestamp": "2026-10-01T16:00:49.880Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "mistralai/mistral-small-2603",
      "model_slug": "mistralai/mistral-small-2603",
      "model_name": "Mistral: Mistral Small 4",
      "provider_name": "auto-routing",
      "accuracy": 0.43666650000000007,
      "accuracy_std_dev": 0.023333500000000007,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.014376880000000002,
      "avg_duration_per_task_ms": 93606.58,
      "avg_output_tokens_per_task": 10159.34,
      "last_run_timestamp": "2026-09-01T13:02:23.335Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-chat-v3-0324",
      "model_slug": "deepseek/deepseek-chat-v3-0324",
      "model_name": "DeepSeek: DeepSeek V3 0324",
      "provider_name": "auto-routing",
      "accuracy": 0.435,
      "accuracy_std_dev": 0.020749805890658335,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.074401825,
      "avg_duration_per_task_ms": 778012.545,
      "avg_output_tokens_per_task": 20313.29,
      "last_run_timestamp": "2026-08-29T16:11:46.632Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-32b-04-28",
      "model_slug": "qwen/qwen3-32b",
      "model_name": "Qwen: Qwen3 32B",
      "provider_name": "Tenstorrent",
      "accuracy": 0.43434374747474747,
      "accuracy_std_dev": 0.012448999999999988,
      "run_count": 2,
      "total_questions": 99,
      "avg_cost_per_task_usd": 0.023558444444444444,
      "avg_duration_per_task_ms": 1871236.9090909092,
      "avg_output_tokens_per_task": 14653.222222222223,
      "last_run_timestamp": "2026-10-13T10:40:58.207Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-chat-v3",
      "model_slug": "deepseek/deepseek-chat",
      "model_name": "DeepSeek: DeepSeek V3",
      "provider_name": "DeepInfra",
      "accuracy": 0.43434354545454545,
      "accuracy_std_dev": 0.022037645087440712,
      "run_count": 4,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.06906736868686869,
      "avg_duration_per_task_ms": 323353.22222222225,
      "avg_output_tokens_per_task": 3378.3585858585857,
      "last_run_timestamp": "2026-10-08T22:54:33.845Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4o",
      "model_slug": "openai/gpt-4o",
      "model_name": "OpenAI: GPT-4o",
      "provider_name": "OpenAI",
      "accuracy": 0.43416662500000003,
      "accuracy_std_dev": 0.01613729505630899,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.2170375225,
      "avg_duration_per_task_ms": 46329.7025,
      "avg_output_tokens_per_task": 2265.4175,
      "last_run_timestamp": "2026-10-02T23:10:05.271Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-next-80b-a3b-instruct-2509",
      "model_slug": "qwen/qwen3-next-80b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Next 80B A3B Instruct",
      "provider_name": "Novita",
      "accuracy": 0.4333331666666667,
      "accuracy_std_dev": 0.016330067701601517,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.03858049,
      "avg_duration_per_task_ms": 137271.52,
      "avg_output_tokens_per_task": 3487.26,
      "last_run_timestamp": "2026-10-03T16:18:20.377Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "CoreWeave",
      "accuracy": 0.4328860536912752,
      "accuracy_std_dev": 0.30245117330901533,
      "run_count": 6,
      "total_questions": 298,
      "avg_cost_per_task_usd": 0.006059127516778524,
      "avg_duration_per_task_ms": 401903.8154362416,
      "avg_output_tokens_per_task": 13127.23154362416,
      "last_run_timestamp": "2026-10-24T19:32:27.020Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "Venice",
      "accuracy": 0.4327958548387097,
      "accuracy_std_dev": 0.02500116609760433,
      "run_count": 5,
      "total_questions": 248,
      "avg_cost_per_task_usd": 0.016440024193548387,
      "avg_duration_per_task_ms": 245781.88709677418,
      "avg_output_tokens_per_task": 2789.4475806451615,
      "last_run_timestamp": "2026-10-03T16:14:29.926Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-chat-v3-0324",
      "model_slug": "deepseek/deepseek-chat-v3-0324",
      "model_name": "DeepSeek: DeepSeek V3 0324",
      "provider_name": "DeepInfra",
      "accuracy": 0.4321305463917526,
      "accuracy_std_dev": 0.024017877565471933,
      "run_count": 4,
      "total_questions": 194,
      "avg_cost_per_task_usd": 0.03244629381443299,
      "avg_duration_per_task_ms": 504849.24742268043,
      "avg_output_tokens_per_task": 4749.7525773195875,
      "last_run_timestamp": "2026-08-22T16:13:43.431Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Coder 30B A3B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.4311113333333333,
      "accuracy_std_dev": 0.006285236475706817,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.023601506666666664,
      "avg_duration_per_task_ms": 216917.58666666667,
      "avg_output_tokens_per_task": 3146.213333333333,
      "last_run_timestamp": "2026-08-15T08:02:48.904Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "Alibaba",
      "accuracy": 0.43111116666666666,
      "accuracy_std_dev": 0.04833021190524159,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.4555577033333333,
      "avg_duration_per_task_ms": 109357.17666666667,
      "avg_output_tokens_per_task": 3283.1,
      "last_run_timestamp": "2026-10-03T16:14:53.189Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-14b-04-28",
      "model_slug": "qwen/qwen3-14b",
      "model_name": "Qwen: Qwen3 14B",
      "provider_name": "auto-routing",
      "accuracy": 0.431111,
      "accuracy_std_dev": 0.013698950130089036,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.029769013333333334,
      "avg_duration_per_task_ms": 630152.1266666667,
      "avg_output_tokens_per_task": 18760.18666666667,
      "last_run_timestamp": "2026-10-01T16:00:49.878Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta-llama/llama-4-maverick-17b-128e-instruct",
      "model_slug": "meta-llama/llama-4-maverick",
      "model_name": "Meta: Llama 4 Maverick",
      "provider_name": "Parasail",
      "accuracy": 0.43089426016260163,
      "accuracy_std_dev": 0.03483555910043644,
      "run_count": 5,
      "total_questions": 246,
      "avg_cost_per_task_usd": 0.04202976422764227,
      "avg_duration_per_task_ms": 58240.52032520325,
      "avg_output_tokens_per_task": 2134.682926829268,
      "last_run_timestamp": "2026-08-29T16:52:55.844Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-14b-04-28",
      "model_slug": "qwen/qwen3-14b",
      "model_name": "Qwen: Qwen3 14B",
      "provider_name": "DeepInfra",
      "accuracy": 0.43083325,
      "accuracy_std_dev": 0.031213896924086562,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.029063365,
      "avg_duration_per_task_ms": 761962.965,
      "avg_output_tokens_per_task": 19001.86,
      "last_run_timestamp": "2026-10-08T22:41:10.664Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4o-2024-08-06",
      "model_slug": "openai/gpt-4o-2024-08-06",
      "model_name": "OpenAI: GPT-4o (2024-08-06)",
      "provider_name": "OpenAI",
      "accuracy": 0.43,
      "accuracy_std_dev": 0.0033330000000000026,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.24811575000000002,
      "avg_duration_per_task_ms": 46287.18,
      "avg_output_tokens_per_task": 2565.8,
      "last_run_timestamp": "2026-10-02T23:09:02.629Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta-llama/llama-4-maverick-17b-128e-instruct",
      "model_slug": "meta-llama/llama-4-maverick",
      "model_name": "Meta: Llama 4 Maverick",
      "provider_name": "auto-routing",
      "accuracy": 0.4293334,
      "accuracy_std_dev": 0.013727396775791108,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.046574660000000004,
      "avg_duration_per_task_ms": 138519.112,
      "avg_output_tokens_per_task": 2139.824,
      "last_run_timestamp": "2026-08-29T16:33:29.885Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "DeepInfra",
      "accuracy": 0.42916675,
      "accuracy_std_dev": 0.025967332577442378,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.027418319999999996,
      "avg_duration_per_task_ms": 84447.825,
      "avg_output_tokens_per_task": 2784.11,
      "last_run_timestamp": "2026-10-03T16:12:20.034Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "Io Net",
      "accuracy": 0.42795377521613837,
      "accuracy_std_dev": 0.029431010421121517,
      "run_count": 7,
      "total_questions": 347,
      "avg_cost_per_task_usd": 0.05299580115273775,
      "avg_duration_per_task_ms": 466627.5331412104,
      "avg_output_tokens_per_task": 21231.06916426513,
      "last_run_timestamp": "2026-09-19T16:00:30.864Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Novita",
      "accuracy": 0.42555566666666667,
      "accuracy_std_dev": 0.013698558095248165,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.02489071,
      "avg_duration_per_task_ms": 220083.76333333334,
      "avg_output_tokens_per_task": 3565.5666666666666,
      "last_run_timestamp": "2026-10-03T16:13:22.758Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "DeepInfra",
      "accuracy": 0.4226667,
      "accuracy_std_dev": 0.03237967195031475,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.024325878,
      "avg_duration_per_task_ms": 394315.086,
      "avg_output_tokens_per_task": 3598.632,
      "last_run_timestamp": "2026-10-24T19:25:31.387Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 VL 30B A3B Instruct",
      "provider_name": "Novita",
      "accuracy": 0.42006044796380093,
      "accuracy_std_dev": 0.026085942399861443,
      "run_count": 9,
      "total_questions": 442,
      "avg_cost_per_task_usd": 0.05284868552036199,
      "avg_duration_per_task_ms": 192763.79411764705,
      "avg_output_tokens_per_task": 4184.2692307692305,
      "last_run_timestamp": "2026-10-03T08:12:06.170Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "AtlasCloud",
      "accuracy": 0.41733359999999997,
      "accuracy_std_dev": 0.04591776804506073,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.054456984,
      "avg_duration_per_task_ms": 140659.208,
      "avg_output_tokens_per_task": 3410.888,
      "last_run_timestamp": "2026-08-29T16:12:10.467Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "DeepInfra",
      "accuracy": 0.41666675909090906,
      "accuracy_std_dev": 0.02610295346611798,
      "run_count": 9,
      "total_questions": 440,
      "avg_cost_per_task_usd": 0.03160882727272728,
      "avg_duration_per_task_ms": 641960.8022727272,
      "avg_output_tokens_per_task": 6320.127272727273,
      "last_run_timestamp": "2026-10-24T19:18:23.891Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-14b-04-28",
      "model_slug": "qwen/qwen3-14b",
      "model_name": "Qwen: Qwen3 14B",
      "provider_name": "Alibaba",
      "accuracy": 0.413333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.06003468,
      "avg_duration_per_task_ms": 392895.94,
      "avg_output_tokens_per_task": 17058.96,
      "last_run_timestamp": "2026-08-01T08:06:29.769Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-30b-a3b-04-28",
      "model_slug": "qwen/qwen3-30b-a3b",
      "model_name": "Qwen: Qwen3 30B A3B",
      "provider_name": "Alibaba",
      "accuracy": 0.413333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.03394864,
      "avg_duration_per_task_ms": 276713.14,
      "avg_output_tokens_per_task": 20490.4,
      "last_run_timestamp": "2026-08-01T17:38:29.532Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "mistralai/mistral-small-2603",
      "model_slug": "mistralai/mistral-small-2603",
      "model_name": "Mistral: Mistral Small 4",
      "provider_name": "Mistral",
      "accuracy": 0.41142871428571426,
      "accuracy_std_dev": 0.022737753675420126,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.014496525714285714,
      "avg_duration_per_task_ms": 78866.04571428572,
      "avg_output_tokens_per_task": 9061.385714285714,
      "last_run_timestamp": "2026-09-17T02:10:06.482Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Parasail",
      "accuracy": 0.410000125,
      "accuracy_std_dev": 0.05535228340239791,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.007014935,
      "avg_duration_per_task_ms": 163231.8775,
      "avg_output_tokens_per_task": 4221.6925,
      "last_run_timestamp": "2026-10-24T19:16:50.878Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_slug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_name": "Qwen: Qwen3 VL 235B A22B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.4099998333333334,
      "accuracy_std_dev": 0.01953144663029912,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.06387897999999999,
      "avg_duration_per_task_ms": 359150.7866666667,
      "avg_output_tokens_per_task": 9115.776666666667,
      "last_run_timestamp": "2026-10-03T16:05:51.296Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "StreamLake",
      "accuracy": 0.4088235294117647,
      "accuracy_std_dev": 0.023166950988907642,
      "run_count": 7,
      "total_questions": 340,
      "avg_cost_per_task_usd": 0.011171617647058823,
      "avg_duration_per_task_ms": 111563.3294117647,
      "avg_output_tokens_per_task": 3213.9382352941175,
      "last_run_timestamp": "2026-10-03T16:17:21.424Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4.1-mini-2025-04-14",
      "model_slug": "openai/gpt-4.1-mini",
      "model_name": "OpenAI: GPT-4.1 Mini",
      "provider_name": "Azure",
      "accuracy": 0.40777783333333334,
      "accuracy_std_dev": 0.01781244190087991,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.025491406666666664,
      "avg_duration_per_task_ms": 73232.91333333333,
      "avg_output_tokens_per_task": 2860.7966666666666,
      "last_run_timestamp": "2026-10-01T19:05:00.133Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.406667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.06754938,
      "avg_duration_per_task_ms": 114529.78,
      "avg_output_tokens_per_task": 3789.7,
      "last_run_timestamp": "2026-10-04T03:00:16.402Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-chat-v3",
      "model_slug": "deepseek/deepseek-chat",
      "model_name": "DeepSeek: DeepSeek V3",
      "provider_name": "StreamLake",
      "accuracy": 0.4053599698492462,
      "accuracy_std_dev": 0.008101744842933528,
      "run_count": 4,
      "total_questions": 199,
      "avg_cost_per_task_usd": 0.06672347236180905,
      "avg_duration_per_task_ms": 212067.50753768845,
      "avg_output_tokens_per_task": 8080.718592964824,
      "last_run_timestamp": "2026-10-02T15:44:48.231Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "Novita",
      "accuracy": 0.40389273722627733,
      "accuracy_std_dev": 0.01580206672840261,
      "run_count": 3,
      "total_questions": 137,
      "avg_cost_per_task_usd": 0.008183423357664232,
      "avg_duration_per_task_ms": 89294.24087591241,
      "avg_output_tokens_per_task": 1500.5474452554745,
      "last_run_timestamp": "2026-10-03T16:15:51.302Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_slug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_name": "Qwen: Qwen3 VL 235B A22B Instruct",
      "provider_name": "Venice",
      "accuracy": 0.4033333333333333,
      "accuracy_std_dev": 0.032603090061458215,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.07132451,
      "avg_duration_per_task_ms": 337606.7366666667,
      "avg_output_tokens_per_task": 13788.606666666667,
      "last_run_timestamp": "2026-10-03T16:12:29.905Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "SiliconFlow",
      "accuracy": 0.4000001666666666,
      "accuracy_std_dev": 0.025819845942069878,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.027428286666666666,
      "avg_duration_per_task_ms": 327766.04,
      "avg_output_tokens_per_task": 4214.8533333333335,
      "last_run_timestamp": "2026-10-03T16:13:56.972Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4o-2024-05-13",
      "model_slug": "openai/gpt-4o-2024-05-13",
      "model_name": "OpenAI: GPT-4o (2024-05-13)",
      "provider_name": "auto-routing",
      "accuracy": 0.4,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 1.1315643,
      "avg_duration_per_task_ms": 71996.84,
      "avg_output_tokens_per_task": 3249.82,
      "last_run_timestamp": "2026-08-01T08:06:02.002Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Coder 30B A3B Instruct",
      "provider_name": "SiliconFlow",
      "accuracy": 0.4,
      "accuracy_std_dev": 0.019626323972325204,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.020620846666666668,
      "avg_duration_per_task_ms": 194089.08666666667,
      "avg_output_tokens_per_task": 3830.9066666666668,
      "last_run_timestamp": "2026-08-15T08:02:44.458Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_slug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_name": "Qwen: Qwen3 VL 235B A22B Instruct",
      "provider_name": "Novita",
      "accuracy": 0.3999998333333334,
      "accuracy_std_dev": 0.027755373188007314,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.07644987333333333,
      "avg_duration_per_task_ms": 118340.34,
      "avg_output_tokens_per_task": 3090.1866666666665,
      "last_run_timestamp": "2026-10-03T16:12:07.495Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Chutes",
      "accuracy": 0.3990239504504504,
      "accuracy_std_dev": 0.05792485239888057,
      "run_count": 9,
      "total_questions": 444,
      "avg_cost_per_task_usd": 0.004533889639639639,
      "avg_duration_per_task_ms": 307259.5382882883,
      "avg_output_tokens_per_task": 4442.5472972972975,
      "last_run_timestamp": "2026-10-24T19:54:38.220Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_slug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_name": "Qwen: Qwen3 VL 235B A22B Instruct",
      "provider_name": "Alibaba",
      "accuracy": 0.3983335,
      "accuracy_std_dev": 0.01960272562468461,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.062113629999999996,
      "avg_duration_per_task_ms": 98902.40333333334,
      "avg_output_tokens_per_task": 2914.5433333333335,
      "last_run_timestamp": "2026-10-03T16:12:33.500Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Alibaba",
      "accuracy": 0.39555566666666664,
      "accuracy_std_dev": 0.058203726918662366,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.05201909999999999,
      "avg_duration_per_task_ms": 91142.84333333334,
      "avg_output_tokens_per_task": 3810.85,
      "last_run_timestamp": "2026-10-03T16:05:51.296Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_slug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_name": "Qwen: Qwen3 VL 235B A22B Instruct",
      "provider_name": "Parasail",
      "accuracy": 0.39555533333333337,
      "accuracy_std_dev": 0.012570885432440924,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.03699347666666667,
      "avg_duration_per_task_ms": 190794.29666666666,
      "avg_output_tokens_per_task": 3701.5133333333333,
      "last_run_timestamp": "2026-10-03T16:08:50.061Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "auto-routing",
      "accuracy": 0.3948148888888889,
      "accuracy_std_dev": 0.1204017290208487,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.023600613333333336,
      "avg_duration_per_task_ms": 193026.64,
      "avg_output_tokens_per_task": 3368.5844444444447,
      "last_run_timestamp": "2026-10-24T19:36:27.834Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-chat-v3",
      "model_slug": "deepseek/deepseek-chat",
      "model_name": "DeepSeek: DeepSeek V3",
      "provider_name": "auto-routing",
      "accuracy": 0.3944446666666666,
      "accuracy_std_dev": 0.015947545189011245,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.07351260000000001,
      "avg_duration_per_task_ms": 297595.3333333333,
      "avg_output_tokens_per_task": 7950.953333333333,
      "last_run_timestamp": "2026-10-01T19:04:42.463Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "CoreWeave",
      "accuracy": 0.39210027167630057,
      "accuracy_std_dev": 0.08499299536756556,
      "run_count": 7,
      "total_questions": 346,
      "avg_cost_per_task_usd": 0.3060756098265896,
      "avg_duration_per_task_ms": 324526.6416184971,
      "avg_output_tokens_per_task": 12470.086705202311,
      "last_run_timestamp": "2026-10-10T18:31:45.008Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "auto-routing",
      "accuracy": 0.39,
      "accuracy_std_dev": 0.019148600183825446,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.025881283333333335,
      "avg_duration_per_task_ms": 240617.67666666667,
      "avg_output_tokens_per_task": 4326.496666666667,
      "last_run_timestamp": "2026-10-03T16:03:16.197Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "Nebius",
      "accuracy": 0.3893334,
      "accuracy_std_dev": 0.009043086986201126,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.014810912,
      "avg_duration_per_task_ms": 96843.704,
      "avg_output_tokens_per_task": 601.908,
      "last_run_timestamp": "2026-08-29T17:06:47.617Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_slug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_name": "Qwen: Qwen3 VL 235B A22B Instruct",
      "provider_name": "DeepInfra",
      "accuracy": 0.3868015260115607,
      "accuracy_std_dev": 0.026163281770060737,
      "run_count": 7,
      "total_questions": 346,
      "avg_cost_per_task_usd": 0.05894507803468208,
      "avg_duration_per_task_ms": 398637.9768786127,
      "avg_output_tokens_per_task": 3843.208092485549,
      "last_run_timestamp": "2026-10-08T23:21:09.489Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "Ambient",
      "accuracy": 0.386667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.014510460000000001,
      "avg_duration_per_task_ms": 64332.54,
      "avg_output_tokens_per_task": 4106.96,
      "last_run_timestamp": "2026-10-09T15:37:38.473Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 VL 30B A3B Instruct",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.386667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.0592903,
      "avg_duration_per_task_ms": 203536.16,
      "avg_output_tokens_per_task": 5170.96,
      "last_run_timestamp": "2026-10-04T03:00:16.402Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "CoreWeave",
      "accuracy": 0.384,
      "accuracy_std_dev": 0.006798758063058267,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.028780152000000003,
      "avg_duration_per_task_ms": 81353.024,
      "avg_output_tokens_per_task": 4284.144,
      "last_run_timestamp": "2026-08-29T16:13:12.633Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.3759873436018957,
      "accuracy_std_dev": 0.1400803380925799,
      "run_count": 17,
      "total_questions": 844,
      "avg_cost_per_task_usd": 0.03816363151658768,
      "avg_duration_per_task_ms": 580869.5746445497,
      "avg_output_tokens_per_task": 145902.23341232227,
      "last_run_timestamp": "2026-10-24T19:41:04.164Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "BaseTen",
      "accuracy": 0.3737958670520231,
      "accuracy_std_dev": 0.32322783310601233,
      "run_count": 7,
      "total_questions": 346,
      "avg_cost_per_task_usd": 0.012894294797687859,
      "avg_duration_per_task_ms": 78253.61560693642,
      "avg_output_tokens_per_task": 13434.153179190751,
      "last_run_timestamp": "2026-10-24T19:09:36.304Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "AkashML",
      "accuracy": 0.37205161949265686,
      "accuracy_std_dev": 0.021985652621046604,
      "run_count": 15,
      "total_questions": 749,
      "avg_cost_per_task_usd": 0.013301118825100131,
      "avg_duration_per_task_ms": 38508.31642189586,
      "avg_output_tokens_per_task": 742.2656875834446,
      "last_run_timestamp": "2026-10-24T18:51:12.161Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "Alibaba",
      "accuracy": 0.36999983333333336,
      "accuracy_std_dev": 0.032375053268510444,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.05485019333333334,
      "avg_duration_per_task_ms": 70677.64666666667,
      "avg_output_tokens_per_task": 4390.706666666667,
      "last_run_timestamp": "2026-10-03T16:05:34.166Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "Nebius",
      "accuracy": 0.3688886666666667,
      "accuracy_std_dev": 0.01791611454963257,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.03368235333333333,
      "avg_duration_per_task_ms": 150468.61333333334,
      "avg_output_tokens_per_task": 4460.053333333333,
      "last_run_timestamp": "2026-10-23T08:14:36.871Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "GMICloud",
      "accuracy": 0.36833325,
      "accuracy_std_dev": 0.17853565236861657,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.015193219999999999,
      "avg_duration_per_task_ms": 127682.09,
      "avg_output_tokens_per_task": 2917.505,
      "last_run_timestamp": "2026-10-03T16:12:29.879Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4o-2024-05-13",
      "model_slug": "openai/gpt-4o-2024-05-13",
      "model_name": "OpenAI: GPT-4o (2024-05-13)",
      "provider_name": "Azure",
      "accuracy": 0.36666699999999997,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.80181,
      "avg_duration_per_task_ms": 63430.68,
      "avg_output_tokens_per_task": 3064.6,
      "last_run_timestamp": "2026-08-01T08:06:07.633Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-vl-8b-instruct",
      "model_slug": "qwen/qwen3-vl-8b-instruct",
      "model_name": "Qwen: Qwen3 VL 8B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.363333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.03867836,
      "avg_duration_per_task_ms": 117421.08,
      "avg_output_tokens_per_task": 4043.36,
      "last_run_timestamp": "2026-08-01T17:38:36.457Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.36,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.041000339999999996,
      "avg_duration_per_task_ms": 322345.3,
      "avg_output_tokens_per_task": 4212.74,
      "last_run_timestamp": "2026-10-04T03:00:10.237Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 VL 30B A3B Instruct",
      "provider_name": "DeepInfra",
      "accuracy": 0.35752687096774194,
      "accuracy_std_dev": 0.02846485415894485,
      "run_count": 10,
      "total_questions": 496,
      "avg_cost_per_task_usd": 0.07167372177419354,
      "avg_duration_per_task_ms": 382176.28830645164,
      "avg_output_tokens_per_task": 7777.157258064516,
      "last_run_timestamp": "2026-10-08T23:34:52.306Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "CoreWeave",
      "accuracy": 0.3555554444444445,
      "accuracy_std_dev": 0.0331101789589147,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.08666491777777777,
      "avg_duration_per_task_ms": 26232.928888888888,
      "avg_output_tokens_per_task": 1004.2177777777778,
      "last_run_timestamp": "2026-10-24T18:19:37.955Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 VL 30B A3B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.3519345982142857,
      "accuracy_std_dev": 0.02099059885040837,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.07758646428571427,
      "avg_duration_per_task_ms": 205456.72098214287,
      "avg_output_tokens_per_task": 7279.627232142857,
      "last_run_timestamp": "2026-10-03T08:12:20.019Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-chat-v3",
      "model_slug": "deepseek/deepseek-chat",
      "model_name": "DeepSeek: DeepSeek V3",
      "provider_name": "Novita",
      "accuracy": 0.34058,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 46,
      "avg_cost_per_task_usd": 0.11353834782608695,
      "avg_duration_per_task_ms": 397172.5,
      "avg_output_tokens_per_task": 14067.434782608696,
      "last_run_timestamp": "2026-08-01T08:05:57.857Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.333333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.051354699999999996,
      "avg_duration_per_task_ms": 358714.54,
      "avg_output_tokens_per_task": 69285.14,
      "last_run_timestamp": "2026-10-04T03:00:10.240Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-vl-8b-instruct",
      "model_slug": "qwen/qwen3-vl-8b-instruct",
      "model_name": "Qwen: Qwen3 VL 8B Instruct",
      "provider_name": "Alibaba",
      "accuracy": 0.333333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.03045268,
      "avg_duration_per_task_ms": 77570.34,
      "avg_output_tokens_per_task": 3577.76,
      "last_run_timestamp": "2026-08-01T17:38:37.779Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-vl-8b-instruct",
      "model_slug": "qwen/qwen3-vl-8b-instruct",
      "model_name": "Qwen: Qwen3 VL 8B Instruct",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.326667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.02872976,
      "avg_duration_per_task_ms": 133695.86,
      "avg_output_tokens_per_task": 3408.56,
      "last_run_timestamp": "2026-10-04T03:00:16.402Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.32660695515695065,
      "accuracy_std_dev": 0.04329336143048513,
      "run_count": 9,
      "total_questions": 446,
      "avg_cost_per_task_usd": 0.04019819282511211,
      "avg_duration_per_task_ms": 95760.30493273542,
      "avg_output_tokens_per_task": 1304.5695067264573,
      "last_run_timestamp": "2026-10-24T18:56:17.284Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 VL 30B A3B Instruct",
      "provider_name": "Alibaba",
      "accuracy": 0.325926,
      "accuracy_std_dev": 0.02483419696484847,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.07731095111111112,
      "avg_duration_per_task_ms": 138830.34,
      "avg_output_tokens_per_task": 7809.953333333333,
      "last_run_timestamp": "2026-10-03T08:23:09.856Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.313333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.15186704,
      "avg_duration_per_task_ms": 32635.52,
      "avg_output_tokens_per_task": 1294.54,
      "last_run_timestamp": "2026-10-04T03:00:16.406Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "DeepInfra",
      "accuracy": 0.30874319672131145,
      "accuracy_std_dev": 0.03698276832472119,
      "run_count": 5,
      "total_questions": 244,
      "avg_cost_per_task_usd": 0.01523452049180328,
      "avg_duration_per_task_ms": 155396.16393442624,
      "avg_output_tokens_per_task": 1405.610655737705,
      "last_run_timestamp": "2026-08-29T16:43:42.449Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "Together",
      "accuracy": 0.29333325000000005,
      "accuracy_std_dev": 0.03543382721619978,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.1708829475,
      "avg_duration_per_task_ms": 140129.865,
      "avg_output_tokens_per_task": 1891.19,
      "last_run_timestamp": "2026-10-17T19:19:08.724Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4o-mini",
      "model_slug": "openai/gpt-4o-mini",
      "model_name": "OpenAI: GPT-4o-mini",
      "provider_name": "auto-routing",
      "accuracy": 0.2844443333333333,
      "accuracy_std_dev": 0.031894788459698065,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.02081186,
      "avg_duration_per_task_ms": 62184.12666666666,
      "avg_output_tokens_per_task": 3804.7533333333336,
      "last_run_timestamp": "2026-10-01T19:04:58.075Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4o-mini",
      "model_slug": "openai/gpt-4o-mini",
      "model_name": "OpenAI: GPT-4o-mini",
      "provider_name": "OpenAI",
      "accuracy": 0.28000025,
      "accuracy_std_dev": 0.027080189561513406,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.02535222,
      "avg_duration_per_task_ms": 67904.845,
      "avg_output_tokens_per_task": 4023.05,
      "last_run_timestamp": "2026-10-02T23:13:47.090Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen-2.5-72b-instruct",
      "model_slug": "qwen/qwen-2.5-72b-instruct",
      "model_name": "Qwen2.5 72B Instruct",
      "provider_name": "Novita",
      "accuracy": 0.2622223333333333,
      "accuracy_std_dev": 0.01911642045177101,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.08337879999999999,
      "avg_duration_per_task_ms": 576772.3666666667,
      "avg_output_tokens_per_task": 5220.566666666667,
      "last_run_timestamp": "2026-10-01T16:00:45.063Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 VL 30B A3B Instruct",
      "provider_name": "SiliconFlow",
      "accuracy": 0.25879402512562816,
      "accuracy_std_dev": 0.19158936601948973,
      "run_count": 8,
      "total_questions": 398,
      "avg_cost_per_task_usd": 0.1755232135678392,
      "avg_duration_per_task_ms": 1051691.3015075377,
      "avg_output_tokens_per_task": 6720.771356783919,
      "last_run_timestamp": "2026-10-03T08:09:23.857Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "CoreWeave",
      "accuracy": 0.2550507121212121,
      "accuracy_std_dev": 0.24463597214378424,
      "run_count": 8,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.010489681818181816,
      "avg_duration_per_task_ms": 1138281.7373737374,
      "avg_output_tokens_per_task": 70411.75252525252,
      "last_run_timestamp": "2026-10-24T19:33:50.345Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Nebius",
      "accuracy": 0.2513333,
      "accuracy_std_dev": 0.2375252245788013,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.079624326,
      "avg_duration_per_task_ms": 183949.264,
      "avg_output_tokens_per_task": 23168.86,
      "last_run_timestamp": "2026-10-24T19:07:57.340Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen-2.5-72b-instruct",
      "model_slug": "qwen/qwen-2.5-72b-instruct",
      "model_name": "Qwen2.5 72B Instruct",
      "provider_name": "DeepInfra",
      "accuracy": 0.25,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.0713485,
      "avg_duration_per_task_ms": 660516.14,
      "avg_output_tokens_per_task": 6662.1,
      "last_run_timestamp": "2026-09-01T12:59:14.290Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen-2.5-72b-instruct",
      "model_slug": "qwen/qwen-2.5-72b-instruct",
      "model_name": "Qwen2.5 72B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.24200919178082192,
      "accuracy_std_dev": 0.027381797668443095,
      "run_count": 3,
      "total_questions": 146,
      "avg_cost_per_task_usd": 0.07592458219178082,
      "avg_duration_per_task_ms": 576526.0342465753,
      "avg_output_tokens_per_task": 5262.068493150685,
      "last_run_timestamp": "2026-10-01T16:00:49.856Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-next-2025-02-03",
      "model_slug": "qwen/qwen3-coder-next",
      "model_name": "Qwen: Qwen3 Coder Next",
      "provider_name": "Novita",
      "accuracy": 0.22555566666666663,
      "accuracy_std_dev": 0.06975256621722879,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.01210011333333333,
      "avg_duration_per_task_ms": 20112.93,
      "avg_output_tokens_per_task": 1002.5833333333334,
      "last_run_timestamp": "2026-10-03T16:13:42.255Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta-llama/llama-3.1-8b-instruct",
      "model_slug": "meta-llama/llama-3.1-8b-instruct",
      "model_name": "Meta: Llama 3.1 8B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.22244883673469384,
      "accuracy_std_dev": 0.060049660254159636,
      "run_count": 5,
      "total_questions": 245,
      "avg_cost_per_task_usd": 0.004490742857142857,
      "avg_duration_per_task_ms": 16706.061224489797,
      "avg_output_tokens_per_task": 1551.4979591836734,
      "last_run_timestamp": "2026-10-24T19:30:38.210Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "Venice",
      "accuracy": 0.22242427272727275,
      "accuracy_std_dev": 0.29734520032174755,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.010851310909090908,
      "avg_duration_per_task_ms": 29224.176363636365,
      "avg_output_tokens_per_task": 5625.745454545455,
      "last_run_timestamp": "2026-10-17T19:48:24.252Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.1-terminus",
      "model_slug": "deepseek/deepseek-v3.1-terminus",
      "model_name": "DeepSeek: DeepSeek V3.1 Terminus",
      "provider_name": "AtlasCloud",
      "accuracy": 0.2213332,
      "accuracy_std_dev": 0.11472869795025131,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.017538988,
      "avg_duration_per_task_ms": 77819.968,
      "avg_output_tokens_per_task": 1398.036,
      "last_run_timestamp": "2026-08-29T16:17:08.780Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "AtlasCloud",
      "accuracy": 0.2066666,
      "accuracy_std_dev": 0.10724846748480837,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.19412482799999997,
      "avg_duration_per_task_ms": 581641.784,
      "avg_output_tokens_per_task": 11075.26,
      "last_run_timestamp": "2026-08-29T16:22:37.844Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "amazon/nova-micro-v1",
      "model_slug": "amazon/nova-micro-v1",
      "model_name": "Amazon: Nova Micro 1.0",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.18766967073170732,
      "accuracy_std_dev": 0.018396157170452745,
      "run_count": 5,
      "total_questions": 246,
      "avg_cost_per_task_usd": 0.011055426829268291,
      "avg_duration_per_task_ms": 138140.34552845528,
      "avg_output_tokens_per_task": 5390.59756097561,
      "last_run_timestamp": "2026-10-01T19:04:52.650Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-next-2025-02-03",
      "model_slug": "qwen/qwen3-coder-next",
      "model_name": "Qwen: Qwen3 Coder Next",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.186667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.01693508,
      "avg_duration_per_task_ms": 26941.4,
      "avg_output_tokens_per_task": 1041.58,
      "last_run_timestamp": "2026-10-04T03:00:16.405Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen-2.5-7b-instruct",
      "model_slug": "qwen/qwen-2.5-7b-instruct",
      "model_name": "Qwen: Qwen2.5 7B Instruct",
      "provider_name": "Phala",
      "accuracy": 0.166667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.01877496,
      "avg_duration_per_task_ms": 63137.36,
      "avg_output_tokens_per_task": 2797.32,
      "last_run_timestamp": "2026-08-01T08:06:16.826Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen-2.5-7b-instruct",
      "model_slug": "qwen/qwen-2.5-7b-instruct",
      "model_name": "Qwen: Qwen2.5 7B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.166667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.016978299999999998,
      "avg_duration_per_task_ms": 57790.52,
      "avg_output_tokens_per_task": 2492.08,
      "last_run_timestamp": "2026-08-01T08:06:07.057Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4.1-nano-2025-04-14",
      "model_slug": "openai/gpt-4.1-nano",
      "model_name": "OpenAI: GPT-4.1 Nano",
      "provider_name": "Azure",
      "accuracy": 0.13333333333333333,
      "accuracy_std_dev": 0.021430473285072877,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.007961373333333334,
      "avg_duration_per_task_ms": 80728.30333333333,
      "avg_output_tokens_per_task": 2469.17,
      "last_run_timestamp": "2026-10-01T19:04:58.093Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4.1-nano-2025-04-14",
      "model_slug": "openai/gpt-4.1-nano",
      "model_name": "OpenAI: GPT-4.1 Nano",
      "provider_name": "OpenAI",
      "accuracy": 0.11833325,
      "accuracy_std_dev": 0.018484159885899604,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.007158584999999999,
      "avg_duration_per_task_ms": 55526.67,
      "avg_output_tokens_per_task": 2353.445,
      "last_run_timestamp": "2026-10-02T23:07:59.112Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4.1-nano-2025-04-14",
      "model_slug": "openai/gpt-4.1-nano",
      "model_name": "OpenAI: GPT-4.1 Nano",
      "provider_name": "auto-routing",
      "accuracy": 0.108889,
      "accuracy_std_dev": 0.008314764498569193,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.006775626666666668,
      "avg_duration_per_task_ms": 60620.74,
      "avg_output_tokens_per_task": 2327.0066666666667,
      "last_run_timestamp": "2026-10-01T19:04:36.194Z"
    }
  ]
}
