{
  "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": "2026-10-18",
  "retrieved_at": "2026-10-18T01:45:34.625Z",
  "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": 2148,
  "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-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": "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": 0.9305554166666667,
      "accuracy_std_dev": 0.007749155464289131,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.026224210437710435,
      "avg_duration_per_task_ms": 18736.835016835015,
      "avg_output_tokens_per_task": 2232.077441077441,
      "last_run_timestamp": "2026-10-17T08:21:08.761Z"
    },
    {
      "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": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Parasail",
      "accuracy": 0.9295867436026091,
      "accuracy_std_dev": 0.01014563057765174,
      "run_count": 11,
      "total_questions": 1993,
      "avg_cost_per_task_usd": 0.12675964425489214,
      "avg_duration_per_task_ms": 135519.84947315603,
      "avg_output_tokens_per_task": 8382.747114902157,
      "last_run_timestamp": "2026-10-17T08:31:19.762Z"
    },
    {
      "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": 0.92929275,
      "accuracy_std_dev": 0.003571007097094602,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.024821165404040405,
      "avg_duration_per_task_ms": 19487.328282828283,
      "avg_output_tokens_per_task": 2241.7411616161617,
      "last_run_timestamp": "2026-10-17T08:18:07.776Z"
    },
    {
      "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": "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": 0.9280299999999999,
      "accuracy_std_dev": 0.0024899934738870384,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.025444652777777778,
      "avg_duration_per_task_ms": 19201.565656565657,
      "avg_output_tokens_per_task": 2305.0138888888887,
      "last_run_timestamp": "2026-10-17T08:03:37.220Z"
    },
    {
      "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": 0.9276092499999999,
      "accuracy_std_dev": 0.0031494064817835167,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.025043602272727273,
      "avg_duration_per_task_ms": 20424.821969696968,
      "avg_output_tokens_per_task": 2271.911616161616,
      "last_run_timestamp": "2026-10-17T08:18:07.015Z"
    },
    {
      "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": "Fireworks",
      "accuracy": 0.9264645740259738,
      "accuracy_std_dev": 0.008111437376199883,
      "run_count": 31,
      "total_questions": 5775,
      "avg_cost_per_task_usd": 0.17016577662337665,
      "avg_duration_per_task_ms": 157879.1478787879,
      "avg_output_tokens_per_task": 9136.811948051947,
      "last_run_timestamp": "2026-10-17T08:23:33.018Z"
    },
    {
      "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": 0.9262631428571431,
      "accuracy_std_dev": 0.007525054247550093,
      "run_count": 7,
      "total_questions": 1372,
      "avg_cost_per_task_usd": 0.11394943294460641,
      "avg_duration_per_task_ms": 190369.23032069972,
      "avg_output_tokens_per_task": 9499.49052478134,
      "last_run_timestamp": "2026-10-17T08:26:17.826Z"
    },
    {
      "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": 0.9262129285099052,
      "accuracy_std_dev": 0.013787528526068137,
      "run_count": 7,
      "total_questions": 1161,
      "avg_cost_per_task_usd": 0.10748236175710593,
      "avg_duration_per_task_ms": 226093.6391042205,
      "avg_output_tokens_per_task": 8296.273040482343,
      "last_run_timestamp": "2026-10-17T08:22:58.505Z"
    },
    {
      "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": "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": "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": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Decart",
      "accuracy": 0.924032,
      "accuracy_std_dev": 0.002938362554212806,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.11472872474747475,
      "avg_duration_per_task_ms": 192645.5883838384,
      "avg_output_tokens_per_task": 9486.564393939394,
      "last_run_timestamp": "2026-10-17T08:19:42.160Z"
    },
    {
      "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": 0.923681,
      "accuracy_std_dev": 0.0028613344206273234,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.025073777777777778,
      "avg_duration_per_task_ms": 20021.311447811448,
      "avg_output_tokens_per_task": 2270.7323232323233,
      "last_run_timestamp": "2026-10-17T08:04:11.182Z"
    },
    {
      "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": "Amazon Bedrock",
      "accuracy": 0.92312,
      "accuracy_std_dev": 0.0028613344206273234,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.1503178552188552,
      "avg_duration_per_task_ms": 99928.72895622895,
      "avg_output_tokens_per_task": 9768,
      "last_run_timestamp": "2026-10-17T08:20:36.823Z"
    },
    {
      "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": "Modal",
      "accuracy": 0.9214533832285114,
      "accuracy_std_dev": 0.007123977352126375,
      "run_count": 13,
      "total_questions": 2385,
      "avg_cost_per_task_usd": 0.14641879832285115,
      "avg_duration_per_task_ms": 129257.1857442348,
      "avg_output_tokens_per_task": 9613.907337526205,
      "last_run_timestamp": "2026-10-17T08:23:01.607Z"
    },
    {
      "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": 0.9212961249999999,
      "accuracy_std_dev": 0.006614649375392113,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.1696066616161616,
      "avg_duration_per_task_ms": 303586.860479798,
      "avg_output_tokens_per_task": 9644.771464646465,
      "last_run_timestamp": "2026-10-17T08:02:36.487Z"
    },
    {
      "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": "Sail Research",
      "accuracy": 0.9200982307692309,
      "accuracy_std_dev": 0.007282007290405241,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.14751564180264182,
      "avg_duration_per_task_ms": 191954.80303030304,
      "avg_output_tokens_per_task": 11844.022144522145,
      "last_run_timestamp": "2026-10-17T08:26:18.894Z"
    },
    {
      "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": 0.9196950955414013,
      "accuracy_std_dev": 0.03543852389220188,
      "run_count": 14,
      "total_questions": 2669,
      "avg_cost_per_task_usd": 0.12421539677781943,
      "avg_duration_per_task_ms": 189674.27538403898,
      "avg_output_tokens_per_task": 9508.503559385537,
      "last_run_timestamp": "2026-10-17T08:31:19.880Z"
    },
    {
      "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": 0.9194757752808989,
      "accuracy_std_dev": 0.005468000000000028,
      "run_count": 2,
      "total_questions": 356,
      "avg_cost_per_task_usd": 0.12247723595505616,
      "avg_duration_per_task_ms": 247597.07584269662,
      "avg_output_tokens_per_task": 32087.120786516854,
      "last_run_timestamp": "2026-10-17T08:23:02.364Z"
    },
    {
      "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": 0.9193136803377565,
      "accuracy_std_dev": 0.012748055372296913,
      "run_count": 13,
      "total_questions": 2487,
      "avg_cost_per_task_usd": 0.1257693297145155,
      "avg_duration_per_task_ms": 119458.23924406916,
      "avg_output_tokens_per_task": 8396.139123441897,
      "last_run_timestamp": "2026-10-17T08:19:33.656Z"
    },
    {
      "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": 0.9192753646408839,
      "accuracy_std_dev": 0.01148102022995612,
      "run_count": 11,
      "total_questions": 2172,
      "avg_cost_per_task_usd": 0.13229984714548804,
      "avg_duration_per_task_ms": 285821.6473296501,
      "avg_output_tokens_per_task": 10235.589779005524,
      "last_run_timestamp": "2026-10-17T08:02:36.128Z"
    },
    {
      "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": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Wafer",
      "accuracy": 0.9189512857142857,
      "accuracy_std_dev": 0.006699457328007703,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.08570814935064935,
      "avg_duration_per_task_ms": 308581.3556998557,
      "avg_output_tokens_per_task": 24794.714285714286,
      "last_run_timestamp": "2026-10-17T08:18:07.429Z"
    },
    {
      "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": 0.9187128800685322,
      "accuracy_std_dev": 0.014979707918825845,
      "run_count": 9,
      "total_questions": 1751,
      "avg_cost_per_task_usd": 0.24826981896059397,
      "avg_duration_per_task_ms": 409248.07595659624,
      "avg_output_tokens_per_task": 36709.836664762995,
      "last_run_timestamp": "2026-10-17T08:26:18.450Z"
    },
    {
      "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": 0.9166664999999998,
      "accuracy_std_dev": 0.003669332602803965,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.025965300505050505,
      "avg_duration_per_task_ms": 19379.844065656565,
      "avg_output_tokens_per_task": 2237.6660353535353,
      "last_run_timestamp": "2026-10-17T08:19:39.352Z"
    },
    {
      "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": 0.9159637816307403,
      "accuracy_std_dev": 0.011684355360373981,
      "run_count": 8,
      "total_questions": 1067,
      "avg_cost_per_task_usd": 0.11537374320524835,
      "avg_duration_per_task_ms": 187895.38612933457,
      "avg_output_tokens_per_task": 7839.027179006561,
      "last_run_timestamp": "2026-10-17T08:04:21.303Z"
    },
    {
      "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": 0.9157357830756013,
      "accuracy_std_dev": 0.007025144934744547,
      "run_count": 12,
      "total_questions": 2328,
      "avg_cost_per_task_usd": 0.14151731056701028,
      "avg_duration_per_task_ms": 289404.89347079035,
      "avg_output_tokens_per_task": 9270.16881443299,
      "last_run_timestamp": "2026-10-17T08:25:47.682Z"
    },
    {
      "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": 0.915614375,
      "accuracy_std_dev": 0.0073109223928568175,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.10001511047979798,
      "avg_duration_per_task_ms": 653810.1868686868,
      "avg_output_tokens_per_task": 40376.643308080806,
      "last_run_timestamp": "2026-10-17T08:21:09.717Z"
    },
    {
      "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": 0.9151233333333333,
      "accuracy_std_dev": 0.012774219763083605,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.03489141035353535,
      "avg_duration_per_task_ms": 231436.15993265994,
      "avg_output_tokens_per_task": 28899.603535353537,
      "last_run_timestamp": "2026-10-17T16:03:42.289Z"
    },
    {
      "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": 0.9132906754665453,
      "accuracy_std_dev": 0.029785895483567967,
      "run_count": 23,
      "total_questions": 4394,
      "avg_cost_per_task_usd": 0.14536340896677283,
      "avg_duration_per_task_ms": 401099.01661356393,
      "avg_output_tokens_per_task": 8988.59285389167,
      "last_run_timestamp": "2026-10-18T00:58:02.224Z"
    },
    {
      "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": "AtlasCloud",
      "accuracy": 0.9121771666666666,
      "accuracy_std_dev": 0.008877109240750762,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.15960165235690235,
      "avg_duration_per_task_ms": 534122.271885522,
      "avg_output_tokens_per_task": 39986.65572390572,
      "last_run_timestamp": "2026-10-17T08:20:04.684Z"
    },
    {
      "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": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Chutes",
      "accuracy": 0.910344342048864,
      "accuracy_std_dev": 0.014501798751218183,
      "run_count": 12,
      "total_questions": 2333,
      "avg_cost_per_task_usd": 0.17792582340334334,
      "avg_duration_per_task_ms": 507774.8448349764,
      "avg_output_tokens_per_task": 12020.794256322331,
      "last_run_timestamp": "2026-10-17T08:19:39.445Z"
    },
    {
      "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": 0.9102327931985295,
      "accuracy_std_dev": 0.004925353358133757,
      "run_count": 11,
      "total_questions": 2176,
      "avg_cost_per_task_usd": 0.03013682904411765,
      "avg_duration_per_task_ms": 430481.5505514706,
      "avg_output_tokens_per_task": 31169.346966911766,
      "last_run_timestamp": "2026-10-17T16:03:21.036Z"
    },
    {
      "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": "Alibaba",
      "accuracy": 0.9093497692307693,
      "accuracy_std_dev": 0.015665798833784123,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.1382119296814297,
      "avg_duration_per_task_ms": 558863.4650349651,
      "avg_output_tokens_per_task": 40145.44832944833,
      "last_run_timestamp": "2026-10-17T08:21:36.009Z"
    },
    {
      "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": "GMICloud",
      "accuracy": 0.9083225551330798,
      "accuracy_std_dev": 0.025262661104000026,
      "run_count": 12,
      "total_questions": 2367,
      "avg_cost_per_task_usd": 0.037087138994507815,
      "avg_duration_per_task_ms": 242369.35530207012,
      "avg_output_tokens_per_task": 30868.7237008872,
      "last_run_timestamp": "2026-10-17T16:03:14.482Z"
    },
    {
      "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": 0.9073230999999999,
      "accuracy_std_dev": 0.016488125499583027,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.03575355858585859,
      "avg_duration_per_task_ms": 312749.41262626264,
      "avg_output_tokens_per_task": 29726.523737373736,
      "last_run_timestamp": "2026-10-17T16:07:08.883Z"
    },
    {
      "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": 0.907143,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 140,
      "avg_cost_per_task_usd": 0.012309828571428571,
      "avg_duration_per_task_ms": 26262.2,
      "avg_output_tokens_per_task": 10166.814285714287,
      "last_run_timestamp": "2026-10-17T21:00:07.590Z"
    },
    {
      "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": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "StreamLake",
      "accuracy": 0.9069864999999999,
      "accuracy_std_dev": 0.027094409526382632,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.04235188005050505,
      "avg_duration_per_task_ms": 291380.1031144781,
      "avg_output_tokens_per_task": 32881.04797979798,
      "last_run_timestamp": "2026-10-17T16:05:02.601Z"
    },
    {
      "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": 0.9067949999999999,
      "accuracy_std_dev": 0.021859135265271935,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.03907181726354454,
      "avg_duration_per_task_ms": 267831.84113865934,
      "avg_output_tokens_per_task": 32752.338842975205,
      "last_run_timestamp": "2026-10-17T16:03:30.556Z"
    },
    {
      "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": 0.9067522535599448,
      "accuracy_std_dev": 0.05130336735714156,
      "run_count": 11,
      "total_questions": 2177,
      "avg_cost_per_task_usd": 0.1400334451079467,
      "avg_duration_per_task_ms": 309819.79834634817,
      "avg_output_tokens_per_task": 10155.972439136427,
      "last_run_timestamp": "2026-10-17T08:18:04.827Z"
    },
    {
      "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": 0.9063747011952191,
      "accuracy_std_dev": 0.0020234348684023858,
      "run_count": 3,
      "total_questions": 502,
      "avg_cost_per_task_usd": 0.11477266334661354,
      "avg_duration_per_task_ms": 440554.38844621513,
      "avg_output_tokens_per_task": 40023.181274900395,
      "last_run_timestamp": "2026-10-17T08:23:47.110Z"
    },
    {
      "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": 0.9062234806560137,
      "accuracy_std_dev": 0.022290388983096536,
      "run_count": 13,
      "total_questions": 2378,
      "avg_cost_per_task_usd": 0.29330102901597976,
      "avg_duration_per_task_ms": 441358.91000841046,
      "avg_output_tokens_per_task": 50972.094617325485,
      "last_run_timestamp": "2026-10-17T16:06:14.668Z"
    },
    {
      "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": 0.9060298181818183,
      "accuracy_std_dev": 0.02266404860253019,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.15440814876033057,
      "avg_duration_per_task_ms": 587765.4531680441,
      "avg_output_tokens_per_task": 38971.41505968779,
      "last_run_timestamp": "2026-10-17T08:23:03.040Z"
    },
    {
      "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": 0.905724,
      "accuracy_std_dev": 0.006734000000000018,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.008862866161616162,
      "avg_duration_per_task_ms": 144443.7373737374,
      "avg_output_tokens_per_task": 15989.888888888889,
      "last_run_timestamp": "2026-10-18T01:07:53.160Z"
    },
    {
      "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": 0.905724,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.06914223232323233,
      "avg_duration_per_task_ms": 293387.91414141416,
      "avg_output_tokens_per_task": 16031.914141414141,
      "last_run_timestamp": "2026-10-17T21:00:07.591Z"
    },
    {
      "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": 0.9054832857142857,
      "accuracy_std_dev": 0.008882188721807021,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.021842018037518035,
      "avg_duration_per_task_ms": 208032.45382395384,
      "avg_output_tokens_per_task": 29051.58008658009,
      "last_run_timestamp": "2026-10-17T08:22:58.205Z"
    },
    {
      "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": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "DeepInfra",
      "accuracy": 0.9052889151291513,
      "accuracy_std_dev": 0.017795645091488572,
      "run_count": 3,
      "total_questions": 542,
      "avg_cost_per_task_usd": 0.12811947416974168,
      "avg_duration_per_task_ms": 1694586.1826568267,
      "avg_output_tokens_per_task": 51002.60147601476,
      "last_run_timestamp": "2026-10-17T08:18:07.328Z"
    },
    {
      "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": 0.9049022275600507,
      "accuracy_std_dev": 0.03924926725076816,
      "run_count": 12,
      "total_questions": 2373,
      "avg_cost_per_task_usd": 0.14131352844500636,
      "avg_duration_per_task_ms": 216359.54614412136,
      "avg_output_tokens_per_task": 9500.680994521703,
      "last_run_timestamp": "2026-10-17T08:18:07.439Z"
    },
    {
      "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": 0.9048820000000001,
      "accuracy_std_dev": 0.0025253333443329903,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.03441120328282828,
      "avg_duration_per_task_ms": 137911.47095959596,
      "avg_output_tokens_per_task": 28491.76767676768,
      "last_run_timestamp": "2026-10-17T08:04:21.355Z"
    },
    {
      "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": 0.904882,
      "accuracy_std_dev": 0.00840358433447696,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.03428501851851851,
      "avg_duration_per_task_ms": 234623.59848484848,
      "avg_output_tokens_per_task": 28392.512626262625,
      "last_run_timestamp": "2026-10-17T08:18:06.503Z"
    },
    {
      "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": 0.9047746274509804,
      "accuracy_std_dev": 0.022192155496030565,
      "run_count": 10,
      "total_questions": 1836,
      "avg_cost_per_task_usd": 0.15225744008714598,
      "avg_duration_per_task_ms": 561798.8300653595,
      "avg_output_tokens_per_task": 37859.14270152505,
      "last_run_timestamp": "2026-10-17T08:25:53.287Z"
    },
    {
      "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": 0.9046852723932015,
      "accuracy_std_dev": 0.030917632257017453,
      "run_count": 11,
      "total_questions": 2177,
      "avg_cost_per_task_usd": 0.10195215893431328,
      "avg_duration_per_task_ms": 567988.1175930179,
      "avg_output_tokens_per_task": 39285.040881947636,
      "last_run_timestamp": "2026-10-17T08:19:16.608Z"
    },
    {
      "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": 0.9046213388116308,
      "accuracy_std_dev": 0.01859967095960451,
      "run_count": 12,
      "total_questions": 2373,
      "avg_cost_per_task_usd": 0.03434321828908554,
      "avg_duration_per_task_ms": 322804.26085124316,
      "avg_output_tokens_per_task": 30148.002528445006,
      "last_run_timestamp": "2026-10-17T16:03:16.411Z"
    },
    {
      "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": 0.90446125,
      "accuracy_std_dev": 0.004353383590668304,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.05431423611111111,
      "avg_duration_per_task_ms": 32466.23232323232,
      "avg_output_tokens_per_task": 2506.719696969697,
      "last_run_timestamp": "2026-10-17T08:18:07.104Z"
    },
    {
      "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": 0.9043450497466605,
      "accuracy_std_dev": 0.011580572970737592,
      "run_count": 11,
      "total_questions": 2171,
      "avg_cost_per_task_usd": 0.027795939198526028,
      "avg_duration_per_task_ms": 304566.4965453708,
      "avg_output_tokens_per_task": 23011.022570244128,
      "last_run_timestamp": "2026-10-17T16:03:30.490Z"
    },
    {
      "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": 0.9042808571428572,
      "accuracy_std_dev": 0.009792912093208319,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.03522179004329004,
      "avg_duration_per_task_ms": 198105.57215007214,
      "avg_output_tokens_per_task": 29184.305916305915,
      "last_run_timestamp": "2026-10-17T08:21:10.226Z"
    },
    {
      "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": 0.9040400000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.012735267676767678,
      "avg_duration_per_task_ms": 67216.78787878787,
      "avg_output_tokens_per_task": 20958.237373737375,
      "last_run_timestamp": "2026-10-18T00:00:02.679Z"
    },
    {
      "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": 0.9038162150401354,
      "accuracy_std_dev": 0.012954254555025118,
      "run_count": 12,
      "total_questions": 2367,
      "avg_cost_per_task_usd": 0.026641588508660753,
      "avg_duration_per_task_ms": 188585.9387410224,
      "avg_output_tokens_per_task": 25722.437262357413,
      "last_run_timestamp": "2026-10-17T08:18:07.017Z"
    },
    {
      "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": 0.9030148159392788,
      "accuracy_std_dev": 0.008815415925383199,
      "run_count": 8,
      "total_questions": 1581,
      "avg_cost_per_task_usd": 0.01550765211891208,
      "avg_duration_per_task_ms": 411171.4794433903,
      "avg_output_tokens_per_task": 28853.977229601518,
      "last_run_timestamp": "2026-10-17T08:30:38.610Z"
    },
    {
      "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": 0.9028881689530687,
      "accuracy_std_dev": 0.007076458182464086,
      "run_count": 7,
      "total_questions": 1385,
      "avg_cost_per_task_usd": 0.029951561732851982,
      "avg_duration_per_task_ms": 161666.90108303248,
      "avg_output_tokens_per_task": 28961.749458483755,
      "last_run_timestamp": "2026-10-17T08:19:16.690Z"
    },
    {
      "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": 0.9027143672230654,
      "accuracy_std_dev": 0.017487582904735587,
      "run_count": 10,
      "total_questions": 1977,
      "avg_cost_per_task_usd": 0.11330698988366211,
      "avg_duration_per_task_ms": 583845.9398077896,
      "avg_output_tokens_per_task": 36138.885179565,
      "last_run_timestamp": "2026-10-17T08:18:06.586Z"
    },
    {
      "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": 0.9023565,
      "accuracy_std_dev": 0.005050499999999958,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.11881703282828283,
      "avg_duration_per_task_ms": 830086.2297979798,
      "avg_output_tokens_per_task": 47333.19696969697,
      "last_run_timestamp": "2026-10-17T08:18:06.299Z"
    },
    {
      "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": 0.9023565,
      "accuracy_std_dev": 0.0037644204401208687,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.03250985984848485,
      "avg_duration_per_task_ms": 841972.3194444445,
      "avg_output_tokens_per_task": 37015.48358585859,
      "last_run_timestamp": "2026-10-17T08:18:07.239Z"
    },
    {
      "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": 0.9020199999999999,
      "accuracy_std_dev": 0.0038974287934483223,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.053951265656565654,
      "avg_duration_per_task_ms": 27050.535353535353,
      "avg_output_tokens_per_task": 2481.211111111111,
      "last_run_timestamp": "2026-10-17T08:20:30.948Z"
    },
    {
      "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": 0.9017853214285716,
      "accuracy_std_dev": 0.00452549999999996,
      "run_count": 2,
      "total_questions": 392,
      "avg_cost_per_task_usd": 0.03336421428571429,
      "avg_duration_per_task_ms": 106866.74489795919,
      "avg_output_tokens_per_task": 27637.576530612245,
      "last_run_timestamp": "2026-10-17T08:22:12.182Z"
    },
    {
      "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": 0.901208909090909,
      "accuracy_std_dev": 0.005840261800709478,
      "run_count": 22,
      "total_questions": 4356,
      "avg_cost_per_task_usd": 0.20488930509641878,
      "avg_duration_per_task_ms": 97845.36914600551,
      "avg_output_tokens_per_task": 7828.393709825528,
      "last_run_timestamp": "2026-10-17T08:02:02.330Z"
    },
    {
      "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": 0.901203,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 194,
      "avg_cost_per_task_usd": 0.01318992268041237,
      "avg_duration_per_task_ms": 207259.9175257732,
      "avg_output_tokens_per_task": 21902.201030927834,
      "last_run_timestamp": "2026-10-17T15:00:06.591Z"
    },
    {
      "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": 0.900973165044435,
      "accuracy_std_dev": 0.019246299929394788,
      "run_count": 12,
      "total_questions": 2363,
      "avg_cost_per_task_usd": 0.11851335124841302,
      "avg_duration_per_task_ms": 360984.45366060094,
      "avg_output_tokens_per_task": 8109.957257723233,
      "last_run_timestamp": "2026-10-17T08:21:12.247Z"
    },
    {
      "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": 0.9008713529411765,
      "accuracy_std_dev": 0.005828517351993652,
      "run_count": 34,
      "total_questions": 6732,
      "avg_cost_per_task_usd": 0.21723382456922163,
      "avg_duration_per_task_ms": 94455.50430778372,
      "avg_output_tokens_per_task": 7921.044711824124,
      "last_run_timestamp": "2026-10-17T08:18:12.401Z"
    },
    {
      "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": 0.9000179188034189,
      "accuracy_std_dev": 0.00894793549668819,
      "run_count": 11,
      "total_questions": 1872,
      "avg_cost_per_task_usd": 0.02389172542735043,
      "avg_duration_per_task_ms": 163758.47329059828,
      "avg_output_tokens_per_task": 19761.337606837606,
      "last_run_timestamp": "2026-10-17T16:03:32.458Z"
    },
    {
      "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": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Wafer",
      "accuracy": 0.8999134797927462,
      "accuracy_std_dev": 0.016737025742048665,
      "run_count": 10,
      "total_questions": 1930,
      "avg_cost_per_task_usd": 0.014986389637305699,
      "avg_duration_per_task_ms": 258557.6274611399,
      "avg_output_tokens_per_task": 18920.399481865286,
      "last_run_timestamp": "2026-10-17T08:26:18.134Z"
    },
    {
      "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": 0.8998876666666668,
      "accuracy_std_dev": 0.004716533340871803,
      "run_count": 15,
      "total_questions": 2970,
      "avg_cost_per_task_usd": 0.05738907373737373,
      "avg_duration_per_task_ms": 25168.736026936025,
      "avg_output_tokens_per_task": 2481.9760942760945,
      "last_run_timestamp": "2026-10-17T08:25:41.793Z"
    },
    {
      "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": "anthropic/claude-4.8-opus-20260528",
      "model_slug": "anthropic/claude-opus-4.8",
      "model_name": "Anthropic: Claude Opus 4.8",
      "provider_name": "Google",
      "accuracy": 0.8995188857142861,
      "accuracy_std_dev": 0.006231455637198027,
      "run_count": 35,
      "total_questions": 6930,
      "avg_cost_per_task_usd": 0.2226963189033189,
      "avg_duration_per_task_ms": 94364.85238095238,
      "avg_output_tokens_per_task": 8111.008658008658,
      "last_run_timestamp": "2026-10-17T08:26:18.152Z"
    },
    {
      "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": 0.8995076050632912,
      "accuracy_std_dev": 0.0054905559391406865,
      "run_count": 12,
      "total_questions": 2370,
      "avg_cost_per_task_usd": 0.013410476793248943,
      "avg_duration_per_task_ms": 179927.08818565402,
      "avg_output_tokens_per_task": 26352.98523206751,
      "last_run_timestamp": "2026-10-17T08:18:05.090Z"
    },
    {
      "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": "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": "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": 0.8992142666666664,
      "accuracy_std_dev": 0.004384375584073756,
      "run_count": 15,
      "total_questions": 2970,
      "avg_cost_per_task_usd": 0.05852382592592592,
      "avg_duration_per_task_ms": 39927.462626262626,
      "avg_output_tokens_per_task": 2529.2070707070707,
      "last_run_timestamp": "2026-10-17T08:18:06.772Z"
    },
    {
      "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": "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": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "Xiaomi",
      "accuracy": 0.8989897499999999,
      "accuracy_std_dev": 0.0020616538718950853,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.03381236616161616,
      "avg_duration_per_task_ms": 869027.9406565657,
      "avg_output_tokens_per_task": 38502.118686868685,
      "last_run_timestamp": "2026-10-17T08:20:32.672Z"
    },
    {
      "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": 0.8988212999999999,
      "accuracy_std_dev": 0.004847094842274054,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.22962485303030306,
      "avg_duration_per_task_ms": 110453.92777777778,
      "avg_output_tokens_per_task": 8973.717676767677,
      "last_run_timestamp": "2026-10-17T08:21:05.846Z"
    },
    {
      "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": 0.898697092413793,
      "accuracy_std_dev": 0.020363329103607237,
      "run_count": 11,
      "total_questions": 2175,
      "avg_cost_per_task_usd": 0.12840642850574713,
      "avg_duration_per_task_ms": 515985.4285057471,
      "avg_output_tokens_per_task": 38420.705747126434,
      "last_run_timestamp": "2026-10-17T08:18:00.494Z"
    },
    {
      "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": 0.8986529999999999,
      "accuracy_std_dev": 0.006334877031166399,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.057629573232323233,
      "avg_duration_per_task_ms": 25925.42575757576,
      "avg_output_tokens_per_task": 2526.4747474747473,
      "last_run_timestamp": "2026-10-17T08:19:36.998Z"
    },
    {
      "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": 0.8985086428571428,
      "accuracy_std_dev": 0.02010583105045876,
      "run_count": 14,
      "total_questions": 2772,
      "avg_cost_per_task_usd": 0.08497059956709957,
      "avg_duration_per_task_ms": 213919.4072871573,
      "avg_output_tokens_per_task": 29834.7113997114,
      "last_run_timestamp": "2026-10-17T16:03:45.988Z"
    },
    {
      "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": "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": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8977269999999999,
      "accuracy_std_dev": 0.008938227620731064,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.12509009848484848,
      "avg_duration_per_task_ms": 997729.1325757576,
      "avg_output_tokens_per_task": 49776.08964646464,
      "last_run_timestamp": "2026-10-17T08:18:07.879Z"
    },
    {
      "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": 0.8974747000000001,
      "accuracy_std_dev": 0.008159314922835103,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.19849825303030305,
      "avg_duration_per_task_ms": 95555.76262626263,
      "avg_output_tokens_per_task": 7731.669191919192,
      "last_run_timestamp": "2026-10-17T08:22:57.373Z"
    },
    {
      "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": 0.8973243579175705,
      "accuracy_std_dev": 0.006747288047801109,
      "run_count": 7,
      "total_questions": 1383,
      "avg_cost_per_task_usd": 0.018802611713665943,
      "avg_duration_per_task_ms": 445979.23065798986,
      "avg_output_tokens_per_task": 20953.845986984816,
      "last_run_timestamp": "2026-10-17T08:23:01.486Z"
    },
    {
      "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": 0.8973063333333335,
      "accuracy_std_dev": 0.007142249894504881,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.11625130303030304,
      "avg_duration_per_task_ms": 577263.1498316498,
      "avg_output_tokens_per_task": 46324.27272727273,
      "last_run_timestamp": "2026-10-17T08:04:10.998Z"
    },
    {
      "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": 0.897306,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.010453515151515153,
      "avg_duration_per_task_ms": 272743.00505050505,
      "avg_output_tokens_per_task": 20780.954545454544,
      "last_run_timestamp": "2026-10-17T15:00:06.591Z"
    },
    {
      "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": 0.8972450554493306,
      "accuracy_std_dev": 0.007483975735232924,
      "run_count": 24,
      "total_questions": 4707,
      "avg_cost_per_task_usd": 0.12556499532611007,
      "avg_duration_per_task_ms": 57794.63182494158,
      "avg_output_tokens_per_task": 4708.337582324198,
      "last_run_timestamp": "2026-10-17T08:20:12.208Z"
    },
    {
      "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": 0.896873431174089,
      "accuracy_std_dev": 0.016764191472169333,
      "run_count": 15,
      "total_questions": 2964,
      "avg_cost_per_task_usd": 0.023734335695006745,
      "avg_duration_per_task_ms": 240258.34311740892,
      "avg_output_tokens_per_task": 26795.730094466937,
      "last_run_timestamp": "2026-10-17T08:21:29.877Z"
    },
    {
      "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": "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": 0.896387909090909,
      "accuracy_std_dev": 0.011448796653356942,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.20013578328741968,
      "avg_duration_per_task_ms": 96239.94214876032,
      "avg_output_tokens_per_task": 7787.9586776859505,
      "last_run_timestamp": "2026-10-17T08:18:06.121Z"
    },
    {
      "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": "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": "anthropic/claude-4.8-opus-20260528",
      "model_slug": "anthropic/claude-opus-4.8",
      "model_name": "Anthropic: Claude Opus 4.8",
      "provider_name": "Anthropic",
      "accuracy": 0.8959834999999999,
      "accuracy_std_dev": 0.006828979737120304,
      "run_count": 14,
      "total_questions": 2772,
      "avg_cost_per_task_usd": 0.2064866551226551,
      "avg_duration_per_task_ms": 100409.85894660895,
      "avg_output_tokens_per_task": 8045.632034632034,
      "last_run_timestamp": "2026-10-17T08:13:13.631Z"
    },
    {
      "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": 0.8957629166666667,
      "accuracy_std_dev": 0.0073529979425439275,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.12094736405723905,
      "avg_duration_per_task_ms": 58110.515993265995,
      "avg_output_tokens_per_task": 4618.655303030303,
      "last_run_timestamp": "2026-10-17T08:18:57.786Z"
    },
    {
      "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": 0.8952784126849893,
      "accuracy_std_dev": 0.033903762384218605,
      "run_count": 12,
      "total_questions": 2365,
      "avg_cost_per_task_usd": 0.14386759830866808,
      "avg_duration_per_task_ms": 806266.6080338267,
      "avg_output_tokens_per_task": 36028.42875264271,
      "last_run_timestamp": "2026-10-17T08:18:07.271Z"
    },
    {
      "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": 0.8951084117647057,
      "accuracy_std_dev": 0.010289170146741347,
      "run_count": 16,
      "total_questions": 3162,
      "avg_cost_per_task_usd": 0.013490567046173307,
      "avg_duration_per_task_ms": 308688.6669829222,
      "avg_output_tokens_per_task": 27618.48513598988,
      "last_run_timestamp": "2026-10-17T08:20:07.047Z"
    },
    {
      "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": 0.8951083888888889,
      "accuracy_std_dev": 0.007379038657943986,
      "run_count": 36,
      "total_questions": 7128,
      "avg_cost_per_task_usd": 0.20431740810886648,
      "avg_duration_per_task_ms": 70463.12612233446,
      "avg_output_tokens_per_task": 7440.589225589226,
      "last_run_timestamp": "2026-10-17T08:23:14.990Z"
    },
    {
      "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": 0.8950693704171935,
      "accuracy_std_dev": 0.008974559925376093,
      "run_count": 12,
      "total_questions": 2373,
      "avg_cost_per_task_usd": 0.22676957437842393,
      "avg_duration_per_task_ms": 108718.53687315634,
      "avg_output_tokens_per_task": 8847.257901390645,
      "last_run_timestamp": "2026-10-17T08:04:11.021Z"
    },
    {
      "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": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Cohere",
      "accuracy": 0.8949492,
      "accuracy_std_dev": 0.011447658824405965,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.008737155555555556,
      "avg_duration_per_task_ms": 172419.46262626263,
      "avg_output_tokens_per_task": 30724.964646464647,
      "last_run_timestamp": "2026-10-17T08:26:18.166Z"
    },
    {
      "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": 0.8942760000000001,
      "accuracy_std_dev": 0.004441544821343102,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.05451896666666667,
      "avg_duration_per_task_ms": 31770.72121212121,
      "avg_output_tokens_per_task": 2522.4050505050504,
      "last_run_timestamp": "2026-10-17T08:20:03.210Z"
    },
    {
      "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": 0.8942210023995201,
      "accuracy_std_dev": 0.013642788754376415,
      "run_count": 17,
      "total_questions": 3334,
      "avg_cost_per_task_usd": 0.003915166466706659,
      "avg_duration_per_task_ms": 322301.0131973605,
      "avg_output_tokens_per_task": 21462.84853029394,
      "last_run_timestamp": "2026-10-17T08:21:32.578Z"
    },
    {
      "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": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "Novita",
      "accuracy": 0.89393925,
      "accuracy_std_dev": 0.007241007297848803,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.1140800441919192,
      "avg_duration_per_task_ms": 871452.9570707071,
      "avg_output_tokens_per_task": 45451.416666666664,
      "last_run_timestamp": "2026-10-17T08:18:04.691Z"
    },
    {
      "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": "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": 0.892375117408907,
      "accuracy_std_dev": 0.01166327801187222,
      "run_count": 15,
      "total_questions": 2964,
      "avg_cost_per_task_usd": 0.20794853846153846,
      "avg_duration_per_task_ms": 100204.1423751687,
      "avg_output_tokens_per_task": 8089.7665317139,
      "last_run_timestamp": "2026-10-17T08:04:25.515Z"
    },
    {
      "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": 0.8922560000000002,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.026633252525252525,
      "avg_duration_per_task_ms": 171876.1313131313,
      "avg_output_tokens_per_task": 22031.722222222223,
      "last_run_timestamp": "2026-10-17T19:00:28.848Z"
    },
    {
      "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": 0.8921505000000001,
      "accuracy_std_dev": 0.013990169329747226,
      "run_count": 16,
      "total_questions": 3168,
      "avg_cost_per_task_usd": 0.014801192550505052,
      "avg_duration_per_task_ms": 209365.545770202,
      "avg_output_tokens_per_task": 24981.652777777777,
      "last_run_timestamp": "2026-10-17T08:25:25.032Z"
    },
    {
      "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": 0.8917747142857142,
      "accuracy_std_dev": 0.02007570557176212,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.026639956709956707,
      "avg_duration_per_task_ms": 151711.15223665224,
      "avg_output_tokens_per_task": 28656.00072150072,
      "last_run_timestamp": "2026-10-17T08:26:19.460Z"
    },
    {
      "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": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Venice",
      "accuracy": 0.8914076107512954,
      "accuracy_std_dev": 0.04401694299340168,
      "run_count": 8,
      "total_questions": 1544,
      "avg_cost_per_task_usd": 0.05118517422279792,
      "avg_duration_per_task_ms": 524717.9449481865,
      "avg_output_tokens_per_task": 16600.09261658031,
      "last_run_timestamp": "2026-10-17T16:06:05.333Z"
    },
    {
      "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": "Phala",
      "accuracy": 0.8910815631578948,
      "accuracy_std_dev": 0.017995824624525182,
      "run_count": 6,
      "total_questions": 1140,
      "avg_cost_per_task_usd": 0.03522420701754386,
      "avg_duration_per_task_ms": 244594.20964912281,
      "avg_output_tokens_per_task": 33383.53684210526,
      "last_run_timestamp": "2026-10-17T08:21:10.129Z"
    },
    {
      "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": 0.8909229583333332,
      "accuracy_std_dev": 0.008679762547055289,
      "run_count": 24,
      "total_questions": 4752,
      "avg_cost_per_task_usd": 0.2421371898148148,
      "avg_duration_per_task_ms": 85114.81292087543,
      "avg_output_tokens_per_task": 9005.400883838383,
      "last_run_timestamp": "2026-10-17T08:19:37.698Z"
    },
    {
      "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": 0.890909,
      "accuracy_std_dev": 0.009121956215637054,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.02464470909090909,
      "avg_duration_per_task_ms": 256040.4707070707,
      "avg_output_tokens_per_task": 30465.680808080808,
      "last_run_timestamp": "2026-10-17T08:18:02.091Z"
    },
    {
      "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": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "GMICloud",
      "accuracy": 0.8905722857142856,
      "accuracy_std_dev": 0.010947483868963876,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.005104790764790765,
      "avg_duration_per_task_ms": 226150.0036075036,
      "avg_output_tokens_per_task": 28015.68181818182,
      "last_run_timestamp": "2026-10-17T16:06:11.358Z"
    },
    {
      "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": 0.8905610702341137,
      "accuracy_std_dev": 0.005541939285123924,
      "run_count": 5,
      "total_questions": 897,
      "avg_cost_per_task_usd": 0.011493041248606466,
      "avg_duration_per_task_ms": 313351.39576365665,
      "avg_output_tokens_per_task": 24048.177257525083,
      "last_run_timestamp": "2026-10-17T08:26:17.873Z"
    },
    {
      "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": 0.8900633006342494,
      "accuracy_std_dev": 0.009369877075486213,
      "run_count": 36,
      "total_questions": 7095,
      "avg_cost_per_task_usd": 0.1335988453840733,
      "avg_duration_per_task_ms": 59373.30049330514,
      "avg_output_tokens_per_task": 4777.861874559549,
      "last_run_timestamp": "2026-10-17T08:25:53.452Z"
    },
    {
      "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": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8900111666666666,
      "accuracy_std_dev": 0.01947983538103498,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.02698784511784512,
      "avg_duration_per_task_ms": 210861.35185185185,
      "avg_output_tokens_per_task": 27490.997474747473,
      "last_run_timestamp": "2026-10-17T08:19:16.770Z"
    },
    {
      "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-4.7-opus-20260416",
      "model_slug": "anthropic/claude-opus-4.7",
      "model_name": "Anthropic: Claude Opus 4.7",
      "provider_name": "Azure",
      "accuracy": 0.8895921012658227,
      "accuracy_std_dev": 0.009804308362766958,
      "run_count": 12,
      "total_questions": 2370,
      "avg_cost_per_task_usd": 0.20471323291139243,
      "avg_duration_per_task_ms": 99592.4729957806,
      "avg_output_tokens_per_task": 7961.961603375527,
      "last_run_timestamp": "2026-10-17T08:06:16.301Z"
    },
    {
      "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": 0.8895694008908687,
      "accuracy_std_dev": 0.0030747750877096573,
      "run_count": 5,
      "total_questions": 898,
      "avg_cost_per_task_usd": 0.03201956347438753,
      "avg_duration_per_task_ms": 793837.6236080178,
      "avg_output_tokens_per_task": 36607.681514476615,
      "last_run_timestamp": "2026-10-17T08:18:02.261Z"
    },
    {
      "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": 0.8890626292530308,
      "accuracy_std_dev": 0.008344399302285123,
      "run_count": 13,
      "total_questions": 2557,
      "avg_cost_per_task_usd": 0.015432386781384432,
      "avg_duration_per_task_ms": 167027.2585060618,
      "avg_output_tokens_per_task": 26079.107547907704,
      "last_run_timestamp": "2026-10-17T08:18:06.855Z"
    },
    {
      "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": 0.8886081666666666,
      "accuracy_std_dev": 0.008877425310239943,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.01216526936026936,
      "avg_duration_per_task_ms": 321417.5909090909,
      "avg_output_tokens_per_task": 22394.61531986532,
      "last_run_timestamp": "2026-10-18T01:11:24.163Z"
    },
    {
      "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": 0.8878433783257624,
      "accuracy_std_dev": 0.020080093052184153,
      "run_count": 8,
      "total_questions": 1541,
      "avg_cost_per_task_usd": 0.155311282284231,
      "avg_duration_per_task_ms": 605967.6262167423,
      "avg_output_tokens_per_task": 42247.91953277093,
      "last_run_timestamp": "2026-10-10T08:22:51.630Z"
    },
    {
      "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": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Inceptron",
      "accuracy": 0.8875035477214103,
      "accuracy_std_dev": 0.007722797339695002,
      "run_count": 12,
      "total_questions": 2326,
      "avg_cost_per_task_usd": 0.09761406061908857,
      "avg_duration_per_task_ms": 475010.2897678418,
      "avg_output_tokens_per_task": 29146.092863284608,
      "last_run_timestamp": "2026-10-17T16:03:24.606Z"
    },
    {
      "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": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8871941263616558,
      "accuracy_std_dev": 0.01904533103531935,
      "run_count": 7,
      "total_questions": 1377,
      "avg_cost_per_task_usd": 0.18794971750181555,
      "avg_duration_per_task_ms": 1470311.4488017429,
      "avg_output_tokens_per_task": 55171.43355119826,
      "last_run_timestamp": "2026-10-17T16:06:45.211Z"
    },
    {
      "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": 0.8867924431103489,
      "accuracy_std_dev": 0.030089157796681178,
      "run_count": 9,
      "total_questions": 1749,
      "avg_cost_per_task_usd": 0.009832666666666667,
      "avg_duration_per_task_ms": 319227.98799313896,
      "avg_output_tokens_per_task": 26332.455117209833,
      "last_run_timestamp": "2026-10-17T08:31:20.808Z"
    },
    {
      "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": "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": 0.8866412831858407,
      "accuracy_std_dev": 0.003953963770698987,
      "run_count": 4,
      "total_questions": 791,
      "avg_cost_per_task_usd": 0.032885610619469034,
      "avg_duration_per_task_ms": 793000.0455120101,
      "avg_output_tokens_per_task": 37709.38053097345,
      "last_run_timestamp": "2026-10-17T08:18:06.333Z"
    },
    {
      "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": 0.8863250000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 195,
      "avg_cost_per_task_usd": 0.01984508717948718,
      "avg_duration_per_task_ms": 89498.22564102564,
      "avg_output_tokens_per_task": 39493.94871794872,
      "last_run_timestamp": "2026-10-17T08:20:36.550Z"
    },
    {
      "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": "AtlasCloud",
      "accuracy": 0.8856619999999997,
      "accuracy_std_dev": 0.01945709939413717,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.02667407154882155,
      "avg_duration_per_task_ms": 212253.46506734006,
      "avg_output_tokens_per_task": 25588.583333333332,
      "last_run_timestamp": "2026-10-17T08:02:36.567Z"
    },
    {
      "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": 0.8856209104124406,
      "accuracy_std_dev": 0.01242517693752844,
      "run_count": 15,
      "total_questions": 2958,
      "avg_cost_per_task_usd": 0.007487724137931034,
      "avg_duration_per_task_ms": 435881.5027045301,
      "avg_output_tokens_per_task": 26714.433062880325,
      "last_run_timestamp": "2026-10-17T08:21:08.586Z"
    },
    {
      "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": "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": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Ionstream",
      "accuracy": 0.8855217999999999,
      "accuracy_std_dev": 0.01517039377735464,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.09142711111111111,
      "avg_duration_per_task_ms": 370562.0878787879,
      "avg_output_tokens_per_task": 32276.01313131313,
      "last_run_timestamp": "2026-10-17T08:18:07.021Z"
    },
    {
      "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": 0.8855217500000001,
      "accuracy_std_dev": 0.01007759093339443,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.02490623021885522,
      "avg_duration_per_task_ms": 342831.7175925926,
      "avg_output_tokens_per_task": 26359.463804713803,
      "last_run_timestamp": "2026-10-17T08:26:01.434Z"
    },
    {
      "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": 0.8853686363636363,
      "accuracy_std_dev": 0.014979663555347446,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.11657416161616162,
      "avg_duration_per_task_ms": 588624.5243342516,
      "avg_output_tokens_per_task": 30454.895316804406,
      "last_run_timestamp": "2026-10-17T16:04:59.290Z"
    },
    {
      "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-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Novita",
      "accuracy": 0.8852812857142857,
      "accuracy_std_dev": 0.029564788253762207,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.1024163391053391,
      "avg_duration_per_task_ms": 456497.3621933622,
      "avg_output_tokens_per_task": 36161.62914862915,
      "last_run_timestamp": "2026-10-17T16:07:02.630Z"
    },
    {
      "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-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Moonshot AI",
      "accuracy": 0.8847767229341582,
      "accuracy_std_dev": 0.013060988043230358,
      "run_count": 23,
      "total_questions": 4526,
      "avg_cost_per_task_usd": 0.19209964803358373,
      "avg_duration_per_task_ms": 368161.0671674768,
      "avg_output_tokens_per_task": 31278.666813963766,
      "last_run_timestamp": "2026-10-17T16:06:12.145Z"
    },
    {
      "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": 0.8845977213793105,
      "accuracy_std_dev": 0.003213065076760805,
      "run_count": 4,
      "total_questions": 725,
      "avg_cost_per_task_usd": 0.12423838068965518,
      "avg_duration_per_task_ms": 424980.20137931034,
      "avg_output_tokens_per_task": 36853.06344827586,
      "last_run_timestamp": "2026-10-17T16:06:05.085Z"
    },
    {
      "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": 0.8843901497975709,
      "accuracy_std_dev": 0.009140983682706173,
      "run_count": 4,
      "total_questions": 741,
      "avg_cost_per_task_usd": 0.006282346828609987,
      "avg_duration_per_task_ms": 497094.5789473684,
      "avg_output_tokens_per_task": 34511.78542510122,
      "last_run_timestamp": "2026-10-17T08:18:00.112Z"
    },
    {
      "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": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Together",
      "accuracy": 0.8838381666666666,
      "accuracy_std_dev": 0.004007391562960742,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.029480208754208752,
      "avg_duration_per_task_ms": 87846.02777777778,
      "avg_output_tokens_per_task": 24419.969696969696,
      "last_run_timestamp": "2026-10-17T08:26:03.843Z"
    },
    {
      "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": 0.8838380000000001,
      "accuracy_std_dev": 0.003367000000000009,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.03186083585858586,
      "avg_duration_per_task_ms": 197855.25252525252,
      "avg_output_tokens_per_task": 26300.560606060608,
      "last_run_timestamp": "2026-10-17T08:21:34.197Z"
    },
    {
      "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": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Wafer",
      "accuracy": 0.8833967254326782,
      "accuracy_std_dev": 0.14603517698007254,
      "run_count": 19,
      "total_questions": 3409,
      "avg_cost_per_task_usd": 0.15225260780287475,
      "avg_duration_per_task_ms": 248086.39131710178,
      "avg_output_tokens_per_task": 10842.60457612203,
      "last_run_timestamp": "2026-10-17T08:31:20.054Z"
    },
    {
      "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": 0.8833942549450549,
      "accuracy_std_dev": 0.07014329085756449,
      "run_count": 15,
      "total_questions": 2730,
      "avg_cost_per_task_usd": 0.009541483150183149,
      "avg_duration_per_task_ms": 150939.43516483516,
      "avg_output_tokens_per_task": 24093.52454212454,
      "last_run_timestamp": "2026-10-17T08:31:21.314Z"
    },
    {
      "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": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Inceptron",
      "accuracy": 0.8830752947678703,
      "accuracy_std_dev": 0.01790514986459119,
      "run_count": 7,
      "total_questions": 1357,
      "avg_cost_per_task_usd": 0.16143720854826824,
      "avg_duration_per_task_ms": 733318.2763448784,
      "avg_output_tokens_per_task": 51996.267501842296,
      "last_run_timestamp": "2026-10-17T16:06:15.854Z"
    },
    {
      "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": 0.882979,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 188,
      "avg_cost_per_task_usd": 0.030796239361702126,
      "avg_duration_per_task_ms": 313329.39361702127,
      "avg_output_tokens_per_task": 61352.47340425532,
      "last_run_timestamp": "2026-10-17T08:18:00.765Z"
    },
    {
      "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": 0.882767,
      "accuracy_std_dev": 0.036073332414049926,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.11128402341597796,
      "avg_duration_per_task_ms": 517321.9977043159,
      "avg_output_tokens_per_task": 35762.50183654729,
      "last_run_timestamp": "2026-10-17T08:30:31.481Z"
    },
    {
      "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": 0.882724073089701,
      "accuracy_std_dev": 0.011353269859384124,
      "run_count": 8,
      "total_questions": 1505,
      "avg_cost_per_task_usd": 0.0675646491694352,
      "avg_duration_per_task_ms": 310422.11893687706,
      "avg_output_tokens_per_task": 18670.514950166114,
      "last_run_timestamp": "2026-10-17T16:07:07.333Z"
    },
    {
      "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-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "StreamLake",
      "accuracy": 0.8821546153846153,
      "accuracy_std_dev": 0.011719582183012278,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.00583898445998446,
      "avg_duration_per_task_ms": 396669.965034965,
      "avg_output_tokens_per_task": 29643.133644133643,
      "last_run_timestamp": "2026-10-17T08:18:07.710Z"
    },
    {
      "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": 0.8819863000000001,
      "accuracy_std_dev": 0.01391203304373593,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.18384018737373733,
      "avg_duration_per_task_ms": 327070.09444444446,
      "avg_output_tokens_per_task": 46141.83181818182,
      "last_run_timestamp": "2026-10-17T08:26:18.154Z"
    },
    {
      "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": 0.8819856633763972,
      "accuracy_std_dev": 0.010038555521850644,
      "run_count": 36,
      "total_questions": 6889,
      "avg_cost_per_task_usd": 0.13809856582958338,
      "avg_duration_per_task_ms": 62825.102482218026,
      "avg_output_tokens_per_task": 5003.652634634926,
      "last_run_timestamp": "2026-10-17T08:19:33.742Z"
    },
    {
      "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": 0.88173375,
      "accuracy_std_dev": 0.004667830993887862,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.041546487373737374,
      "avg_duration_per_task_ms": 232821.30555555556,
      "avg_output_tokens_per_task": 36792.617424242424,
      "last_run_timestamp": "2026-10-17T08:23:33.827Z"
    },
    {
      "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-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "InferenceNet",
      "accuracy": 0.8813526750788644,
      "accuracy_std_dev": 0.020871709863832417,
      "run_count": 5,
      "total_questions": 951,
      "avg_cost_per_task_usd": 0.017862513144058886,
      "avg_duration_per_task_ms": 275442.094637224,
      "avg_output_tokens_per_task": 29043.33123028391,
      "last_run_timestamp": "2026-10-17T08:21:08.302Z"
    },
    {
      "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": 0.8813129999999999,
      "accuracy_std_dev": 0.008841579779654784,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.031503367003367005,
      "avg_duration_per_task_ms": 145405.51430976432,
      "avg_output_tokens_per_task": 26097.064814814814,
      "last_run_timestamp": "2026-10-17T08:23:28.544Z"
    },
    {
      "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": 0.8812964663774403,
      "accuracy_std_dev": 0.01721654338129465,
      "run_count": 7,
      "total_questions": 1383,
      "avg_cost_per_task_usd": 0.06070962472885033,
      "avg_duration_per_task_ms": 528815.1655820679,
      "avg_output_tokens_per_task": 35118.7838033261,
      "last_run_timestamp": "2026-10-17T16:06:14.202Z"
    },
    {
      "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": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "DigitalOcean",
      "accuracy": 0.8809836805006587,
      "accuracy_std_dev": 0.022722285591462856,
      "run_count": 8,
      "total_questions": 1518,
      "avg_cost_per_task_usd": 0.09808666139657445,
      "avg_duration_per_task_ms": 594163.1073781291,
      "avg_output_tokens_per_task": 28545.295125164692,
      "last_run_timestamp": "2026-10-17T08:25:53.370Z"
    },
    {
      "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": 0.8808310763358778,
      "accuracy_std_dev": 0.009797982004176503,
      "run_count": 12,
      "total_questions": 2358,
      "avg_cost_per_task_usd": 0.006506368956743002,
      "avg_duration_per_task_ms": 376147.7417302799,
      "avg_output_tokens_per_task": 18935.907972858353,
      "last_run_timestamp": "2026-10-17T08:25:56.789Z"
    },
    {
      "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": 0.8804712857142858,
      "accuracy_std_dev": 0.01954996565370957,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.00817126911976912,
      "avg_duration_per_task_ms": 383711.4632034632,
      "avg_output_tokens_per_task": 28807.10028860029,
      "last_run_timestamp": "2026-10-17T16:06:05.497Z"
    },
    {
      "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": 0.880471,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.02275759595959596,
      "avg_duration_per_task_ms": 316937.4393939394,
      "avg_output_tokens_per_task": 45749.0101010101,
      "last_run_timestamp": "2026-10-17T08:18:06.019Z"
    },
    {
      "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": "anthropic/claude-opus-5-20260723",
      "model_slug": "anthropic/claude-opus-5",
      "model_name": "Anthropic: Claude Opus 5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.8804433325415677,
      "accuracy_std_dev": 0.01817374661786,
      "run_count": 13,
      "total_questions": 2526,
      "avg_cost_per_task_usd": 0.12863515320665084,
      "avg_duration_per_task_ms": 62895.29176563737,
      "avg_output_tokens_per_task": 4919.452098178939,
      "last_run_timestamp": "2026-10-17T08:21:06.398Z"
    },
    {
      "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": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Moonshot AI",
      "accuracy": 0.8799901428571427,
      "accuracy_std_dev": 0.018718710398411293,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.22771173593073596,
      "avg_duration_per_task_ms": 1000338.3607503608,
      "avg_output_tokens_per_task": 57115.55772005772,
      "last_run_timestamp": "2026-10-17T16:04:06.917Z"
    },
    {
      "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": "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": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Reka",
      "accuracy": 0.879189,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 189,
      "avg_cost_per_task_usd": 0.020723571428571428,
      "avg_duration_per_task_ms": 141095.54497354498,
      "avg_output_tokens_per_task": 17734.21164021164,
      "last_run_timestamp": "2026-09-15T00:33:26.568Z"
    },
    {
      "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": 0.8790683333333332,
      "accuracy_std_dev": 0.0073650451382791405,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.031043784511784515,
      "avg_duration_per_task_ms": 126272.08838383839,
      "avg_output_tokens_per_task": 25679.83080808081,
      "last_run_timestamp": "2026-10-17T08:23:03.620Z"
    },
    {
      "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": 0.8790613422382673,
      "accuracy_std_dev": 0.0067570990696255865,
      "run_count": 7,
      "total_questions": 1385,
      "avg_cost_per_task_usd": 0.01656789963898917,
      "avg_duration_per_task_ms": 310715.13862815883,
      "avg_output_tokens_per_task": 29176.26859205776,
      "last_run_timestamp": "2026-10-17T08:23:40.017Z"
    },
    {
      "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": 0.8790281428571429,
      "accuracy_std_dev": 0.029905434553551217,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.19772237878787877,
      "avg_duration_per_task_ms": 1010773.2835497835,
      "avg_output_tokens_per_task": 58578.94011544011,
      "last_run_timestamp": "2026-10-17T16:10:08.969Z"
    },
    {
      "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": 0.8788940912152681,
      "accuracy_std_dev": 0.03197805161052743,
      "run_count": 18,
      "total_questions": 3563,
      "avg_cost_per_task_usd": 0.14574238170081394,
      "avg_duration_per_task_ms": 381005.49284310977,
      "avg_output_tokens_per_task": 11030.008139208532,
      "last_run_timestamp": "2026-10-17T08:24:42.243Z"
    },
    {
      "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": 0.8788589973154363,
      "accuracy_std_dev": 0.03288382573712152,
      "run_count": 8,
      "total_questions": 1490,
      "avg_cost_per_task_usd": 0.011980427516778525,
      "avg_duration_per_task_ms": 369898.5476510067,
      "avg_output_tokens_per_task": 24175.64765100671,
      "last_run_timestamp": "2026-10-17T08:31:21.156Z"
    },
    {
      "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": "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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "DeepInfra",
      "accuracy": 0.8785058930232558,
      "accuracy_std_dev": 0.05081114611234429,
      "run_count": 12,
      "total_questions": 2365,
      "avg_cost_per_task_usd": 0.0752388731501057,
      "avg_duration_per_task_ms": 545552.287103594,
      "avg_output_tokens_per_task": 39460.21268498943,
      "last_run_timestamp": "2026-10-17T16:06:05.608Z"
    },
    {
      "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": 0.8783141613793103,
      "accuracy_std_dev": 0.007309229033068505,
      "run_count": 11,
      "total_questions": 2175,
      "avg_cost_per_task_usd": 0.11746212551724139,
      "avg_duration_per_task_ms": 746972.4225287356,
      "avg_output_tokens_per_task": 30792.3508045977,
      "last_run_timestamp": "2026-10-17T16:07:11.144Z"
    },
    {
      "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": 0.8779042927756654,
      "accuracy_std_dev": 0.017419794249287197,
      "run_count": 4,
      "total_questions": 789,
      "avg_cost_per_task_usd": 0.03390838529784537,
      "avg_duration_per_task_ms": 751391.9290240811,
      "avg_output_tokens_per_task": 39097.844106463876,
      "last_run_timestamp": "2026-10-17T08:18:06.916Z"
    },
    {
      "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": 0.8776864563843236,
      "accuracy_std_dev": 0.005946450694269239,
      "run_count": 8,
      "total_questions": 1582,
      "avg_cost_per_task_usd": 0.12355401517067002,
      "avg_duration_per_task_ms": 906793.7718078382,
      "avg_output_tokens_per_task": 43362.45132743363,
      "last_run_timestamp": "2026-10-17T16:03:16.411Z"
    },
    {
      "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": 0.8773939383465671,
      "accuracy_std_dev": 0.06611258607459759,
      "run_count": 11,
      "total_questions": 2141,
      "avg_cost_per_task_usd": 0.011637005137786084,
      "avg_duration_per_task_ms": 466296.29145259224,
      "avg_output_tokens_per_task": 22336.07753386268,
      "last_run_timestamp": "2026-10-17T08:31:20.282Z"
    },
    {
      "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": "anthropic/claude-opus-5-20260723",
      "model_slug": "anthropic/claude-opus-5",
      "model_name": "Anthropic: Claude Opus 5",
      "provider_name": "auto-routing",
      "accuracy": 0.8771042727272729,
      "accuracy_std_dev": 0.017480755975388337,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.13006525436179983,
      "avg_duration_per_task_ms": 63287.73140495868,
      "avg_output_tokens_per_task": 4980.730027548209,
      "last_run_timestamp": "2026-10-17T08:04:10.999Z"
    },
    {
      "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": 0.877104,
      "accuracy_std_dev": 0.010101000000000027,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.007626962121212121,
      "avg_duration_per_task_ms": 696482.2146464647,
      "avg_output_tokens_per_task": 43429.71717171717,
      "last_run_timestamp": "2026-10-17T16:06:09.089Z"
    },
    {
      "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": "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": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Parasail",
      "accuracy": 0.8759949037313433,
      "accuracy_std_dev": 0.017795644369637542,
      "run_count": 7,
      "total_questions": 1340,
      "avg_cost_per_task_usd": 0.18478927313432836,
      "avg_duration_per_task_ms": 777556.180597015,
      "avg_output_tokens_per_task": 53320.65,
      "last_run_timestamp": "2026-10-17T16:05:59.065Z"
    },
    {
      "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": 0.8755271341772152,
      "accuracy_std_dev": 0.059505259340153195,
      "run_count": 6,
      "total_questions": 1185,
      "avg_cost_per_task_usd": 0.08042012827004219,
      "avg_duration_per_task_ms": 714967.8244725738,
      "avg_output_tokens_per_task": 30676.93164556962,
      "last_run_timestamp": "2026-10-17T16:06:09.716Z"
    },
    {
      "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": 0.87542075,
      "accuracy_std_dev": 0.004610637205148564,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.0796655340909091,
      "avg_duration_per_task_ms": 373448.2607323232,
      "avg_output_tokens_per_task": 22652.47537878788,
      "last_run_timestamp": "2026-10-17T16:07:05.604Z"
    },
    {
      "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": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "CoreWeave",
      "accuracy": 0.8751617692307694,
      "accuracy_std_dev": 0.014328220895270583,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.11090025951825952,
      "avg_duration_per_task_ms": 290979.16355866357,
      "avg_output_tokens_per_task": 31543.16822066822,
      "last_run_timestamp": "2026-10-17T16:03:21.009Z"
    },
    {
      "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": 0.8750232023608768,
      "accuracy_std_dev": 0.011627604386879436,
      "run_count": 9,
      "total_questions": 1779,
      "avg_cost_per_task_usd": 0.009395388420460931,
      "avg_duration_per_task_ms": 434284.4266441821,
      "avg_output_tokens_per_task": 22384.09162450815,
      "last_run_timestamp": "2026-10-17T08:31:21.223Z"
    },
    {
      "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": "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": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Alibaba",
      "accuracy": 0.8744278861209964,
      "accuracy_std_dev": 0.02518065250226848,
      "run_count": 10,
      "total_questions": 1967,
      "avg_cost_per_task_usd": 0.08236600203355363,
      "avg_duration_per_task_ms": 405834.9110320285,
      "avg_output_tokens_per_task": 31974.98373157092,
      "last_run_timestamp": "2026-10-17T16:03:24.607Z"
    },
    {
      "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": 0.8742985,
      "accuracy_std_dev": 0.016485202058715166,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.03491330218855219,
      "avg_duration_per_task_ms": 143971.7685185185,
      "avg_output_tokens_per_task": 26443.413299663298,
      "last_run_timestamp": "2026-10-17T08:21:10.318Z"
    },
    {
      "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": 0.8741963636363635,
      "accuracy_std_dev": 0.03122712595651627,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.16790697474747474,
      "avg_duration_per_task_ms": 351374.1510560147,
      "avg_output_tokens_per_task": 38330.46235078053,
      "last_run_timestamp": "2026-10-17T16:07:34.585Z"
    },
    {
      "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": 0.8741024915384615,
      "accuracy_std_dev": 0.03167681555707161,
      "run_count": 7,
      "total_questions": 1300,
      "avg_cost_per_task_usd": 0.20500766230769232,
      "avg_duration_per_task_ms": 1300362.693076923,
      "avg_output_tokens_per_task": 51104.55923076923,
      "last_run_timestamp": "2026-10-17T16:05:57.014Z"
    },
    {
      "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-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "StreamLake",
      "accuracy": 0.8739477499999999,
      "accuracy_std_dev": 0.015161742633599224,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.020946954545454546,
      "avg_duration_per_task_ms": 723886.8560606061,
      "avg_output_tokens_per_task": 35575.18244949495,
      "last_run_timestamp": "2026-10-17T16:05:38.303Z"
    },
    {
      "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": 0.873737,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.01707317171717172,
      "avg_duration_per_task_ms": 108581.24747474748,
      "avg_output_tokens_per_task": 42482.4898989899,
      "last_run_timestamp": "2026-10-17T08:04:11.014Z"
    },
    {
      "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": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Venice",
      "accuracy": 0.8733865833333335,
      "accuracy_std_dev": 0.012291988600021385,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.0063178345959595955,
      "avg_duration_per_task_ms": 267411.5660774411,
      "avg_output_tokens_per_task": 18551.484427609426,
      "last_run_timestamp": "2026-10-17T08:18:06.388Z"
    },
    {
      "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": 0.8733164166666666,
      "accuracy_std_dev": 0.045719398832184864,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.09441799915824915,
      "avg_duration_per_task_ms": 629233.5349326599,
      "avg_output_tokens_per_task": 41849.718434343435,
      "last_run_timestamp": "2026-10-17T16:07:29.391Z"
    },
    {
      "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": 0.8732769164556962,
      "accuracy_std_dev": 0.0036989960650004694,
      "run_count": 6,
      "total_questions": 1185,
      "avg_cost_per_task_usd": 0.04468822784810126,
      "avg_duration_per_task_ms": 182929.39156118143,
      "avg_output_tokens_per_task": 38802.67004219409,
      "last_run_timestamp": "2026-10-17T08:22:53.877Z"
    },
    {
      "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": 0.8732727052578362,
      "accuracy_std_dev": 0.015296519341340365,
      "run_count": 10,
      "total_questions": 1978,
      "avg_cost_per_task_usd": 0.09049616683518706,
      "avg_duration_per_task_ms": 212094.93225480284,
      "avg_output_tokens_per_task": 22916.149140546007,
      "last_run_timestamp": "2026-10-17T08:21:14.647Z"
    },
    {
      "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": 0.8731974290076335,
      "accuracy_std_dev": 0.011081761001303017,
      "run_count": 10,
      "total_questions": 1965,
      "avg_cost_per_task_usd": 0.04679652417302798,
      "avg_duration_per_task_ms": 227836.92264631044,
      "avg_output_tokens_per_task": 34673.968956743,
      "last_run_timestamp": "2026-10-17T08:26:00.771Z"
    },
    {
      "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": 0.873136,
      "accuracy_std_dev": 0.06760829277325589,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.12835601948051947,
      "avg_duration_per_task_ms": 1515824.0468975469,
      "avg_output_tokens_per_task": 55051.631313131315,
      "last_run_timestamp": "2026-10-17T16:05:37.599Z"
    },
    {
      "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": 0.8728152090966519,
      "accuracy_std_dev": 0.007321385674003263,
      "run_count": 8,
      "total_questions": 1583,
      "avg_cost_per_task_usd": 0.04953204042956412,
      "avg_duration_per_task_ms": 276172.8900821225,
      "avg_output_tokens_per_task": 21040.667719519897,
      "last_run_timestamp": "2026-10-17T16:03:30.492Z"
    },
    {
      "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": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8724903310344827,
      "accuracy_std_dev": 0.10049535477303755,
      "run_count": 11,
      "total_questions": 2175,
      "avg_cost_per_task_usd": 0.02776077609195403,
      "avg_duration_per_task_ms": 206156.86804597703,
      "avg_output_tokens_per_task": 23904.91816091954,
      "last_run_timestamp": "2026-10-17T16:03:16.413Z"
    },
    {
      "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": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Cloudflare",
      "accuracy": 0.872399359223301,
      "accuracy_std_dev": 0.037525025908999386,
      "run_count": 11,
      "total_questions": 2163,
      "avg_cost_per_task_usd": 0.10075985020804439,
      "avg_duration_per_task_ms": 612085.7415626445,
      "avg_output_tokens_per_task": 25038.095238095237,
      "last_run_timestamp": "2026-10-17T16:03:14.425Z"
    },
    {
      "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": "CoreWeave",
      "accuracy": 0.8721831538461537,
      "accuracy_std_dev": 0.016072645308321247,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.008282733100233098,
      "avg_duration_per_task_ms": 224627.168997669,
      "avg_output_tokens_per_task": 29323.653457653458,
      "last_run_timestamp": "2026-10-17T08:23:45.259Z"
    },
    {
      "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": "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": 0.8720540000000001,
      "accuracy_std_dev": 0.002749143987983661,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.004076121212121212,
      "avg_duration_per_task_ms": 39248.412457912455,
      "avg_output_tokens_per_task": 7919.883838383838,
      "last_run_timestamp": "2026-10-17T08:19:02.944Z"
    },
    {
      "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": 0.8720537777777777,
      "accuracy_std_dev": 0.005143031980322249,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.004249169472502806,
      "avg_duration_per_task_ms": 39560.52076318743,
      "avg_output_tokens_per_task": 7734.946689113356,
      "last_run_timestamp": "2026-10-17T08:18:06.937Z"
    },
    {
      "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": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Baidu",
      "accuracy": 0.8718132857142857,
      "accuracy_std_dev": 0.025315326948111786,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.14662903607503608,
      "avg_duration_per_task_ms": 951998.4913419914,
      "avg_output_tokens_per_task": 61296.518037518035,
      "last_run_timestamp": "2026-10-17T16:03:24.607Z"
    },
    {
      "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": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Mancer 2",
      "accuracy": 0.8713910913385827,
      "accuracy_std_dev": 0.02908788258380426,
      "run_count": 14,
      "total_questions": 2540,
      "avg_cost_per_task_usd": 0.011632457874015747,
      "avg_duration_per_task_ms": 341476.78307086613,
      "avg_output_tokens_per_task": 20755.323622047243,
      "last_run_timestamp": "2026-10-17T08:18:00.507Z"
    },
    {
      "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-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Cloudflare",
      "accuracy": 0.8709067008816122,
      "accuracy_std_dev": 0.03126206183140546,
      "run_count": 9,
      "total_questions": 1588,
      "avg_cost_per_task_usd": 0.024239201511335014,
      "avg_duration_per_task_ms": 368035.88853904285,
      "avg_output_tokens_per_task": 24202.87468513854,
      "last_run_timestamp": "2026-10-17T16:03:19.122Z"
    },
    {
      "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": 0.870791,
      "accuracy_std_dev": 0.011807038716375923,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.08650412815656565,
      "avg_duration_per_task_ms": 201299.91603535353,
      "avg_output_tokens_per_task": 14085.986742424242,
      "last_run_timestamp": "2026-10-17T08:18:06.671Z"
    },
    {
      "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": 0.8706306660808436,
      "accuracy_std_dev": 0.013549251556023127,
      "run_count": 9,
      "total_questions": 1707,
      "avg_cost_per_task_usd": 0.008233070298769771,
      "avg_duration_per_task_ms": 238186.8295254833,
      "avg_output_tokens_per_task": 29547.44698301113,
      "last_run_timestamp": "2026-10-17T16:03:22.653Z"
    },
    {
      "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": 0.8703702499999999,
      "accuracy_std_dev": 0.009143745481885412,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.07626149684343433,
      "avg_duration_per_task_ms": 105007.26452020202,
      "avg_output_tokens_per_task": 12387.4273989899,
      "last_run_timestamp": "2026-10-17T08:03:37.482Z"
    },
    {
      "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-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Reka",
      "accuracy": 0.8702065335018964,
      "accuracy_std_dev": 0.007070779465518657,
      "run_count": 4,
      "total_questions": 791,
      "avg_cost_per_task_usd": 0.31242882300884955,
      "avg_duration_per_task_ms": 1310479.2250316055,
      "avg_output_tokens_per_task": 52210.361567635904,
      "last_run_timestamp": "2026-10-17T16:03:45.624Z"
    },
    {
      "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": 0.870157356321839,
      "accuracy_std_dev": 0.01848232110693894,
      "run_count": 8,
      "total_questions": 1566,
      "avg_cost_per_task_usd": 0.07430513729246488,
      "avg_duration_per_task_ms": 401768.99042145594,
      "avg_output_tokens_per_task": 23634.381864623243,
      "last_run_timestamp": "2026-10-17T08:20:36.736Z"
    },
    {
      "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": 0.8699560401315789,
      "accuracy_std_dev": 0.0072051253688084314,
      "run_count": 9,
      "total_questions": 1520,
      "avg_cost_per_task_usd": 0.05408648092105263,
      "avg_duration_per_task_ms": 430159.67631578946,
      "avg_output_tokens_per_task": 17930.499342105264,
      "last_run_timestamp": "2026-10-17T16:03:54.454Z"
    },
    {
      "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-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Parasail",
      "accuracy": 0.8698892857142857,
      "accuracy_std_dev": 0.00522506975248824,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.008162125541125541,
      "avg_duration_per_task_ms": 376919.626984127,
      "avg_output_tokens_per_task": 28806.24891774892,
      "last_run_timestamp": "2026-10-17T16:06:10.645Z"
    },
    {
      "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": "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": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "Anthropic",
      "accuracy": 0.8692479999999999,
      "accuracy_std_dev": 0.006780740569190578,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.003940161616161617,
      "avg_duration_per_task_ms": 37433.5505050505,
      "avg_output_tokens_per_task": 7640.225589225589,
      "last_run_timestamp": "2026-10-17T08:20:54.898Z"
    },
    {
      "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": "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": "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": "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": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "CoreWeave",
      "accuracy": 0.86847625,
      "accuracy_std_dev": 0.018488540753328793,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.19820480492424242,
      "avg_duration_per_task_ms": 528317.1729797979,
      "avg_output_tokens_per_task": 57987.417297979795,
      "last_run_timestamp": "2026-10-17T16:03:44.011Z"
    },
    {
      "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": 0.8684461428571428,
      "accuracy_std_dev": 0.017689481203961483,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.037552243867243866,
      "avg_duration_per_task_ms": 502237.20707070705,
      "avg_output_tokens_per_task": 34877.69985569986,
      "last_run_timestamp": "2026-10-17T16:07:06.799Z"
    },
    {
      "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": "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": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Crusoe",
      "accuracy": 0.8681256666666667,
      "accuracy_std_dev": 0.00619849967510062,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.04747902525252525,
      "avg_duration_per_task_ms": 225031.7861952862,
      "avg_output_tokens_per_task": 39421.97306397306,
      "last_run_timestamp": "2026-10-17T08:26:17.952Z"
    },
    {
      "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": 0.8679384444444443,
      "accuracy_std_dev": 0.0043547631932328865,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.0705021952861953,
      "avg_duration_per_task_ms": 234442.81986531988,
      "avg_output_tokens_per_task": 11483.022446689114,
      "last_run_timestamp": "2026-10-17T08:04:10.927Z"
    },
    {
      "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": 0.8679118212141426,
      "accuracy_std_dev": 0.029083773117496287,
      "run_count": 8,
      "total_questions": 1499,
      "avg_cost_per_task_usd": 0.011126581054036026,
      "avg_duration_per_task_ms": 399635.85390260175,
      "avg_output_tokens_per_task": 24844.47164776518,
      "last_run_timestamp": "2026-10-17T08:26:17.731Z"
    },
    {
      "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": 0.8677515555555556,
      "accuracy_std_dev": 0.008841170600109865,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.0042902659932659935,
      "avg_duration_per_task_ms": 35789.490460157125,
      "avg_output_tokens_per_task": 7820.8361391694725,
      "last_run_timestamp": "2026-10-17T08:26:18.244Z"
    },
    {
      "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": 0.8677245714285714,
      "accuracy_std_dev": 0.02218605746767963,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.11909902236652237,
      "avg_duration_per_task_ms": 600755.6313131313,
      "avg_output_tokens_per_task": 35283.31168831169,
      "last_run_timestamp": "2026-10-17T16:04:59.319Z"
    },
    {
      "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": "Azure",
      "accuracy": 0.8672838333333334,
      "accuracy_std_dev": 0.005618689361309189,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.004177218855218855,
      "avg_duration_per_task_ms": 35769.420875420874,
      "avg_output_tokens_per_task": 7720.202861952862,
      "last_run_timestamp": "2026-10-17T08:26:01.760Z"
    },
    {
      "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": 0.8663297,
      "accuracy_std_dev": 0.04012380805220261,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.1796413303030303,
      "avg_duration_per_task_ms": 45593.2398989899,
      "avg_output_tokens_per_task": 3390.838888888889,
      "last_run_timestamp": "2026-10-17T08:18:03.745Z"
    },
    {
      "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": 0.8661615,
      "accuracy_std_dev": 0.004208500000000004,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.0040270404040404035,
      "avg_duration_per_task_ms": 38264.56565656565,
      "avg_output_tokens_per_task": 7821.176767676768,
      "last_run_timestamp": "2026-10-17T08:17:57.645Z"
    },
    {
      "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": 0.865880181448332,
      "accuracy_std_dev": 0.018202091806230666,
      "run_count": 7,
      "total_questions": 1229,
      "avg_cost_per_task_usd": 0.005052348250610252,
      "avg_duration_per_task_ms": 670227.6916192026,
      "avg_output_tokens_per_task": 27781.558991049635,
      "last_run_timestamp": "2026-10-17T16:03:21.031Z"
    },
    {
      "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-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "auto-routing",
      "accuracy": 0.8655302499999998,
      "accuracy_std_dev": 0.017156845031575593,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.038863957702020197,
      "avg_duration_per_task_ms": 732278.3465909091,
      "avg_output_tokens_per_task": 34126.10921717172,
      "last_run_timestamp": "2026-10-17T16:03:30.488Z"
    },
    {
      "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": 0.8654881000000001,
      "accuracy_std_dev": 0.027275302500430684,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.007844937878787879,
      "avg_duration_per_task_ms": 473405.9202020202,
      "avg_output_tokens_per_task": 25350.743434343436,
      "last_run_timestamp": "2026-10-17T08:18:05.440Z"
    },
    {
      "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": 0.8653196666666667,
      "accuracy_std_dev": 0.008584133982075423,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.07009036139169472,
      "avg_duration_per_task_ms": 231708.10998877665,
      "avg_output_tokens_per_task": 11412.423681257014,
      "last_run_timestamp": "2026-10-17T08:19:03.028Z"
    },
    {
      "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": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8648990000000001,
      "accuracy_std_dev": 0.017388735477314038,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.07359775946969697,
      "avg_duration_per_task_ms": 279722.81755050505,
      "avg_output_tokens_per_task": 20944.539772727272,
      "last_run_timestamp": "2026-10-17T16:03:30.487Z"
    },
    {
      "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": 0.8646462,
      "accuracy_std_dev": 0.005985377628855152,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.023131965656565653,
      "avg_duration_per_task_ms": 316631.1070707071,
      "avg_output_tokens_per_task": 37590.83535353535,
      "last_run_timestamp": "2026-10-17T08:22:58.767Z"
    },
    {
      "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": 0.8640821969539633,
      "accuracy_std_dev": 0.024524497231589105,
      "run_count": 15,
      "total_questions": 2889,
      "avg_cost_per_task_usd": 0.03844685289027345,
      "avg_duration_per_task_ms": 78089.81827622015,
      "avg_output_tokens_per_task": 24843.67185877466,
      "last_run_timestamp": "2026-10-17T08:19:08.093Z"
    },
    {
      "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": "Z.AI",
      "accuracy": 0.8639263412162163,
      "accuracy_std_dev": 0.027057887036664362,
      "run_count": 9,
      "total_questions": 1776,
      "avg_cost_per_task_usd": 0.10281788851351349,
      "avg_duration_per_task_ms": 359495.01407657657,
      "avg_output_tokens_per_task": 23127.624436936938,
      "last_run_timestamp": "2026-10-17T08:31:21.198Z"
    },
    {
      "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": 0.8637523318442885,
      "accuracy_std_dev": 0.013965236766252836,
      "run_count": 8,
      "total_questions": 1567,
      "avg_cost_per_task_usd": 0.07525077600510531,
      "avg_duration_per_task_ms": 273604.328653478,
      "avg_output_tokens_per_task": 20778.284620293554,
      "last_run_timestamp": "2026-10-17T16:07:17.170Z"
    },
    {
      "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": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "CoreWeave",
      "accuracy": 0.8634711176470588,
      "accuracy_std_dev": 0.012638435637372207,
      "run_count": 5,
      "total_questions": 918,
      "avg_cost_per_task_usd": 0.02367367320261438,
      "avg_duration_per_task_ms": 131839.07298474945,
      "avg_output_tokens_per_task": 36232.9880174292,
      "last_run_timestamp": "2026-10-17T08:21:32.495Z"
    },
    {
      "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": "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": 0.8630240909090909,
      "accuracy_std_dev": 0.013961714569457724,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.7496591460055096,
      "avg_duration_per_task_ms": 494920.16850321396,
      "avg_output_tokens_per_task": 29823.61570247934,
      "last_run_timestamp": "2026-10-17T08:18:05.667Z"
    },
    {
      "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": 0.8624119090909091,
      "accuracy_std_dev": 0.008735668278069101,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.8035566230486685,
      "avg_duration_per_task_ms": 460560.63590449956,
      "avg_output_tokens_per_task": 31970.262626262625,
      "last_run_timestamp": "2026-10-17T08:02:36.421Z"
    },
    {
      "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": 0.8621932857142858,
      "accuracy_std_dev": 0.007166829677943584,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.004471386002886003,
      "avg_duration_per_task_ms": 249807.05122655124,
      "avg_output_tokens_per_task": 29262.619047619046,
      "last_run_timestamp": "2026-10-17T16:06:17.280Z"
    },
    {
      "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": "anthropic/claude-sonnet-5-20260630",
      "model_slug": "anthropic/claude-sonnet-5",
      "model_name": "Anthropic: Claude Sonnet 5",
      "provider_name": "Azure",
      "accuracy": 0.8617332608695653,
      "accuracy_std_dev": 0.02687779012852092,
      "run_count": 23,
      "total_questions": 4554,
      "avg_cost_per_task_usd": 0.1588332134387352,
      "avg_duration_per_task_ms": 153168.5540184453,
      "avg_output_tokens_per_task": 15408.788537549408,
      "last_run_timestamp": "2026-10-17T08:22:59.016Z"
    },
    {
      "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": 0.8616467272727273,
      "accuracy_std_dev": 0.00836578286388189,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.8426584527089072,
      "avg_duration_per_task_ms": 413263.7479338843,
      "avg_output_tokens_per_task": 33515.24885215794,
      "last_run_timestamp": "2026-10-17T08:03:37.448Z"
    },
    {
      "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": 0.8616161,
      "accuracy_std_dev": 0.0180033848064746,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.09928383787878788,
      "avg_duration_per_task_ms": 481966.9338383838,
      "avg_output_tokens_per_task": 26359.05101010101,
      "last_run_timestamp": "2026-10-17T16:03:19.114Z"
    },
    {
      "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": 0.8615345381097562,
      "accuracy_std_dev": 0.03207393262697919,
      "run_count": 10,
      "total_questions": 1968,
      "avg_cost_per_task_usd": 0.13308645274390243,
      "avg_duration_per_task_ms": 545136.4395325204,
      "avg_output_tokens_per_task": 42522.091971544716,
      "last_run_timestamp": "2026-10-17T16:06:11.440Z"
    },
    {
      "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": 0.8615317499999998,
      "accuracy_std_dev": 0.035989945114631224,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.1836229675925926,
      "avg_duration_per_task_ms": 45563.180555555555,
      "avg_output_tokens_per_task": 3466.4579124579122,
      "last_run_timestamp": "2026-10-17T08:18:07.803Z"
    },
    {
      "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": 0.8614476,
      "accuracy_std_dev": 0.0162708383754495,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.08960124141414141,
      "avg_duration_per_task_ms": 187539.3025252525,
      "avg_output_tokens_per_task": 24757.361616161616,
      "last_run_timestamp": "2026-10-17T16:04:59.259Z"
    },
    {
      "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": 0.8613916666666668,
      "accuracy_std_dev": 0.05599875111504226,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.10855281649831651,
      "avg_duration_per_task_ms": 469370.49326599325,
      "avg_output_tokens_per_task": 27476.757575757576,
      "last_run_timestamp": "2026-10-17T16:03:19.114Z"
    },
    {
      "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": 0.8613916666666666,
      "accuracy_std_dev": 0.0411070840839981,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.17600762920875418,
      "avg_duration_per_task_ms": 43642.34217171717,
      "avg_output_tokens_per_task": 3313.9747474747473,
      "last_run_timestamp": "2026-10-17T08:04:21.178Z"
    },
    {
      "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": "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": "anthropic/claude-4.5-opus-20251124",
      "model_slug": "anthropic/claude-opus-4.5",
      "model_name": "Anthropic: Claude Opus 4.5",
      "provider_name": "Anthropic",
      "accuracy": 0.8610461538461541,
      "accuracy_std_dev": 0.012042697039132513,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.8432189724164725,
      "avg_duration_per_task_ms": 455003.31235431234,
      "avg_output_tokens_per_task": 33550.2871017871,
      "last_run_timestamp": "2026-10-17T08:20:29.360Z"
    },
    {
      "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": 0.8610345454545455,
      "accuracy_std_dev": 0.007875541064975502,
      "run_count": 22,
      "total_questions": 4356,
      "avg_cost_per_task_usd": 0.8829572610192835,
      "avg_duration_per_task_ms": 415451.3753443526,
      "avg_output_tokens_per_task": 33445.12649219467,
      "last_run_timestamp": "2026-10-17T08:02:36.220Z"
    },
    {
      "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": 0.8609005,
      "accuracy_std_dev": 0.01279172761983305,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.009262198232323233,
      "avg_duration_per_task_ms": 325442.8162878788,
      "avg_output_tokens_per_task": 32760.926767676767,
      "last_run_timestamp": "2026-10-17T16:03:24.606Z"
    },
    {
      "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": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Relace",
      "accuracy": 0.8606902500000001,
      "accuracy_std_dev": 0.008858447447916569,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.07241029419191919,
      "avg_duration_per_task_ms": 429769.84722222225,
      "avg_output_tokens_per_task": 56551.03535353536,
      "last_run_timestamp": "2026-10-17T16:07:34.379Z"
    },
    {
      "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": "anthropic/claude-4.5-opus-20251124",
      "model_slug": "anthropic/claude-opus-4.5",
      "model_name": "Anthropic: Claude Opus 4.5",
      "provider_name": "Google",
      "accuracy": 0.859963090909091,
      "accuracy_std_dev": 0.010045046128266804,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.8526076652892564,
      "avg_duration_per_task_ms": 482177.8953168044,
      "avg_output_tokens_per_task": 33913.51193755739,
      "last_run_timestamp": "2026-10-17T08:18:00.936Z"
    },
    {
      "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": 0.8598952222222221,
      "accuracy_std_dev": 0.0069538031605058765,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.06597872390572392,
      "avg_duration_per_task_ms": 219441.50561167227,
      "avg_output_tokens_per_task": 10728.320426487093,
      "last_run_timestamp": "2026-10-17T08:18:06.079Z"
    },
    {
      "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": 0.8597882857142857,
      "accuracy_std_dev": 0.02839544075936078,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.12813852813852813,
      "avg_duration_per_task_ms": 499282.8939393939,
      "avg_output_tokens_per_task": 33659.728715728714,
      "last_run_timestamp": "2026-10-17T16:03:22.662Z"
    },
    {
      "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": 0.8596491924664603,
      "accuracy_std_dev": 0.037215045235596586,
      "run_count": 20,
      "total_questions": 3876,
      "avg_cost_per_task_usd": 0.004361001805985552,
      "avg_duration_per_task_ms": 88288.53689370485,
      "avg_output_tokens_per_task": 16394.82894736842,
      "last_run_timestamp": "2026-10-17T08:02:02.308Z"
    },
    {
      "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": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "StreamLake",
      "accuracy": 0.8592169999999999,
      "accuracy_std_dev": 0.023807429874936097,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.0034952645202020204,
      "avg_duration_per_task_ms": 466987.4917929293,
      "avg_output_tokens_per_task": 33328.60227272727,
      "last_run_timestamp": "2026-10-17T16:06:34.618Z"
    },
    {
      "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": 0.8591320634715027,
      "accuracy_std_dev": 0.0170592919090301,
      "run_count": 8,
      "total_questions": 1544,
      "avg_cost_per_task_usd": 0.08052118911917099,
      "avg_duration_per_task_ms": 345606.61269430054,
      "avg_output_tokens_per_task": 27437.18847150259,
      "last_run_timestamp": "2026-10-17T08:18:00.291Z"
    },
    {
      "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": 0.8589818529411762,
      "accuracy_std_dev": 0.027213277180268808,
      "run_count": 34,
      "total_questions": 6732,
      "avg_cost_per_task_usd": 0.1721233055555556,
      "avg_duration_per_task_ms": 158539.5931372549,
      "avg_output_tokens_per_task": 16076.290106951872,
      "last_run_timestamp": "2026-10-17T08:22:59.081Z"
    },
    {
      "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": "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": "anthropic/claude-4.6-opus-20260205",
      "model_slug": "anthropic/claude-opus-4.6",
      "model_name": "Anthropic: Claude Opus 4.6",
      "provider_name": "Azure",
      "accuracy": 0.8585856363636363,
      "accuracy_std_dev": 0.019208303437802044,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.7565819077134986,
      "avg_duration_per_task_ms": 502366.63636363635,
      "avg_output_tokens_per_task": 30138.17401285583,
      "last_run_timestamp": "2026-10-17T08:04:11.096Z"
    },
    {
      "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": 0.8583754485087226,
      "accuracy_std_dev": 0.028677200239136932,
      "run_count": 9,
      "total_questions": 1777,
      "avg_cost_per_task_usd": 0.00778828474957794,
      "avg_duration_per_task_ms": 416499.75858187955,
      "avg_output_tokens_per_task": 25916.474957794035,
      "last_run_timestamp": "2026-10-17T08:25:56.871Z"
    },
    {
      "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": 0.8580637958148383,
      "accuracy_std_dev": 0.014779169880591208,
      "run_count": 8,
      "total_questions": 1577,
      "avg_cost_per_task_usd": 0.08286067279644893,
      "avg_duration_per_task_ms": 195379.60684844642,
      "avg_output_tokens_per_task": 25387.850348763473,
      "last_run_timestamp": "2026-10-17T08:26:18.335Z"
    },
    {
      "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": 0.8576645647133758,
      "accuracy_std_dev": 0.020977893625656868,
      "run_count": 40,
      "total_questions": 7850,
      "avg_cost_per_task_usd": 0.27657329388535035,
      "avg_duration_per_task_ms": 335319.63796178345,
      "avg_output_tokens_per_task": 49792.269044585984,
      "last_run_timestamp": "2026-10-17T16:06:05.523Z"
    },
    {
      "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": 0.8576035833333334,
      "accuracy_std_dev": 0.030166058535652105,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.1556503026094276,
      "avg_duration_per_task_ms": 151996.30134680134,
      "avg_output_tokens_per_task": 15348.492003367004,
      "last_run_timestamp": "2026-10-17T08:19:33.462Z"
    },
    {
      "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": 0.8574634166666667,
      "accuracy_std_dev": 0.039912043761790195,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.18122077314814813,
      "avg_duration_per_task_ms": 46037.724326599324,
      "avg_output_tokens_per_task": 3416.7558922558924,
      "last_run_timestamp": "2026-10-17T08:19:36.445Z"
    },
    {
      "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": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "AtlasCloud",
      "accuracy": 0.857416679028133,
      "accuracy_std_dev": 0.024059550515011203,
      "run_count": 8,
      "total_questions": 1564,
      "avg_cost_per_task_usd": 0.1341864226342711,
      "avg_duration_per_task_ms": 371153.3861892583,
      "avg_output_tokens_per_task": 30475.74872122762,
      "last_run_timestamp": "2026-10-17T08:20:15.622Z"
    },
    {
      "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": 0.8573675040431266,
      "accuracy_std_dev": 0.08566512385407106,
      "run_count": 15,
      "total_questions": 2968,
      "avg_cost_per_task_usd": 0.2818780539083558,
      "avg_duration_per_task_ms": 484741.4484501348,
      "avg_output_tokens_per_task": 65941.04043126684,
      "last_run_timestamp": "2026-10-17T08:31:19.989Z"
    },
    {
      "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": 0.8571427142857143,
      "accuracy_std_dev": 0.022648008400073402,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.15408665367965368,
      "avg_duration_per_task_ms": 1261266.8744588743,
      "avg_output_tokens_per_task": 52164.505772005774,
      "last_run_timestamp": "2026-10-17T16:03:34.729Z"
    },
    {
      "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": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "Minimax",
      "accuracy": 0.8570973081339713,
      "accuracy_std_dev": 0.00813419198867309,
      "run_count": 16,
      "total_questions": 3135,
      "avg_cost_per_task_usd": 0.0628228870813397,
      "avg_duration_per_task_ms": 581826.4315789473,
      "avg_output_tokens_per_task": 34631.35693779904,
      "last_run_timestamp": "2026-10-17T16:03:42.525Z"
    },
    {
      "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": 0.8569022,
      "accuracy_std_dev": 0.028089651189717544,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.8333996262626261,
      "avg_duration_per_task_ms": 432174.9904040404,
      "avg_output_tokens_per_task": 33238.25858585859,
      "last_run_timestamp": "2026-10-17T08:18:07.358Z"
    },
    {
      "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": "anthropic/claude-4.6-opus-20260205",
      "model_slug": "anthropic/claude-opus-4.6",
      "model_name": "Anthropic: Claude Opus 4.6",
      "provider_name": "Anthropic",
      "accuracy": 0.856643153846154,
      "accuracy_std_dev": 0.017013380403246205,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.7557811247086247,
      "avg_duration_per_task_ms": 501748.9289044289,
      "avg_output_tokens_per_task": 30126.986013986014,
      "last_run_timestamp": "2026-10-17T08:02:36.101Z"
    },
    {
      "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": 0.8564731094373864,
      "accuracy_std_dev": 0.036898649079361114,
      "run_count": 32,
      "total_questions": 5510,
      "avg_cost_per_task_usd": 0.014921334301270416,
      "avg_duration_per_task_ms": 403064.5290381125,
      "avg_output_tokens_per_task": 17553.35952813067,
      "last_run_timestamp": "2026-10-17T08:21:06.310Z"
    },
    {
      "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": 0.8562900909090908,
      "accuracy_std_dev": 0.01747002531847331,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.8390210973370065,
      "avg_duration_per_task_ms": 455775.94260789716,
      "avg_output_tokens_per_task": 33386.61432506887,
      "last_run_timestamp": "2026-10-17T08:18:06.418Z"
    },
    {
      "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": 0.8562671637010675,
      "accuracy_std_dev": 0.055964205439230925,
      "run_count": 9,
      "total_questions": 1686,
      "avg_cost_per_task_usd": 0.010985191577698696,
      "avg_duration_per_task_ms": 499242.5960854092,
      "avg_output_tokens_per_task": 22245.52846975089,
      "last_run_timestamp": "2026-10-17T08:17:59.904Z"
    },
    {
      "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": 0.8561369999999999,
      "accuracy_std_dev": 0.07519723650870061,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.007035742883379248,
      "avg_duration_per_task_ms": 646531.86271809,
      "avg_output_tokens_per_task": 28996.6391184573,
      "last_run_timestamp": "2026-10-17T08:26:19.397Z"
    },
    {
      "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": 0.8561369999999997,
      "accuracy_std_dev": 0.01741226929756455,
      "run_count": 22,
      "total_questions": 4356,
      "avg_cost_per_task_usd": 0.8047317961432507,
      "avg_duration_per_task_ms": 480071.0406336088,
      "avg_output_tokens_per_task": 30506.34618916437,
      "last_run_timestamp": "2026-10-17T08:02:36.548Z"
    },
    {
      "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": 0.8560605,
      "accuracy_std_dev": 0.017676499999999984,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.04304399747474747,
      "avg_duration_per_task_ms": 130697.23232323233,
      "avg_output_tokens_per_task": 34274.84595959596,
      "last_run_timestamp": "2026-10-17T08:22:58.392Z"
    },
    {
      "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": 0.8559402857142857,
      "accuracy_std_dev": 0.016118870642061436,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.13142277200577202,
      "avg_duration_per_task_ms": 818557.549062049,
      "avg_output_tokens_per_task": 41821.506493506495,
      "last_run_timestamp": "2026-10-17T16:06:33.244Z"
    },
    {
      "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": 0.8558689848156183,
      "accuracy_std_dev": 0.06273596781165733,
      "run_count": 7,
      "total_questions": 1383,
      "avg_cost_per_task_usd": 0.2468844273318872,
      "avg_duration_per_task_ms": 1264380.503976862,
      "avg_output_tokens_per_task": 55086.25668835864,
      "last_run_timestamp": "2026-10-17T16:05:25.640Z"
    },
    {
      "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": 0.8558500762766644,
      "accuracy_std_dev": 0.04264423738619322,
      "run_count": 8,
      "total_questions": 1547,
      "avg_cost_per_task_usd": 0.01680374466709761,
      "avg_duration_per_task_ms": 536166.6296056884,
      "avg_output_tokens_per_task": 41929.81641887524,
      "last_run_timestamp": "2026-10-17T08:26:18.533Z"
    },
    {
      "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": 0.8553733300546447,
      "accuracy_std_dev": 0.04403852970379714,
      "run_count": 6,
      "total_questions": 915,
      "avg_cost_per_task_usd": 0.021436923497267762,
      "avg_duration_per_task_ms": 455969.31147540984,
      "avg_output_tokens_per_task": 35599.86010928962,
      "last_run_timestamp": "2026-10-17T08:19:17.013Z"
    },
    {
      "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": 0.8553548880866427,
      "accuracy_std_dev": 0.040121466593740725,
      "run_count": 7,
      "total_questions": 1385,
      "avg_cost_per_task_usd": 0.16188800866425992,
      "avg_duration_per_task_ms": 1012449.7552346571,
      "avg_output_tokens_per_task": 54542.67653429603,
      "last_run_timestamp": "2026-10-17T16:03:32.427Z"
    },
    {
      "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": 0.8552632087114336,
      "accuracy_std_dev": 0.016551219440862706,
      "run_count": 7,
      "total_questions": 1102,
      "avg_cost_per_task_usd": 0.004268463702359346,
      "avg_duration_per_task_ms": 367204.3484573503,
      "avg_output_tokens_per_task": 9380.986388384756,
      "last_run_timestamp": "2026-10-17T08:21:41.875Z"
    },
    {
      "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": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8548968432122371,
      "accuracy_std_dev": 0.022342831477623858,
      "run_count": 8,
      "total_questions": 1569,
      "avg_cost_per_task_usd": 0.043010272147864875,
      "avg_duration_per_task_ms": 646616.7437858508,
      "avg_output_tokens_per_task": 35899.3862332696,
      "last_run_timestamp": "2026-10-17T16:05:53.852Z"
    },
    {
      "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": 0.8547321572398191,
      "accuracy_std_dev": 0.056526030735110595,
      "run_count": 9,
      "total_questions": 1768,
      "avg_cost_per_task_usd": 0.07859683484162898,
      "avg_duration_per_task_ms": 524667.8444570135,
      "avg_output_tokens_per_task": 21406.976244343892,
      "last_run_timestamp": "2026-10-17T08:31:20.873Z"
    },
    {
      "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-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Baidu",
      "accuracy": 0.853815888888889,
      "accuracy_std_dev": 0.010182476570008156,
      "run_count": 18,
      "total_questions": 3564,
      "avg_cost_per_task_usd": 0.1524213243546577,
      "avg_duration_per_task_ms": 432116.63383838383,
      "avg_output_tokens_per_task": 44104.40488215488,
      "last_run_timestamp": "2026-10-17T16:07:03.542Z"
    },
    {
      "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": "anthropic/claude-5-fable-20260609",
      "model_slug": "anthropic/claude-fable-5",
      "model_name": "Anthropic: Claude Fable 5",
      "provider_name": "Google",
      "accuracy": 0.8533156956521739,
      "accuracy_std_dev": 0.05116508602250153,
      "run_count": 23,
      "total_questions": 4554,
      "avg_cost_per_task_usd": 0.1877648739569609,
      "avg_duration_per_task_ms": 41332.029863855954,
      "avg_output_tokens_per_task": 3381.6989459815545,
      "last_run_timestamp": "2026-10-17T08:20:36.641Z"
    },
    {
      "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": 0.8532947142857142,
      "accuracy_std_dev": 0.03432999370277621,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.015808059163059163,
      "avg_duration_per_task_ms": 355212.3253968254,
      "avg_output_tokens_per_task": 28357.400432900435,
      "last_run_timestamp": "2026-10-17T16:05:30.936Z"
    },
    {
      "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": 0.8531986,
      "accuracy_std_dev": 0.008610563944365058,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.00511969898989899,
      "avg_duration_per_task_ms": 561176.8353535354,
      "avg_output_tokens_per_task": 29257.441414141413,
      "last_run_timestamp": "2026-10-17T16:03:16.417Z"
    },
    {
      "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": "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": "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": 0.8522727499999999,
      "accuracy_std_dev": 0.013435082822130278,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.10542726830808083,
      "avg_duration_per_task_ms": 509657.5732323232,
      "avg_output_tokens_per_task": 30518.208333333332,
      "last_run_timestamp": "2026-10-17T08:18:05.929Z"
    },
    {
      "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": 0.8522533600867679,
      "accuracy_std_dev": 0.014343520343232845,
      "run_count": 7,
      "total_questions": 1383,
      "avg_cost_per_task_usd": 0.09758216124367318,
      "avg_duration_per_task_ms": 298428.6804049169,
      "avg_output_tokens_per_task": 29748.560375994217,
      "last_run_timestamp": "2026-10-17T08:18:07.608Z"
    },
    {
      "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": "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": "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": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "BaseTen",
      "accuracy": 0.8515657312005884,
      "accuracy_std_dev": 0.01705133083065113,
      "run_count": 29,
      "total_questions": 5439,
      "avg_cost_per_task_usd": 0.16341508769994484,
      "avg_duration_per_task_ms": 193665.13403199118,
      "avg_output_tokens_per_task": 29005.72347858062,
      "last_run_timestamp": "2026-10-17T08:21:34.007Z"
    },
    {
      "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": 0.8509335454545452,
      "accuracy_std_dev": 0.009587031319478457,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.03928296464646465,
      "avg_duration_per_task_ms": 395425.2134986226,
      "avg_output_tokens_per_task": 39492.5404040404,
      "last_run_timestamp": "2026-10-17T16:06:26.427Z"
    },
    {
      "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": 0.8507791959910913,
      "accuracy_std_dev": 0.007724931968526794,
      "run_count": 7,
      "total_questions": 1347,
      "avg_cost_per_task_usd": 0.09850432145508536,
      "avg_duration_per_task_ms": 703086.4394951744,
      "avg_output_tokens_per_task": 38320.09428359317,
      "last_run_timestamp": "2026-09-19T08:55:10.654Z"
    },
    {
      "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": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Venice",
      "accuracy": 0.8505509090909091,
      "accuracy_std_dev": 0.01036149420125517,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.09833260835629017,
      "avg_duration_per_task_ms": 575134.6868686868,
      "avg_output_tokens_per_task": 30653.659779614325,
      "last_run_timestamp": "2026-10-17T16:03:32.463Z"
    },
    {
      "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": 0.8504434125475285,
      "accuracy_std_dev": 0.018109786799890826,
      "run_count": 8,
      "total_questions": 1578,
      "avg_cost_per_task_usd": 0.0379920773130545,
      "avg_duration_per_task_ms": 600970.3225602028,
      "avg_output_tokens_per_task": 35466.75095057034,
      "last_run_timestamp": "2026-10-17T16:05:47.230Z"
    },
    {
      "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": "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": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Nebius",
      "accuracy": 0.8501683333333333,
      "accuracy_std_dev": 0.09737163729181557,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.11478661195286197,
      "avg_duration_per_task_ms": 1534934.6153198653,
      "avg_output_tokens_per_task": 28551.546296296296,
      "last_run_timestamp": "2026-10-17T16:06:04.370Z"
    },
    {
      "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": 0.8499579999999999,
      "accuracy_std_dev": 0.02833895562648702,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.07370817487373738,
      "avg_duration_per_task_ms": 515044.92234848486,
      "avg_output_tokens_per_task": 22619.678661616163,
      "last_run_timestamp": "2026-10-17T16:03:30.488Z"
    },
    {
      "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": 0.8499179397894735,
      "accuracy_std_dev": 0.03519465908137437,
      "run_count": 36,
      "total_questions": 7125,
      "avg_cost_per_task_usd": 0.16504187410526316,
      "avg_duration_per_task_ms": 150293.6529122807,
      "avg_output_tokens_per_task": 15256.248701754386,
      "last_run_timestamp": "2026-10-17T08:25:18.147Z"
    },
    {
      "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": "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": 0.8491582000000001,
      "accuracy_std_dev": 0.03436330697648291,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.7065395328282827,
      "avg_duration_per_task_ms": 477955.9121212121,
      "avg_output_tokens_per_task": 28278.689393939392,
      "last_run_timestamp": "2026-10-17T08:18:06.246Z"
    },
    {
      "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": 0.8489898,
      "accuracy_std_dev": 0.014438872440741338,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.020101904040404038,
      "avg_duration_per_task_ms": 416446.1484848485,
      "avg_output_tokens_per_task": 39974.569696969695,
      "last_run_timestamp": "2026-10-17T08:20:29.447Z"
    },
    {
      "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": 0.8489013956043955,
      "accuracy_std_dev": 0.0032350000000000434,
      "run_count": 2,
      "total_questions": 364,
      "avg_cost_per_task_usd": 0.06277839285714286,
      "avg_duration_per_task_ms": 182157.82417582418,
      "avg_output_tokens_per_task": 21243.684065934067,
      "last_run_timestamp": "2026-10-17T08:18:00.399Z"
    },
    {
      "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": 0.8488907957153787,
      "accuracy_std_dev": 0.012007467883961828,
      "run_count": 8,
      "total_questions": 1307,
      "avg_cost_per_task_usd": 0.006448847742922722,
      "avg_duration_per_task_ms": 456157.4078041316,
      "avg_output_tokens_per_task": 15889.71155317521,
      "last_run_timestamp": "2026-10-17T16:03:10.734Z"
    },
    {
      "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": 0.8488849929577466,
      "accuracy_std_dev": 0.010126981561924332,
      "run_count": 3,
      "total_questions": 568,
      "avg_cost_per_task_usd": 0.1339099982394366,
      "avg_duration_per_task_ms": 312089.95246478874,
      "avg_output_tokens_per_task": 20114.045774647886,
      "last_run_timestamp": "2026-10-17T08:18:12.239Z"
    },
    {
      "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": 0.848859786319218,
      "accuracy_std_dev": 0.027189382347480782,
      "run_count": 8,
      "total_questions": 1535,
      "avg_cost_per_task_usd": 0.10762727947882736,
      "avg_duration_per_task_ms": 510720.01042345277,
      "avg_output_tokens_per_task": 33822.35960912052,
      "last_run_timestamp": "2026-10-17T08:23:18.493Z"
    },
    {
      "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": 0.8486092684824902,
      "accuracy_std_dev": 0.00933062433006971,
      "run_count": 12,
      "total_questions": 2313,
      "avg_cost_per_task_usd": 0.06752393212278426,
      "avg_duration_per_task_ms": 201231.72676178123,
      "avg_output_tokens_per_task": 21002.852140077823,
      "last_run_timestamp": "2026-10-17T08:18:07.125Z"
    },
    {
      "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/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8481527067003793,
      "accuracy_std_dev": 0.016458626327349835,
      "run_count": 12,
      "total_questions": 2373,
      "avg_cost_per_task_usd": 0.05132803666245259,
      "avg_duration_per_task_ms": 276383.4787189212,
      "avg_output_tokens_per_task": 45526.28360724821,
      "last_run_timestamp": "2026-10-17T16:03:21.007Z"
    },
    {
      "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": 0.8481498894762135,
      "accuracy_std_dev": 0.02068849935160807,
      "run_count": 11,
      "total_questions": 2081,
      "avg_cost_per_task_usd": 0.13852242095146566,
      "avg_duration_per_task_ms": 288258.8217203268,
      "avg_output_tokens_per_task": 27356.0004805382,
      "last_run_timestamp": "2026-10-17T08:31:19.960Z"
    },
    {
      "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": 0.8478604104595879,
      "accuracy_std_dev": 0.05225305319504382,
      "run_count": 7,
      "total_questions": 1262,
      "avg_cost_per_task_usd": 0.17750794611727413,
      "avg_duration_per_task_ms": 1641892.1576862123,
      "avg_output_tokens_per_task": 51877.93106180666,
      "last_run_timestamp": "2026-10-17T16:05:03.658Z"
    },
    {
      "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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Decart",
      "accuracy": 0.8476008905764583,
      "accuracy_std_dev": 0.021332890049144525,
      "run_count": 15,
      "total_questions": 2897,
      "avg_cost_per_task_usd": 0.19357334760096653,
      "avg_duration_per_task_ms": 273006.10079392476,
      "avg_output_tokens_per_task": 43495.62892647566,
      "last_run_timestamp": "2026-10-17T16:06:16.125Z"
    },
    {
      "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": 0.8475975236486487,
      "accuracy_std_dev": 0.01609637191005063,
      "run_count": 9,
      "total_questions": 1776,
      "avg_cost_per_task_usd": 0.02014133502252253,
      "avg_duration_per_task_ms": 329056.98986486485,
      "avg_output_tokens_per_task": 40068.135698198195,
      "last_run_timestamp": "2026-10-17T08:23:24.777Z"
    },
    {
      "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": 0.8472923333333333,
      "accuracy_std_dev": 0.01746311726340848,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.12316125799663302,
      "avg_duration_per_task_ms": 547448.6262626263,
      "avg_output_tokens_per_task": 46449.42003367003,
      "last_run_timestamp": "2026-10-17T16:03:16.417Z"
    },
    {
      "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": 0.8466919554707381,
      "accuracy_std_dev": 0.010384248473499426,
      "run_count": 8,
      "total_questions": 1572,
      "avg_cost_per_task_usd": 0.007098816157760814,
      "avg_duration_per_task_ms": 559077.6908396947,
      "avg_output_tokens_per_task": 17567.215012722645,
      "last_run_timestamp": "2026-10-17T16:09:39.221Z"
    },
    {
      "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": 0.8466410969581748,
      "accuracy_std_dev": 0.019128826891222965,
      "run_count": 8,
      "total_questions": 1578,
      "avg_cost_per_task_usd": 0.06905485171102661,
      "avg_duration_per_task_ms": 249860.40684410647,
      "avg_output_tokens_per_task": 19302.24651457541,
      "last_run_timestamp": "2026-10-17T08:23:01.048Z"
    },
    {
      "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": 0.8465608571428571,
      "accuracy_std_dev": 0.04257476003263826,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.11109369480519479,
      "avg_duration_per_task_ms": 1258901.4877344877,
      "avg_output_tokens_per_task": 44202.00216450216,
      "last_run_timestamp": "2026-10-17T16:03:19.108Z"
    },
    {
      "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": 0.8464300532319391,
      "accuracy_std_dev": 0.023377056934898808,
      "run_count": 8,
      "total_questions": 1578,
      "avg_cost_per_task_usd": 0.04723669264892269,
      "avg_duration_per_task_ms": 616426.6958174905,
      "avg_output_tokens_per_task": 37996.90114068441,
      "last_run_timestamp": "2026-10-17T16:03:34.738Z"
    },
    {
      "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": 0.8463673576923078,
      "accuracy_std_dev": 0.018423266495368705,
      "run_count": 8,
      "total_questions": 1560,
      "avg_cost_per_task_usd": 0.03161102948717948,
      "avg_duration_per_task_ms": 529309.1025641026,
      "avg_output_tokens_per_task": 32981.63846153846,
      "last_run_timestamp": "2026-10-17T16:03:22.630Z"
    },
    {
      "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": 0.8461700000000001,
      "accuracy_std_dev": 0.023688085037630202,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.01534476388888889,
      "avg_duration_per_task_ms": 558635.6464646464,
      "avg_output_tokens_per_task": 37681.07765151515,
      "last_run_timestamp": "2026-10-17T08:04:11.085Z"
    },
    {
      "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": 0.8459081790540539,
      "accuracy_std_dev": 0.02890872574154731,
      "run_count": 9,
      "total_questions": 1776,
      "avg_cost_per_task_usd": 0.013105496621621622,
      "avg_duration_per_task_ms": 598888.7685810811,
      "avg_output_tokens_per_task": 39477.123873873876,
      "last_run_timestamp": "2026-10-17T08:23:38.535Z"
    },
    {
      "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": 0.8458659444444445,
      "accuracy_std_dev": 0.009899875765057874,
      "run_count": 18,
      "total_questions": 3564,
      "avg_cost_per_task_usd": 0.14837817227833897,
      "avg_duration_per_task_ms": 168870.18967452302,
      "avg_output_tokens_per_task": 32660.4012345679,
      "last_run_timestamp": "2026-10-17T08:26:17.688Z"
    },
    {
      "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": 0.845748875,
      "accuracy_std_dev": 0.019359049895575332,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.1413926414141414,
      "avg_duration_per_task_ms": 904128.5599747475,
      "avg_output_tokens_per_task": 55516.943813131315,
      "last_run_timestamp": "2026-10-17T16:06:34.743Z"
    },
    {
      "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": 0.8453582857142857,
      "accuracy_std_dev": 0.00738922036703828,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.14179908585858583,
      "avg_duration_per_task_ms": 149963.48196248195,
      "avg_output_tokens_per_task": 32052.386724386724,
      "last_run_timestamp": "2026-10-17T08:23:53.645Z"
    },
    {
      "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": 0.8453046666666667,
      "accuracy_std_dev": 0.012280214547528612,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.10640831088664424,
      "avg_duration_per_task_ms": 345181.29517396184,
      "avg_output_tokens_per_task": 30203.571829405162,
      "last_run_timestamp": "2026-10-17T08:18:07.794Z"
    },
    {
      "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": 0.8452150751445088,
      "accuracy_std_dev": 0.013405246218831629,
      "run_count": 8,
      "total_questions": 1557,
      "avg_cost_per_task_usd": 0.08661877906229928,
      "avg_duration_per_task_ms": 1008472.3018625563,
      "avg_output_tokens_per_task": 43824.03468208093,
      "last_run_timestamp": "2026-10-17T16:06:14.503Z"
    },
    {
      "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": 0.8450553324808183,
      "accuracy_std_dev": 0.0035741898382710214,
      "run_count": 4,
      "total_questions": 782,
      "avg_cost_per_task_usd": 0.10689804731457801,
      "avg_duration_per_task_ms": 244684.97186700767,
      "avg_output_tokens_per_task": 30978.448849104858,
      "last_run_timestamp": "2026-10-17T08:31:10.752Z"
    },
    {
      "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": 0.8449956870229007,
      "accuracy_std_dev": 0.02085458095209539,
      "run_count": 8,
      "total_questions": 1572,
      "avg_cost_per_task_usd": 0.10591242111959287,
      "avg_duration_per_task_ms": 201111.52290076335,
      "avg_output_tokens_per_task": 30365.776081424938,
      "last_run_timestamp": "2026-10-17T08:22:58.906Z"
    },
    {
      "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": 0.8445056363636364,
      "accuracy_std_dev": 0.014307964827470473,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.13059047199265383,
      "avg_duration_per_task_ms": 392574.217630854,
      "avg_output_tokens_per_task": 38166.979338842975,
      "last_run_timestamp": "2026-10-17T16:05:05.958Z"
    },
    {
      "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": 0.8444203561384532,
      "accuracy_std_dev": 0.09907049002682446,
      "run_count": 19,
      "total_questions": 3698,
      "avg_cost_per_task_usd": 0.008343435640886965,
      "avg_duration_per_task_ms": 607566.9085992428,
      "avg_output_tokens_per_task": 27823.074364521362,
      "last_run_timestamp": "2026-10-17T08:31:21.367Z"
    },
    {
      "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-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Alibaba",
      "accuracy": 0.8442990780261495,
      "accuracy_std_dev": 0.014090709122552674,
      "run_count": 12,
      "total_questions": 2371,
      "avg_cost_per_task_usd": 0.08855882834247152,
      "avg_duration_per_task_ms": 405312.7507380852,
      "avg_output_tokens_per_task": 22020.992408266553,
      "last_run_timestamp": "2026-10-17T16:03:21.028Z"
    },
    {
      "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": 0.8442945100174215,
      "accuracy_std_dev": 0.015586349631435004,
      "run_count": 12,
      "total_questions": 2296,
      "avg_cost_per_task_usd": 0.02107914808362369,
      "avg_duration_per_task_ms": 263998.6659407666,
      "avg_output_tokens_per_task": 37097.8218641115,
      "last_run_timestamp": "2026-10-17T08:22:57.139Z"
    },
    {
      "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": 0.8440790973451325,
      "accuracy_std_dev": 0.04414326163627795,
      "run_count": 12,
      "total_questions": 2373,
      "avg_cost_per_task_usd": 0.16686697387273494,
      "avg_duration_per_task_ms": 164309.61483354404,
      "avg_output_tokens_per_task": 16467.685208596715,
      "last_run_timestamp": "2026-10-17T08:02:36.464Z"
    },
    {
      "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": "Sail Research",
      "accuracy": 0.8434823631462101,
      "accuracy_std_dev": 0.011605077252105817,
      "run_count": 16,
      "total_questions": 3153,
      "avg_cost_per_task_usd": 0.11780266603235014,
      "avg_duration_per_task_ms": 324853.04471931496,
      "avg_output_tokens_per_task": 34347.03774183318,
      "last_run_timestamp": "2026-10-17T08:23:25.036Z"
    },
    {
      "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": 0.8433546514886163,
      "accuracy_std_dev": 0.019866857652736186,
      "run_count": 26,
      "total_questions": 5139,
      "avg_cost_per_task_usd": 0.24385491282350655,
      "avg_duration_per_task_ms": 503356.83965752093,
      "avg_output_tokens_per_task": 47257.87332165791,
      "last_run_timestamp": "2026-10-17T16:06:05.782Z"
    },
    {
      "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": 0.8432919898534386,
      "accuracy_std_dev": 0.016413678415754425,
      "run_count": 9,
      "total_questions": 1774,
      "avg_cost_per_task_usd": 0.009925733934611049,
      "avg_duration_per_task_ms": 452405.15670800454,
      "avg_output_tokens_per_task": 31483.80439684329,
      "last_run_timestamp": "2026-10-17T08:31:20.138Z"
    },
    {
      "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": 0.8430333386243387,
      "accuracy_std_dev": 0.009092680280725797,
      "run_count": 8,
      "total_questions": 1323,
      "avg_cost_per_task_usd": 0.13841274300831444,
      "avg_duration_per_task_ms": 126887.37868480726,
      "avg_output_tokens_per_task": 31223.573696145126,
      "last_run_timestamp": "2026-10-17T08:26:17.949Z"
    },
    {
      "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": 0.8427329166666666,
      "accuracy_std_dev": 0.033483254955420955,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.05271298779461279,
      "avg_duration_per_task_ms": 212758.7373737374,
      "avg_output_tokens_per_task": 21862.314814814814,
      "last_run_timestamp": "2026-10-17T08:20:51.734Z"
    },
    {
      "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-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Together",
      "accuracy": 0.8421298152723189,
      "accuracy_std_dev": 0.012413964315426865,
      "run_count": 9,
      "total_questions": 1781,
      "avg_cost_per_task_usd": 0.13897624312184165,
      "avg_duration_per_task_ms": 200873.80123526108,
      "avg_output_tokens_per_task": 31356.491297024142,
      "last_run_timestamp": "2026-10-17T08:26:31.170Z"
    },
    {
      "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": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "Z.AI",
      "accuracy": 0.8416337193675889,
      "accuracy_std_dev": 0.027997748302318905,
      "run_count": 7,
      "total_questions": 1265,
      "avg_cost_per_task_usd": 0.09091536363636364,
      "avg_duration_per_task_ms": 1534224.6964426877,
      "avg_output_tokens_per_task": 41867.99051383399,
      "last_run_timestamp": "2026-10-17T16:06:12.818Z"
    },
    {
      "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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Nebius",
      "accuracy": 0.8406283333333334,
      "accuracy_std_dev": 0.014825937482069168,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.2733430589225589,
      "avg_duration_per_task_ms": 397861.94276094274,
      "avg_output_tokens_per_task": 61869.08417508417,
      "last_run_timestamp": "2026-10-17T16:03:34.735Z"
    },
    {
      "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": 0.8405724,
      "accuracy_std_dev": 0.051235218214037107,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.1625054691919192,
      "avg_duration_per_task_ms": 161406.5494949495,
      "avg_output_tokens_per_task": 16082.878282828282,
      "last_run_timestamp": "2026-10-17T08:20:51.327Z"
    },
    {
      "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": "Reka",
      "accuracy": 0.8401923283835383,
      "accuracy_std_dev": 0.010616197274749037,
      "run_count": 12,
      "total_questions": 2357,
      "avg_cost_per_task_usd": 0.07210711752227408,
      "avg_duration_per_task_ms": 388115.4365719134,
      "avg_output_tokens_per_task": 21949.04836656767,
      "last_run_timestamp": "2026-10-17T08:18:57.702Z"
    },
    {
      "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": 0.839947,
      "accuracy_std_dev": 0.010969701701895635,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.006264753968253967,
      "avg_duration_per_task_ms": 399380.0064935065,
      "avg_output_tokens_per_task": 25886.00432900433,
      "last_run_timestamp": "2026-10-17T16:05:38.823Z"
    },
    {
      "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": 0.8398760924440692,
      "accuracy_std_dev": 0.02907933486267448,
      "run_count": 12,
      "total_questions": 2369,
      "avg_cost_per_task_usd": 0.01278074419586323,
      "avg_duration_per_task_ms": 235302.36935415788,
      "avg_output_tokens_per_task": 14226.744195863233,
      "last_run_timestamp": "2026-10-17T08:26:18.302Z"
    },
    {
      "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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Novita",
      "accuracy": 0.8397788936464089,
      "accuracy_std_dev": 0.056010342147862845,
      "run_count": 11,
      "total_questions": 2172,
      "avg_cost_per_task_usd": 0.0787299079189687,
      "avg_duration_per_task_ms": 572401.2780847145,
      "avg_output_tokens_per_task": 44061.1864640884,
      "last_run_timestamp": "2026-10-17T16:07:16.346Z"
    },
    {
      "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": 0.8395436603119584,
      "accuracy_std_dev": 0.010997134534013122,
      "run_count": 7,
      "total_questions": 1154,
      "avg_cost_per_task_usd": 0.006737056325823224,
      "avg_duration_per_task_ms": 622594.8396880416,
      "avg_output_tokens_per_task": 16640.68890814558,
      "last_run_timestamp": "2026-10-17T16:06:04.564Z"
    },
    {
      "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": 0.8395171528268551,
      "accuracy_std_dev": 0.020340296193762753,
      "run_count": 8,
      "total_questions": 1132,
      "avg_cost_per_task_usd": 0.016312568021201415,
      "avg_duration_per_task_ms": 87770.31007067137,
      "avg_output_tokens_per_task": 13991.570671378091,
      "last_run_timestamp": "2026-10-17T16:05:36.869Z"
    },
    {
      "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": 0.8384655804464972,
      "accuracy_std_dev": 0.017858184675894877,
      "run_count": 7,
      "total_questions": 1299,
      "avg_cost_per_task_usd": 0.15115439722863744,
      "avg_duration_per_task_ms": 297429.6466512702,
      "avg_output_tokens_per_task": 35258.00923787529,
      "last_run_timestamp": "2026-10-17T08:21:08.719Z"
    },
    {
      "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": 0.8383907751724139,
      "accuracy_std_dev": 0.014934618806956231,
      "run_count": 11,
      "total_questions": 2175,
      "avg_cost_per_task_usd": 0.22592824689655172,
      "avg_duration_per_task_ms": 279235.183908046,
      "avg_output_tokens_per_task": 51209.457931034485,
      "last_run_timestamp": "2026-10-17T16:05:55.392Z"
    },
    {
      "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": 0.8378307503892062,
      "accuracy_std_dev": 0.03618934998158991,
      "run_count": 10,
      "total_questions": 1927,
      "avg_cost_per_task_usd": 0.029675656979761288,
      "avg_duration_per_task_ms": 291090.3741567203,
      "avg_output_tokens_per_task": 9165.982355993772,
      "last_run_timestamp": "2026-10-17T08:18:00.022Z"
    },
    {
      "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": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "StreamLake",
      "accuracy": 0.8378130592547341,
      "accuracy_std_dev": 0.02159011628523284,
      "run_count": 9,
      "total_questions": 1637,
      "avg_cost_per_task_usd": 0.07307316065974342,
      "avg_duration_per_task_ms": 388815.5760537569,
      "avg_output_tokens_per_task": 24094.505803298718,
      "last_run_timestamp": "2026-10-17T16:05:55.624Z"
    },
    {
      "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": 0.8371951481707317,
      "accuracy_std_dev": 0.02877358996641808,
      "run_count": 9,
      "total_questions": 1640,
      "avg_cost_per_task_usd": 0.011173640853658538,
      "avg_duration_per_task_ms": 373621.1006097561,
      "avg_output_tokens_per_task": 22130.590853658538,
      "last_run_timestamp": "2026-10-17T08:20:32.592Z"
    },
    {
      "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": 0.836700375,
      "accuracy_std_dev": 0.05226306970973264,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.09356622285353536,
      "avg_duration_per_task_ms": 408622.6205808081,
      "avg_output_tokens_per_task": 21042.78724747475,
      "last_run_timestamp": "2026-10-17T16:03:34.731Z"
    },
    {
      "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": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Chutes",
      "accuracy": 0.8365635357637288,
      "accuracy_std_dev": 0.030013675310646196,
      "run_count": 11,
      "total_questions": 2167,
      "avg_cost_per_task_usd": 0.06634258514074758,
      "avg_duration_per_task_ms": 1054714.62113521,
      "avg_output_tokens_per_task": 33093.440239963085,
      "last_run_timestamp": "2026-10-17T16:03:24.591Z"
    },
    {
      "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": 0.8364136921606119,
      "accuracy_std_dev": 0.02209302746873717,
      "run_count": 8,
      "total_questions": 1569,
      "avg_cost_per_task_usd": 0.09194236265137029,
      "avg_duration_per_task_ms": 1028125.6876991715,
      "avg_output_tokens_per_task": 42157.6908859146,
      "last_run_timestamp": "2026-10-17T16:07:08.166Z"
    },
    {
      "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": 0.8364074347826086,
      "accuracy_std_dev": 0.0120242022634283,
      "run_count": 23,
      "total_questions": 4554,
      "avg_cost_per_task_usd": 0.26040066996047434,
      "avg_duration_per_task_ms": 271238.8100570927,
      "avg_output_tokens_per_task": 56706.782608695656,
      "last_run_timestamp": "2026-10-17T16:06:15.929Z"
    },
    {
      "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": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "SiliconFlow",
      "accuracy": 0.836088,
      "accuracy_std_dev": 0.012542317105050277,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.08354139898989901,
      "avg_duration_per_task_ms": 748339.285123967,
      "avg_output_tokens_per_task": 48591.798898071625,
      "last_run_timestamp": "2026-10-17T16:06:53.742Z"
    },
    {
      "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": 0.836088,
      "accuracy_std_dev": 0.011196343721379466,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.24613506887052342,
      "avg_duration_per_task_ms": 366252.56749311293,
      "avg_output_tokens_per_task": 39848.296143250685,
      "last_run_timestamp": "2026-10-17T08:20:12.299Z"
    },
    {
      "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": 0.8360502124789209,
      "accuracy_std_dev": 0.025222744943188383,
      "run_count": 9,
      "total_questions": 1779,
      "avg_cost_per_task_usd": 0.15836738055087127,
      "avg_duration_per_task_ms": 914252.053400787,
      "avg_output_tokens_per_task": 52022.07419898819,
      "last_run_timestamp": "2026-10-17T16:04:06.832Z"
    },
    {
      "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": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Nebius",
      "accuracy": 0.8358585,
      "accuracy_std_dev": 0.014603658180401272,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.15205888257575756,
      "avg_duration_per_task_ms": 235007.2601010101,
      "avg_output_tokens_per_task": 34304.49873737374,
      "last_run_timestamp": "2026-10-17T08:23:37.906Z"
    },
    {
      "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": 0.8357648888888889,
      "accuracy_std_dev": 0.023650418015776994,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.11090620426487094,
      "avg_duration_per_task_ms": 397711.55667789,
      "avg_output_tokens_per_task": 36242.744668911335,
      "last_run_timestamp": "2026-10-17T08:23:46.532Z"
    },
    {
      "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": 0.8357119447513812,
      "accuracy_std_dev": 0.030735575803944197,
      "run_count": 11,
      "total_questions": 2172,
      "avg_cost_per_task_usd": 0.01723706399631676,
      "avg_duration_per_task_ms": 153182.64871086556,
      "avg_output_tokens_per_task": 18067.730662983424,
      "last_run_timestamp": "2026-10-17T16:05:57.017Z"
    },
    {
      "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": 0.8356714444444444,
      "accuracy_std_dev": 0.013858410451668465,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.14028259147025812,
      "avg_duration_per_task_ms": 183217.68013468012,
      "avg_output_tokens_per_task": 31651.560044893376,
      "last_run_timestamp": "2026-10-17T08:23:38.082Z"
    },
    {
      "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": 0.8356163613013698,
      "accuracy_std_dev": 0.008544206082681626,
      "run_count": 6,
      "total_questions": 1168,
      "avg_cost_per_task_usd": 0.02781799058219178,
      "avg_duration_per_task_ms": 605091.2380136986,
      "avg_output_tokens_per_task": 14771.433219178081,
      "last_run_timestamp": "2026-10-17T08:19:50.290Z"
    },
    {
      "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/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "DekaLLM",
      "accuracy": 0.8345362173215719,
      "accuracy_std_dev": 0.009596989300434438,
      "run_count": 7,
      "total_questions": 1247,
      "avg_cost_per_task_usd": 0.045831162790697676,
      "avg_duration_per_task_ms": 89512.38492381717,
      "avg_output_tokens_per_task": 15115.87570168404,
      "last_run_timestamp": "2026-10-17T08:23:46.620Z"
    },
    {
      "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": 0.8343871437605396,
      "accuracy_std_dev": 0.02214593854208312,
      "run_count": 12,
      "total_questions": 2372,
      "avg_cost_per_task_usd": 0.024504262225969645,
      "avg_duration_per_task_ms": 383677.22807757166,
      "avg_output_tokens_per_task": 45255.54173693086,
      "last_run_timestamp": "2026-10-17T08:18:07.356Z"
    },
    {
      "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": "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": "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": 0.8340346666666668,
      "accuracy_std_dev": 0.033680117491217625,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.05396650210437711,
      "avg_duration_per_task_ms": 255758.39562289562,
      "avg_output_tokens_per_task": 22970.753367003366,
      "last_run_timestamp": "2026-10-17T08:02:36.130Z"
    },
    {
      "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": 0.8340288847157016,
      "accuracy_std_dev": 0.01861503359435616,
      "run_count": 11,
      "total_questions": 1917,
      "avg_cost_per_task_usd": 0.09581860772039645,
      "avg_duration_per_task_ms": 622289.2008346375,
      "avg_output_tokens_per_task": 29867.270213875847,
      "last_run_timestamp": "2026-10-17T16:03:34.668Z"
    },
    {
      "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-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Friendli",
      "accuracy": 0.8337731531353135,
      "accuracy_std_dev": 0.013206984544172061,
      "run_count": 8,
      "total_questions": 1515,
      "avg_cost_per_task_usd": 0.22087456369636965,
      "avg_duration_per_task_ms": 445733.97095709573,
      "avg_output_tokens_per_task": 50080.22772277228,
      "last_run_timestamp": "2026-10-17T16:03:34.732Z"
    },
    {
      "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": 0.83367,
      "accuracy_std_dev": 0.01234350860169023,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.0391534101010101,
      "avg_duration_per_task_ms": 225964.9994949495,
      "avg_output_tokens_per_task": 14531.921212121211,
      "last_run_timestamp": "2026-10-17T08:02:36.310Z"
    },
    {
      "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": 0.8336229634994206,
      "accuracy_std_dev": 0.021738738072566568,
      "run_count": 9,
      "total_questions": 1726,
      "avg_cost_per_task_usd": 0.020348153534183084,
      "avg_duration_per_task_ms": 236255.9015063731,
      "avg_output_tokens_per_task": 40476.10834298957,
      "last_run_timestamp": "2026-10-17T08:31:21.023Z"
    },
    {
      "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": 0.8334862727272728,
      "accuracy_std_dev": 0.06978429003083315,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.20374753076216714,
      "avg_duration_per_task_ms": 537318.2309458219,
      "avg_output_tokens_per_task": 46322.687786960516,
      "last_run_timestamp": "2026-10-17T16:06:55.201Z"
    },
    {
      "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": 0.8333332476370511,
      "accuracy_std_dev": 0.03496673477303106,
      "run_count": 4,
      "total_questions": 529,
      "avg_cost_per_task_usd": 0.09262325141776938,
      "avg_duration_per_task_ms": 324581.95463138,
      "avg_output_tokens_per_task": 38346.877126654064,
      "last_run_timestamp": "2026-10-17T16:03:14.481Z"
    },
    {
      "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-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "AkashML",
      "accuracy": 0.8332576446056212,
      "accuracy_std_dev": 0.012460556042274078,
      "run_count": 14,
      "total_questions": 2206,
      "avg_cost_per_task_usd": 0.11502464823209428,
      "avg_duration_per_task_ms": 224648.71078875792,
      "avg_output_tokens_per_task": 27449.897552130555,
      "last_run_timestamp": "2026-10-17T08:03:37.237Z"
    },
    {
      "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-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Cloudflare",
      "accuracy": 0.8328165280620154,
      "accuracy_std_dev": 0.017273176086970456,
      "run_count": 17,
      "total_questions": 3225,
      "avg_cost_per_task_usd": 0.23915214232558138,
      "avg_duration_per_task_ms": 374125.24062015506,
      "avg_output_tokens_per_task": 42949.212093023256,
      "last_run_timestamp": "2026-10-17T16:06:00.836Z"
    },
    {
      "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": 0.832801601148692,
      "accuracy_std_dev": 0.01796862582217962,
      "run_count": 8,
      "total_questions": 1567,
      "avg_cost_per_task_usd": 0.015661317804722397,
      "avg_duration_per_task_ms": 438016.0848755584,
      "avg_output_tokens_per_task": 40296.81493299298,
      "last_run_timestamp": "2026-10-17T08:23:38.619Z"
    },
    {
      "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": 0.8327431107738998,
      "accuracy_std_dev": 0.03389053866730361,
      "run_count": 10,
      "total_questions": 1977,
      "avg_cost_per_task_usd": 0.10782147395042994,
      "avg_duration_per_task_ms": 1015295.3773394031,
      "avg_output_tokens_per_task": 33688.65048052605,
      "last_run_timestamp": "2026-10-17T16:03:19.092Z"
    },
    {
      "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": 0.8321570888335409,
      "accuracy_std_dev": 0.0106410052060675,
      "run_count": 15,
      "total_questions": 2409,
      "avg_cost_per_task_usd": 0.01938649107513491,
      "avg_duration_per_task_ms": 213034.18430884185,
      "avg_output_tokens_per_task": 38582.406392694065,
      "last_run_timestamp": "2026-10-17T08:20:36.345Z"
    },
    {
      "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": 0.8320707500000001,
      "accuracy_std_dev": 0.008392927678756304,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.08125586531986531,
      "avg_duration_per_task_ms": 697034.4953703703,
      "avg_output_tokens_per_task": 56943.53914141414,
      "last_run_timestamp": "2026-10-17T16:07:34.171Z"
    },
    {
      "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": "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": "Cerebras",
      "accuracy": 0.8316496666666667,
      "accuracy_std_dev": 0.009013484577133449,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.02192354882154882,
      "avg_duration_per_task_ms": 12405.745791245792,
      "avg_output_tokens_per_task": 14165.61616161616,
      "last_run_timestamp": "2026-10-17T08:04:11.170Z"
    },
    {
      "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": 0.8315218454968942,
      "accuracy_std_dev": 0.015272735801327252,
      "run_count": 7,
      "total_questions": 1288,
      "avg_cost_per_task_usd": 0.022463093167701866,
      "avg_duration_per_task_ms": 302352.67624223605,
      "avg_output_tokens_per_task": 44704.51242236025,
      "last_run_timestamp": "2026-10-17T08:21:10.036Z"
    },
    {
      "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": "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": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "CoreWeave",
      "accuracy": 0.8308479888164026,
      "accuracy_std_dev": 0.015453198666473883,
      "run_count": 14,
      "total_questions": 2146,
      "avg_cost_per_task_usd": 0.037517430102516305,
      "avg_duration_per_task_ms": 263231.2437092265,
      "avg_output_tokens_per_task": 29907.408201304752,
      "last_run_timestamp": "2026-10-17T16:03:24.620Z"
    },
    {
      "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": 0.8307303478466634,
      "accuracy_std_dev": 0.010455075072751015,
      "run_count": 11,
      "total_questions": 2113,
      "avg_cost_per_task_usd": 0.22723341410317086,
      "avg_duration_per_task_ms": 294845.04164694744,
      "avg_output_tokens_per_task": 51477.59678182679,
      "last_run_timestamp": "2026-10-17T16:05:16.008Z"
    },
    {
      "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": "Wafer",
      "accuracy": 0.8305271666666666,
      "accuracy_std_dev": 0.004924217515391559,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.05513053872053872,
      "avg_duration_per_task_ms": 198190.63888888888,
      "avg_output_tokens_per_task": 15260.853535353535,
      "last_run_timestamp": "2026-10-17T08:22:58.669Z"
    },
    {
      "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": 0.8304393877813504,
      "accuracy_std_dev": 0.01344638632449253,
      "run_count": 8,
      "total_questions": 1555,
      "avg_cost_per_task_usd": 0.1521527459807074,
      "avg_duration_per_task_ms": 258681.17106109325,
      "avg_output_tokens_per_task": 34390.49389067524,
      "last_run_timestamp": "2026-10-17T08:31:20.051Z"
    },
    {
      "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": 0.8302596014344261,
      "accuracy_std_dev": 0.00648568653960398,
      "run_count": 10,
      "total_questions": 1952,
      "avg_cost_per_task_usd": 0.03510589190573771,
      "avg_duration_per_task_ms": 205657.60348360657,
      "avg_output_tokens_per_task": 14193.383709016394,
      "last_run_timestamp": "2026-10-17T08:30:44.166Z"
    },
    {
      "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": 0.8286928213457077,
      "accuracy_std_dev": 0.0309478649146514,
      "run_count": 7,
      "total_questions": 1293,
      "avg_cost_per_task_usd": 0.006722985305491105,
      "avg_duration_per_task_ms": 672885.3240525909,
      "avg_output_tokens_per_task": 18073.616395978344,
      "last_run_timestamp": "2026-10-17T16:03:30.487Z"
    },
    {
      "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-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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Friendli",
      "accuracy": 0.8282827272727273,
      "accuracy_std_dev": 0.019724543474799548,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.2436875546372819,
      "avg_duration_per_task_ms": 366201.29155188246,
      "avg_output_tokens_per_task": 55218.82966023875,
      "last_run_timestamp": "2026-10-17T16:06:12.901Z"
    },
    {
      "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": 0.8279999858823529,
      "accuracy_std_dev": 0.022369675456957054,
      "run_count": 11,
      "total_questions": 2125,
      "avg_cost_per_task_usd": 0.02507603670588236,
      "avg_duration_per_task_ms": 878373.8503529412,
      "avg_output_tokens_per_task": 46965.459294117645,
      "last_run_timestamp": "2026-10-17T08:23:38.171Z"
    },
    {
      "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": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "Baidu",
      "accuracy": 0.8278469506242907,
      "accuracy_std_dev": 0.008508360090646084,
      "run_count": 9,
      "total_questions": 1762,
      "avg_cost_per_task_usd": 0.007694182179341657,
      "avg_duration_per_task_ms": 337110.9665153235,
      "avg_output_tokens_per_task": 19055.32803632236,
      "last_run_timestamp": "2026-10-17T16:06:33.562Z"
    },
    {
      "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": 0.8276515833333333,
      "accuracy_std_dev": 0.024461777935036868,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.038061598905723915,
      "avg_duration_per_task_ms": 345575.1603535354,
      "avg_output_tokens_per_task": 39167.65488215488,
      "last_run_timestamp": "2026-10-17T16:06:09.632Z"
    },
    {
      "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": 0.8274411249999999,
      "accuracy_std_dev": 0.01202255100256912,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.03823982765151515,
      "avg_duration_per_task_ms": 264254.53851010103,
      "avg_output_tokens_per_task": 15517.1773989899,
      "last_run_timestamp": "2026-10-17T08:18:05.671Z"
    },
    {
      "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": 0.8273644545454546,
      "accuracy_std_dev": 0.02235000962929887,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.13616948393021117,
      "avg_duration_per_task_ms": 546895.1822773187,
      "avg_output_tokens_per_task": 43108.64462809917,
      "last_run_timestamp": "2026-10-17T16:03:32.459Z"
    },
    {
      "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": 0.8273474444444444,
      "accuracy_std_dev": 0.006968947666631396,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.13947370089786754,
      "avg_duration_per_task_ms": 169790.02132435466,
      "avg_output_tokens_per_task": 34460.64870931538,
      "last_run_timestamp": "2026-10-17T08:18:00.675Z"
    },
    {
      "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": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "AkashML",
      "accuracy": 0.8272847449686571,
      "accuracy_std_dev": 0.021180314014064436,
      "run_count": 16,
      "total_questions": 3031,
      "avg_cost_per_task_usd": 0.04213694556252062,
      "avg_duration_per_task_ms": 414089.77202243486,
      "avg_output_tokens_per_task": 45769.54503464203,
      "last_run_timestamp": "2026-10-17T16:03:24.608Z"
    },
    {
      "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": 0.8271603333333334,
      "accuracy_std_dev": 0.018009865870313054,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.12849964365881034,
      "avg_duration_per_task_ms": 324347.8726150393,
      "avg_output_tokens_per_task": 53019.92480359147,
      "last_run_timestamp": "2026-10-17T16:06:38.059Z"
    },
    {
      "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": "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": "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": 0.8264122222222224,
      "accuracy_std_dev": 0.010177010309491996,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.04327315263748596,
      "avg_duration_per_task_ms": 229752.91245791246,
      "avg_output_tokens_per_task": 14308.176767676769,
      "last_run_timestamp": "2026-10-17T08:20:04.890Z"
    },
    {
      "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": 0.8262167272727274,
      "accuracy_std_dev": 0.05836834255140663,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.05152247704315887,
      "avg_duration_per_task_ms": 533182.7640036731,
      "avg_output_tokens_per_task": 24667.44031221304,
      "last_run_timestamp": "2026-10-17T08:02:36.555Z"
    },
    {
      "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": 0.8261783333333332,
      "accuracy_std_dev": 0.014287137159774959,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.06237826599326599,
      "avg_duration_per_task_ms": 538208.1666666666,
      "avg_output_tokens_per_task": 49120.39562289562,
      "last_run_timestamp": "2026-10-17T16:07:11.061Z"
    },
    {
      "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": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "CoreWeave",
      "accuracy": 0.8258873707224335,
      "accuracy_std_dev": 0.009676826643527784,
      "run_count": 6,
      "total_questions": 1052,
      "avg_cost_per_task_usd": 0.025350776615969583,
      "avg_duration_per_task_ms": 120122.08840304182,
      "avg_output_tokens_per_task": 20151.38783269962,
      "last_run_timestamp": "2026-08-29T18:29:13.946Z"
    },
    {
      "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": 0.8254771666666667,
      "accuracy_std_dev": 0.00940686748102449,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.14361566498316497,
      "avg_duration_per_task_ms": 426029.37457912456,
      "avg_output_tokens_per_task": 28741.069865319867,
      "last_run_timestamp": "2026-10-17T08:21:06.486Z"
    },
    {
      "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": 0.8254611153846155,
      "accuracy_std_dev": 0.015660417265753186,
      "run_count": 8,
      "total_questions": 1482,
      "avg_cost_per_task_usd": 0.024416639676113364,
      "avg_duration_per_task_ms": 282189.7125506073,
      "avg_output_tokens_per_task": 48622.69095816464,
      "last_run_timestamp": "2026-10-17T08:18:07.522Z"
    },
    {
      "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": 0.8249891001283697,
      "accuracy_std_dev": 0.05097064513194498,
      "run_count": 12,
      "total_questions": 2337,
      "avg_cost_per_task_usd": 0.09995508172871204,
      "avg_duration_per_task_ms": 541954.0753102268,
      "avg_output_tokens_per_task": 29446.99828840394,
      "last_run_timestamp": "2026-10-18T01:36:16.673Z"
    },
    {
      "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": 0.8249519000000001,
      "accuracy_std_dev": 0.01363772648859694,
      "run_count": 9,
      "total_questions": 1730,
      "avg_cost_per_task_usd": 0.024441546242774562,
      "avg_duration_per_task_ms": 449419.536416185,
      "avg_output_tokens_per_task": 24400.59768786127,
      "last_run_timestamp": "2026-10-17T16:07:00.439Z"
    },
    {
      "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": 0.824916,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.022821656565656566,
      "avg_duration_per_task_ms": 516406.73232323234,
      "avg_output_tokens_per_task": 45403.40909090909,
      "last_run_timestamp": "2026-10-04T03:04:38.670Z"
    },
    {
      "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": 0.8245299743589743,
      "accuracy_std_dev": 0.015168242439056692,
      "run_count": 10,
      "total_questions": 1950,
      "avg_cost_per_task_usd": 0.1439138517948718,
      "avg_duration_per_task_ms": 294902.17794871796,
      "avg_output_tokens_per_task": 32491.769743589743,
      "last_run_timestamp": "2026-10-17T08:19:50.385Z"
    },
    {
      "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": "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": "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": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "AkashML",
      "accuracy": 0.8237281106367841,
      "accuracy_std_dev": 0.017561497390472025,
      "run_count": 15,
      "total_questions": 2811,
      "avg_cost_per_task_usd": 0.03114107612949128,
      "avg_duration_per_task_ms": 263866.4489505514,
      "avg_output_tokens_per_task": 14017.920313055853,
      "last_run_timestamp": "2026-10-17T08:18:05.906Z"
    },
    {
      "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": 0.8236842684210525,
      "accuracy_std_dev": 0.00005100000000002325,
      "run_count": 2,
      "total_questions": 380,
      "avg_cost_per_task_usd": 0.11858948157894737,
      "avg_duration_per_task_ms": 197797.2657894737,
      "avg_output_tokens_per_task": 29495,
      "last_run_timestamp": "2026-10-17T08:18:07.270Z"
    },
    {
      "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": 0.8236161787072243,
      "accuracy_std_dev": 0.007591822309852583,
      "run_count": 8,
      "total_questions": 1578,
      "avg_cost_per_task_usd": 0.007953439797211661,
      "avg_duration_per_task_ms": 483107.97084917617,
      "avg_output_tokens_per_task": 20467.734474017743,
      "last_run_timestamp": "2026-10-17T16:03:30.494Z"
    },
    {
      "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": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8234428749999999,
      "accuracy_std_dev": 0.008223286849513092,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.04775078093434344,
      "avg_duration_per_task_ms": 439035.39835858584,
      "avg_output_tokens_per_task": 26449.10542929293,
      "last_run_timestamp": "2026-10-17T16:06:45.443Z"
    },
    {
      "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": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Chutes",
      "accuracy": 0.8232322999999999,
      "accuracy_std_dev": 0.020878107083976735,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.03450791515151515,
      "avg_duration_per_task_ms": 265271.5378787879,
      "avg_output_tokens_per_task": 14464.131818181819,
      "last_run_timestamp": "2026-10-17T08:19:16.529Z"
    },
    {
      "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": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8222643415559772,
      "accuracy_std_dev": 0.014072964998868393,
      "run_count": 8,
      "total_questions": 1581,
      "avg_cost_per_task_usd": 0.1947691688804554,
      "avg_duration_per_task_ms": 969270.2144212524,
      "avg_output_tokens_per_task": 51840.38583175206,
      "last_run_timestamp": "2026-10-17T16:06:43.912Z"
    },
    {
      "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": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "auto-routing",
      "accuracy": 0.8220689337931035,
      "accuracy_std_dev": 0.04423033551106625,
      "run_count": 11,
      "total_questions": 2175,
      "avg_cost_per_task_usd": 0.085449148045977,
      "avg_duration_per_task_ms": 780476.7949425287,
      "avg_output_tokens_per_task": 31176.493793103447,
      "last_run_timestamp": "2026-10-17T16:03:16.416Z"
    },
    {
      "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": "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": "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": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "Novita",
      "accuracy": 0.8203461428571428,
      "accuracy_std_dev": 0.030285949285476416,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.09649056854256852,
      "avg_duration_per_task_ms": 1458719.1861471862,
      "avg_output_tokens_per_task": 44637.40115440115,
      "last_run_timestamp": "2026-10-17T16:06:17.114Z"
    },
    {
      "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": 0.8201458333333331,
      "accuracy_std_dev": 0.013691135080976862,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.314472340909091,
      "avg_duration_per_task_ms": 289534.5925925926,
      "avg_output_tokens_per_task": 20775.11153198653,
      "last_run_timestamp": "2026-10-17T08:04:11.104Z"
    },
    {
      "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": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "CoreWeave",
      "accuracy": 0.8197122727272727,
      "accuracy_std_dev": 0.018601160931956084,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.042628178145087244,
      "avg_duration_per_task_ms": 234970.43434343435,
      "avg_output_tokens_per_task": 14213.333333333334,
      "last_run_timestamp": "2026-10-17T08:22:58.890Z"
    },
    {
      "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": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Mancer 2",
      "accuracy": 0.8188131249999999,
      "accuracy_std_dev": 0.0780194627391741,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.01174183143939394,
      "avg_duration_per_task_ms": 760343.6117424242,
      "avg_output_tokens_per_task": 25978.727272727272,
      "last_run_timestamp": "2026-10-17T16:11:29.249Z"
    },
    {
      "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": 0.8187420045395589,
      "accuracy_std_dev": 0.05002500478245231,
      "run_count": 8,
      "total_questions": 1542,
      "avg_cost_per_task_usd": 0.02404102204928664,
      "avg_duration_per_task_ms": 913283.17769131,
      "avg_output_tokens_per_task": 46875.09792477302,
      "last_run_timestamp": "2026-10-17T08:26:17.943Z"
    },
    {
      "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": 0.8186202474674383,
      "accuracy_std_dev": 0.008502623002248598,
      "run_count": 7,
      "total_questions": 1382,
      "avg_cost_per_task_usd": 0.03867190303907381,
      "avg_duration_per_task_ms": 207333.16425470333,
      "avg_output_tokens_per_task": 14772.572358900145,
      "last_run_timestamp": "2026-10-17T08:18:02.516Z"
    },
    {
      "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": 0.8185119845094664,
      "accuracy_std_dev": 0.016376070492197817,
      "run_count": 9,
      "total_questions": 1743,
      "avg_cost_per_task_usd": 0.018653541594951233,
      "avg_duration_per_task_ms": 606686.3998852553,
      "avg_output_tokens_per_task": 20040.86230636833,
      "last_run_timestamp": "2026-10-17T16:06:52.537Z"
    },
    {
      "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": 0.8183687777777777,
      "accuracy_std_dev": 0.014898891142764636,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.03439966722783389,
      "avg_duration_per_task_ms": 222983.87822671156,
      "avg_output_tokens_per_task": 13813.295735129068,
      "last_run_timestamp": "2026-10-17T08:18:01.964Z"
    },
    {
      "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": "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": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Alibaba",
      "accuracy": 0.8175460387794024,
      "accuracy_std_dev": 0.054722959283415716,
      "run_count": 8,
      "total_questions": 1573,
      "avg_cost_per_task_usd": 0.1901260775588048,
      "avg_duration_per_task_ms": 615386.0864589956,
      "avg_output_tokens_per_task": 45248.49332485696,
      "last_run_timestamp": "2026-10-17T16:04:35.937Z"
    },
    {
      "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": "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": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8168566996197718,
      "accuracy_std_dev": 0.004401271634425658,
      "run_count": 8,
      "total_questions": 1578,
      "avg_cost_per_task_usd": 0.009628157160963245,
      "avg_duration_per_task_ms": 739454.9480354879,
      "avg_output_tokens_per_task": 18256.190114068442,
      "last_run_timestamp": "2026-10-17T16:06:46.647Z"
    },
    {
      "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": 0.8165709598662209,
      "accuracy_std_dev": 0.061324306686151714,
      "run_count": 22,
      "total_questions": 4186,
      "avg_cost_per_task_usd": 0.0074439737219302434,
      "avg_duration_per_task_ms": 498928.61681796465,
      "avg_output_tokens_per_task": 17412.904204491162,
      "last_run_timestamp": "2026-10-17T16:06:04.476Z"
    },
    {
      "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": 0.81649825,
      "accuracy_std_dev": 0.019083748339555836,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.006335979166666667,
      "avg_duration_per_task_ms": 484646.10353535356,
      "avg_output_tokens_per_task": 18725.272727272728,
      "last_run_timestamp": "2026-10-17T16:03:10.763Z"
    },
    {
      "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": 0.81607725,
      "accuracy_std_dev": 0.00955115719101616,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.03420622222222222,
      "avg_duration_per_task_ms": 636556.8263888889,
      "avg_output_tokens_per_task": 16058.70707070707,
      "last_run_timestamp": "2026-10-17T08:19:36.361Z"
    },
    {
      "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": "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": 0.8159994814814814,
      "accuracy_std_dev": 0.01147137633051603,
      "run_count": 27,
      "total_questions": 5346,
      "avg_cost_per_task_usd": 0.3335022566404788,
      "avg_duration_per_task_ms": 298311.0364758698,
      "avg_output_tokens_per_task": 21104.580621025067,
      "last_run_timestamp": "2026-10-17T08:30:40.050Z"
    },
    {
      "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": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "Venice",
      "accuracy": 0.8156565,
      "accuracy_std_dev": 0.006011157355285233,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.028145230429292924,
      "avg_duration_per_task_ms": 190452.8042929293,
      "avg_output_tokens_per_task": 23529.740530303032,
      "last_run_timestamp": "2026-10-17T16:06:46.963Z"
    },
    {
      "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": 0.8155935315487571,
      "accuracy_std_dev": 0.02579145123843122,
      "run_count": 8,
      "total_questions": 1569,
      "avg_cost_per_task_usd": 0.2145485825366476,
      "avg_duration_per_task_ms": 977552.3288718929,
      "avg_output_tokens_per_task": 57980.598470363286,
      "last_run_timestamp": "2026-10-17T16:03:22.660Z"
    },
    {
      "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": "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": 0.8140431250000001,
      "accuracy_std_dev": 0.010669274703217085,
      "run_count": 24,
      "total_questions": 4752,
      "avg_cost_per_task_usd": 0.36383448947811453,
      "avg_duration_per_task_ms": 358426.1837121212,
      "avg_output_tokens_per_task": 22914.688552188552,
      "last_run_timestamp": "2026-10-17T08:22:56.641Z"
    },
    {
      "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": 0.814024,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 164,
      "avg_cost_per_task_usd": 0.010100225609756097,
      "avg_duration_per_task_ms": 515065.23780487804,
      "avg_output_tokens_per_task": 33245.84146341463,
      "last_run_timestamp": "2026-09-28T09:46:11.365Z"
    },
    {
      "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": "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": 0.8132656083650188,
      "accuracy_std_dev": 0.029437471710177493,
      "run_count": 8,
      "total_questions": 1578,
      "avg_cost_per_task_usd": 0.08244444803548796,
      "avg_duration_per_task_ms": 931196.6774397972,
      "avg_output_tokens_per_task": 40999.92395437262,
      "last_run_timestamp": "2026-10-17T16:03:45.986Z"
    },
    {
      "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": "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": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "auto-routing",
      "accuracy": 0.8128908571428571,
      "accuracy_std_dev": 0.014743239462097008,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.007136471861471861,
      "avg_duration_per_task_ms": 527522.6536796537,
      "avg_output_tokens_per_task": 18109.714285714286,
      "last_run_timestamp": "2026-10-17T16:03:10.732Z"
    },
    {
      "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": 0.8124998749999999,
      "accuracy_std_dev": 0.02094152012413078,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.04008979356060606,
      "avg_duration_per_task_ms": 181327.2178030303,
      "avg_output_tokens_per_task": 22170.065025252527,
      "last_run_timestamp": "2026-10-17T16:04:35.857Z"
    },
    {
      "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": "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": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "Venice",
      "accuracy": 0.8119721658218682,
      "accuracy_std_dev": 0.11936196199179458,
      "run_count": 7,
      "total_questions": 1381,
      "avg_cost_per_task_usd": 0.12433009341057204,
      "avg_duration_per_task_ms": 711621.3982621289,
      "avg_output_tokens_per_task": 38654.58653149891,
      "last_run_timestamp": "2026-10-17T16:04:59.484Z"
    },
    {
      "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": "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": 0.8115756996197718,
      "accuracy_std_dev": 0.04867340225790031,
      "run_count": 8,
      "total_questions": 1578,
      "avg_cost_per_task_usd": 0.2293247072243346,
      "avg_duration_per_task_ms": 1102307.0380228136,
      "avg_output_tokens_per_task": 52484.05069708492,
      "last_run_timestamp": "2026-10-17T16:06:39.412Z"
    },
    {
      "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": 0.8112779399727148,
      "accuracy_std_dev": 0.010023193016698797,
      "run_count": 4,
      "total_questions": 733,
      "avg_cost_per_task_usd": 0.04357236971350614,
      "avg_duration_per_task_ms": 179585.35061391542,
      "avg_output_tokens_per_task": 17263.93315143247,
      "last_run_timestamp": "2026-10-17T08:18:01.976Z"
    },
    {
      "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": 0.8111110666666667,
      "accuracy_std_dev": 0.017159518883957357,
      "run_count": 15,
      "total_questions": 2970,
      "avg_cost_per_task_usd": 0.34046912828282827,
      "avg_duration_per_task_ms": 354667.00067340065,
      "avg_output_tokens_per_task": 22526.755218855218,
      "last_run_timestamp": "2026-10-17T08:02:36.395Z"
    },
    {
      "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": 0.8110267499999999,
      "accuracy_std_dev": 0.02579051758762317,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.008924909090909091,
      "avg_duration_per_task_ms": 574972.908459596,
      "avg_output_tokens_per_task": 18934.403409090908,
      "last_run_timestamp": "2026-10-17T16:03:42.306Z"
    },
    {
      "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": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Novita",
      "accuracy": 0.8098576959847037,
      "accuracy_std_dev": 0.02469341221579907,
      "run_count": 8,
      "total_questions": 1569,
      "avg_cost_per_task_usd": 0.25613483683875077,
      "avg_duration_per_task_ms": 954692.9738687061,
      "avg_output_tokens_per_task": 58301.671128107075,
      "last_run_timestamp": "2026-10-17T16:05:55.712Z"
    },
    {
      "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": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "DigitalOcean",
      "accuracy": 0.8089225000000001,
      "accuracy_std_dev": 0.05752497129290896,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.003809875420875421,
      "avg_duration_per_task_ms": 924304.0791245791,
      "avg_output_tokens_per_task": 22694.977272727272,
      "last_run_timestamp": "2026-10-17T16:06:12.644Z"
    },
    {
      "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": 0.8084547777777777,
      "accuracy_std_dev": 0.02465025412885807,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.11369982884399552,
      "avg_duration_per_task_ms": 767797.1189674523,
      "avg_output_tokens_per_task": 59144.13187429854,
      "last_run_timestamp": "2026-10-17T16:08:52.854Z"
    },
    {
      "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": "anthropic/claude-4.5-sonnet-20250929",
      "model_slug": "anthropic/claude-sonnet-4.5",
      "model_name": "Anthropic: Claude Sonnet 4.5",
      "provider_name": "Azure",
      "accuracy": 0.80780025,
      "accuracy_std_dev": 0.020912351996387992,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.34150990530303027,
      "avg_duration_per_task_ms": 356238.11531986535,
      "avg_output_tokens_per_task": 22606.00042087542,
      "last_run_timestamp": "2026-10-17T08:20:12.481Z"
    },
    {
      "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": 0.8075196666666667,
      "accuracy_std_dev": 0.03714760001519464,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.025326144781144783,
      "avg_duration_per_task_ms": 166431.35521885523,
      "avg_output_tokens_per_task": 13572.52356902357,
      "last_run_timestamp": "2026-10-17T08:21:10.969Z"
    },
    {
      "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": 0.8073469346289753,
      "accuracy_std_dev": 0.10831830089636926,
      "run_count": 12,
      "total_questions": 2264,
      "avg_cost_per_task_usd": 0.04317355300353357,
      "avg_duration_per_task_ms": 164819.90106007067,
      "avg_output_tokens_per_task": 20660.478798586573,
      "last_run_timestamp": "2026-10-17T08:17:08.134Z"
    },
    {
      "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": 0.806818,
      "accuracy_std_dev": 0.005877369607230755,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.01103122095959596,
      "avg_duration_per_task_ms": 646747.8522727273,
      "avg_output_tokens_per_task": 11881.43686868687,
      "last_run_timestamp": "2026-10-17T08:23:15.092Z"
    },
    {
      "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": "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": 0.8050504,
      "accuracy_std_dev": 0.021411457634640384,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.3368577297979798,
      "avg_duration_per_task_ms": 353947.41616161616,
      "avg_output_tokens_per_task": 22328.277777777777,
      "last_run_timestamp": "2026-10-17T08:20:36.170Z"
    },
    {
      "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": 0.8050312243186584,
      "accuracy_std_dev": 0.024402124224901022,
      "run_count": 4,
      "total_questions": 477,
      "avg_cost_per_task_usd": 0.0531789392033543,
      "avg_duration_per_task_ms": 296093.35429769394,
      "avg_output_tokens_per_task": 21712.16142557652,
      "last_run_timestamp": "2026-10-17T16:03:24.612Z"
    },
    {
      "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/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "DigitalOcean",
      "accuracy": 0.8045648025477706,
      "accuracy_std_dev": 0.09929505313576492,
      "run_count": 8,
      "total_questions": 1570,
      "avg_cost_per_task_usd": 0.05945578917197451,
      "avg_duration_per_task_ms": 1750115.5261146496,
      "avg_output_tokens_per_task": 19217.150955414014,
      "last_run_timestamp": "2026-10-17T16:03:32.459Z"
    },
    {
      "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": 0.8035647794928336,
      "accuracy_std_dev": 0.02820532797579602,
      "run_count": 6,
      "total_questions": 907,
      "avg_cost_per_task_usd": 0.06920113671444322,
      "avg_duration_per_task_ms": 1464272.5214994487,
      "avg_output_tokens_per_task": 39612.85556780596,
      "last_run_timestamp": "2026-10-17T16:03:32.444Z"
    },
    {
      "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": 0.8033669999999999,
      "accuracy_std_dev": 0.005573328377190791,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.048433724242424245,
      "avg_duration_per_task_ms": 54676.08484848485,
      "avg_output_tokens_per_task": 15399.329292929293,
      "last_run_timestamp": "2026-10-17T08:18:07.897Z"
    },
    {
      "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": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Alibaba",
      "accuracy": 0.801499818181818,
      "accuracy_std_dev": 0.1018190627747603,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.09262380762167126,
      "avg_duration_per_task_ms": 515813.015610652,
      "avg_output_tokens_per_task": 34167.608815427,
      "last_run_timestamp": "2026-10-17T16:03:43.963Z"
    },
    {
      "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": "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": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Chutes",
      "accuracy": 0.7991169867549668,
      "accuracy_std_dev": 0.04305133271491851,
      "run_count": 7,
      "total_questions": 1359,
      "avg_cost_per_task_usd": 0.14345824724061812,
      "avg_duration_per_task_ms": 1569797.9367181752,
      "avg_output_tokens_per_task": 47091.807947019865,
      "last_run_timestamp": "2026-10-17T16:05:55.538Z"
    },
    {
      "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": 0.7989417142857143,
      "accuracy_std_dev": 0.13167886968217607,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.15234512626262628,
      "avg_duration_per_task_ms": 2001837.8780663782,
      "avg_output_tokens_per_task": 54356.18470418471,
      "last_run_timestamp": "2026-10-17T16:08:20.692Z"
    },
    {
      "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": "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": 0.7969274999999999,
      "accuracy_std_dev": 0.041893006701596396,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.026948493055555557,
      "avg_duration_per_task_ms": 347172.4311868687,
      "avg_output_tokens_per_task": 25834.242424242424,
      "last_run_timestamp": "2026-10-17T16:03:42.311Z"
    },
    {
      "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": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "AtlasCloud",
      "accuracy": 0.79541113957935,
      "accuracy_std_dev": 0.024018187027945608,
      "run_count": 8,
      "total_questions": 1569,
      "avg_cost_per_task_usd": 0.2794700968769917,
      "avg_duration_per_task_ms": 930887.5340981517,
      "avg_output_tokens_per_task": 70418.68706182281,
      "last_run_timestamp": "2026-10-17T16:03:21.031Z"
    },
    {
      "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": 0.79523250996483,
      "accuracy_std_dev": 0.10177729700753506,
      "run_count": 5,
      "total_questions": 853,
      "avg_cost_per_task_usd": 0.014341097303634235,
      "avg_duration_per_task_ms": 277183.26846424385,
      "avg_output_tokens_per_task": 30270.443141852287,
      "last_run_timestamp": "2026-10-17T08:23:52.399Z"
    },
    {
      "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": "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": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Venice",
      "accuracy": 0.7948342770467838,
      "accuracy_std_dev": 0.09161678915984332,
      "run_count": 7,
      "total_questions": 1368,
      "avg_cost_per_task_usd": 0.09734662865497076,
      "avg_duration_per_task_ms": 830552.8106725146,
      "avg_output_tokens_per_task": 29487.959795321636,
      "last_run_timestamp": "2026-10-17T16:06:10.732Z"
    },
    {
      "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": 0.7937710000000001,
      "accuracy_std_dev": 0.014359120716464491,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.01779377714646465,
      "avg_duration_per_task_ms": 59921.61174242424,
      "avg_output_tokens_per_task": 5776.905303030303,
      "last_run_timestamp": "2026-10-17T16:06:09.008Z"
    },
    {
      "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": 0.7936945454545455,
      "accuracy_std_dev": 0.03387125348987259,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.11027911937557391,
      "avg_duration_per_task_ms": 415904.7438016529,
      "avg_output_tokens_per_task": 30768.579889807163,
      "last_run_timestamp": "2026-10-17T16:06:51.142Z"
    },
    {
      "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": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7919672857142857,
      "accuracy_std_dev": 0.008482616505271861,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.19117722655122654,
      "avg_duration_per_task_ms": 824934.1883116884,
      "avg_output_tokens_per_task": 47794.95887445888,
      "last_run_timestamp": "2026-10-17T16:03:24.604Z"
    },
    {
      "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": 0.7919196568825911,
      "accuracy_std_dev": 0.020933482740337288,
      "run_count": 10,
      "total_questions": 1976,
      "avg_cost_per_task_usd": 0.06286805870445344,
      "avg_duration_per_task_ms": 174709.42408906884,
      "avg_output_tokens_per_task": 19709.32439271255,
      "last_run_timestamp": "2026-10-17T08:25:47.232Z"
    },
    {
      "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": 0.7911756164047152,
      "accuracy_std_dev": 0.01560584292357742,
      "run_count": 11,
      "total_questions": 2036,
      "avg_cost_per_task_usd": 0.0046643629666011785,
      "avg_duration_per_task_ms": 1348073.47151277,
      "avg_output_tokens_per_task": 30640.777013752457,
      "last_run_timestamp": "2026-10-17T16:03:22.656Z"
    },
    {
      "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/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "auto-routing",
      "accuracy": 0.7896387272727272,
      "accuracy_std_dev": 0.03757319952038085,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.04347873507805325,
      "avg_duration_per_task_ms": 494226.878328742,
      "avg_output_tokens_per_task": 42653.03902662994,
      "last_run_timestamp": "2026-10-17T16:06:11.609Z"
    },
    {
      "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": "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": 0.7886248285229202,
      "accuracy_std_dev": 0.005658817544328477,
      "run_count": 3,
      "total_questions": 589,
      "avg_cost_per_task_usd": 0.011348117147707979,
      "avg_duration_per_task_ms": 785846.1052631579,
      "avg_output_tokens_per_task": 38718.01867572156,
      "last_run_timestamp": "2026-10-17T16:06:33.477Z"
    },
    {
      "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": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "Together",
      "accuracy": 0.787996238141336,
      "accuracy_std_dev": 0.01502120070857429,
      "run_count": 11,
      "total_questions": 2066,
      "avg_cost_per_task_usd": 0.00856900435624395,
      "avg_duration_per_task_ms": 373289.4123910939,
      "avg_output_tokens_per_task": 33861.243949661184,
      "last_run_timestamp": "2026-10-17T16:07:01.286Z"
    },
    {
      "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": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "DekaLLM",
      "accuracy": 0.785340410122164,
      "accuracy_std_dev": 0.005585783044678892,
      "run_count": 3,
      "total_questions": 573,
      "avg_cost_per_task_usd": 0.037465542757417106,
      "avg_duration_per_task_ms": 218978.30017452006,
      "avg_output_tokens_per_task": 37410.67888307155,
      "last_run_timestamp": "2026-10-17T16:06:05.435Z"
    },
    {
      "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": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "Darkbloom",
      "accuracy": 0.7845117999999999,
      "accuracy_std_dev": 0.012186465958595225,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.02083039292929293,
      "avg_duration_per_task_ms": 462638.8303030303,
      "avg_output_tokens_per_task": 27689.152525252524,
      "last_run_timestamp": "2026-10-17T16:06:16.035Z"
    },
    {
      "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": 0.7843049551569508,
      "accuracy_std_dev": 0.09329604363444359,
      "run_count": 10,
      "total_questions": 1115,
      "avg_cost_per_task_usd": 0.005419434080717489,
      "avg_duration_per_task_ms": 254231.5470852018,
      "avg_output_tokens_per_task": 21881.403587443947,
      "last_run_timestamp": "2026-10-17T16:03:19.088Z"
    },
    {
      "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": "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": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "CoreWeave",
      "accuracy": 0.7828282222222223,
      "accuracy_std_dev": 0.011110583660814166,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.02110236139169473,
      "avg_duration_per_task_ms": 267338.53815937147,
      "avg_output_tokens_per_task": 12550.465207631874,
      "last_run_timestamp": "2026-10-17T16:03:19.114Z"
    },
    {
      "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": 0.7827002835443038,
      "accuracy_std_dev": 0.025807959273061464,
      "run_count": 5,
      "total_questions": 790,
      "avg_cost_per_task_usd": 0.010936145569620253,
      "avg_duration_per_task_ms": 1215812.0873417722,
      "avg_output_tokens_per_task": 39346.51392405063,
      "last_run_timestamp": "2026-10-17T08:02:36.403Z"
    },
    {
      "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": 0.7812311687388988,
      "accuracy_std_dev": 0.010295797880688991,
      "run_count": 3,
      "total_questions": 563,
      "avg_cost_per_task_usd": 0.001959150976909414,
      "avg_duration_per_task_ms": 798438.0159857904,
      "avg_output_tokens_per_task": 12884.238010657193,
      "last_run_timestamp": "2026-10-17T16:05:11.871Z"
    },
    {
      "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": 0.7807238749999998,
      "accuracy_std_dev": 0.039711866691322574,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.1739194539141414,
      "avg_duration_per_task_ms": 546996.5315656565,
      "avg_output_tokens_per_task": 54147.905303030304,
      "last_run_timestamp": "2026-10-17T16:03:24.611Z"
    },
    {
      "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": "Parasail",
      "accuracy": 0.7805356401045069,
      "accuracy_std_dev": 0.13188409621496255,
      "run_count": 8,
      "total_questions": 1531,
      "avg_cost_per_task_usd": 0.023446755062050944,
      "avg_duration_per_task_ms": 365068.11038536904,
      "avg_output_tokens_per_task": 23915.368386675374,
      "last_run_timestamp": "2026-10-17T16:05:01.165Z"
    },
    {
      "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": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "auto-routing",
      "accuracy": 0.7799202727272726,
      "accuracy_std_dev": 0.017481939064962863,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.0056759981634527095,
      "avg_duration_per_task_ms": 639660.3314967861,
      "avg_output_tokens_per_task": 35506.24150596878,
      "last_run_timestamp": "2026-10-17T16:03:16.417Z"
    },
    {
      "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": 0.7796947507042253,
      "accuracy_std_dev": 0.11796703504024969,
      "run_count": 8,
      "total_questions": 1420,
      "avg_cost_per_task_usd": 0.03984405704225352,
      "avg_duration_per_task_ms": 145067.09014084507,
      "avg_output_tokens_per_task": 30331.589436619717,
      "last_run_timestamp": "2026-10-17T16:04:40.679Z"
    },
    {
      "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": "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": 0.7782009841269842,
      "accuracy_std_dev": 0.02997378382749665,
      "run_count": 8,
      "total_questions": 1575,
      "avg_cost_per_task_usd": 0.011125198095238096,
      "avg_duration_per_task_ms": 566928,
      "avg_output_tokens_per_task": 36841.713015873014,
      "last_run_timestamp": "2026-10-17T16:03:22.655Z"
    },
    {
      "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": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "Venice",
      "accuracy": 0.7771783756345177,
      "accuracy_std_dev": 0.059728767079685954,
      "run_count": 8,
      "total_questions": 1576,
      "avg_cost_per_task_usd": 0.007288401015228427,
      "avg_duration_per_task_ms": 899378.7354060913,
      "avg_output_tokens_per_task": 18512.973350253807,
      "last_run_timestamp": "2026-10-17T16:05:43.347Z"
    },
    {
      "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": 0.7766553333333335,
      "accuracy_std_dev": 0.10513747605555944,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.04385584792368125,
      "avg_duration_per_task_ms": 1356759.8063973065,
      "avg_output_tokens_per_task": 25066.917508417508,
      "last_run_timestamp": "2026-10-17T16:07:07.180Z"
    },
    {
      "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": 0.776515,
      "accuracy_std_dev": 0.0045135097762162586,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.005179621212121213,
      "avg_duration_per_task_ms": 982954.0037878788,
      "avg_output_tokens_per_task": 28613.36111111111,
      "last_run_timestamp": "2026-10-17T16:08:37.371Z"
    },
    {
      "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": 0.7762077565217391,
      "accuracy_std_dev": 0.01894682591908638,
      "run_count": 7,
      "total_questions": 1380,
      "avg_cost_per_task_usd": 0.022503373188405798,
      "avg_duration_per_task_ms": 924139.8405797102,
      "avg_output_tokens_per_task": 38821.189855072465,
      "last_run_timestamp": "2026-10-17T16:08:08.338Z"
    },
    {
      "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": 0.775813521906226,
      "accuracy_std_dev": 0.014886571160397556,
      "run_count": 9,
      "total_questions": 1301,
      "avg_cost_per_task_usd": 0.047757321291314374,
      "avg_duration_per_task_ms": 944137.5941583398,
      "avg_output_tokens_per_task": 24378.20368946964,
      "last_run_timestamp": "2026-10-17T16:03:42.289Z"
    },
    {
      "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-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "auto-routing",
      "accuracy": 0.7749893918918921,
      "accuracy_std_dev": 0.055734405977703745,
      "run_count": 8,
      "total_questions": 1554,
      "avg_cost_per_task_usd": 0.15352950514800515,
      "avg_duration_per_task_ms": 1000120.4671814672,
      "avg_output_tokens_per_task": 63630.38416988417,
      "last_run_timestamp": "2026-10-17T16:03:14.481Z"
    },
    {
      "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": "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": 0.7713499090909091,
      "accuracy_std_dev": 0.010348443040156889,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.014670483011937557,
      "avg_duration_per_task_ms": 135015.37970615242,
      "avg_output_tokens_per_task": 29281.215794306703,
      "last_run_timestamp": "2026-10-17T16:03:42.602Z"
    },
    {
      "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": 0.7708977708978327,
      "accuracy_std_dev": 0.0364570846062103,
      "run_count": 8,
      "total_questions": 969,
      "avg_cost_per_task_usd": 0.012726153766769865,
      "avg_duration_per_task_ms": 52839.89783281734,
      "avg_output_tokens_per_task": 16873.008255933953,
      "last_run_timestamp": "2026-10-17T16:06:07.257Z"
    },
    {
      "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": 0.7707982259154932,
      "accuracy_std_dev": 0.0072861311144724605,
      "run_count": 9,
      "total_questions": 1775,
      "avg_cost_per_task_usd": 0.011357562816901407,
      "avg_duration_per_task_ms": 995701.5797183098,
      "avg_output_tokens_per_task": 11816.376338028169,
      "last_run_timestamp": "2026-10-17T16:06:12.733Z"
    },
    {
      "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Parasail",
      "accuracy": 0.7691922196969697,
      "accuracy_std_dev": 0.017375228086595713,
      "run_count": 7,
      "total_questions": 1320,
      "avg_cost_per_task_usd": 0.0214991303030303,
      "avg_duration_per_task_ms": 238124.97348484848,
      "avg_output_tokens_per_task": 28579.5696969697,
      "last_run_timestamp": "2026-10-17T16:05:11.736Z"
    },
    {
      "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": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Reka",
      "accuracy": 0.7676531051401869,
      "accuracy_std_dev": 0.011937876359567938,
      "run_count": 13,
      "total_questions": 2568,
      "avg_cost_per_task_usd": 0.01795761292834891,
      "avg_duration_per_task_ms": 102616.90381619938,
      "avg_output_tokens_per_task": 5949.5965732087225,
      "last_run_timestamp": "2026-10-17T08:25:18.584Z"
    },
    {
      "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": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "GMICloud",
      "accuracy": 0.7671156666666666,
      "accuracy_std_dev": 0.03093018545405481,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.013063415824915825,
      "avg_duration_per_task_ms": 936092.1515151515,
      "avg_output_tokens_per_task": 46359.02693602694,
      "last_run_timestamp": "2026-10-10T08:15:34.059Z"
    },
    {
      "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": "AkashML",
      "accuracy": 0.7664968651509212,
      "accuracy_std_dev": 0.016788365025716064,
      "run_count": 13,
      "total_questions": 2551,
      "avg_cost_per_task_usd": 0.007439355154841239,
      "avg_duration_per_task_ms": 466577.11054488434,
      "avg_output_tokens_per_task": 27687.860446883577,
      "last_run_timestamp": "2026-10-17T16:03:42.289Z"
    },
    {
      "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": 0.7664264643091028,
      "accuracy_std_dev": 0.008541751796323752,
      "run_count": 8,
      "total_questions": 1527,
      "avg_cost_per_task_usd": 0.024611972495088407,
      "avg_duration_per_task_ms": 71012.95219384413,
      "avg_output_tokens_per_task": 25773.081859855927,
      "last_run_timestamp": "2026-10-17T16:06:09.019Z"
    },
    {
      "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": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Nebius",
      "accuracy": 0.7660061679586564,
      "accuracy_std_dev": 0.07448120643864166,
      "run_count": 6,
      "total_questions": 1161,
      "avg_cost_per_task_usd": 0.18327362360034452,
      "avg_duration_per_task_ms": 1579588.0456503015,
      "avg_output_tokens_per_task": 43785.853574504734,
      "last_run_timestamp": "2026-10-17T16:07:01.369Z"
    },
    {
      "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": "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": 0.7659312647058824,
      "accuracy_std_dev": 0.018460106716917977,
      "run_count": 4,
      "total_questions": 544,
      "avg_cost_per_task_usd": 0.008997586397058823,
      "avg_duration_per_task_ms": 678227.3382352941,
      "avg_output_tokens_per_task": 29866.545955882353,
      "last_run_timestamp": "2026-10-17T16:03:22.651Z"
    },
    {
      "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": "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Near AI",
      "accuracy": 0.7630475000000001,
      "accuracy_std_dev": 0.0012624999999999997,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.004469671717171717,
      "avg_duration_per_task_ms": 567471.6388888889,
      "avg_output_tokens_per_task": 29671.171717171717,
      "last_run_timestamp": "2026-10-17T16:08:23.187Z"
    },
    {
      "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": "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": 0.7622055000000001,
      "accuracy_std_dev": 0.008018844134287667,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.020353424242424242,
      "avg_duration_per_task_ms": 1224475.143939394,
      "avg_output_tokens_per_task": 61537.31691919192,
      "last_run_timestamp": "2026-10-17T16:06:04.991Z"
    },
    {
      "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": 0.7620653333333334,
      "accuracy_std_dev": 0.00883726150393256,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.007579895622895623,
      "avg_duration_per_task_ms": 139557.0117845118,
      "avg_output_tokens_per_task": 30124.809764309764,
      "last_run_timestamp": "2026-10-17T16:06:03.146Z"
    },
    {
      "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": 0.7619047142857143,
      "accuracy_std_dev": 0.011624017325155306,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.020688753246753247,
      "avg_duration_per_task_ms": 1247623.8001443,
      "avg_output_tokens_per_task": 51654.42640692641,
      "last_run_timestamp": "2026-10-17T16:03:24.588Z"
    },
    {
      "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": 0.7606508265895953,
      "accuracy_std_dev": 0.044831136424469266,
      "run_count": 8,
      "total_questions": 1557,
      "avg_cost_per_task_usd": 0.27017780154142584,
      "avg_duration_per_task_ms": 889465.8985228003,
      "avg_output_tokens_per_task": 74206.37700706488,
      "last_run_timestamp": "2026-10-17T16:03:34.730Z"
    },
    {
      "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": "Groq",
      "accuracy": 0.75925925,
      "accuracy_std_dev": 0.006627946132664344,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.017339029040404044,
      "avg_duration_per_task_ms": 75536.6938131313,
      "avg_output_tokens_per_task": 28657.905303030304,
      "last_run_timestamp": "2026-10-17T16:06:10.556Z"
    },
    {
      "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": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "DeepInfra",
      "accuracy": 0.7587235606060606,
      "accuracy_std_dev": 0.03366533728388146,
      "run_count": 8,
      "total_questions": 1452,
      "avg_cost_per_task_usd": 0.007390471074380165,
      "avg_duration_per_task_ms": 981052.2692837466,
      "avg_output_tokens_per_task": 19052.973829201102,
      "last_run_timestamp": "2026-10-17T16:06:14.585Z"
    },
    {
      "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": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Xiaomi",
      "accuracy": 0.758698,
      "accuracy_std_dev": 0.0082855039677741,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.015277286195286194,
      "avg_duration_per_task_ms": 1138394.335016835,
      "avg_output_tokens_per_task": 54274.19528619529,
      "last_run_timestamp": "2026-10-17T08:20:47.658Z"
    },
    {
      "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": 0.7586279375000001,
      "accuracy_std_dev": 0.013868153578742699,
      "run_count": 16,
      "total_questions": 3168,
      "avg_cost_per_task_usd": 0.018340981376262628,
      "avg_duration_per_task_ms": 158098.73516414143,
      "avg_output_tokens_per_task": 30346.9375,
      "last_run_timestamp": "2026-10-17T16:04:59.074Z"
    },
    {
      "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": 0.7584345607287448,
      "accuracy_std_dev": 0.07943384181065398,
      "run_count": 5,
      "total_questions": 988,
      "avg_cost_per_task_usd": 0.01775063663967611,
      "avg_duration_per_task_ms": 1537506.3319838056,
      "avg_output_tokens_per_task": 17162.693319838058,
      "last_run_timestamp": "2026-10-17T16:14:35.207Z"
    },
    {
      "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": 0.7578456145038168,
      "accuracy_std_dev": 0.01544961858701939,
      "run_count": 8,
      "total_questions": 1572,
      "avg_cost_per_task_usd": 0.01670373155216285,
      "avg_duration_per_task_ms": 254619.6227735369,
      "avg_output_tokens_per_task": 28145.37468193384,
      "last_run_timestamp": "2026-10-17T16:11:05.478Z"
    },
    {
      "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": 0.7576013994334279,
      "accuracy_std_dev": 0.11080526429777046,
      "run_count": 9,
      "total_questions": 1765,
      "avg_cost_per_task_usd": 0.15527972067988668,
      "avg_duration_per_task_ms": 887994.8634560907,
      "avg_output_tokens_per_task": 41127.971104815864,
      "last_run_timestamp": "2026-10-18T01:36:49.342Z"
    },
    {
      "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": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "Baidu",
      "accuracy": 0.7573266413662239,
      "accuracy_std_dev": 0.03112028552949828,
      "run_count": 8,
      "total_questions": 1581,
      "avg_cost_per_task_usd": 0.1659841872232764,
      "avg_duration_per_task_ms": 1040091.2226438962,
      "avg_output_tokens_per_task": 75445.36812144212,
      "last_run_timestamp": "2026-10-17T16:03:32.455Z"
    },
    {
      "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": 0.7571548999999999,
      "accuracy_std_dev": 0.12945420603166974,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.01753179696969697,
      "avg_duration_per_task_ms": 657106.2575757576,
      "avg_output_tokens_per_task": 36803.128282828286,
      "last_run_timestamp": "2026-10-17T08:23:46.892Z"
    },
    {
      "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": 0.7558923750000001,
      "accuracy_std_dev": 0.009411207626249413,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.018470106691919192,
      "avg_duration_per_task_ms": 142999.88257575757,
      "avg_output_tokens_per_task": 30551.185606060608,
      "last_run_timestamp": "2026-10-17T16:07:16.153Z"
    },
    {
      "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": 0.7554704302103251,
      "accuracy_std_dev": 0.05900138631408906,
      "run_count": 8,
      "total_questions": 1569,
      "avg_cost_per_task_usd": 0.2827103932441045,
      "avg_duration_per_task_ms": 965963.304652645,
      "avg_output_tokens_per_task": 81541.58444869344,
      "last_run_timestamp": "2026-10-17T16:03:32.461Z"
    },
    {
      "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": 0.7551755057034221,
      "accuracy_std_dev": 0.031175143900028734,
      "run_count": 8,
      "total_questions": 1578,
      "avg_cost_per_task_usd": 0.004692086185044361,
      "avg_duration_per_task_ms": 725750.8219264892,
      "avg_output_tokens_per_task": 28116.21609632446,
      "last_run_timestamp": "2026-10-17T16:06:08.928Z"
    },
    {
      "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": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "DeepInfra",
      "accuracy": 0.7545171538940811,
      "accuracy_std_dev": 0.03419234805098943,
      "run_count": 10,
      "total_questions": 1605,
      "avg_cost_per_task_usd": 0.04500766791277259,
      "avg_duration_per_task_ms": 265542.445482866,
      "avg_output_tokens_per_task": 16958.05109034268,
      "last_run_timestamp": "2026-10-17T08:22:59.031Z"
    },
    {
      "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": "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/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "Alibaba",
      "accuracy": 0.7500000000000001,
      "accuracy_std_dev": 0.14215609823271738,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.029263358585858587,
      "avg_duration_per_task_ms": 133657.62815656565,
      "avg_output_tokens_per_task": 22413.67297979798,
      "last_run_timestamp": "2026-10-17T16:06:08.676Z"
    },
    {
      "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": 0.75,
      "accuracy_std_dev": 0.038766301716181284,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.05368984785353535,
      "avg_duration_per_task_ms": 976528.220959596,
      "avg_output_tokens_per_task": 27872.888257575756,
      "last_run_timestamp": "2026-10-17T16:03:34.731Z"
    },
    {
      "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": "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": 0.7485289698857738,
      "accuracy_std_dev": 0.027178252463320746,
      "run_count": 5,
      "total_questions": 963,
      "avg_cost_per_task_usd": 0.01670032710280374,
      "avg_duration_per_task_ms": 258104.83489096572,
      "avg_output_tokens_per_task": 53131.220145379026,
      "last_run_timestamp": "2026-10-17T08:04:11.005Z"
    },
    {
      "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": 0.7485157480916032,
      "accuracy_std_dev": 0.032894842349333445,
      "run_count": 4,
      "total_questions": 786,
      "avg_cost_per_task_usd": 0.014386558524173029,
      "avg_duration_per_task_ms": 1033415.2888040713,
      "avg_output_tokens_per_task": 51046.291348600505,
      "last_run_timestamp": "2026-10-17T08:02:02.330Z"
    },
    {
      "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": "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": "qwen/qwen3-coder-next-2025-02-03",
      "model_slug": "qwen/qwen3-coder-next",
      "model_name": "Qwen: Qwen3 Coder Next",
      "provider_name": "auto-routing",
      "accuracy": 0.7467534285714286,
      "accuracy_std_dev": 0.009517321679925722,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.015003951659451658,
      "avg_duration_per_task_ms": 107319.33261183261,
      "avg_output_tokens_per_task": 14469.299422799422,
      "last_run_timestamp": "2026-10-03T16:43:33.673Z"
    },
    {
      "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": 0.7467264444444446,
      "accuracy_std_dev": 0.05912320253901284,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.01319890179573513,
      "avg_duration_per_task_ms": 100190.80527497194,
      "avg_output_tokens_per_task": 5733.753647586981,
      "last_run_timestamp": "2026-10-17T08:31:20.045Z"
    },
    {
      "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": 0.7466665885714285,
      "accuracy_std_dev": 0.012433924485430002,
      "run_count": 8,
      "total_questions": 1575,
      "avg_cost_per_task_usd": 0.11517983936507933,
      "avg_duration_per_task_ms": 766437.2444444444,
      "avg_output_tokens_per_task": 44974.1346031746,
      "last_run_timestamp": "2026-10-17T16:06:03.256Z"
    },
    {
      "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": "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": 0.7457915,
      "accuracy_std_dev": 0.008417500000000022,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.015561896464646466,
      "avg_duration_per_task_ms": 220958.36363636365,
      "avg_output_tokens_per_task": 61761.72474747475,
      "last_run_timestamp": "2026-10-17T16:04:31.255Z"
    },
    {
      "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": 0.744318125,
      "accuracy_std_dev": 0.008684208274758005,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.006803909090909091,
      "avg_duration_per_task_ms": 163196.51830808082,
      "avg_output_tokens_per_task": 7906.570075757576,
      "last_run_timestamp": "2026-10-03T16:58:32.590Z"
    },
    {
      "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": "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": "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": 0.7393379999999999,
      "accuracy_std_dev": 0.01649975226076642,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.016316385521885524,
      "avg_duration_per_task_ms": 860370.430976431,
      "avg_output_tokens_per_task": 57893.19191919192,
      "last_run_timestamp": "2026-10-17T08:03:50.202Z"
    },
    {
      "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": 0.7391490737652341,
      "accuracy_std_dev": 0.06249035866035656,
      "run_count": 8,
      "total_questions": 1559,
      "avg_cost_per_task_usd": 0.2797374688903143,
      "avg_duration_per_task_ms": 1317138.1584348942,
      "avg_output_tokens_per_task": 88993.17318794099,
      "last_run_timestamp": "2026-10-17T16:03:22.651Z"
    },
    {
      "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": 0.7388167142857143,
      "accuracy_std_dev": 0.011146245808910405,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.00482988961038961,
      "avg_duration_per_task_ms": 107324.14213564213,
      "avg_output_tokens_per_task": 7874.049062049062,
      "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-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "GMICloud",
      "accuracy": 0.7385620777988616,
      "accuracy_std_dev": 0.1239043242359175,
      "run_count": 8,
      "total_questions": 1581,
      "avg_cost_per_task_usd": 0.005195311827956989,
      "avg_duration_per_task_ms": 366943.59392789373,
      "avg_output_tokens_per_task": 15666.573055028462,
      "last_run_timestamp": "2026-10-17T16:06:04.908Z"
    },
    {
      "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": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Venice",
      "accuracy": 0.7372685,
      "accuracy_std_dev": 0.06687866148481143,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.11904529797979799,
      "avg_duration_per_task_ms": 603565.316919192,
      "avg_output_tokens_per_task": 35752.68939393939,
      "last_run_timestamp": "2026-10-17T16:06:16.211Z"
    },
    {
      "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": 0.7361669504685407,
      "accuracy_std_dev": 0.0072054719473726965,
      "run_count": 8,
      "total_questions": 1494,
      "avg_cost_per_task_usd": 0.006662820615796521,
      "avg_duration_per_task_ms": 235800.40361445784,
      "avg_output_tokens_per_task": 8241.185408299865,
      "last_run_timestamp": "2026-10-17T16:03:45.724Z"
    },
    {
      "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Novita",
      "accuracy": 0.7343086374045802,
      "accuracy_std_dev": 0.014863350526021203,
      "run_count": 8,
      "total_questions": 1572,
      "avg_cost_per_task_usd": 0.006568683842239185,
      "avg_duration_per_task_ms": 349096.14885496185,
      "avg_output_tokens_per_task": 26082.53117048346,
      "last_run_timestamp": "2026-10-17T16:06:42.988Z"
    },
    {
      "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": 0.7340773735119047,
      "accuracy_std_dev": 0.017792537262244897,
      "run_count": 17,
      "total_questions": 3360,
      "avg_cost_per_task_usd": 0.006654232738095238,
      "avg_duration_per_task_ms": 121514.80714285714,
      "avg_output_tokens_per_task": 6940.3625,
      "last_run_timestamp": "2026-10-17T16:03:44.090Z"
    },
    {
      "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": 0.734007,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0222420303030303,
      "avg_duration_per_task_ms": 980476.9090909091,
      "avg_output_tokens_per_task": 55433.85858585859,
      "last_run_timestamp": "2026-10-17T08:02:36.324Z"
    },
    {
      "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": 0.733048471794872,
      "accuracy_std_dev": 0.05454575358179184,
      "run_count": 6,
      "total_questions": 1170,
      "avg_cost_per_task_usd": 0.017418347008547008,
      "avg_duration_per_task_ms": 1056799.6769230769,
      "avg_output_tokens_per_task": 44136.11538461538,
      "last_run_timestamp": "2026-10-17T16:06:05.349Z"
    },
    {
      "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": 0.7312169142857143,
      "accuracy_std_dev": 0.012725630006007392,
      "run_count": 35,
      "total_questions": 6930,
      "avg_cost_per_task_usd": 0.2044193023088022,
      "avg_duration_per_task_ms": 243801.0481962482,
      "avg_output_tokens_per_task": 38178.343001443005,
      "last_run_timestamp": "2026-10-17T08:18:07.606Z"
    },
    {
      "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": 0.7310605833333333,
      "accuracy_std_dev": 0.015416108968966707,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.1858813265993266,
      "avg_duration_per_task_ms": 297454.41077441076,
      "avg_output_tokens_per_task": 37007.64057239057,
      "last_run_timestamp": "2026-10-17T08:19:37.981Z"
    },
    {
      "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": "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": 0.7297980000000001,
      "accuracy_std_dev": 0.01108520326509943,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.006000632756132756,
      "avg_duration_per_task_ms": 399051.6572871573,
      "avg_output_tokens_per_task": 7941.9047619047615,
      "last_run_timestamp": "2026-10-17T16:07:29.700Z"
    },
    {
      "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": 0.7285024173913045,
      "accuracy_std_dev": 0.053992787395869,
      "run_count": 7,
      "total_questions": 1380,
      "avg_cost_per_task_usd": 0.2980392717391304,
      "avg_duration_per_task_ms": 1335107.6862318842,
      "avg_output_tokens_per_task": 94830.67536231884,
      "last_run_timestamp": "2026-10-17T16:09:49.789Z"
    },
    {
      "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": 0.7282548333333334,
      "accuracy_std_dev": 0.009283646839410062,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.19013492424242423,
      "avg_duration_per_task_ms": 242933.63425925927,
      "avg_output_tokens_per_task": 37836.16708754209,
      "last_run_timestamp": "2026-10-17T08:18:06.163Z"
    },
    {
      "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": "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": "anthropic/claude-4.5-haiku-20251001",
      "model_slug": "anthropic/claude-haiku-4.5",
      "model_name": "Anthropic: Claude Haiku 4.5",
      "provider_name": "Azure",
      "accuracy": 0.7271424333619948,
      "accuracy_std_dev": 0.012738961511398356,
      "run_count": 12,
      "total_questions": 2326,
      "avg_cost_per_task_usd": 0.1868387523645744,
      "avg_duration_per_task_ms": 345994.4789337919,
      "avg_output_tokens_per_task": 37177.46560619088,
      "last_run_timestamp": "2026-10-17T08:18:07.188Z"
    },
    {
      "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": 0.7269411348511384,
      "accuracy_std_dev": 0.05322474791235121,
      "run_count": 6,
      "total_questions": 1142,
      "avg_cost_per_task_usd": 0.0670280718038529,
      "avg_duration_per_task_ms": 1772246.9842381787,
      "avg_output_tokens_per_task": 34290.80998248686,
      "last_run_timestamp": "2026-10-17T16:09:10.586Z"
    },
    {
      "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": 0.726852,
      "accuracy_std_dev": 0.015898785165854648,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.005457933712121213,
      "avg_duration_per_task_ms": 255088.5498737374,
      "avg_output_tokens_per_task": 7917.640782828283,
      "last_run_timestamp": "2026-10-17T16:03:34.729Z"
    },
    {
      "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": "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": 0.7250370931445603,
      "accuracy_std_dev": 0.10506379222881687,
      "run_count": 8,
      "total_questions": 1342,
      "avg_cost_per_task_usd": 0.0044079307004470936,
      "avg_duration_per_task_ms": 545749.8681073026,
      "avg_output_tokens_per_task": 25961.05439642325,
      "last_run_timestamp": "2026-10-17T08:02:36.125Z"
    },
    {
      "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "auto-routing",
      "accuracy": 0.723064,
      "accuracy_std_dev": 0.02312882263540449,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.0057709015151515145,
      "avg_duration_per_task_ms": 342190.50694444444,
      "avg_output_tokens_per_task": 17394.30808080808,
      "last_run_timestamp": "2026-10-17T16:03:21.031Z"
    },
    {
      "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": "anthropic/claude-4.5-haiku-20251001",
      "model_slug": "anthropic/claude-haiku-4.5",
      "model_name": "Anthropic: Claude Haiku 4.5",
      "provider_name": "Google",
      "accuracy": 0.7215208055555558,
      "accuracy_std_dev": 0.013006390565622906,
      "run_count": 36,
      "total_questions": 7128,
      "avg_cost_per_task_usd": 0.197375834315376,
      "avg_duration_per_task_ms": 267372.5809483726,
      "avg_output_tokens_per_task": 36812.16091470258,
      "last_run_timestamp": "2026-10-17T08:26:18.338Z"
    },
    {
      "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": 0.7209128888888889,
      "accuracy_std_dev": 0.010210347185787597,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.00438738608305275,
      "avg_duration_per_task_ms": 649291.1077441077,
      "avg_output_tokens_per_task": 7853.185746352413,
      "last_run_timestamp": "2026-10-17T16:03:10.732Z"
    },
    {
      "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": 0.7173201760037348,
      "accuracy_std_dev": 0.0373953514862056,
      "run_count": 11,
      "total_questions": 2142,
      "avg_cost_per_task_usd": 0.010640089169000933,
      "avg_duration_per_task_ms": 264502.14985994395,
      "avg_output_tokens_per_task": 47791.74136321195,
      "last_run_timestamp": "2026-10-17T08:02:36.374Z"
    },
    {
      "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": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Makora",
      "accuracy": 0.7160493333333333,
      "accuracy_std_dev": 0.011690806112307026,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.018215316498316498,
      "avg_duration_per_task_ms": 375673.00505050505,
      "avg_output_tokens_per_task": 64702.65151515151,
      "last_run_timestamp": "2026-10-17T08:18:00.966Z"
    },
    {
      "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": 0.7154881249999999,
      "accuracy_std_dev": 0.05526776053550001,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.006628352272727273,
      "avg_duration_per_task_ms": 109434.6811868687,
      "avg_output_tokens_per_task": 16300.07133838384,
      "last_run_timestamp": "2026-10-17T16:03:44.040Z"
    },
    {
      "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-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": "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": "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": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "GMICloud",
      "accuracy": 0.7091377451737451,
      "accuracy_std_dev": 0.09862411167716228,
      "run_count": 8,
      "total_questions": 1554,
      "avg_cost_per_task_usd": 0.13626344465894466,
      "avg_duration_per_task_ms": 996229.6081081082,
      "avg_output_tokens_per_task": 72903.59588159589,
      "last_run_timestamp": "2026-10-17T16:03:22.631Z"
    },
    {
      "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": "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": 0.7064695714285715,
      "accuracy_std_dev": 0.05204993264948887,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.016059919913419915,
      "avg_duration_per_task_ms": 997721.5901875902,
      "avg_output_tokens_per_task": 48256.89682539683,
      "last_run_timestamp": "2026-10-17T16:03:19.083Z"
    },
    {
      "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": 0.7059358649303454,
      "accuracy_std_dev": 0.04834769030284051,
      "run_count": 17,
      "total_questions": 3302,
      "avg_cost_per_task_usd": 0.008049274076317382,
      "avg_duration_per_task_ms": 519119.62113870383,
      "avg_output_tokens_per_task": 25312.240763173835,
      "last_run_timestamp": "2026-10-17T16:11:22.823Z"
    },
    {
      "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": "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Mancer 2",
      "accuracy": 0.7037037142857143,
      "accuracy_std_dev": 0.04201543115405606,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.007533322510822511,
      "avg_duration_per_task_ms": 804813.1904761905,
      "avg_output_tokens_per_task": 25640.617604617604,
      "last_run_timestamp": "2026-10-17T16:06:47.630Z"
    },
    {
      "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": "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": "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": 0.7008102604166667,
      "accuracy_std_dev": 0.02755789234312128,
      "run_count": 9,
      "total_questions": 1728,
      "avg_cost_per_task_usd": 0.009351708912037038,
      "avg_duration_per_task_ms": 375939.1927083333,
      "avg_output_tokens_per_task": 45073.12384259259,
      "last_run_timestamp": "2026-10-17T08:02:36.220Z"
    },
    {
      "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": 0.700785612121212,
      "accuracy_std_dev": 0.15090828820097987,
      "run_count": 8,
      "total_questions": 1485,
      "avg_cost_per_task_usd": 0.08803936228956229,
      "avg_duration_per_task_ms": 618855.5521885522,
      "avg_output_tokens_per_task": 27400.163636363635,
      "last_run_timestamp": "2026-10-17T16:05:47.316Z"
    },
    {
      "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": 0.7000472153628653,
      "accuracy_std_dev": 0.1118059188262249,
      "run_count": 12,
      "total_questions": 2122,
      "avg_cost_per_task_usd": 0.03202059802073515,
      "avg_duration_per_task_ms": 1048480.0966069746,
      "avg_output_tokens_per_task": 33620.651743638075,
      "last_run_timestamp": "2026-10-17T16:06:17.196Z"
    },
    {
      "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": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "DekaLLM",
      "accuracy": 0.695286,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.007596338383838384,
      "avg_duration_per_task_ms": 2114924.02020202,
      "avg_output_tokens_per_task": 54058.00505050505,
      "last_run_timestamp": "2026-10-10T16:26:33.881Z"
    },
    {
      "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": "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-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": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "DeepInfra",
      "accuracy": 0.6898100478128179,
      "accuracy_std_dev": 0.04728305630022238,
      "run_count": 10,
      "total_questions": 1966,
      "avg_cost_per_task_usd": 0.007569284333672431,
      "avg_duration_per_task_ms": 285452.8184130214,
      "avg_output_tokens_per_task": 40484.10834181078,
      "last_run_timestamp": "2026-10-17T08:02:36.471Z"
    },
    {
      "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": 0.6894712328966521,
      "accuracy_std_dev": 0.029494685027795925,
      "run_count": 6,
      "total_questions": 687,
      "avg_cost_per_task_usd": 0.009409835516739446,
      "avg_duration_per_task_ms": 56014.804949053854,
      "avg_output_tokens_per_task": 6093.893740902475,
      "last_run_timestamp": "2026-10-17T16:03:30.488Z"
    },
    {
      "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": 0.6848486000000001,
      "accuracy_std_dev": 0.06736876735580072,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.0077578979797979794,
      "avg_duration_per_task_ms": 1345818.3151515152,
      "avg_output_tokens_per_task": 61748.48484848485,
      "last_run_timestamp": "2026-10-17T16:03:44.042Z"
    },
    {
      "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": "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": "Claude Platform on AWS",
      "accuracy": 0.6814815,
      "accuracy_std_dev": 0.13220309825813464,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.6080289767676768,
      "avg_duration_per_task_ms": 627829.4717171717,
      "avg_output_tokens_per_task": 40460.77222222222,
      "last_run_timestamp": "2026-10-17T08:21:26.728Z"
    },
    {
      "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": 0.6790182687747036,
      "accuracy_std_dev": 0.01730772007660164,
      "run_count": 8,
      "total_questions": 1012,
      "avg_cost_per_task_usd": 0.004263678853754942,
      "avg_duration_per_task_ms": 849188.9940711462,
      "avg_output_tokens_per_task": 42493.400197628456,
      "last_run_timestamp": "2026-10-17T16:05:46.660Z"
    },
    {
      "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": 0.6777017260273973,
      "accuracy_std_dev": 0.01773685922201921,
      "run_count": 9,
      "total_questions": 1752,
      "avg_cost_per_task_usd": 0.0021049600456621003,
      "avg_duration_per_task_ms": 197982.70319634702,
      "avg_output_tokens_per_task": 4056.279109589041,
      "last_run_timestamp": "2026-10-17T16:06:14.756Z"
    },
    {
      "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": "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": 0.6747943299556681,
      "accuracy_std_dev": 0.09586293689486765,
      "run_count": 8,
      "total_questions": 1579,
      "avg_cost_per_task_usd": 0.011517654844838505,
      "avg_duration_per_task_ms": 960521.5490816972,
      "avg_output_tokens_per_task": 33713.77960734642,
      "last_run_timestamp": "2026-10-17T16:03:32.458Z"
    },
    {
      "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": "anthropic/claude-4.6-sonnet-20260217",
      "model_slug": "anthropic/claude-sonnet-4.6",
      "model_name": "Anthropic: Claude Sonnet 4.6",
      "provider_name": "Google",
      "accuracy": 0.6714038275698719,
      "accuracy_std_dev": 0.14105760450673585,
      "run_count": 32,
      "total_questions": 6333,
      "avg_cost_per_task_usd": 0.6613663402810673,
      "avg_duration_per_task_ms": 596949.4503394916,
      "avg_output_tokens_per_task": 41226.86957208274,
      "last_run_timestamp": "2026-10-17T08:25:40.799Z"
    },
    {
      "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": 0.6708499999999998,
      "accuracy_std_dev": 0.13629141592324748,
      "run_count": 33,
      "total_questions": 6534,
      "avg_cost_per_task_usd": 0.6572281983471076,
      "avg_duration_per_task_ms": 465247.270278543,
      "avg_output_tokens_per_task": 40899.25971839608,
      "last_run_timestamp": "2026-10-17T08:20:14.789Z"
    },
    {
      "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": 0.6698808181818181,
      "accuracy_std_dev": 0.13895384854941661,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.6170192970615244,
      "avg_duration_per_task_ms": 632348.5018365473,
      "avg_output_tokens_per_task": 40978.85583103765,
      "last_run_timestamp": "2026-10-17T08:03:37.263Z"
    },
    {
      "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": "anthropic/claude-4.6-sonnet-20260217",
      "model_slug": "anthropic/claude-sonnet-4.6",
      "model_name": "Anthropic: Claude Sonnet 4.6",
      "provider_name": "Azure",
      "accuracy": 0.6672790000000001,
      "accuracy_std_dev": 0.1453197674941212,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.6195322176308541,
      "avg_duration_per_task_ms": 633753.1198347107,
      "avg_output_tokens_per_task": 41213.71395775941,
      "last_run_timestamp": "2026-10-17T08:21:05.935Z"
    },
    {
      "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": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Venice",
      "accuracy": 0.6652237142857144,
      "accuracy_std_dev": 0.016133353975221514,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.005764501443001443,
      "avg_duration_per_task_ms": 425162.0974025974,
      "avg_output_tokens_per_task": 15800.678932178933,
      "last_run_timestamp": "2026-10-17T16:05:59.879Z"
    },
    {
      "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": 0.664469578782172,
      "accuracy_std_dev": 0.017175728905408912,
      "run_count": 9,
      "total_questions": 1593,
      "avg_cost_per_task_usd": 0.006976930947897049,
      "avg_duration_per_task_ms": 485628.49654739484,
      "avg_output_tokens_per_task": 49654.3157564344,
      "last_run_timestamp": "2026-10-17T16:12:05.186Z"
    },
    {
      "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": 0.6643207216890595,
      "accuracy_std_dev": 0.02405113813802531,
      "run_count": 8,
      "total_questions": 1563,
      "avg_cost_per_task_usd": 0.027497225847728728,
      "avg_duration_per_task_ms": 839162.8976327576,
      "avg_output_tokens_per_task": 18575.21817018554,
      "last_run_timestamp": "2026-10-17T16:06:25.445Z"
    },
    {
      "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",
      "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": "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": "Parasail",
      "accuracy": 0.6582491666666667,
      "accuracy_std_dev": 0.024492817507837335,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.007911382154882154,
      "avg_duration_per_task_ms": 819880.6978114478,
      "avg_output_tokens_per_task": 54362.767676767675,
      "last_run_timestamp": "2026-10-17T16:06:46.554Z"
    },
    {
      "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": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.6576179999999999,
      "accuracy_std_dev": 0.01601415369290554,
      "run_count": 16,
      "total_questions": 3168,
      "avg_cost_per_task_usd": 0.008870517992424245,
      "avg_duration_per_task_ms": 321713.3065025252,
      "avg_output_tokens_per_task": 58815.57323232323,
      "last_run_timestamp": "2026-10-17T16:03:45.826Z"
    },
    {
      "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": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Venice",
      "accuracy": 0.6560053838120105,
      "accuracy_std_dev": 0.20012163968993332,
      "run_count": 4,
      "total_questions": 766,
      "avg_cost_per_task_usd": 0.07259215535248041,
      "avg_duration_per_task_ms": 1457323.1005221931,
      "avg_output_tokens_per_task": 31100.218015665796,
      "last_run_timestamp": "2026-10-17T16:05:04.854Z"
    },
    {
      "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": 0.6554433333333334,
      "accuracy_std_dev": 0.0216322289497253,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.007501081369248036,
      "avg_duration_per_task_ms": 643870.0549943884,
      "avg_output_tokens_per_task": 58023.832772166104,
      "last_run_timestamp": "2026-10-17T16:06:15.837Z"
    },
    {
      "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": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "auto-routing",
      "accuracy": 0.6550925,
      "accuracy_std_dev": 0.13950088584306553,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.014581559974747476,
      "avg_duration_per_task_ms": 426923.34027777775,
      "avg_output_tokens_per_task": 13349.987373737375,
      "last_run_timestamp": "2026-10-17T16:03:19.089Z"
    },
    {
      "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": "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": "deepseek/deepseek-chat-v3-0324",
      "model_slug": "deepseek/deepseek-chat-v3-0324",
      "model_name": "DeepSeek: DeepSeek V3 0324",
      "provider_name": "Crusoe",
      "accuracy": 0.650363823580786,
      "accuracy_std_dev": 0.007103203700131033,
      "run_count": 6,
      "total_questions": 1145,
      "avg_cost_per_task_usd": 0.0036315545851528385,
      "avg_duration_per_task_ms": 58503.03493449782,
      "avg_output_tokens_per_task": 2202.661135371179,
      "last_run_timestamp": "2026-08-29T16:10:45.295Z"
    },
    {
      "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": 0.6502977410714286,
      "accuracy_std_dev": 0.027563000000000004,
      "run_count": 2,
      "total_questions": 224,
      "avg_cost_per_task_usd": 0.00407284375,
      "avg_duration_per_task_ms": 238632.88839285713,
      "avg_output_tokens_per_task": 13352.90625,
      "last_run_timestamp": "2026-10-09T03:20:25.108Z"
    },
    {
      "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": "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": "anthropic/claude-4.6-sonnet-20260217",
      "model_slug": "anthropic/claude-sonnet-4.6",
      "model_name": "Anthropic: Claude Sonnet 4.6",
      "provider_name": "Anthropic",
      "accuracy": 0.6459467692307692,
      "accuracy_std_dev": 0.14204828227220914,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.6158103986013985,
      "avg_duration_per_task_ms": 628254.9867909868,
      "avg_output_tokens_per_task": 40928.243589743586,
      "last_run_timestamp": "2026-10-17T08:10:38.314Z"
    },
    {
      "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-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": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "auto-routing",
      "accuracy": 0.6420455,
      "accuracy_std_dev": 0.02154191080196928,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.007707547348484848,
      "avg_duration_per_task_ms": 627203.6717171717,
      "avg_output_tokens_per_task": 45830.878787878784,
      "last_run_timestamp": "2026-10-17T16:03:19.117Z"
    },
    {
      "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Phala",
      "accuracy": 0.6417703596127247,
      "accuracy_std_dev": 0.22170503732098648,
      "run_count": 8,
      "total_questions": 1446,
      "avg_cost_per_task_usd": 0.008234107883817428,
      "avg_duration_per_task_ms": 190026.56362378976,
      "avg_output_tokens_per_task": 13566.77316735823,
      "last_run_timestamp": "2026-10-17T16:05:13.856Z"
    },
    {
      "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": 0.6404886237845923,
      "accuracy_std_dev": 0.019812264339163707,
      "run_count": 7,
      "total_questions": 1337,
      "avg_cost_per_task_usd": 0.005841909498878086,
      "avg_duration_per_task_ms": 853617.8878085265,
      "avg_output_tokens_per_task": 35274.41361256545,
      "last_run_timestamp": "2026-10-17T08:18:07.100Z"
    },
    {
      "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": 0.638266870496592,
      "accuracy_std_dev": 0.013425056404805994,
      "run_count": 6,
      "total_questions": 1027,
      "avg_cost_per_task_usd": 0.007463308666017527,
      "avg_duration_per_task_ms": 34605.769230769234,
      "avg_output_tokens_per_task": 12362.268743914314,
      "last_run_timestamp": "2026-08-29T16:23:45.123Z"
    },
    {
      "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": 0.6374194522580645,
      "accuracy_std_dev": 0.03485686397164123,
      "run_count": 8,
      "total_questions": 1550,
      "avg_cost_per_task_usd": 0.059443716774193536,
      "avg_duration_per_task_ms": 1133345.7683870967,
      "avg_output_tokens_per_task": 25152.790967741934,
      "last_run_timestamp": "2026-10-17T16:05:03.429Z"
    },
    {
      "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": "Groq",
      "accuracy": 0.6338480836550836,
      "accuracy_std_dev": 0.014864426946152353,
      "run_count": 8,
      "total_questions": 1554,
      "avg_cost_per_task_usd": 0.016590933719433723,
      "avg_duration_per_task_ms": 90048.76833976834,
      "avg_output_tokens_per_task": 55073.25611325612,
      "last_run_timestamp": "2026-10-17T16:06:05.694Z"
    },
    {
      "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": 0.6336718477157359,
      "accuracy_std_dev": 0.012529818338002093,
      "run_count": 6,
      "total_questions": 1182,
      "avg_cost_per_task_usd": 0.003363994923857868,
      "avg_duration_per_task_ms": 229990.32571912013,
      "avg_output_tokens_per_task": 10944.545685279189,
      "last_run_timestamp": "2026-10-03T16:41:22.866Z"
    },
    {
      "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": "Darkbloom",
      "accuracy": 0.6311448,
      "accuracy_std_dev": 0.007685307265165154,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.0011036777777777777,
      "avg_duration_per_task_ms": 412979.17272727273,
      "avg_output_tokens_per_task": 11552.624242424243,
      "last_run_timestamp": "2026-10-17T16:07:02.132Z"
    },
    {
      "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-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": "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": 0.6272744175576814,
      "accuracy_std_dev": 0.010599488803187111,
      "run_count": 9,
      "total_questions": 1777,
      "avg_cost_per_task_usd": 0.0018089758019133374,
      "avg_duration_per_task_ms": 225302.9572312887,
      "avg_output_tokens_per_task": 9913.687675858188,
      "last_run_timestamp": "2026-10-17T16:06:43.994Z"
    },
    {
      "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": 0.6253005714285714,
      "accuracy_std_dev": 0.007302792065996677,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.024268393217893217,
      "avg_duration_per_task_ms": 1032337.5678210679,
      "avg_output_tokens_per_task": 100917.8658008658,
      "last_run_timestamp": "2026-10-03T17:27:37.016Z"
    },
    {
      "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": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "DeepInfra",
      "accuracy": 0.6200336919191919,
      "accuracy_std_dev": 0.11339640181763927,
      "run_count": 8,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.09412290202020202,
      "avg_duration_per_task_ms": 1259060.601010101,
      "avg_output_tokens_per_task": 26810.946464646466,
      "last_run_timestamp": "2026-10-17T16:09:29.739Z"
    },
    {
      "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": 0.6196141015067407,
      "accuracy_std_dev": 0.08983333259674553,
      "run_count": 8,
      "total_questions": 1261,
      "avg_cost_per_task_usd": 0.0522221165741475,
      "avg_duration_per_task_ms": 323929.8643933386,
      "avg_output_tokens_per_task": 19801.716098334655,
      "last_run_timestamp": "2026-10-17T16:08:25.260Z"
    },
    {
      "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": "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-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": "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": 0.6021274927659573,
      "accuracy_std_dev": 0.0818342643735224,
      "run_count": 6,
      "total_questions": 1175,
      "avg_cost_per_task_usd": 0.012776367659574468,
      "avg_duration_per_task_ms": 1356678.9268085107,
      "avg_output_tokens_per_task": 56278.53021276596,
      "last_run_timestamp": "2026-10-17T16:06:06.458Z"
    },
    {
      "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": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "Darkbloom",
      "accuracy": 0.6003243382761816,
      "accuracy_std_dev": 0.19105188493471462,
      "run_count": 11,
      "total_questions": 2158,
      "avg_cost_per_task_usd": 0.02540118164967563,
      "avg_duration_per_task_ms": 704415.6756255792,
      "avg_output_tokens_per_task": 35888.73818350324,
      "last_run_timestamp": "2026-10-17T16:04:35.717Z"
    },
    {
      "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": 0.5994523122629583,
      "accuracy_std_dev": 0.10371686124425177,
      "run_count": 8,
      "total_questions": 1582,
      "avg_cost_per_task_usd": 0.009700670670037928,
      "avg_duration_per_task_ms": 326884.85082174465,
      "avg_output_tokens_per_task": 27706.066371681416,
      "last_run_timestamp": "2026-10-17T16:05:58.333Z"
    },
    {
      "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "DigitalOcean",
      "accuracy": 0.5947809999999999,
      "accuracy_std_dev": 0.13542944894076767,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.0028764909090909087,
      "avg_duration_per_task_ms": 219197.7287878788,
      "avg_output_tokens_per_task": 6791.85202020202,
      "last_run_timestamp": "2026-10-17T16:05:42.381Z"
    },
    {
      "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": 0.5902780000000002,
      "accuracy_std_dev": 0.009063399582937982,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.010274607954545456,
      "avg_duration_per_task_ms": 11527.349747474747,
      "avg_output_tokens_per_task": 13245.006313131313,
      "last_run_timestamp": "2026-10-17T16:12:37.369Z"
    },
    {
      "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": 0.5696305846528623,
      "accuracy_std_dev": 0.02980472884459779,
      "run_count": 8,
      "total_questions": 821,
      "avg_cost_per_task_usd": 0.03290161875761267,
      "avg_duration_per_task_ms": 61744.40560292327,
      "avg_output_tokens_per_task": 7044.583434835567,
      "last_run_timestamp": "2026-10-17T16:05:03.577Z"
    },
    {
      "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": 0.49831644444444445,
      "accuracy_std_dev": 0.012943393871530763,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.0023515011223344553,
      "avg_duration_per_task_ms": 10501.347362514029,
      "avg_output_tokens_per_task": 2329.722222222222,
      "last_run_timestamp": "2026-10-17T16:05:59.924Z"
    },
    {
      "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": 0.48550299999999996,
      "accuracy_std_dev": 0.005230926367926303,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.002357327721661055,
      "avg_duration_per_task_ms": 14866.148709315376,
      "avg_output_tokens_per_task": 2216.58810325477,
      "last_run_timestamp": "2026-10-17T16:07:34.447Z"
    },
    {
      "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": 0.48224677424483314,
      "accuracy_std_dev": 0.016591290595765484,
      "run_count": 13,
      "total_questions": 2516,
      "avg_cost_per_task_usd": 0.0011952321144674085,
      "avg_duration_per_task_ms": 57709.6359300477,
      "avg_output_tokens_per_task": 2204.5071542130368,
      "last_run_timestamp": "2026-10-17T16:04:59.438Z"
    },
    {
      "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": "CoreWeave",
      "accuracy": 0.48042899999999994,
      "accuracy_std_dev": 0.00882577897411893,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.0025970202020202024,
      "avg_duration_per_task_ms": 51590.09785353536,
      "avg_output_tokens_per_task": 2849.220328282828,
      "last_run_timestamp": "2026-10-17T16:03:42.287Z"
    },
    {
      "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": "Together",
      "accuracy": 0.4758696666666667,
      "accuracy_std_dev": 0.012035777397225144,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.003142085858585859,
      "avg_duration_per_task_ms": 61848.09259259259,
      "avg_output_tokens_per_task": 2215.2059483726152,
      "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": "Parasail",
      "accuracy": 0.4753785,
      "accuracy_std_dev": 0.01205013506148375,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.0013200202020202019,
      "avg_duration_per_task_ms": 46402.35037878788,
      "avg_output_tokens_per_task": 2362.8636363636365,
      "last_run_timestamp": "2026-10-17T16:06:14.836Z"
    },
    {
      "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": 0.4717298835443038,
      "accuracy_std_dev": 0.015344517267985329,
      "run_count": 8,
      "total_questions": 1580,
      "avg_cost_per_task_usd": 0.0013212088607594936,
      "avg_duration_per_task_ms": 90902.26202531645,
      "avg_output_tokens_per_task": 3031.067088607595,
      "last_run_timestamp": "2026-10-17T16:03:46.179Z"
    },
    {
      "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": 0.4715905,
      "accuracy_std_dev": 0.008925495476442747,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.0013712070707070706,
      "avg_duration_per_task_ms": 71157.06376262626,
      "avg_output_tokens_per_task": 2259.597222222222,
      "last_run_timestamp": "2026-10-17T16:03:19.113Z"
    },
    {
      "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": 0.46742483824768327,
      "accuracy_std_dev": 0.006948085005956674,
      "run_count": 6,
      "total_questions": 1187,
      "avg_cost_per_task_usd": 0.002137747262005055,
      "avg_duration_per_task_ms": 34994.02274641955,
      "avg_output_tokens_per_task": 2587.217354675653,
      "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": 0.4634305555555556,
      "accuracy_std_dev": 0.017035579793877875,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.0007985993265993267,
      "avg_duration_per_task_ms": 162610.1588103255,
      "avg_output_tokens_per_task": 2244.9444444444443,
      "last_run_timestamp": "2026-10-17T16:05:59.152Z"
    },
    {
      "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": "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": "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": 0.40225624573378843,
      "accuracy_std_dev": 0.0951137316667801,
      "run_count": 9,
      "total_questions": 1758,
      "avg_cost_per_task_usd": 0.004076047781569966,
      "avg_duration_per_task_ms": 40782.70307167235,
      "avg_output_tokens_per_task": 1714.6023890784984,
      "last_run_timestamp": "2026-10-17T16:03:30.485Z"
    },
    {
      "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": 0.31216928571428576,
      "accuracy_std_dev": 0.09915241498206441,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.008725867965367964,
      "avg_duration_per_task_ms": 624172.6608946609,
      "avg_output_tokens_per_task": 21583.0202020202,
      "last_run_timestamp": "2026-10-17T16:03:19.114Z"
    },
    {
      "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": "auto-routing",
      "accuracy": 0.286826625,
      "accuracy_std_dev": 0.017534761211216273,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.00147922601010101,
      "avg_duration_per_task_ms": 134135.83712121213,
      "avg_output_tokens_per_task": 20106.075757575756,
      "last_run_timestamp": "2026-10-17T16:06:11.525Z"
    },
    {
      "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": 0.280724,
      "accuracy_std_dev": 0.010959107970085884,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.0028003642676767675,
      "avg_duration_per_task_ms": 53003.86363636364,
      "avg_output_tokens_per_task": 35001.6029040404,
      "last_run_timestamp": "2026-10-17T16:03:32.457Z"
    },
    {
      "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": "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": 0.27476285969556585,
      "accuracy_std_dev": 0.016441799164020953,
      "run_count": 8,
      "total_questions": 1511,
      "avg_cost_per_task_usd": 0.008752005956320317,
      "avg_duration_per_task_ms": 357755.06022501655,
      "avg_output_tokens_per_task": 39075.03904698875,
      "last_run_timestamp": "2026-10-17T16:06:52.618Z"
    },
    {
      "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": "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": 0.26941151269035535,
      "accuracy_std_dev": 0.011962128387354623,
      "run_count": 9,
      "total_questions": 1773,
      "avg_cost_per_task_usd": 0.0003863474337281444,
      "avg_duration_per_task_ms": 404376.6192893401,
      "avg_output_tokens_per_task": 9255.347433728144,
      "last_run_timestamp": "2026-10-17T16:03:34.728Z"
    },
    {
      "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": 0.25907233333333335,
      "accuracy_std_dev": 0.0140503088854935,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.0004220207631874298,
      "avg_duration_per_task_ms": 68331.25420875421,
      "avg_output_tokens_per_task": 8116.787878787879,
      "last_run_timestamp": "2026-10-17T16:05:03.496Z"
    },
    {
      "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": 0.22962980000000002,
      "accuracy_std_dev": 0.11415863629423752,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.0012938323232323233,
      "avg_duration_per_task_ms": 537044.9343434343,
      "avg_output_tokens_per_task": 4183.114141414141,
      "last_run_timestamp": "2026-08-29T17:16:14.290Z"
    },
    {
      "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": "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-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Tenstorrent",
      "accuracy": 0.817778,
      "accuracy_std_dev": 0.008314764498569197,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.045532706666666665,
      "avg_duration_per_task_ms": 81385.46,
      "avg_output_tokens_per_task": 5393.9,
      "last_run_timestamp": "2026-10-17T11:16:25.206Z"
    },
    {
      "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": 0.8115940000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 46,
      "avg_cost_per_task_usd": 0.05019476086956522,
      "avg_duration_per_task_ms": 222941.58695652173,
      "avg_output_tokens_per_task": 15139.565217391304,
      "last_run_timestamp": "2026-10-10T19:11:12.554Z"
    },
    {
      "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": 0.8083334999999999,
      "accuracy_std_dev": 0.011666499999999969,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.27321468,
      "avg_duration_per_task_ms": 27944.19,
      "avg_output_tokens_per_task": 12314.55,
      "last_run_timestamp": "2026-10-17T10:26:07.340Z"
    },
    {
      "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": "Azure",
      "accuracy": 0.8054545454545454,
      "accuracy_std_dev": 0.018161693915609792,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.8776261163636364,
      "avg_duration_per_task_ms": 132507.47090909092,
      "avg_output_tokens_per_task": 5799.549090909091,
      "last_run_timestamp": "2026-10-17T10:21:30.047Z"
    },
    {
      "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": 0.805,
      "accuracy_std_dev": 0.0050000000000000044,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.03878186,
      "avg_duration_per_task_ms": 112134.29,
      "avg_output_tokens_per_task": 4615.95,
      "last_run_timestamp": "2026-10-17T10:59:57.558Z"
    },
    {
      "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": 0.8022223333333333,
      "accuracy_std_dev": 0.022222300000974995,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.87912498,
      "avg_duration_per_task_ms": 142260.38,
      "avg_output_tokens_per_task": 5752.837777777778,
      "last_run_timestamp": "2026-10-17T11:00:09.754Z"
    },
    {
      "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": 0.8018181818181818,
      "accuracy_std_dev": 0.026222075532643652,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.9189436672727272,
      "avg_duration_per_task_ms": 125382.45454545454,
      "avg_output_tokens_per_task": 5783.118181818182,
      "last_run_timestamp": "2026-10-17T10:30:43.110Z"
    },
    {
      "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-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Wafer",
      "accuracy": 0.8,
      "accuracy_std_dev": 0.026666999999999996,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01127516,
      "avg_duration_per_task_ms": 269406.38,
      "avg_output_tokens_per_task": 14001.05,
      "last_run_timestamp": "2026-10-17T19:00:44.308Z"
    },
    {
      "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": 0.7983335000000001,
      "accuracy_std_dev": 0.008660350267165889,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.05923889500000001,
      "avg_duration_per_task_ms": 937672.915,
      "avg_output_tokens_per_task": 41561.56,
      "last_run_timestamp": "2026-10-17T11:22:17.754Z"
    },
    {
      "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": 0.7957576363636364,
      "accuracy_std_dev": 0.016641845981920093,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.8206288509090909,
      "avg_duration_per_task_ms": 133285.35818181818,
      "avg_output_tokens_per_task": 5774.987272727272,
      "last_run_timestamp": "2026-10-17T10:25:10.036Z"
    },
    {
      "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": 0.7944445714285714,
      "accuracy_std_dev": 0.021757085246773614,
      "run_count": 21,
      "total_questions": 1050,
      "avg_cost_per_task_usd": 0.9448550457142856,
      "avg_duration_per_task_ms": 95959.28857142857,
      "avg_output_tokens_per_task": 5694.378095238095,
      "last_run_timestamp": "2026-10-17T10:59:43.833Z"
    },
    {
      "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": 0.7942177959183673,
      "accuracy_std_dev": 0.014513999999999971,
      "run_count": 2,
      "total_questions": 98,
      "avg_cost_per_task_usd": 0.06639602040816327,
      "avg_duration_per_task_ms": 463491.29591836734,
      "avg_output_tokens_per_task": 8929.387755102041,
      "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": "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": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "DekaLLM",
      "accuracy": 0.7933334999999999,
      "accuracy_std_dev": 0.013333499999999998,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.11020131,
      "avg_duration_per_task_ms": 320831.42,
      "avg_output_tokens_per_task": 18660.61,
      "last_run_timestamp": "2026-10-17T10:59:44.533Z"
    },
    {
      "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": 0.793333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.01272006,
      "avg_duration_per_task_ms": 187174.32,
      "avg_output_tokens_per_task": 12168.72,
      "last_run_timestamp": "2026-10-18T01:15:30.019Z"
    },
    {
      "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": 0.793333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.01333386,
      "avg_duration_per_task_ms": 71979.7,
      "avg_output_tokens_per_task": 13242.08,
      "last_run_timestamp": "2026-10-18T00:09:45.214Z"
    },
    {
      "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": "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": "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": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Wafer",
      "accuracy": 0.7900492128514055,
      "accuracy_std_dev": 0.014761537132245638,
      "run_count": 15,
      "total_questions": 747,
      "avg_cost_per_task_usd": 0.10236340696117804,
      "avg_duration_per_task_ms": 137138.95448460508,
      "avg_output_tokens_per_task": 11684.20749665328,
      "last_run_timestamp": "2026-10-17T11:15:32.561Z"
    },
    {
      "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": 0.7897823718592966,
      "accuracy_std_dev": 0.03282945408637456,
      "run_count": 8,
      "total_questions": 398,
      "avg_cost_per_task_usd": 0.196944148241206,
      "avg_duration_per_task_ms": 804508.3291457286,
      "avg_output_tokens_per_task": 33320.5351758794,
      "last_run_timestamp": "2026-10-17T19:42:06.819Z"
    },
    {
      "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": 0.788889,
      "accuracy_std_dev": 0.019115994210782406,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.015915486666666666,
      "avg_duration_per_task_ms": 190491.50666666665,
      "avg_output_tokens_per_task": 11843.433333333332,
      "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": "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": 0.7883334999999999,
      "accuracy_std_dev": 0.013642112089042542,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.17747285999999998,
      "avg_duration_per_task_ms": 72082.685,
      "avg_output_tokens_per_task": 6332.95,
      "last_run_timestamp": "2026-10-17T10:24:14.037Z"
    },
    {
      "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": 0.788,
      "accuracy_std_dev": 0.021039613432760593,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.10081700199999999,
      "avg_duration_per_task_ms": 73090.95,
      "avg_output_tokens_per_task": 4784.634,
      "last_run_timestamp": "2026-10-17T11:21:57.972Z"
    },
    {
      "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": 0.7878787272727273,
      "accuracy_std_dev": 0.021801399271568482,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.48459852909090906,
      "avg_duration_per_task_ms": 123998.65090909091,
      "avg_output_tokens_per_task": 7121.34,
      "last_run_timestamp": "2026-10-17T11:00:48.268Z"
    },
    {
      "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": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "GMICloud",
      "accuracy": 0.7866667777777777,
      "accuracy_std_dev": 0.022662232682876603,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.06603865777777777,
      "avg_duration_per_task_ms": 319761.93111111113,
      "avg_output_tokens_per_task": 16377.204444444444,
      "last_run_timestamp": "2026-10-17T20:11:55.905Z"
    },
    {
      "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": 0.7866666666666666,
      "accuracy_std_dev": 0.02880303482081167,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.15989929333333333,
      "avg_duration_per_task_ms": 77372.58666666667,
      "avg_output_tokens_per_task": 6353.453333333333,
      "last_run_timestamp": "2026-10-17T10:25:34.224Z"
    },
    {
      "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": 0.7866666666666666,
      "accuracy_std_dev": 0.014401440259764158,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.015385413333333335,
      "avg_duration_per_task_ms": 462000.64666666667,
      "avg_output_tokens_per_task": 21255.126666666667,
      "last_run_timestamp": "2026-10-17T11:53:52.433Z"
    },
    {
      "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": 0.7863944591836735,
      "accuracy_std_dev": 0.023571561637702326,
      "run_count": 10,
      "total_questions": 490,
      "avg_cost_per_task_usd": 0.0799341387755102,
      "avg_duration_per_task_ms": 1168407.112244898,
      "avg_output_tokens_per_task": 26108.473469387754,
      "last_run_timestamp": "2026-10-17T20:08:31.778Z"
    },
    {
      "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": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Crusoe",
      "accuracy": 0.7859532642140468,
      "accuracy_std_dev": 0.022060806735148097,
      "run_count": 6,
      "total_questions": 299,
      "avg_cost_per_task_usd": 0.028781180602006688,
      "avg_duration_per_task_ms": 346700.6989966555,
      "avg_output_tokens_per_task": 11874.802675585284,
      "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": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Nebius",
      "accuracy": 0.7855944623115578,
      "accuracy_std_dev": 0.024102213207504365,
      "run_count": 4,
      "total_questions": 199,
      "avg_cost_per_task_usd": 0.24837878391959797,
      "avg_duration_per_task_ms": 56668.633165829146,
      "avg_output_tokens_per_task": 4728.195979899498,
      "last_run_timestamp": "2026-10-17T11:02:38.903Z"
    },
    {
      "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": 0.7855555833333333,
      "accuracy_std_dev": 0.01921298590128705,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.18381405,
      "avg_duration_per_task_ms": 65853.755,
      "avg_output_tokens_per_task": 6393.828333333333,
      "last_run_timestamp": "2026-10-17T11:10:53.414Z"
    },
    {
      "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": 0.7853332000000001,
      "accuracy_std_dev": 0.019505181079907966,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.070760672,
      "avg_duration_per_task_ms": 60914.488,
      "avg_output_tokens_per_task": 12425.428,
      "last_run_timestamp": "2026-10-17T11:22:17.486Z"
    },
    {
      "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": 0.785,
      "accuracy_std_dev": 0.012360308339600598,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.1797631525,
      "avg_duration_per_task_ms": 69465.2375,
      "avg_output_tokens_per_task": 6280.4925,
      "last_run_timestamp": "2026-10-17T10:36:28.721Z"
    },
    {
      "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": 0.7844445,
      "accuracy_std_dev": 0.008314898130263933,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.08145889999999999,
      "avg_duration_per_task_ms": 40269.02,
      "avg_output_tokens_per_task": 4762.626666666667,
      "last_run_timestamp": "2026-10-17T11:17:26.577Z"
    },
    {
      "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": 0.7844443333333334,
      "accuracy_std_dev": 0.006285236475706792,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.037741833333333336,
      "avg_duration_per_task_ms": 258684.46,
      "avg_output_tokens_per_task": 13025.313333333334,
      "last_run_timestamp": "2026-10-17T11:00:49.268Z"
    },
    {
      "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": 0.7838094285714285,
      "accuracy_std_dev": 0.017036561992685377,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.0766554,
      "avg_duration_per_task_ms": 47593.05714285714,
      "avg_output_tokens_per_task": 4753.951428571429,
      "last_run_timestamp": "2026-10-17T11:22:27.251Z"
    },
    {
      "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": 0.7836751845597104,
      "accuracy_std_dev": 0.029924488404365016,
      "run_count": 17,
      "total_questions": 829,
      "avg_cost_per_task_usd": 0.0797846031363088,
      "avg_duration_per_task_ms": 80343.72135102534,
      "avg_output_tokens_per_task": 4214.624849215923,
      "last_run_timestamp": "2026-10-17T11:16:45.514Z"
    },
    {
      "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": 0.7833334999999999,
      "accuracy_std_dev": 0.01914828103903846,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.012174070000000002,
      "avg_duration_per_task_ms": 629324.83,
      "avg_output_tokens_per_task": 23420.965,
      "last_run_timestamp": "2026-10-17T12:15:06.096Z"
    },
    {
      "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": 0.7833334999999999,
      "accuracy_std_dev": 0.0033334999999999892,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.12497997999999999,
      "avg_duration_per_task_ms": 155609.03,
      "avg_output_tokens_per_task": 4544.63,
      "last_run_timestamp": "2026-10-17T11:01:25.869Z"
    },
    {
      "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-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Alibaba",
      "accuracy": 0.7830555,
      "accuracy_std_dev": 0.023072206170700996,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.132926985,
      "avg_duration_per_task_ms": 302505.37,
      "avg_output_tokens_per_task": 19619.503333333334,
      "last_run_timestamp": "2026-10-17T11:22:27.630Z"
    },
    {
      "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": 0.7826666,
      "accuracy_std_dev": 0.01372745505328646,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.12898250000000003,
      "avg_duration_per_task_ms": 448868.648,
      "avg_output_tokens_per_task": 12081.028,
      "last_run_timestamp": "2026-10-17T19:47:17.813Z"
    },
    {
      "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": 0.7826580270270269,
      "accuracy_std_dev": 0.01733842438695692,
      "run_count": 3,
      "total_questions": 148,
      "avg_cost_per_task_usd": 0.03042019594594595,
      "avg_duration_per_task_ms": 91460.18243243243,
      "avg_output_tokens_per_task": 15686.162162162162,
      "last_run_timestamp": "2026-10-17T10:23:46.492Z"
    },
    {
      "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": 0.7818182727272727,
      "accuracy_std_dev": 0.022489619278441696,
      "run_count": 22,
      "total_questions": 1100,
      "avg_cost_per_task_usd": 0.4780214336363635,
      "avg_duration_per_task_ms": 107722.86090909092,
      "avg_output_tokens_per_task": 7240.321818181818,
      "last_run_timestamp": "2026-10-17T11:08:24.698Z"
    },
    {
      "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": 0.7811111666666667,
      "accuracy_std_dev": 0.027057157151831192,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.08468537,
      "avg_duration_per_task_ms": 260135.41333333333,
      "avg_output_tokens_per_task": 20074.013333333332,
      "last_run_timestamp": "2026-10-17T11:21:46.230Z"
    },
    {
      "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": 0.7811111111111111,
      "accuracy_std_dev": 0.021998793459058637,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.18190104222222223,
      "avg_duration_per_task_ms": 388430.46,
      "avg_output_tokens_per_task": 28795.273333333334,
      "last_run_timestamp": "2026-10-17T19:40:07.953Z"
    },
    {
      "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": 0.781111,
      "accuracy_std_dev": 0.030651344804646544,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.03906277,
      "avg_duration_per_task_ms": 319274.30666666664,
      "avg_output_tokens_per_task": 12696.91,
      "last_run_timestamp": "2026-10-17T11:08:09.690Z"
    },
    {
      "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": 0.7808335,
      "accuracy_std_dev": 0.0240801728035743,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.056226975,
      "avg_duration_per_task_ms": 913150.26,
      "avg_output_tokens_per_task": 39116.4,
      "last_run_timestamp": "2026-10-17T11:47:31.377Z"
    },
    {
      "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": 0.78083325,
      "accuracy_std_dev": 0.0203954004909808,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.08186078000000001,
      "avg_duration_per_task_ms": 137926.375,
      "avg_output_tokens_per_task": 4672.7925,
      "last_run_timestamp": "2026-10-17T10:59:13.366Z"
    },
    {
      "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": 0.7807406666666666,
      "accuracy_std_dev": 0.02483414725288101,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.10385739777777778,
      "avg_duration_per_task_ms": 312310.43333333335,
      "avg_output_tokens_per_task": 20848.71111111111,
      "last_run_timestamp": "2026-10-17T11:10:01.812Z"
    },
    {
      "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": 0.7806513735632183,
      "accuracy_std_dev": 0.017034729877517862,
      "run_count": 7,
      "total_questions": 348,
      "avg_cost_per_task_usd": 0.1535702011494253,
      "avg_duration_per_task_ms": 329766.867816092,
      "avg_output_tokens_per_task": 20311.916666666668,
      "last_run_timestamp": "2026-10-17T12:02:08.368Z"
    },
    {
      "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": 0.7801960294117648,
      "accuracy_std_dev": 0.024494153043587038,
      "run_count": 34,
      "total_questions": 1700,
      "avg_cost_per_task_usd": 0.548420931764706,
      "avg_duration_per_task_ms": 113328.19352941176,
      "avg_output_tokens_per_task": 7234.435882352941,
      "last_run_timestamp": "2026-10-17T11:01:24.330Z"
    },
    {
      "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": 0.7800268380566802,
      "accuracy_std_dev": 0.020840856341331072,
      "run_count": 5,
      "total_questions": 247,
      "avg_cost_per_task_usd": 0.07761835222672064,
      "avg_duration_per_task_ms": 235647.0890688259,
      "avg_output_tokens_per_task": 11752.493927125506,
      "last_run_timestamp": "2026-10-17T19:18:11.122Z"
    },
    {
      "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": "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": 0.78,
      "accuracy_std_dev": 0.022222200001100003,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.4880859666666667,
      "avg_duration_per_task_ms": 126960.01777777777,
      "avg_output_tokens_per_task": 8014.38,
      "last_run_timestamp": "2026-10-17T11:11:34.526Z"
    },
    {
      "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": 0.78,
      "accuracy_std_dev": 0.025385603971542606,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.17071506,
      "avg_duration_per_task_ms": 87188.75,
      "avg_output_tokens_per_task": 6362.575,
      "last_run_timestamp": "2026-10-17T10:59:47.136Z"
    },
    {
      "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": 0.78,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.00748126,
      "avg_duration_per_task_ms": 121223.72,
      "avg_output_tokens_per_task": 12121.04,
      "last_run_timestamp": "2026-10-17T15:39:37.383Z"
    },
    {
      "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": 0.78,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.1158576,
      "avg_duration_per_task_ms": 173684.7,
      "avg_output_tokens_per_task": 11854.88,
      "last_run_timestamp": "2026-10-17T22:46:23.487Z"
    },
    {
      "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": 0.78,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.028234659999999998,
      "avg_duration_per_task_ms": 113376.08,
      "avg_output_tokens_per_task": 11900.56,
      "last_run_timestamp": "2026-10-17T19:19:16.262Z"
    },
    {
      "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": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "Novita",
      "accuracy": 0.7799996666666668,
      "accuracy_std_dev": 0.024944382578492918,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.05863376,
      "avg_duration_per_task_ms": 922572.7666666667,
      "avg_output_tokens_per_task": 40326.753333333334,
      "last_run_timestamp": "2026-10-17T11:22:08.858Z"
    },
    {
      "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": 0.77916675,
      "accuracy_std_dev": 0.024818790485990654,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0730951225,
      "avg_duration_per_task_ms": 271033.7,
      "avg_output_tokens_per_task": 14317.8375,
      "last_run_timestamp": "2026-10-17T11:02:34.586Z"
    },
    {
      "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": 0.77916675,
      "accuracy_std_dev": 0.020121260868481886,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0690324625,
      "avg_duration_per_task_ms": 39596.8475,
      "avg_output_tokens_per_task": 4762.57,
      "last_run_timestamp": "2026-10-17T11:01:13.760Z"
    },
    {
      "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": 0.7788889999999999,
      "accuracy_std_dev": 0.0178124938830399,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.15423991666666667,
      "avg_duration_per_task_ms": 293986.99666666664,
      "avg_output_tokens_per_task": 19979.47,
      "last_run_timestamp": "2026-10-17T11:22:17.534Z"
    },
    {
      "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": 0.7788888333333334,
      "accuracy_std_dev": 0.019404546369486603,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.09862560333333334,
      "avg_duration_per_task_ms": 369802.0233333333,
      "avg_output_tokens_per_task": 20634.286666666667,
      "last_run_timestamp": "2026-10-17T11:38:21.418Z"
    },
    {
      "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": 0.778383818181818,
      "accuracy_std_dev": 0.020761880948324426,
      "run_count": 33,
      "total_questions": 1650,
      "avg_cost_per_task_usd": 0.4916623981818182,
      "avg_duration_per_task_ms": 123938.08424242424,
      "avg_output_tokens_per_task": 7196.92303030303,
      "last_run_timestamp": "2026-10-17T11:07:54.507Z"
    },
    {
      "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": "DigitalOcean",
      "accuracy": 0.7779999000000001,
      "accuracy_std_dev": 0.04120956253698892,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.11152408,
      "avg_duration_per_task_ms": 189018.608,
      "avg_output_tokens_per_task": 6916.256,
      "last_run_timestamp": "2026-10-17T10:59:44.534Z"
    },
    {
      "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": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "DeepInfra",
      "accuracy": 0.777778,
      "accuracy_std_dev": 0.034995646995971776,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.07155335333333333,
      "avg_duration_per_task_ms": 807590.7333333333,
      "avg_output_tokens_per_task": 13442.86,
      "last_run_timestamp": "2026-10-17T11:32:02.278Z"
    },
    {
      "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": 0.7777778333333333,
      "accuracy_std_dev": 0.019499832225061745,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.09435254333333333,
      "avg_duration_per_task_ms": 128876.18666666666,
      "avg_output_tokens_per_task": 5214.3,
      "last_run_timestamp": "2026-10-17T11:32:49.835Z"
    },
    {
      "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": 0.7777776666666667,
      "accuracy_std_dev": 0.017497908167041606,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.08824630666666666,
      "avg_duration_per_task_ms": 389172.2,
      "avg_output_tokens_per_task": 14887.42,
      "last_run_timestamp": "2026-10-17T10:26:01.481Z"
    },
    {
      "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": 0.7776190714285715,
      "accuracy_std_dev": 0.022762808810062738,
      "run_count": 14,
      "total_questions": 700,
      "avg_cost_per_task_usd": 0.49921174,
      "avg_duration_per_task_ms": 127401.97714285714,
      "avg_output_tokens_per_task": 8107.18,
      "last_run_timestamp": "2026-10-17T10:19:21.229Z"
    },
    {
      "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": 0.7772191809045226,
      "accuracy_std_dev": 0.015717449830920396,
      "run_count": 8,
      "total_questions": 398,
      "avg_cost_per_task_usd": 0.08903428140703519,
      "avg_duration_per_task_ms": 135997.33417085427,
      "avg_output_tokens_per_task": 4429.972361809046,
      "last_run_timestamp": "2026-10-13T19:11:39.344Z"
    },
    {
      "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": 0.7771429999999999,
      "accuracy_std_dev": 0.015474446789096262,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.11423808857142857,
      "avg_duration_per_task_ms": 264332.0714285714,
      "avg_output_tokens_per_task": 23099.01714285714,
      "last_run_timestamp": "2026-10-17T10:59:43.838Z"
    },
    {
      "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": 0.777037,
      "accuracy_std_dev": 0.02882250302184812,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.01873947555555555,
      "avg_duration_per_task_ms": 295769.78,
      "avg_output_tokens_per_task": 11621.488888888889,
      "last_run_timestamp": "2026-10-17T19:51:29.473Z"
    },
    {
      "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": 0.7769696363636364,
      "accuracy_std_dev": 0.015919847448291065,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.12614458545454543,
      "avg_duration_per_task_ms": 91476.05272727273,
      "avg_output_tokens_per_task": 4526.654545454546,
      "last_run_timestamp": "2026-10-17T11:13:41.491Z"
    },
    {
      "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": 0.776825380952381,
      "accuracy_std_dev": 0.02899504966529581,
      "run_count": 21,
      "total_questions": 1050,
      "avg_cost_per_task_usd": 0.5120268733333334,
      "avg_duration_per_task_ms": 129027.85523809523,
      "avg_output_tokens_per_task": 7974.611428571428,
      "last_run_timestamp": "2026-10-17T11:11:30.197Z"
    },
    {
      "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": 0.77666675,
      "accuracy_std_dev": 0.019720364546685406,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.44522166500000004,
      "avg_duration_per_task_ms": 128363.55,
      "avg_output_tokens_per_task": 7253.05,
      "last_run_timestamp": "2026-10-17T10:59:59.565Z"
    },
    {
      "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": "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": "AtlasCloud",
      "accuracy": 0.7762962222222223,
      "accuracy_std_dev": 0.028993197254903236,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.10068874222222222,
      "avg_duration_per_task_ms": 236706.54,
      "avg_output_tokens_per_task": 11538.877777777778,
      "last_run_timestamp": "2026-10-17T19:39:33.608Z"
    },
    {
      "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": 0.7758333749999999,
      "accuracy_std_dev": 0.022592582670743405,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.18276251000000002,
      "avg_duration_per_task_ms": 750721.8,
      "avg_output_tokens_per_task": 29533.0075,
      "last_run_timestamp": "2026-10-17T19:12:23.825Z"
    },
    {
      "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": 0.77583325,
      "accuracy_std_dev": 0.027067136183709917,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.021435382500000003,
      "avg_duration_per_task_ms": 238417.755,
      "avg_output_tokens_per_task": 12509.2125,
      "last_run_timestamp": "2026-10-17T19:35:58.094Z"
    },
    {
      "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": 0.7757142142857143,
      "accuracy_std_dev": 0.01738592688345774,
      "run_count": 14,
      "total_questions": 700,
      "avg_cost_per_task_usd": 0.43694479428571437,
      "avg_duration_per_task_ms": 94452.41714285714,
      "avg_output_tokens_per_task": 5698.648571428572,
      "last_run_timestamp": "2026-10-17T11:00:50.086Z"
    },
    {
      "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": 0.7755556666666668,
      "accuracy_std_dev": 0.030952277087298266,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.11345276666666666,
      "avg_duration_per_task_ms": 300353.7133333333,
      "avg_output_tokens_per_task": 20577.506666666668,
      "last_run_timestamp": "2026-10-17T11:24:14.273Z"
    },
    {
      "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": 0.7755553333333334,
      "accuracy_std_dev": 0.003142853940113798,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.007128806666666667,
      "avg_duration_per_task_ms": 167248.23333333334,
      "avg_output_tokens_per_task": 17580.666666666668,
      "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": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "CoreWeave",
      "accuracy": 0.7755098163265306,
      "accuracy_std_dev": 0.018194500000000002,
      "run_count": 2,
      "total_questions": 98,
      "avg_cost_per_task_usd": 0.06427333673469388,
      "avg_duration_per_task_ms": 82291.80612244898,
      "avg_output_tokens_per_task": 4253.367346938776,
      "last_run_timestamp": "2026-10-17T11:13:56.470Z"
    },
    {
      "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": 0.775245761394102,
      "accuracy_std_dev": 0.03049371819200363,
      "run_count": 15,
      "total_questions": 746,
      "avg_cost_per_task_usd": 0.03492235924932976,
      "avg_duration_per_task_ms": 523697.9343163539,
      "avg_output_tokens_per_task": 12545.84852546917,
      "last_run_timestamp": "2026-10-17T19:49:18.769Z"
    },
    {
      "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": 0.7752381428571429,
      "accuracy_std_dev": 0.024871433587428726,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.05666836000000001,
      "avg_duration_per_task_ms": 135387.59714285715,
      "avg_output_tokens_per_task": 14986.728571428572,
      "last_run_timestamp": "2026-10-17T10:59:42.724Z"
    },
    {
      "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": 0.7752381428571429,
      "accuracy_std_dev": 0.029106105056345494,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.06096673714285714,
      "avg_duration_per_task_ms": 121563.02571428572,
      "avg_output_tokens_per_task": 6310.4457142857145,
      "last_run_timestamp": "2026-10-17T11:15:29.802Z"
    },
    {
      "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": 0.775238,
      "accuracy_std_dev": 0.01807009782565028,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.08216295714285714,
      "avg_duration_per_task_ms": 86031.87428571428,
      "avg_output_tokens_per_task": 4554.505714285714,
      "last_run_timestamp": "2026-10-17T11:15:40.034Z"
    },
    {
      "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": 0.775,
      "accuracy_std_dev": 0.015184019378939179,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.17930137999999998,
      "avg_duration_per_task_ms": 76460.92,
      "avg_output_tokens_per_task": 6444.745,
      "last_run_timestamp": "2026-10-17T10:27:00.078Z"
    },
    {
      "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": 0.7748484090909091,
      "accuracy_std_dev": 0.03107060548706196,
      "run_count": 22,
      "total_questions": 1100,
      "avg_cost_per_task_usd": 0.44300329,
      "avg_duration_per_task_ms": 80178.89,
      "avg_output_tokens_per_task": 5667.305454545454,
      "last_run_timestamp": "2026-10-17T11:01:36.809Z"
    },
    {
      "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": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Friendli",
      "accuracy": 0.7744445,
      "accuracy_std_dev": 0.011166655613178608,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.030003983333333338,
      "avg_duration_per_task_ms": 137881.70333333334,
      "avg_output_tokens_per_task": 12197.636666666667,
      "last_run_timestamp": "2026-10-17T19:40:24.035Z"
    },
    {
      "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": "Parasail",
      "accuracy": 0.7740742222222222,
      "accuracy_std_dev": 0.03402175734169657,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.06915481111111112,
      "avg_duration_per_task_ms": 253932.57111111112,
      "avg_output_tokens_per_task": 12007.897777777778,
      "last_run_timestamp": "2026-10-17T11:00:48.240Z"
    },
    {
      "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": 0.774074,
      "accuracy_std_dev": 0.039433570402217795,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.11824472000000001,
      "avg_duration_per_task_ms": 275582.18888888886,
      "avg_output_tokens_per_task": 12068.497777777779,
      "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": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "AkashML",
      "accuracy": 0.7737493322091062,
      "accuracy_std_dev": 0.02457644899198327,
      "run_count": 12,
      "total_questions": 593,
      "avg_cost_per_task_usd": 0.05600873693086003,
      "avg_duration_per_task_ms": 258548.61720067455,
      "avg_output_tokens_per_task": 11042.026981450254,
      "last_run_timestamp": "2026-10-17T10:29:45.431Z"
    },
    {
      "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": 0.773639194842407,
      "accuracy_std_dev": 0.03650679077617633,
      "run_count": 7,
      "total_questions": 349,
      "avg_cost_per_task_usd": 0.1842622492836676,
      "avg_duration_per_task_ms": 578395.2607449857,
      "avg_output_tokens_per_task": 27822.154727793695,
      "last_run_timestamp": "2026-10-17T19:41:38.122Z"
    },
    {
      "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": "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": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "StreamLake",
      "accuracy": 0.7733332222222222,
      "accuracy_std_dev": 0.02131486469202904,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.08245232666666666,
      "avg_duration_per_task_ms": 336783.29777777777,
      "avg_output_tokens_per_task": 20135.455555555556,
      "last_run_timestamp": "2026-10-17T11:24:16.298Z"
    },
    {
      "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": "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": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "DeepInfra",
      "accuracy": 0.7730925662650602,
      "accuracy_std_dev": 0.018139575154892704,
      "run_count": 5,
      "total_questions": 249,
      "avg_cost_per_task_usd": 0.06739513253012049,
      "avg_duration_per_task_ms": 1083762.6305220884,
      "avg_output_tokens_per_task": 34958.25702811245,
      "last_run_timestamp": "2026-10-17T11:17:33.534Z"
    },
    {
      "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": 0.7729629444444445,
      "accuracy_std_dev": 0.01981045608166506,
      "run_count": 18,
      "total_questions": 900,
      "avg_cost_per_task_usd": 0.07089291888888888,
      "avg_duration_per_task_ms": 76354.76444444444,
      "avg_output_tokens_per_task": 4694.775555555556,
      "last_run_timestamp": "2026-10-17T11:17:01.185Z"
    },
    {
      "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": 0.7727272727272727,
      "accuracy_std_dev": 0.026431515541767755,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.47498244545454554,
      "avg_duration_per_task_ms": 113596.48181818181,
      "avg_output_tokens_per_task": 7235.563636363637,
      "last_run_timestamp": "2026-10-17T11:08:22.019Z"
    },
    {
      "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": 0.7727271818181819,
      "accuracy_std_dev": 0.0191652424922077,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.4342336036363637,
      "avg_duration_per_task_ms": 98810.9909090909,
      "avg_output_tokens_per_task": 5628.896363636363,
      "last_run_timestamp": "2026-10-17T11:12:37.911Z"
    },
    {
      "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": 0.772476937614679,
      "accuracy_std_dev": 0.03305459265948225,
      "run_count": 11,
      "total_questions": 545,
      "avg_cost_per_task_usd": 0.10160879082568808,
      "avg_duration_per_task_ms": 803326.4128440367,
      "avg_output_tokens_per_task": 10658.660550458715,
      "last_run_timestamp": "2026-10-17T19:57:29.431Z"
    },
    {
      "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": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Makora",
      "accuracy": 0.7722221666666667,
      "accuracy_std_dev": 0.014098441384501401,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.08663532333333333,
      "avg_duration_per_task_ms": 61723.59333333333,
      "avg_output_tokens_per_task": 4868.923333333333,
      "last_run_timestamp": "2026-10-17T11:23:57.889Z"
    },
    {
      "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": 0.7722220000000001,
      "accuracy_std_dev": 0.024770452330683562,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.07917609666666667,
      "avg_duration_per_task_ms": 223600.7,
      "avg_output_tokens_per_task": 9330.2,
      "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-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": "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": 0.0203131506142337,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.04409809333333333,
      "avg_duration_per_task_ms": 198272.02666666667,
      "avg_output_tokens_per_task": 13209.757777777777,
      "last_run_timestamp": "2026-10-17T19:51:22.454Z"
    },
    {
      "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": 0.7718517777777778,
      "accuracy_std_dev": 0.037022202956177475,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.06489010666666667,
      "avg_duration_per_task_ms": 307042.8888888889,
      "avg_output_tokens_per_task": 12648.484444444444,
      "last_run_timestamp": "2026-10-17T10:41:37.611Z"
    },
    {
      "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": 0.771666625,
      "accuracy_std_dev": 0.02020731097237767,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0087578875,
      "avg_duration_per_task_ms": 205985.82,
      "avg_output_tokens_per_task": 4280.365,
      "last_run_timestamp": "2026-10-17T11:27:39.717Z"
    },
    {
      "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": 0.7715152727272727,
      "accuracy_std_dev": 0.020516898902172905,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.5359888036363636,
      "avg_duration_per_task_ms": 175110.11272727273,
      "avg_output_tokens_per_task": 10646.576363636364,
      "last_run_timestamp": "2026-10-17T11:13:46.638Z"
    },
    {
      "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": "Decart",
      "accuracy": 0.7714285714285715,
      "accuracy_std_dev": 0.017715327672120907,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.059177931428571424,
      "avg_duration_per_task_ms": 49647.705714285716,
      "avg_output_tokens_per_task": 4699.5028571428575,
      "last_run_timestamp": "2026-10-17T10:58:43.523Z"
    },
    {
      "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": 0.7711111666666667,
      "accuracy_std_dev": 0.025141594085079193,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.04609336666666666,
      "avg_duration_per_task_ms": 106575.85,
      "avg_output_tokens_per_task": 4990.85,
      "last_run_timestamp": "2026-10-17T11:02:44.041Z"
    },
    {
      "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": "Relace",
      "accuracy": 0.77083325,
      "accuracy_std_dev": 0.02099910381867521,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.06492240499999999,
      "avg_duration_per_task_ms": 184655.75,
      "avg_output_tokens_per_task": 10289.75,
      "last_run_timestamp": "2026-10-17T19:13:29.361Z"
    },
    {
      "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": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Cloudflare",
      "accuracy": 0.7704762857142857,
      "accuracy_std_dev": 0.01419017577374559,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.15732603714285717,
      "avg_duration_per_task_ms": 351587.40285714285,
      "avg_output_tokens_per_task": 16725.24,
      "last_run_timestamp": "2026-10-17T21:41:22.284Z"
    },
    {
      "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": 0.7700000714285714,
      "accuracy_std_dev": 0.028701941199817064,
      "run_count": 14,
      "total_questions": 700,
      "avg_cost_per_task_usd": 0.5494887142857142,
      "avg_duration_per_task_ms": 184565.79285714286,
      "avg_output_tokens_per_task": 10546.365714285714,
      "last_run_timestamp": "2026-10-17T10:59:10.050Z"
    },
    {
      "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": 0.7700000312500004,
      "accuracy_std_dev": 0.02798810387298991,
      "run_count": 32,
      "total_questions": 1600,
      "avg_cost_per_task_usd": 0.58152457625,
      "avg_duration_per_task_ms": 117734.71,
      "avg_output_tokens_per_task": 8150.94375,
      "last_run_timestamp": "2026-10-17T11:11:43.637Z"
    },
    {
      "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": 0.77,
      "accuracy_std_dev": 0.024267078752911318,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.044564936000000006,
      "avg_duration_per_task_ms": 210405.052,
      "avg_output_tokens_per_task": 9653.816,
      "last_run_timestamp": "2026-10-17T19:41:23.364Z"
    },
    {
      "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": 0.77,
      "accuracy_std_dev": 0.027888739266353848,
      "run_count": 15,
      "total_questions": 750,
      "avg_cost_per_task_usd": 0.049744766666666676,
      "avg_duration_per_task_ms": 93112.96533333333,
      "avg_output_tokens_per_task": 4363.234666666666,
      "last_run_timestamp": "2026-10-17T11:14:20.497Z"
    },
    {
      "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": 0.7697776666666667,
      "accuracy_std_dev": 0.021342537367634824,
      "run_count": 15,
      "total_questions": 750,
      "avg_cost_per_task_usd": 0.019707239999999997,
      "avg_duration_per_task_ms": 239431.496,
      "avg_output_tokens_per_task": 16872.549333333332,
      "last_run_timestamp": "2026-10-17T11:12:59.582Z"
    },
    {
      "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": 0.7696296666666667,
      "accuracy_std_dev": 0.023749919068867958,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.04528427333333334,
      "avg_duration_per_task_ms": 235347.72222222222,
      "avg_output_tokens_per_task": 12527.615555555556,
      "last_run_timestamp": "2026-10-17T19:58:17.795Z"
    },
    {
      "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": 0.7696296666666667,
      "accuracy_std_dev": 0.02999782525450804,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.15527621777777778,
      "avg_duration_per_task_ms": 612793.8288888889,
      "avg_output_tokens_per_task": 12411.853333333333,
      "last_run_timestamp": "2026-10-17T11:35:37.172Z"
    },
    {
      "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": 0.76958325,
      "accuracy_std_dev": 0.03006639296935201,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.11405574499999999,
      "avg_duration_per_task_ms": 171710.355,
      "avg_output_tokens_per_task": 9478.4625,
      "last_run_timestamp": "2026-10-17T21:04:23.082Z"
    },
    {
      "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": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "Darkbloom",
      "accuracy": 0.7693334,
      "accuracy_std_dev": 0.024801480242921004,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.06763208,
      "avg_duration_per_task_ms": 1137766.732,
      "avg_output_tokens_per_task": 63926.756,
      "last_run_timestamp": "2026-10-17T19:12:07.371Z"
    },
    {
      "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": 0.76916675,
      "accuracy_std_dev": 0.020800176950869918,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0561824625,
      "avg_duration_per_task_ms": 113575.025,
      "avg_output_tokens_per_task": 4743.9025,
      "last_run_timestamp": "2026-10-17T11:00:48.270Z"
    },
    {
      "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": 0.7690910909090909,
      "accuracy_std_dev": 0.0207480154952645,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.4942437709090909,
      "avg_duration_per_task_ms": 129656.94545454545,
      "avg_output_tokens_per_task": 8155.923636363636,
      "last_run_timestamp": "2026-10-17T10:36:13.263Z"
    },
    {
      "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": 0.7688889999999999,
      "accuracy_std_dev": 0.027576954057892455,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.13352325333333334,
      "avg_duration_per_task_ms": 270383.85333333333,
      "avg_output_tokens_per_task": 11724.017777777777,
      "last_run_timestamp": "2026-10-17T19:36:13.358Z"
    },
    {
      "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": 0.7688889999999999,
      "accuracy_std_dev": 0.016629528997138415,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.07698892666666665,
      "avg_duration_per_task_ms": 578204.64,
      "avg_output_tokens_per_task": 15573.933333333332,
      "last_run_timestamp": "2026-10-17T11:11:44.961Z"
    },
    {
      "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": 0.7688889333333333,
      "accuracy_std_dev": 0.020608014743352204,
      "run_count": 15,
      "total_questions": 750,
      "avg_cost_per_task_usd": 0.018404090666666668,
      "avg_duration_per_task_ms": 189791.296,
      "avg_output_tokens_per_task": 18860.122666666666,
      "last_run_timestamp": "2026-10-17T11:10:46.239Z"
    },
    {
      "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": "Inceptron",
      "accuracy": 0.7685714285714286,
      "accuracy_std_dev": 0.024093478197916338,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.04910968857142857,
      "avg_duration_per_task_ms": 125132.30571428571,
      "avg_output_tokens_per_task": 4665.682857142857,
      "last_run_timestamp": "2026-10-17T11:17:33.551Z"
    },
    {
      "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": 0.7684848636363636,
      "accuracy_std_dev": 0.030230338267279323,
      "run_count": 22,
      "total_questions": 1100,
      "avg_cost_per_task_usd": 0.5669869918181819,
      "avg_duration_per_task_ms": 182291.99818181817,
      "avg_output_tokens_per_task": 10636.502727272727,
      "last_run_timestamp": "2026-10-17T11:15:51.671Z"
    },
    {
      "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": 0.7684848181818182,
      "accuracy_std_dev": 0.03062860915673504,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.16836050181818182,
      "avg_duration_per_task_ms": 595336.3309090909,
      "avg_output_tokens_per_task": 35883.39090909091,
      "last_run_timestamp": "2026-10-17T11:01:46.079Z"
    },
    {
      "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": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7683335,
      "accuracy_std_dev": 0.025980601739567167,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.028812655,
      "avg_duration_per_task_ms": 169670.27,
      "avg_output_tokens_per_task": 16080.47,
      "last_run_timestamp": "2026-10-17T11:02:18.962Z"
    },
    {
      "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": 0.768333375,
      "accuracy_std_dev": 0.01755948224562374,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.099579235,
      "avg_duration_per_task_ms": 250087.7225,
      "avg_output_tokens_per_task": 17572.65,
      "last_run_timestamp": "2026-10-17T19:37:12.707Z"
    },
    {
      "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": 0.768333375,
      "accuracy_std_dev": 0.03086715037437658,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0819386525,
      "avg_duration_per_task_ms": 47567.265,
      "avg_output_tokens_per_task": 4891.2325,
      "last_run_timestamp": "2026-10-17T10:41:20.896Z"
    },
    {
      "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": 0.7681481111111111,
      "accuracy_std_dev": 0.03307724426061735,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.43618789555555554,
      "avg_duration_per_task_ms": 93061.98222222223,
      "avg_output_tokens_per_task": 5582.697777777777,
      "last_run_timestamp": "2026-10-17T11:14:22.181Z"
    },
    {
      "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": 0.7681481111111111,
      "accuracy_std_dev": 0.02261391222453041,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.027650264444444447,
      "avg_duration_per_task_ms": 160206.6511111111,
      "avg_output_tokens_per_task": 9294.16,
      "last_run_timestamp": "2026-10-17T19:51:53.679Z"
    },
    {
      "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": 0.7681249375000001,
      "accuracy_std_dev": 0.02039432387322742,
      "run_count": 32,
      "total_questions": 1600,
      "avg_cost_per_task_usd": 0.5345501306250001,
      "avg_duration_per_task_ms": 120332.5425,
      "avg_output_tokens_per_task": 8109.15625,
      "last_run_timestamp": "2026-10-17T11:09:50.077Z"
    },
    {
      "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": 0.7680001999999999,
      "accuracy_std_dev": 0.012927091279943853,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.034976304,
      "avg_duration_per_task_ms": 249863.756,
      "avg_output_tokens_per_task": 38297.836,
      "last_run_timestamp": "2026-10-17T10:47:05.212Z"
    },
    {
      "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": 0.768,
      "accuracy_std_dev": 0.029028767342069495,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.011671187999999999,
      "avg_duration_per_task_ms": 241079.168,
      "avg_output_tokens_per_task": 6128.846,
      "last_run_timestamp": "2026-10-17T10:59:30.134Z"
    },
    {
      "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": 0.7677778333333333,
      "accuracy_std_dev": 0.024469832777092887,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.22403587333333333,
      "avg_duration_per_task_ms": 379432.67333333334,
      "avg_output_tokens_per_task": 17682.066666666666,
      "last_run_timestamp": "2026-10-17T11:24:46.101Z"
    },
    {
      "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": 0.76777775,
      "accuracy_std_dev": 0.024011215230682656,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.09013422333333333,
      "avg_duration_per_task_ms": 187936.75166666668,
      "avg_output_tokens_per_task": 11408.701666666666,
      "last_run_timestamp": "2026-10-17T20:10:05.740Z"
    },
    {
      "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": 0.7677776666666667,
      "accuracy_std_dev": 0.030408559549720344,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.032006433333333334,
      "avg_duration_per_task_ms": 190973.69,
      "avg_output_tokens_per_task": 14850.593333333334,
      "last_run_timestamp": "2026-10-17T10:41:27.539Z"
    },
    {
      "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": 0.7676191428571428,
      "accuracy_std_dev": 0.019657045959063507,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.05359846000000001,
      "avg_duration_per_task_ms": 117300.16571428571,
      "avg_output_tokens_per_task": 4471.66,
      "last_run_timestamp": "2026-10-17T11:03:28.220Z"
    },
    {
      "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": 0.767500125,
      "accuracy_std_dev": 0.029521664597535403,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.14441939499999998,
      "avg_duration_per_task_ms": 322443.2475,
      "avg_output_tokens_per_task": 20266.705,
      "last_run_timestamp": "2026-10-17T12:43:23.056Z"
    },
    {
      "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": 0.7674073333333333,
      "accuracy_std_dev": 0.025423825815779805,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.18668503555555555,
      "avg_duration_per_task_ms": 177779.4688888889,
      "avg_output_tokens_per_task": 9629.98,
      "last_run_timestamp": "2026-10-17T19:18:49.006Z"
    },
    {
      "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": 0.7671692361809045,
      "accuracy_std_dev": 0.02443026293350114,
      "run_count": 8,
      "total_questions": 398,
      "avg_cost_per_task_usd": 0.10522786934673367,
      "avg_duration_per_task_ms": 46016.98994974874,
      "avg_output_tokens_per_task": 4282.550251256282,
      "last_run_timestamp": "2026-10-17T11:11:52.303Z"
    },
    {
      "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": 0.7671130223214285,
      "accuracy_std_dev": 0.019417018578276532,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.10946027455357142,
      "avg_duration_per_task_ms": 324665.375,
      "avg_output_tokens_per_task": 11865.004464285714,
      "last_run_timestamp": "2026-10-17T19:43:35.939Z"
    },
    {
      "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": 0.7670682771084337,
      "accuracy_std_dev": 0.016486696678231213,
      "run_count": 5,
      "total_questions": 249,
      "avg_cost_per_task_usd": 0.02956908032128514,
      "avg_duration_per_task_ms": 107195.45381526105,
      "avg_output_tokens_per_task": 18350.590361445782,
      "last_run_timestamp": "2026-10-17T11:00:50.996Z"
    },
    {
      "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": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Reka",
      "accuracy": 0.7666669999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.025223939999999997,
      "avg_duration_per_task_ms": 165137.1,
      "avg_output_tokens_per_task": 14240.82,
      "last_run_timestamp": "2026-09-15T00:00:05.846Z"
    },
    {
      "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": 0.7666669999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.010828599999999999,
      "avg_duration_per_task_ms": 269370.42,
      "avg_output_tokens_per_task": 14956.26,
      "last_run_timestamp": "2026-10-17T15:38:31.686Z"
    },
    {
      "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": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7666667142857143,
      "accuracy_std_dev": 0.019189911372640464,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.06833657142857143,
      "avg_duration_per_task_ms": 120543.64,
      "avg_output_tokens_per_task": 4633.334285714286,
      "last_run_timestamp": "2026-10-17T10:58:54.174Z"
    },
    {
      "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": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "DigitalOcean",
      "accuracy": 0.766666625,
      "accuracy_std_dev": 0.022360828847213497,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.010823125,
      "avg_duration_per_task_ms": 1144392.885,
      "avg_output_tokens_per_task": 13299.99,
      "last_run_timestamp": "2026-10-17T19:43:24.819Z"
    },
    {
      "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": 0.7666665714285714,
      "accuracy_std_dev": 0.041861494979983496,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.009251074285714285,
      "avg_duration_per_task_ms": 197957.5657142857,
      "avg_output_tokens_per_task": 15755.4,
      "last_run_timestamp": "2026-10-17T19:15:12.932Z"
    },
    {
      "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": "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": 0.7666665,
      "accuracy_std_dev": 0.05270466983342177,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.066754105,
      "avg_duration_per_task_ms": 1006890.58,
      "avg_output_tokens_per_task": 42879.28,
      "last_run_timestamp": "2026-10-17T11:45:34.968Z"
    },
    {
      "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": 0.7663313768844222,
      "accuracy_std_dev": 0.021610498297991634,
      "run_count": 4,
      "total_questions": 199,
      "avg_cost_per_task_usd": 0.06149351256281408,
      "avg_duration_per_task_ms": 193004.91959798994,
      "avg_output_tokens_per_task": 7834.050251256282,
      "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": "anthropic/claude-4.7-opus-20260416",
      "model_slug": "anthropic/claude-opus-4.7",
      "model_name": "Anthropic: Claude Opus 4.7",
      "provider_name": "Google",
      "accuracy": 0.7660606363636364,
      "accuracy_std_dev": 0.02228458723765413,
      "run_count": 33,
      "total_questions": 1650,
      "avg_cost_per_task_usd": 0.6243563309090909,
      "avg_duration_per_task_ms": 72757.64666666667,
      "avg_output_tokens_per_task": 5695.7472727272725,
      "last_run_timestamp": "2026-10-17T11:03:34.141Z"
    },
    {
      "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": 0.7660606363636363,
      "accuracy_std_dev": 0.024029451627203914,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.4185906163636364,
      "avg_duration_per_task_ms": 96036.28,
      "avg_output_tokens_per_task": 5622.494545454545,
      "last_run_timestamp": "2026-10-17T10:59:55.542Z"
    },
    {
      "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": 0.76583325,
      "accuracy_std_dev": 0.03340613508305772,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0064167,
      "avg_duration_per_task_ms": 121615.31,
      "avg_output_tokens_per_task": 4277.8575,
      "last_run_timestamp": "2026-10-17T11:37:18.038Z"
    },
    {
      "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": 0.765625,
      "accuracy_std_dev": 0.00520799999999999,
      "run_count": 2,
      "total_questions": 96,
      "avg_cost_per_task_usd": 0.06623564583333334,
      "avg_duration_per_task_ms": 151056.27083333334,
      "avg_output_tokens_per_task": 10137.447916666666,
      "last_run_timestamp": "2026-10-17T19:11:05.297Z"
    },
    {
      "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": 0.7656249866071428,
      "accuracy_std_dev": 0.0414301585804351,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.028515656250000004,
      "avg_duration_per_task_ms": 494142.7745535714,
      "avg_output_tokens_per_task": 12506.10044642857,
      "last_run_timestamp": "2026-10-17T19:35:17.391Z"
    },
    {
      "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": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "GMICloud",
      "accuracy": 0.7653333999999999,
      "accuracy_std_dev": 0.0185713806121139,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.020849908,
      "avg_duration_per_task_ms": 114504.516,
      "avg_output_tokens_per_task": 8739.532,
      "last_run_timestamp": "2026-10-17T19:15:12.182Z"
    },
    {
      "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": 0.7653332,
      "accuracy_std_dev": 0.014236029409916211,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.05721586,
      "avg_duration_per_task_ms": 73979.412,
      "avg_output_tokens_per_task": 4468.456,
      "last_run_timestamp": "2026-10-17T10:59:41.246Z"
    },
    {
      "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": 0.7652466315789472,
      "accuracy_std_dev": 0.0204935205903598,
      "run_count": 8,
      "total_questions": 399,
      "avg_cost_per_task_usd": 0.01140678947368421,
      "avg_duration_per_task_ms": 287101.95238095237,
      "avg_output_tokens_per_task": 12923.513784461153,
      "last_run_timestamp": "2026-10-17T19:01:11.679Z"
    },
    {
      "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": 0.7651005838926175,
      "accuracy_std_dev": 0.022563981273013175,
      "run_count": 3,
      "total_questions": 149,
      "avg_cost_per_task_usd": 0.037864174496644296,
      "avg_duration_per_task_ms": 157541.35570469798,
      "avg_output_tokens_per_task": 25184.140939597317,
      "last_run_timestamp": "2026-10-17T11:16:26.297Z"
    },
    {
      "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": 0.7650601887550201,
      "accuracy_std_dev": 0.019886756371012337,
      "run_count": 5,
      "total_questions": 249,
      "avg_cost_per_task_usd": 0.014014425702811244,
      "avg_duration_per_task_ms": 335075.9437751004,
      "avg_output_tokens_per_task": 16676.168674698794,
      "last_run_timestamp": "2026-10-17T11:01:25.260Z"
    },
    {
      "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": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "OpenInference",
      "accuracy": 0.765,
      "accuracy_std_dev": 0.01997680136224682,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.019566853333333332,
      "avg_duration_per_task_ms": 556627.0233333333,
      "avg_output_tokens_per_task": 13187.933333333332,
      "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": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Sail Research",
      "accuracy": 0.7649999375000001,
      "accuracy_std_dev": 0.017559488177580636,
      "run_count": 16,
      "total_questions": 800,
      "avg_cost_per_task_usd": 0.00967452625,
      "avg_duration_per_task_ms": 342838.47625,
      "avg_output_tokens_per_task": 14737.495,
      "last_run_timestamp": "2026-10-17T11:21:51.557Z"
    },
    {
      "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": 0.7649122421052632,
      "accuracy_std_dev": 0.0039184999999999914,
      "run_count": 2,
      "total_questions": 95,
      "avg_cost_per_task_usd": 0.25060463157894736,
      "avg_duration_per_task_ms": 358896.81052631576,
      "avg_output_tokens_per_task": 6743.5684210526315,
      "last_run_timestamp": "2026-10-10T11:20:06.376Z"
    },
    {
      "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": 0.764846632183908,
      "accuracy_std_dev": 0.017727680655646182,
      "run_count": 7,
      "total_questions": 348,
      "avg_cost_per_task_usd": 0.01970142528735632,
      "avg_duration_per_task_ms": 118406.38793103448,
      "avg_output_tokens_per_task": 14534.735632183909,
      "last_run_timestamp": "2026-10-17T11:01:24.414Z"
    },
    {
      "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": 0.7647619999999999,
      "accuracy_std_dev": 0.014998270500294357,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.17264639428571427,
      "avg_duration_per_task_ms": 205532.82,
      "avg_output_tokens_per_task": 11242.805714285714,
      "last_run_timestamp": "2026-10-17T19:36:29.795Z"
    },
    {
      "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": 0.7644443333333334,
      "accuracy_std_dev": 0.02424157229270045,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.01896127,
      "avg_duration_per_task_ms": 326667.0966666667,
      "avg_output_tokens_per_task": 14093.796666666667,
      "last_run_timestamp": "2026-10-17T10:27:00.952Z"
    },
    {
      "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": 0.7644443333333334,
      "accuracy_std_dev": 0.012570472951413688,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.02870791333333333,
      "avg_duration_per_task_ms": 1126711.7933333332,
      "avg_output_tokens_per_task": 60958.193333333336,
      "last_run_timestamp": "2026-10-10T12:23:30.561Z"
    },
    {
      "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": 0.7638095,
      "accuracy_std_dev": 0.02710519944256874,
      "run_count": 14,
      "total_questions": 700,
      "avg_cost_per_task_usd": 0.5202140842857143,
      "avg_duration_per_task_ms": 218955.78857142857,
      "avg_output_tokens_per_task": 10161.987142857142,
      "last_run_timestamp": "2026-10-17T11:16:23.075Z"
    },
    {
      "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": 0.7637036666666667,
      "accuracy_std_dev": 0.02456770717832659,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.5482636911111111,
      "avg_duration_per_task_ms": 180041.81555555554,
      "avg_output_tokens_per_task": 10794.96,
      "last_run_timestamp": "2026-10-17T11:22:39.138Z"
    },
    {
      "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": 0.7636362727272727,
      "accuracy_std_dev": 0.0307959485974459,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.5473690363636363,
      "avg_duration_per_task_ms": 191557.85636363635,
      "avg_output_tokens_per_task": 10669.91090909091,
      "last_run_timestamp": "2026-10-17T11:08:54.958Z"
    },
    {
      "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": 0.7633818649635037,
      "accuracy_std_dev": 0.03905021498570464,
      "run_count": 11,
      "total_questions": 548,
      "avg_cost_per_task_usd": 0.10277552189781022,
      "avg_duration_per_task_ms": 338883.84854014596,
      "avg_output_tokens_per_task": 17120.268248175184,
      "last_run_timestamp": "2026-10-18T01:42:21.659Z"
    },
    {
      "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": 0.7633335,
      "accuracy_std_dev": 0.01666650000000003,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01259681,
      "avg_duration_per_task_ms": 657803.07,
      "avg_output_tokens_per_task": 25182.87,
      "last_run_timestamp": "2026-10-03T08:05:49.905Z"
    },
    {
      "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": 0.7633332500000001,
      "accuracy_std_dev": 0.0074535971978300555,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.030563419999999997,
      "avg_duration_per_task_ms": 58422.895,
      "avg_output_tokens_per_task": 4829.025,
      "last_run_timestamp": "2026-10-17T12:23:26.697Z"
    },
    {
      "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": 0.7630522329317269,
      "accuracy_std_dev": 0.017240119992621882,
      "run_count": 5,
      "total_questions": 249,
      "avg_cost_per_task_usd": 0.16262659437751,
      "avg_duration_per_task_ms": 831035.3815261044,
      "avg_output_tokens_per_task": 31809.98795180723,
      "last_run_timestamp": "2026-10-10T18:46:35.608Z"
    },
    {
      "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": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Alibaba",
      "accuracy": 0.7627176144578314,
      "accuracy_std_dev": 0.038350734650199336,
      "run_count": 10,
      "total_questions": 498,
      "avg_cost_per_task_usd": 0.19653478714859438,
      "avg_duration_per_task_ms": 478839.8875502008,
      "avg_output_tokens_per_task": 27836.25702811245,
      "last_run_timestamp": "2026-10-17T19:45:29.549Z"
    },
    {
      "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": 0.7626666,
      "accuracy_std_dev": 0.03227647784749755,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.146569104,
      "avg_duration_per_task_ms": 373419.58,
      "avg_output_tokens_per_task": 24831.152,
      "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": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "SiliconFlow",
      "accuracy": 0.762500125,
      "accuracy_std_dev": 0.01823820585637127,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.014795629999999999,
      "avg_duration_per_task_ms": 251230.48,
      "avg_output_tokens_per_task": 15189.695,
      "last_run_timestamp": "2026-10-17T19:16:55.430Z"
    },
    {
      "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": 0.7625,
      "accuracy_std_dev": 0.027271719174815526,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.08931772499999999,
      "avg_duration_per_task_ms": 299563.8,
      "avg_output_tokens_per_task": 12112.5275,
      "last_run_timestamp": "2026-10-17T11:07:54.500Z"
    },
    {
      "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": 0.762499875,
      "accuracy_std_dev": 0.030900828154102528,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.00903259,
      "avg_duration_per_task_ms": 197139.9675,
      "avg_output_tokens_per_task": 4490.1425,
      "last_run_timestamp": "2026-10-17T11:12:20.710Z"
    },
    {
      "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": 0.7624549738058551,
      "accuracy_std_dev": 0.028041605618591807,
      "run_count": 13,
      "total_questions": 649,
      "avg_cost_per_task_usd": 0.08517802465331278,
      "avg_duration_per_task_ms": 145563.85516178736,
      "avg_output_tokens_per_task": 12573.135593220339,
      "last_run_timestamp": "2026-10-17T19:06:47.400Z"
    },
    {
      "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": 0.7624242727272728,
      "accuracy_std_dev": 0.022611815825484268,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.048589845454545456,
      "avg_duration_per_task_ms": 136956.19272727272,
      "avg_output_tokens_per_task": 4598.596363636364,
      "last_run_timestamp": "2026-10-16T01:58:40.721Z"
    },
    {
      "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": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "auto-routing",
      "accuracy": 0.7622223333333333,
      "accuracy_std_dev": 0.039377702534415165,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.13604349999999998,
      "avg_duration_per_task_ms": 1230822.4133333333,
      "avg_output_tokens_per_task": 30337.52,
      "last_run_timestamp": "2026-10-17T11:00:12.961Z"
    },
    {
      "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": 0.7622222222222222,
      "accuracy_std_dev": 0.03325913222886738,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.04489894888888889,
      "avg_duration_per_task_ms": 204045.26444444444,
      "avg_output_tokens_per_task": 9550.304444444444,
      "last_run_timestamp": "2026-10-17T19:42:12.724Z"
    },
    {
      "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": 0.7622101098901098,
      "accuracy_std_dev": 0.024541643374153495,
      "run_count": 11,
      "total_questions": 546,
      "avg_cost_per_task_usd": 0.2003756904761905,
      "avg_duration_per_task_ms": 486645.7838827839,
      "avg_output_tokens_per_task": 20463.4358974359,
      "last_run_timestamp": "2026-10-17T11:13:24.327Z"
    },
    {
      "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-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "DeepInfra",
      "accuracy": 0.7617447986577182,
      "accuracy_std_dev": 0.00897518183195566,
      "run_count": 6,
      "total_questions": 298,
      "avg_cost_per_task_usd": 0.0072220302013422805,
      "avg_duration_per_task_ms": 277237.70805369125,
      "avg_output_tokens_per_task": 12532.48322147651,
      "last_run_timestamp": "2026-10-17T19:46:31.388Z"
    },
    {
      "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": "InferenceNet",
      "accuracy": 0.7616665,
      "accuracy_std_dev": 0.01364229534389284,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.03678678,
      "avg_duration_per_task_ms": 78488.155,
      "avg_output_tokens_per_task": 4646.9025,
      "last_run_timestamp": "2026-10-17T11:01:16.024Z"
    },
    {
      "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": 0.7611212248628885,
      "accuracy_std_dev": 0.04241023232664495,
      "run_count": 11,
      "total_questions": 547,
      "avg_cost_per_task_usd": 0.07025861060329068,
      "avg_duration_per_task_ms": 209956.5265082267,
      "avg_output_tokens_per_task": 12430.747714808043,
      "last_run_timestamp": "2026-10-18T01:41:51.190Z"
    },
    {
      "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": 0.7611110833333333,
      "accuracy_std_dev": 0.019404866071247587,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.006715375000000001,
      "avg_duration_per_task_ms": 256863.22,
      "avg_output_tokens_per_task": 15279.346666666666,
      "last_run_timestamp": "2026-10-17T11:16:57.833Z"
    },
    {
      "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": 0.7609522857142857,
      "accuracy_std_dev": 0.029261543259342417,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.025921108571428573,
      "avg_duration_per_task_ms": 269473.77714285714,
      "avg_output_tokens_per_task": 13250.331428571428,
      "last_run_timestamp": "2026-10-17T19:03:42.106Z"
    },
    {
      "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": 0.7609090909090909,
      "accuracy_std_dev": 0.01608635842996141,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.3294631345454546,
      "avg_duration_per_task_ms": 293594.13636363635,
      "avg_output_tokens_per_task": 10568.594545454545,
      "last_run_timestamp": "2026-10-17T11:17:24.553Z"
    },
    {
      "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": 0.7607481658291457,
      "accuracy_std_dev": 0.027232714686338207,
      "run_count": 12,
      "total_questions": 597,
      "avg_cost_per_task_usd": 0.061689276381909554,
      "avg_duration_per_task_ms": 257829.17755443885,
      "avg_output_tokens_per_task": 11960.412060301507,
      "last_run_timestamp": "2026-10-17T10:59:41.251Z"
    },
    {
      "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": 0.76046883919598,
      "accuracy_std_dev": 0.03279282531286379,
      "run_count": 4,
      "total_questions": 199,
      "avg_cost_per_task_usd": 0.07501306030150753,
      "avg_duration_per_task_ms": 437868.7336683417,
      "avg_output_tokens_per_task": 62802.35175879397,
      "last_run_timestamp": "2026-10-17T19:45:57.690Z"
    },
    {
      "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": 0.7600536219839142,
      "accuracy_std_dev": 0.024783783422409292,
      "run_count": 15,
      "total_questions": 746,
      "avg_cost_per_task_usd": 0.020043907506702412,
      "avg_duration_per_task_ms": 441729.95174262737,
      "avg_output_tokens_per_task": 16926.498659517427,
      "last_run_timestamp": "2026-10-17T10:58:50.948Z"
    },
    {
      "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": 0.76,
      "accuracy_std_dev": 0.0066669999999999785,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.0113985,
      "avg_duration_per_task_ms": 89553.33,
      "avg_output_tokens_per_task": 10995.52,
      "last_run_timestamp": "2026-10-17T11:01:12.758Z"
    },
    {
      "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": 0.76,
      "accuracy_std_dev": 0.038559978206331065,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.1978464090909091,
      "avg_duration_per_task_ms": 103130.56181818181,
      "avg_output_tokens_per_task": 7676.363636363636,
      "last_run_timestamp": "2026-10-17T11:00:48.240Z"
    },
    {
      "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": "Nebius",
      "accuracy": 0.76,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.11891346,
      "avg_duration_per_task_ms": 226293.89,
      "avg_output_tokens_per_task": 28257,
      "last_run_timestamp": "2026-10-17T10:59:28.226Z"
    },
    {
      "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": "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": 0.76,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.019457739999999998,
      "avg_duration_per_task_ms": 87744.74,
      "avg_output_tokens_per_task": 11443.04,
      "last_run_timestamp": "2026-10-17T19:53:56.764Z"
    },
    {
      "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": "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": "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": 0.76,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.0448255,
      "avg_duration_per_task_ms": 126685.96,
      "avg_output_tokens_per_task": 15758.14,
      "last_run_timestamp": "2026-10-17T10:19:11.543Z"
    },
    {
      "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": 0.76,
      "accuracy_std_dev": 0.014401594564491825,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.012781093333333333,
      "avg_duration_per_task_ms": 706143.8066666666,
      "avg_output_tokens_per_task": 24864.26,
      "last_run_timestamp": "2026-10-17T12:54:32.457Z"
    },
    {
      "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-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "SiliconFlow",
      "accuracy": 0.76,
      "accuracy_std_dev": 0.025531142930416043,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.013140583333333336,
      "avg_duration_per_task_ms": 222263.59333333332,
      "avg_output_tokens_per_task": 5092.44,
      "last_run_timestamp": "2026-10-17T11:12:55.230Z"
    },
    {
      "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": 0.7597916562500001,
      "accuracy_std_dev": 0.03048148907596357,
      "run_count": 32,
      "total_questions": 1600,
      "avg_cost_per_task_usd": 0.20304805999999995,
      "avg_duration_per_task_ms": 119095.471875,
      "avg_output_tokens_per_task": 7815.745625,
      "last_run_timestamp": "2026-10-17T11:12:58.092Z"
    },
    {
      "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": 0.7596329412844037,
      "accuracy_std_dev": 0.03240988921571945,
      "run_count": 11,
      "total_questions": 545,
      "avg_cost_per_task_usd": 0.1936376660550459,
      "avg_duration_per_task_ms": 554065.5724770642,
      "avg_output_tokens_per_task": 27060.469724770643,
      "last_run_timestamp": "2026-10-17T11:01:59.994Z"
    },
    {
      "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": 0.7596315477386935,
      "accuracy_std_dev": 0.034079836369912346,
      "run_count": 8,
      "total_questions": 398,
      "avg_cost_per_task_usd": 0.011217208542713568,
      "avg_duration_per_task_ms": 145495.03015075377,
      "avg_output_tokens_per_task": 13403.010050251256,
      "last_run_timestamp": "2026-10-17T19:18:31.538Z"
    },
    {
      "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/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": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Alibaba",
      "accuracy": 0.7592592222222222,
      "accuracy_std_dev": 0.03086336097769501,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.10491929333333334,
      "avg_duration_per_task_ms": 277405.59777777776,
      "avg_output_tokens_per_task": 12223.173333333334,
      "last_run_timestamp": "2026-10-17T11:16:04.165Z"
    },
    {
      "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": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Mancer 2",
      "accuracy": 0.7590475714285715,
      "accuracy_std_dev": 0.036934642282423845,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.09028300285714286,
      "avg_duration_per_task_ms": 562815.2714285714,
      "avg_output_tokens_per_task": 12727.568571428572,
      "last_run_timestamp": "2026-10-17T11:01:50.724Z"
    },
    {
      "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": 0.7590475714285715,
      "accuracy_std_dev": 0.019331153995980878,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.007369005714285714,
      "avg_duration_per_task_ms": 125780.48857142858,
      "avg_output_tokens_per_task": 4298.177142857143,
      "last_run_timestamp": "2026-10-17T11:22:24.281Z"
    },
    {
      "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": 0.7588888888888888,
      "accuracy_std_dev": 0.04049087993732375,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.15803105555555555,
      "avg_duration_per_task_ms": 321070.18888888886,
      "avg_output_tokens_per_task": 22494.742222222223,
      "last_run_timestamp": "2026-10-17T10:25:46.397Z"
    },
    {
      "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": 0.7588888333333333,
      "accuracy_std_dev": 0.010482195371464689,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.061315760000000004,
      "avg_duration_per_task_ms": 281940.7266666667,
      "avg_output_tokens_per_task": 13996.163333333334,
      "last_run_timestamp": "2026-10-17T10:26:47.850Z"
    },
    {
      "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": 0.7587879090909091,
      "accuracy_std_dev": 0.029103513053970746,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.20085546545454547,
      "avg_duration_per_task_ms": 125448.63818181818,
      "avg_output_tokens_per_task": 7874.443636363636,
      "last_run_timestamp": "2026-10-17T11:03:07.076Z"
    },
    {
      "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": 0.7586728009049774,
      "accuracy_std_dev": 0.02217328649078593,
      "run_count": 9,
      "total_questions": 442,
      "avg_cost_per_task_usd": 0.025404156108597285,
      "avg_duration_per_task_ms": 400240.5045248869,
      "avg_output_tokens_per_task": 12533.12443438914,
      "last_run_timestamp": "2026-10-17T19:19:50.674Z"
    },
    {
      "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": 0.7585186666666667,
      "accuracy_std_dev": 0.031234283510982678,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.08685644444444443,
      "avg_duration_per_task_ms": 69302.62222222223,
      "avg_output_tokens_per_task": 12174.302222222223,
      "last_run_timestamp": "2026-10-17T19:41:24.910Z"
    },
    {
      "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": 0.7583335,
      "accuracy_std_dev": 0.023273625990592902,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.28355265,
      "avg_duration_per_task_ms": 276032.65,
      "avg_output_tokens_per_task": 10176.66,
      "last_run_timestamp": "2026-10-17T11:01:26.330Z"
    },
    {
      "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": "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": 0.758333375,
      "accuracy_std_dev": 0.023273652844673413,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.007575310000000001,
      "avg_duration_per_task_ms": 189684.715,
      "avg_output_tokens_per_task": 4442.28,
      "last_run_timestamp": "2026-10-17T11:47:44.868Z"
    },
    {
      "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": 0.7583333333333333,
      "accuracy_std_dev": 0.021963715871611736,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.022033749999999998,
      "avg_duration_per_task_ms": 126674.91333333333,
      "avg_output_tokens_per_task": 17337.105,
      "last_run_timestamp": "2026-10-17T11:08:52.154Z"
    },
    {
      "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": 0.7583332500000001,
      "accuracy_std_dev": 0.02074966532712037,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.02415933,
      "avg_duration_per_task_ms": 139345.795,
      "avg_output_tokens_per_task": 13798.13,
      "last_run_timestamp": "2026-10-17T19:37:45.712Z"
    },
    {
      "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": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Crusoe",
      "accuracy": 0.7583331666666667,
      "accuracy_std_dev": 0.013844473427047428,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.09591996333333334,
      "avg_duration_per_task_ms": 214673.9,
      "avg_output_tokens_per_task": 9689.54,
      "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": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Alibaba",
      "accuracy": 0.7582582837837838,
      "accuracy_std_dev": 0.01616414926104476,
      "run_count": 9,
      "total_questions": 444,
      "avg_cost_per_task_usd": 0.03309702927927928,
      "avg_duration_per_task_ms": 172017.87837837837,
      "avg_output_tokens_per_task": 16666.704954954956,
      "last_run_timestamp": "2026-10-17T11:14:13.206Z"
    },
    {
      "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": 0.758189028901734,
      "accuracy_std_dev": 0.033845822723590796,
      "run_count": 7,
      "total_questions": 346,
      "avg_cost_per_task_usd": 0.057105187861271674,
      "avg_duration_per_task_ms": 394678.4624277457,
      "avg_output_tokens_per_task": 12961.20520231214,
      "last_run_timestamp": "2026-10-17T11:07:54.528Z"
    },
    {
      "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": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "Anthropic",
      "accuracy": 0.7577776666666667,
      "accuracy_std_dev": 0.016629618082071397,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.012770106666666666,
      "avg_duration_per_task_ms": 94420.36666666667,
      "avg_output_tokens_per_task": 11193.713333333333,
      "last_run_timestamp": "2026-10-17T10:59:45.577Z"
    },
    {
      "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": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Novita",
      "accuracy": 0.7576190714285714,
      "accuracy_std_dev": 0.02276274605096257,
      "run_count": 14,
      "total_questions": 700,
      "avg_cost_per_task_usd": 0.028185067142857137,
      "avg_duration_per_task_ms": 173417.5,
      "avg_output_tokens_per_task": 17841.255714285715,
      "last_run_timestamp": "2026-10-17T11:01:48.296Z"
    },
    {
      "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": 0.7575758181818182,
      "accuracy_std_dev": 0.03306223866318624,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.020269487272727277,
      "avg_duration_per_task_ms": 191902.98727272727,
      "avg_output_tokens_per_task": 14622.787272727273,
      "last_run_timestamp": "2026-10-17T11:03:04.803Z"
    },
    {
      "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": 0.75750025,
      "accuracy_std_dev": 0.017618680900893247,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.14844741250000001,
      "avg_duration_per_task_ms": 168776.3125,
      "avg_output_tokens_per_task": 20104.8775,
      "last_run_timestamp": "2026-10-17T11:13:05.682Z"
    },
    {
      "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": 0.7575,
      "accuracy_std_dev": 0.024022628223198242,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.069431845,
      "avg_duration_per_task_ms": 168602.24,
      "avg_output_tokens_per_task": 4679.8025,
      "last_run_timestamp": "2026-10-17T11:23:18.111Z"
    },
    {
      "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-v4.1-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": "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": 0.7573333999999999,
      "accuracy_std_dev": 0.05359938971518238,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.023817484,
      "avg_duration_per_task_ms": 360909.7,
      "avg_output_tokens_per_task": 13950.552,
      "last_run_timestamp": "2026-10-17T10:59:33.404Z"
    },
    {
      "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": 0.7573333,
      "accuracy_std_dev": 0.021333193750819417,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.017451074,
      "avg_duration_per_task_ms": 122650.494,
      "avg_output_tokens_per_task": 4554.818,
      "last_run_timestamp": "2026-10-17T11:17:33.222Z"
    },
    {
      "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": 0.757143,
      "accuracy_std_dev": 0.03282614272627407,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.014302074285714289,
      "avg_duration_per_task_ms": 167054.57142857142,
      "avg_output_tokens_per_task": 14974.468571428572,
      "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": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "GMICloud",
      "accuracy": 0.7571428571428571,
      "accuracy_std_dev": 0.017405469989374624,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.02240780857142857,
      "avg_duration_per_task_ms": 143693.9342857143,
      "avg_output_tokens_per_task": 16873.04,
      "last_run_timestamp": "2026-10-17T11:13:54.412Z"
    },
    {
      "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": 0.7569232307692307,
      "accuracy_std_dev": 0.022282632551121173,
      "run_count": 13,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.25500098,
      "avg_duration_per_task_ms": 282733.35846153845,
      "avg_output_tokens_per_task": 10086.226153846153,
      "last_run_timestamp": "2026-10-17T11:10:01.607Z"
    },
    {
      "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": 0.7566667499999999,
      "accuracy_std_dev": 0.026874078797002546,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.07064113500000001,
      "avg_duration_per_task_ms": 324361.97,
      "avg_output_tokens_per_task": 14598.985,
      "last_run_timestamp": "2026-10-17T11:22:26.174Z"
    },
    {
      "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": 0.7566667,
      "accuracy_std_dev": 0.036301371263493636,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.090532718,
      "avg_duration_per_task_ms": 236615.216,
      "avg_output_tokens_per_task": 12357.186,
      "last_run_timestamp": "2026-10-17T10:27:20.763Z"
    },
    {
      "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": 0.7566666666666667,
      "accuracy_std_dev": 0.02285220631337136,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.023255454999999998,
      "avg_duration_per_task_ms": 118711.59,
      "avg_output_tokens_per_task": 15018.058333333332,
      "last_run_timestamp": "2026-10-17T10:59:43.836Z"
    },
    {
      "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": 0.756666625,
      "accuracy_std_dev": 0.028867614496774323,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.022124855000000002,
      "avg_duration_per_task_ms": 493245.72,
      "avg_output_tokens_per_task": 14533.4,
      "last_run_timestamp": "2026-10-17T19:02:12.105Z"
    },
    {
      "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": 0.7565656363636365,
      "accuracy_std_dev": 0.03454844880508337,
      "run_count": 33,
      "total_questions": 1650,
      "avg_cost_per_task_usd": 0.22266426666666672,
      "avg_duration_per_task_ms": 96431.12363636364,
      "avg_output_tokens_per_task": 7566.764242424242,
      "last_run_timestamp": "2026-10-17T11:22:38.996Z"
    },
    {
      "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": 0.7564362073170733,
      "accuracy_std_dev": 0.024511456002449167,
      "run_count": 10,
      "total_questions": 492,
      "avg_cost_per_task_usd": 0.014174674796747964,
      "avg_duration_per_task_ms": 53986.221544715445,
      "avg_output_tokens_per_task": 4039.2662601626016,
      "last_run_timestamp": "2026-10-17T11:16:33.462Z"
    },
    {
      "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": "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": 0.013569731111111111,
      "avg_duration_per_task_ms": 218757.14444444445,
      "avg_output_tokens_per_task": 14705.06,
      "last_run_timestamp": "2026-10-17T19:58:35.302Z"
    },
    {
      "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": 0.7561905714285714,
      "accuracy_std_dev": 0.01326515904656301,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.19363898571428573,
      "avg_duration_per_task_ms": 310234.1457142857,
      "avg_output_tokens_per_task": 12005.822857142857,
      "last_run_timestamp": "2026-10-17T10:45:58.767Z"
    },
    {
      "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": 0.7561904285714286,
      "accuracy_std_dev": 0.019633948342873183,
      "run_count": 14,
      "total_questions": 700,
      "avg_cost_per_task_usd": 0.026500512857142854,
      "avg_duration_per_task_ms": 291184.5557142857,
      "avg_output_tokens_per_task": 17123.52857142857,
      "last_run_timestamp": "2026-10-17T11:00:59.508Z"
    },
    {
      "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": 0.756,
      "accuracy_std_dev": 0.024073867807230306,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.351000588,
      "avg_duration_per_task_ms": 102672.34,
      "avg_output_tokens_per_task": 7151.472,
      "last_run_timestamp": "2026-10-17T11:07:54.497Z"
    },
    {
      "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": "Venice",
      "accuracy": 0.7558921212121212,
      "accuracy_std_dev": 0.026509964125541553,
      "run_count": 8,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.0215872904040404,
      "avg_duration_per_task_ms": 248439.8383838384,
      "avg_output_tokens_per_task": 4242.020202020202,
      "last_run_timestamp": "2026-10-17T11:01:48.514Z"
    },
    {
      "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": 0.75583325,
      "accuracy_std_dev": 0.016308889023704216,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.09814360999999999,
      "avg_duration_per_task_ms": 308482.07,
      "avg_output_tokens_per_task": 12874.305,
      "last_run_timestamp": "2026-10-17T19:56:39.413Z"
    },
    {
      "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": 0.75583325,
      "accuracy_std_dev": 0.017300429508757886,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.17775563000000003,
      "avg_duration_per_task_ms": 323022.9725,
      "avg_output_tokens_per_task": 10875.7,
      "last_run_timestamp": "2026-10-17T10:27:08.906Z"
    },
    {
      "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": 0.7557576363636364,
      "accuracy_std_dev": 0.027008915986704473,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.5367455872727274,
      "avg_duration_per_task_ms": 190847.75454545455,
      "avg_output_tokens_per_task": 10370.42909090909,
      "last_run_timestamp": "2026-10-17T10:58:42.069Z"
    },
    {
      "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": 0.7557079178082191,
      "accuracy_std_dev": 0.01607886717264481,
      "run_count": 3,
      "total_questions": 146,
      "avg_cost_per_task_usd": 0.2608585205479452,
      "avg_duration_per_task_ms": 125025.47260273973,
      "avg_output_tokens_per_task": 7152.054794520548,
      "last_run_timestamp": "2026-10-03T08:05:09.514Z"
    },
    {
      "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": 0.7555555555555555,
      "accuracy_std_dev": 0.016329992101182772,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.06037513111111111,
      "avg_duration_per_task_ms": 215849.39333333334,
      "avg_output_tokens_per_task": 10590.155555555555,
      "last_run_timestamp": "2026-10-17T19:53:48.237Z"
    },
    {
      "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": 0.7555553333333334,
      "accuracy_std_dev": 0.013698571612479267,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.010344493333333333,
      "avg_duration_per_task_ms": 251680.81333333332,
      "avg_output_tokens_per_task": 16163.846666666666,
      "last_run_timestamp": "2026-10-17T11:43:44.655Z"
    },
    {
      "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": 0.7550001666666666,
      "accuracy_std_dev": 0.026925751821237763,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.014109759999999997,
      "avg_duration_per_task_ms": 255723.91333333333,
      "avg_output_tokens_per_task": 16860.226666666666,
      "last_run_timestamp": "2026-10-17T11:01:23.939Z"
    },
    {
      "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": 0.755,
      "accuracy_std_dev": 0.019649908116833523,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.33917411000000003,
      "avg_duration_per_task_ms": 125094.895,
      "avg_output_tokens_per_task": 7173.1,
      "last_run_timestamp": "2026-10-17T10:36:11.487Z"
    },
    {
      "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": 0.755,
      "accuracy_std_dev": 0.021666999999999992,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.03593898,
      "avg_duration_per_task_ms": 72072.58,
      "avg_output_tokens_per_task": 16785.94,
      "last_run_timestamp": "2026-10-17T11:01:14.141Z"
    },
    {
      "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": 0.7548148888888889,
      "accuracy_std_dev": 0.027716145949987135,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.33142048444444444,
      "avg_duration_per_task_ms": 603214.5822222222,
      "avg_output_tokens_per_task": 27052.68,
      "last_run_timestamp": "2026-10-17T19:11:02.176Z"
    },
    {
      "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": 0.7546065175879397,
      "accuracy_std_dev": 0.07084324947628658,
      "run_count": 8,
      "total_questions": 398,
      "avg_cost_per_task_usd": 0.25537817085427134,
      "avg_duration_per_task_ms": 457516.3442211055,
      "avg_output_tokens_per_task": 29053.731155778896,
      "last_run_timestamp": "2026-10-17T19:15:48.233Z"
    },
    {
      "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": 0.7546032380952381,
      "accuracy_std_dev": 0.03353078090196666,
      "run_count": 21,
      "total_questions": 1050,
      "avg_cost_per_task_usd": 0.20101115333333333,
      "avg_duration_per_task_ms": 102086.14857142857,
      "avg_output_tokens_per_task": 7692.0952380952385,
      "last_run_timestamp": "2026-10-17T11:23:18.195Z"
    },
    {
      "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": 0.7545454545454545,
      "accuracy_std_dev": 0.020659486891294522,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.34660061999999997,
      "avg_duration_per_task_ms": 220288.40363636363,
      "avg_output_tokens_per_task": 11981.236363636364,
      "last_run_timestamp": "2026-10-17T11:15:56.338Z"
    },
    {
      "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": 0.7544342825688073,
      "accuracy_std_dev": 0.02504935915479907,
      "run_count": 11,
      "total_questions": 545,
      "avg_cost_per_task_usd": 0.1000418,
      "avg_duration_per_task_ms": 796849.0642201835,
      "avg_output_tokens_per_task": 21296.05137614679,
      "last_run_timestamp": "2026-10-10T11:45:05.221Z"
    },
    {
      "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": "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": 0.75416675,
      "accuracy_std_dev": 0.023555451717702226,
      "run_count": 32,
      "total_questions": 1600,
      "avg_cost_per_task_usd": 0.37240653,
      "avg_duration_per_task_ms": 164105.02375,
      "avg_output_tokens_per_task": 11891.904375,
      "last_run_timestamp": "2026-10-17T11:22:25.890Z"
    },
    {
      "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": 0.754166625,
      "accuracy_std_dev": 0.01898479106375351,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.00794965,
      "avg_duration_per_task_ms": 155474.15,
      "avg_output_tokens_per_task": 13162.275,
      "last_run_timestamp": "2026-10-17T19:15:01.096Z"
    },
    {
      "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": 0.7539394545454545,
      "accuracy_std_dev": 0.027185124626810275,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.4974660309090909,
      "avg_duration_per_task_ms": 195466.43818181817,
      "avg_output_tokens_per_task": 10115.303636363637,
      "last_run_timestamp": "2026-10-17T11:26:23.900Z"
    },
    {
      "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": 0.7538460512820513,
      "accuracy_std_dev": 0.013548520491182818,
      "run_count": 4,
      "total_questions": 195,
      "avg_cost_per_task_usd": 0.0682567846153846,
      "avg_duration_per_task_ms": 219433.69743589743,
      "avg_output_tokens_per_task": 7166.764102564102,
      "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": "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": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "DeepInfra",
      "accuracy": 0.7537594536340853,
      "accuracy_std_dev": 0.022172530138312475,
      "run_count": 16,
      "total_questions": 798,
      "avg_cost_per_task_usd": 0.007178401002506266,
      "avg_duration_per_task_ms": 202699.93609022556,
      "avg_output_tokens_per_task": 13304.931077694235,
      "last_run_timestamp": "2026-10-17T11:00:48.272Z"
    },
    {
      "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": 0.7534247123287672,
      "accuracy_std_dev": 0.009625530542376443,
      "run_count": 3,
      "total_questions": 146,
      "avg_cost_per_task_usd": 0.033538157534246577,
      "avg_duration_per_task_ms": 519040.26712328766,
      "avg_output_tokens_per_task": 11776.294520547945,
      "last_run_timestamp": "2026-10-17T11:01:17.325Z"
    },
    {
      "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": "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": 0.753333375,
      "accuracy_std_dev": 0.03349946931645897,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.007990629999999999,
      "avg_duration_per_task_ms": 168858.505,
      "avg_output_tokens_per_task": 4418.365,
      "last_run_timestamp": "2026-10-17T11:15:43.073Z"
    },
    {
      "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": "Cloudflare",
      "accuracy": 0.7533333333333333,
      "accuracy_std_dev": 0.027755466596690478,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.02379494888888889,
      "avg_duration_per_task_ms": 218862.64444444445,
      "avg_output_tokens_per_task": 13387.833333333334,
      "last_run_timestamp": "2026-10-17T19:57:48.068Z"
    },
    {
      "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": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7533333333333333,
      "accuracy_std_dev": 0.009428090415820642,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.30728538666666666,
      "avg_duration_per_task_ms": 307577.23333333334,
      "avg_output_tokens_per_task": 10871.333333333334,
      "last_run_timestamp": "2026-10-17T11:22:31.443Z"
    },
    {
      "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-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Near AI",
      "accuracy": 0.7533333333333333,
      "accuracy_std_dev": 0.04354662039954679,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.008277793333333333,
      "avg_duration_per_task_ms": 330908.58666666667,
      "avg_output_tokens_per_task": 4160.446666666667,
      "last_run_timestamp": "2026-10-17T11:02:35.316Z"
    },
    {
      "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": "Phala",
      "accuracy": 0.7533331666666667,
      "accuracy_std_dev": 0.0249444271225503,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.01177491,
      "avg_duration_per_task_ms": 456509.82666666666,
      "avg_output_tokens_per_task": 5746.43,
      "last_run_timestamp": "2026-10-17T11:16:27.809Z"
    },
    {
      "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": "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": 0.753333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.04351718,
      "avg_duration_per_task_ms": 151205.1,
      "avg_output_tokens_per_task": 15714.7,
      "last_run_timestamp": "2026-10-17T11:21:27.566Z"
    },
    {
      "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": 0.753333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.027578619999999998,
      "avg_duration_per_task_ms": 79703.36,
      "avg_output_tokens_per_task": 19558.22,
      "last_run_timestamp": "2026-10-17T11:01:12.710Z"
    },
    {
      "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": "anthropic/claude-4.5-sonnet-20250929",
      "model_slug": "anthropic/claude-sonnet-4.5",
      "model_name": "Anthropic: Claude Sonnet 4.5",
      "provider_name": "Anthropic",
      "accuracy": 0.7528572857142857,
      "accuracy_std_dev": 0.029543055172381347,
      "run_count": 14,
      "total_questions": 700,
      "avg_cost_per_task_usd": 0.35331957714285717,
      "avg_duration_per_task_ms": 231631.49,
      "avg_output_tokens_per_task": 10894.82142857143,
      "last_run_timestamp": "2026-10-17T10:59:10.049Z"
    },
    {
      "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": 0.7528390961923848,
      "accuracy_std_dev": 0.021447321416904284,
      "run_count": 10,
      "total_questions": 499,
      "avg_cost_per_task_usd": 0.028104985971943892,
      "avg_duration_per_task_ms": 567640.0821643287,
      "avg_output_tokens_per_task": 22541.43887775551,
      "last_run_timestamp": "2026-10-17T19:46:06.750Z"
    },
    {
      "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": 0.7526667,
      "accuracy_std_dev": 0.03133329149020257,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.014175738,
      "avg_duration_per_task_ms": 140877.726,
      "avg_output_tokens_per_task": 16423.73,
      "last_run_timestamp": "2026-10-17T11:13:51.811Z"
    },
    {
      "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": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Ionstream",
      "accuracy": 0.7525252929292929,
      "accuracy_std_dev": 0.03312506718986897,
      "run_count": 6,
      "total_questions": 297,
      "avg_cost_per_task_usd": 0.10546999326599327,
      "avg_duration_per_task_ms": 422614.45791245793,
      "avg_output_tokens_per_task": 27716.097643097644,
      "last_run_timestamp": "2026-10-17T11:01:37.513Z"
    },
    {
      "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": 0.752381,
      "accuracy_std_dev": 0.012046696749376112,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.16778733428571427,
      "avg_duration_per_task_ms": 474167.0485714286,
      "avg_output_tokens_per_task": 9790.025714285714,
      "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": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "StreamLake",
      "accuracy": 0.7522322901785714,
      "accuracy_std_dev": 0.028704344916087695,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.005231792410714286,
      "avg_duration_per_task_ms": 267977.26339285716,
      "avg_output_tokens_per_task": 12520.285714285714,
      "last_run_timestamp": "2026-10-17T19:42:09.889Z"
    },
    {
      "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-opus-20260205",
      "model_slug": "anthropic/claude-opus-4.6",
      "model_name": "Anthropic: Claude Opus 4.6",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7521212727272727,
      "accuracy_std_dev": 0.030855459711638198,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.5465765,
      "avg_duration_per_task_ms": 183245.5,
      "avg_output_tokens_per_task": 10464.481818181817,
      "last_run_timestamp": "2026-10-17T11:16:11.369Z"
    },
    {
      "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": 0.752121090909091,
      "accuracy_std_dev": 0.019450635878619636,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.037303981818181814,
      "avg_duration_per_task_ms": 289240.86,
      "avg_output_tokens_per_task": 22102.725454545456,
      "last_run_timestamp": "2026-10-17T19:41:19.572Z"
    },
    {
      "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": 0.7520080321285141,
      "accuracy_std_dev": 0.02504223543536002,
      "run_count": 10,
      "total_questions": 498,
      "avg_cost_per_task_usd": 0.044945560240963855,
      "avg_duration_per_task_ms": 491918.7791164659,
      "avg_output_tokens_per_task": 23149.75502008032,
      "last_run_timestamp": "2026-10-17T19:15:16.916Z"
    },
    {
      "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": 0.752,
      "accuracy_std_dev": 0.022469515152757562,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.019928192,
      "avg_duration_per_task_ms": 74879.332,
      "avg_output_tokens_per_task": 4253.16,
      "last_run_timestamp": "2026-10-10T11:24:18.065Z"
    },
    {
      "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": 0.7519999333333334,
      "accuracy_std_dev": 0.015625282459171393,
      "run_count": 15,
      "total_questions": 750,
      "avg_cost_per_task_usd": 0.36818162,
      "avg_duration_per_task_ms": 160625.57733333332,
      "avg_output_tokens_per_task": 7185.578666666666,
      "last_run_timestamp": "2026-10-17T10:59:56.018Z"
    },
    {
      "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": 0.7516665,
      "accuracy_std_dev": 0.0327870655403926,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.024614285,
      "avg_duration_per_task_ms": 136581.2,
      "avg_output_tokens_per_task": 8213.13,
      "last_run_timestamp": "2026-10-17T10:30:18.157Z"
    },
    {
      "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": 0.7512820769230769,
      "accuracy_std_dev": 0.022093110003108842,
      "run_count": 13,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.009568580000000002,
      "avg_duration_per_task_ms": 227674.30153846153,
      "avg_output_tokens_per_task": 4275.484615384616,
      "last_run_timestamp": "2026-10-17T11:16:25.425Z"
    },
    {
      "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": 0.7512120909090909,
      "accuracy_std_dev": 0.028024962309453366,
      "run_count": 22,
      "total_questions": 1100,
      "avg_cost_per_task_usd": 0.5612838363636363,
      "avg_duration_per_task_ms": 202150.13818181818,
      "avg_output_tokens_per_task": 9898.827272727272,
      "last_run_timestamp": "2026-10-17T11:07:54.531Z"
    },
    {
      "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-sonnet-5-20260630",
      "model_slug": "anthropic/claude-sonnet-5",
      "model_name": "Anthropic: Claude Sonnet 5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.7511112222222222,
      "accuracy_std_dev": 0.024343305923476197,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.19633602444444442,
      "avg_duration_per_task_ms": 119418.28666666667,
      "avg_output_tokens_per_task": 7693.504444444445,
      "last_run_timestamp": "2026-10-17T10:24:17.909Z"
    },
    {
      "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": 0.751111,
      "accuracy_std_dev": 0.03691848879476096,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.07572861111111111,
      "avg_duration_per_task_ms": 225640.9911111111,
      "avg_output_tokens_per_task": 9556.191111111111,
      "last_run_timestamp": "2026-10-17T19:39:27.161Z"
    },
    {
      "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": "Inceptron",
      "accuracy": 0.7510040843373493,
      "accuracy_std_dev": 0.03597470626982242,
      "run_count": 5,
      "total_questions": 249,
      "avg_cost_per_task_usd": 0.01730639357429719,
      "avg_duration_per_task_ms": 141614.70682730924,
      "avg_output_tokens_per_task": 4001.2329317269077,
      "last_run_timestamp": "2026-10-17T11:16:03.104Z"
    },
    {
      "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": 0.7509090909090909,
      "accuracy_std_dev": 0.021699979478224694,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.32347144,
      "avg_duration_per_task_ms": 197653.47454545455,
      "avg_output_tokens_per_task": 11502.629090909091,
      "last_run_timestamp": "2026-10-17T11:00:48.264Z"
    },
    {
      "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": 0.7506668,
      "accuracy_std_dev": 0.025508095447524096,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.354436884,
      "avg_duration_per_task_ms": 88439.744,
      "avg_output_tokens_per_task": 7011.238,
      "last_run_timestamp": "2026-10-17T10:22:34.532Z"
    },
    {
      "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": 0.7505556666666666,
      "accuracy_std_dev": 0.0348230358702716,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.014334536666666668,
      "avg_duration_per_task_ms": 41852.60666666667,
      "avg_output_tokens_per_task": 4084.32,
      "last_run_timestamp": "2026-10-17T11:16:45.385Z"
    },
    {
      "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": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7503703333333334,
      "accuracy_std_dev": 0.03357129710200534,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.16148964666666668,
      "avg_duration_per_task_ms": 232801.40444444446,
      "avg_output_tokens_per_task": 9705.164444444445,
      "last_run_timestamp": "2026-10-17T19:58:33.819Z"
    },
    {
      "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": 0.7500001,
      "accuracy_std_dev": 0.020275803862979145,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.079513518,
      "avg_duration_per_task_ms": 279626.384,
      "avg_output_tokens_per_task": 11402.63,
      "last_run_timestamp": "2026-10-17T23:15:35.483Z"
    },
    {
      "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": "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": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "auto-routing",
      "accuracy": 0.749697,
      "accuracy_std_dev": 0.04827223001986659,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.1384416090909091,
      "avg_duration_per_task_ms": 852920.6018181818,
      "avg_output_tokens_per_task": 32022.93090909091,
      "last_run_timestamp": "2026-10-17T19:47:55.902Z"
    },
    {
      "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": 0.7496969090909091,
      "accuracy_std_dev": 0.03721236699235288,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.031196743636363634,
      "avg_duration_per_task_ms": 263478.21454545454,
      "avg_output_tokens_per_task": 17098.916363636363,
      "last_run_timestamp": "2026-10-17T10:59:43.837Z"
    },
    {
      "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": 0.7496295555555555,
      "accuracy_std_dev": 0.03543190381597766,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.09387901111111112,
      "avg_duration_per_task_ms": 235629.06666666668,
      "avg_output_tokens_per_task": 12019.075555555555,
      "last_run_timestamp": "2026-10-17T20:23:53.468Z"
    },
    {
      "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": 0.749166625,
      "accuracy_std_dev": 0.02559034513218561,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.17263283,
      "avg_duration_per_task_ms": 326399.945,
      "avg_output_tokens_per_task": 11094.7075,
      "last_run_timestamp": "2026-10-17T11:01:52.082Z"
    },
    {
      "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": 0.7490475714285715,
      "accuracy_std_dev": 0.026500335856939866,
      "run_count": 14,
      "total_questions": 700,
      "avg_cost_per_task_usd": 0.34032975285714284,
      "avg_duration_per_task_ms": 209096.78714285715,
      "avg_output_tokens_per_task": 11852.83,
      "last_run_timestamp": "2026-10-17T11:11:39.553Z"
    },
    {
      "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": 0.748889,
      "accuracy_std_dev": 0.013698950130089036,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.012787326666666666,
      "avg_duration_per_task_ms": 96341.12,
      "avg_output_tokens_per_task": 11111.46,
      "last_run_timestamp": "2026-10-17T10:28:17.794Z"
    },
    {
      "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": 0.7488888333333333,
      "accuracy_std_dev": 0.03166158633853052,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.021630793333333332,
      "avg_duration_per_task_ms": 180688.53666666665,
      "avg_output_tokens_per_task": 8425.083333333334,
      "last_run_timestamp": "2026-10-17T19:19:35.443Z"
    },
    {
      "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": "Phala",
      "accuracy": 0.748484909090909,
      "accuracy_std_dev": 0.025280931741512393,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.08331156545454546,
      "avg_duration_per_task_ms": 839586.4345454546,
      "avg_output_tokens_per_task": 36324.976363636364,
      "last_run_timestamp": "2026-10-17T19:46:44.648Z"
    },
    {
      "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": 0.7484848181818182,
      "accuracy_std_dev": 0.027574050564115497,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.5215495545454547,
      "avg_duration_per_task_ms": 226716.27272727274,
      "avg_output_tokens_per_task": 10061.14,
      "last_run_timestamp": "2026-10-17T10:58:52.040Z"
    },
    {
      "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-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Venice",
      "accuracy": 0.7483332500000001,
      "accuracy_std_dev": 0.03362363173703132,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.10647082,
      "avg_duration_per_task_ms": 364506.8875,
      "avg_output_tokens_per_task": 8022.99,
      "last_run_timestamp": "2026-10-17T19:40:08.644Z"
    },
    {
      "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": 0.7482990408163266,
      "accuracy_std_dev": 0.02555550000000001,
      "run_count": 2,
      "total_questions": 98,
      "avg_cost_per_task_usd": 0.015089622448979592,
      "avg_duration_per_task_ms": 65496.448979591834,
      "avg_output_tokens_per_task": 4181.989795918367,
      "last_run_timestamp": "2026-10-17T10:59:16.472Z"
    },
    {
      "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": 0.7481482222222222,
      "accuracy_std_dev": 0.025293761601442692,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.34738337555555554,
      "avg_duration_per_task_ms": 200384.05333333334,
      "avg_output_tokens_per_task": 11864.717777777778,
      "last_run_timestamp": "2026-10-17T11:13:53.871Z"
    },
    {
      "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": 0.7481481111111111,
      "accuracy_std_dev": 0.038332810160036944,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.11396319111111111,
      "avg_duration_per_task_ms": 269132.9111111111,
      "avg_output_tokens_per_task": 9467.16888888889,
      "last_run_timestamp": "2026-10-17T19:41:15.762Z"
    },
    {
      "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": 0.7481398080357142,
      "accuracy_std_dev": 0.020714088201028814,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.11667296428571428,
      "avg_duration_per_task_ms": 119172.515625,
      "avg_output_tokens_per_task": 10392.310267857143,
      "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": "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-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7479999,
      "accuracy_std_dev": 0.04171853092200154,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.041866538,
      "avg_duration_per_task_ms": 196873.67,
      "avg_output_tokens_per_task": 19363.92,
      "last_run_timestamp": "2026-10-17T11:00:50.913Z"
    },
    {
      "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": 0.7477778333333334,
      "accuracy_std_dev": 0.02878187658589265,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.023901795,
      "avg_duration_per_task_ms": 159481.33333333334,
      "avg_output_tokens_per_task": 7971.18,
      "last_run_timestamp": "2026-10-17T19:25:51.481Z"
    },
    {
      "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-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "DeepInfra",
      "accuracy": 0.7475,
      "accuracy_std_dev": 0.016811884025890747,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.01307641,
      "avg_duration_per_task_ms": 361470.8875,
      "avg_output_tokens_per_task": 16341.3725,
      "last_run_timestamp": "2026-10-17T11:13:02.351Z"
    },
    {
      "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": 0.7474074444444444,
      "accuracy_std_dev": 0.030216386511336353,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.07081323111111111,
      "avg_duration_per_task_ms": 255490.43777777778,
      "avg_output_tokens_per_task": 9828.68888888889,
      "last_run_timestamp": "2026-10-17T19:51:09.246Z"
    },
    {
      "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": 0.7474073333333333,
      "accuracy_std_dev": 0.013496814644780307,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.013488784444444445,
      "avg_duration_per_task_ms": 102984.29111111112,
      "avg_output_tokens_per_task": 10999.817777777778,
      "last_run_timestamp": "2026-10-17T11:00:48.268Z"
    },
    {
      "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": 0.7473841614349777,
      "accuracy_std_dev": 0.04596493608054828,
      "run_count": 9,
      "total_questions": 446,
      "avg_cost_per_task_usd": 0.07238068609865472,
      "avg_duration_per_task_ms": 340205.40134529147,
      "avg_output_tokens_per_task": 12677.688340807175,
      "last_run_timestamp": "2026-10-17T20:02:16.814Z"
    },
    {
      "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": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Reka",
      "accuracy": 0.7472220833333334,
      "accuracy_std_dev": 0.022187493002659323,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.013829399999999999,
      "avg_duration_per_task_ms": 128044.86833333333,
      "avg_output_tokens_per_task": 15864.645,
      "last_run_timestamp": "2026-10-17T10:46:05.123Z"
    },
    {
      "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-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "Venice",
      "accuracy": 0.747037111111111,
      "accuracy_std_dev": 0.03619039918555579,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.07854862444444444,
      "avg_duration_per_task_ms": 352663.33777777775,
      "avg_output_tokens_per_task": 9208.286666666667,
      "last_run_timestamp": "2026-10-17T19:38:35.311Z"
    },
    {
      "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": 0.7468716348122866,
      "accuracy_std_dev": 0.026415158764534374,
      "run_count": 6,
      "total_questions": 293,
      "avg_cost_per_task_usd": 0.1360147542662116,
      "avg_duration_per_task_ms": 219042.409556314,
      "avg_output_tokens_per_task": 11992.477815699658,
      "last_run_timestamp": "2026-10-17T19:36:10.889Z"
    },
    {
      "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": 0.746667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.05210678,
      "avg_duration_per_task_ms": 128609.46,
      "avg_output_tokens_per_task": 18103.84,
      "last_run_timestamp": "2026-10-17T11:13:14.747Z"
    },
    {
      "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": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7466667499999999,
      "accuracy_std_dev": 0.02687417182328601,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.07283523750000001,
      "avg_duration_per_task_ms": 314827.96,
      "avg_output_tokens_per_task": 11539.3075,
      "last_run_timestamp": "2026-10-17T23:23:01.052Z"
    },
    {
      "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": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "DeepInfra",
      "accuracy": 0.746666625,
      "accuracy_std_dev": 0.03179809568172874,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.4209529125,
      "avg_duration_per_task_ms": 657671.5425,
      "avg_output_tokens_per_task": 9874.79,
      "last_run_timestamp": "2026-10-17T10:59:15.031Z"
    },
    {
      "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": "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": 0.7466666000000001,
      "accuracy_std_dev": 0.012649321461643697,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.35012448,
      "avg_duration_per_task_ms": 129259.12,
      "avg_output_tokens_per_task": 7227.552,
      "last_run_timestamp": "2026-10-17T10:38:33.392Z"
    },
    {
      "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": 0.7466666000000001,
      "accuracy_std_dev": 0.04226904304618216,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.011620872,
      "avg_duration_per_task_ms": 319057.722,
      "avg_output_tokens_per_task": 16204.506,
      "last_run_timestamp": "2026-10-17T11:16:46.243Z"
    },
    {
      "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": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "Google",
      "accuracy": 0.7466665555555556,
      "accuracy_std_dev": 0.019116420452416828,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.013255473333333333,
      "avg_duration_per_task_ms": 72684.63333333333,
      "avg_output_tokens_per_task": 11005.913333333334,
      "last_run_timestamp": "2026-10-17T11:00:48.240Z"
    },
    {
      "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": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "DeepInfra",
      "accuracy": 0.7464314105793451,
      "accuracy_std_dev": 0.030785849417538806,
      "run_count": 8,
      "total_questions": 397,
      "avg_cost_per_task_usd": 0.06752257934508815,
      "avg_duration_per_task_ms": 443836.2191435768,
      "avg_output_tokens_per_task": 10462.246851385391,
      "last_run_timestamp": "2026-10-10T18:57:24.715Z"
    },
    {
      "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": "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": 0.7461111666666667,
      "accuracy_std_dev": 0.018147215208002467,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.016089411666666668,
      "avg_duration_per_task_ms": 175782.82666666666,
      "avg_output_tokens_per_task": 4282.083333333333,
      "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": "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": 0.7459256666666667,
      "accuracy_std_dev": 0.02542365100103096,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.08970129333333333,
      "avg_duration_per_task_ms": 241877.4,
      "avg_output_tokens_per_task": 10154.835555555555,
      "last_run_timestamp": "2026-10-17T20:11:09.874Z"
    },
    {
      "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": 0.745833375,
      "accuracy_std_dev": 0.03357197181778836,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.1318732225,
      "avg_duration_per_task_ms": 359914.9725,
      "avg_output_tokens_per_task": 11863.2075,
      "last_run_timestamp": "2026-10-17T19:38:35.577Z"
    },
    {
      "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": 0.7458124924623115,
      "accuracy_std_dev": 0.02659090084943297,
      "run_count": 8,
      "total_questions": 398,
      "avg_cost_per_task_usd": 0.0608610351758794,
      "avg_duration_per_task_ms": 201989.4798994975,
      "avg_output_tokens_per_task": 32682.85175879397,
      "last_run_timestamp": "2026-10-17T10:49:21.111Z"
    },
    {
      "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": 0.7456411538461538,
      "accuracy_std_dev": 0.027280880543103812,
      "run_count": 13,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.0642081323076923,
      "avg_duration_per_task_ms": 143388.6953846154,
      "avg_output_tokens_per_task": 9780.298461538461,
      "last_run_timestamp": "2026-10-17T19:39:04.582Z"
    },
    {
      "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": "Google",
      "accuracy": 0.7453763225806453,
      "accuracy_std_dev": 0.03210277114898852,
      "run_count": 31,
      "total_questions": 1550,
      "avg_cost_per_task_usd": 0.3755844303225807,
      "avg_duration_per_task_ms": 200838.47032258066,
      "avg_output_tokens_per_task": 11986.134838709677,
      "last_run_timestamp": "2026-10-17T11:25:08.770Z"
    },
    {
      "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": 0.7452106666666666,
      "accuracy_std_dev": 0.020011549577426443,
      "run_count": 7,
      "total_questions": 348,
      "avg_cost_per_task_usd": 0.047182301724137926,
      "avg_duration_per_task_ms": 787118.643678161,
      "avg_output_tokens_per_task": 14354.25,
      "last_run_timestamp": "2026-10-17T19:16:29.038Z"
    },
    {
      "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": 0.7450001666666666,
      "accuracy_std_dev": 0.024324256298440514,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.017958773333333337,
      "avg_duration_per_task_ms": 482242.6666666667,
      "avg_output_tokens_per_task": 25086.3,
      "last_run_timestamp": "2026-10-17T11:07:54.526Z"
    },
    {
      "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": 0.745,
      "accuracy_std_dev": 0.02020721818435187,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.010433062500000001,
      "avg_duration_per_task_ms": 212069.0525,
      "avg_output_tokens_per_task": 14202.0475,
      "last_run_timestamp": "2026-10-17T19:19:49.453Z"
    },
    {
      "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": 0.745,
      "accuracy_std_dev": 0.015184019378939179,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.044957634999999996,
      "avg_duration_per_task_ms": 186177.835,
      "avg_output_tokens_per_task": 30974.63,
      "last_run_timestamp": "2026-10-17T11:15:59.791Z"
    },
    {
      "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": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "DeepInfra",
      "accuracy": 0.745,
      "accuracy_std_dev": 0.015122928387055215,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.03447282333333334,
      "avg_duration_per_task_ms": 421378.29333333333,
      "avg_output_tokens_per_task": 9247.876666666667,
      "last_run_timestamp": "2026-10-09T02:05:38.689Z"
    },
    {
      "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": 0.744993393858478,
      "accuracy_std_dev": 0.01640917934761584,
      "run_count": 15,
      "total_questions": 749,
      "avg_cost_per_task_usd": 0.04752114152202937,
      "avg_duration_per_task_ms": 285856.17356475303,
      "avg_output_tokens_per_task": 27467.751668891855,
      "last_run_timestamp": "2026-10-17T19:35:48.022Z"
    },
    {
      "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": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Parasail",
      "accuracy": 0.7444864177693761,
      "accuracy_std_dev": 0.0319646756633704,
      "run_count": 11,
      "total_questions": 529,
      "avg_cost_per_task_usd": 0.02029702079395085,
      "avg_duration_per_task_ms": 91063.5595463138,
      "avg_output_tokens_per_task": 6679.818525519849,
      "last_run_timestamp": "2026-10-17T18:34:28.093Z"
    },
    {
      "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": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "Azure",
      "accuracy": 0.7444443333333334,
      "accuracy_std_dev": 0.010657241210661543,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.012899976666666667,
      "avg_duration_per_task_ms": 76869.22333333333,
      "avg_output_tokens_per_task": 10868.43,
      "last_run_timestamp": "2026-10-17T10:31:56.734Z"
    },
    {
      "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": "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": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Makora",
      "accuracy": 0.7441076363636364,
      "accuracy_std_dev": 0.008720982312790204,
      "run_count": 4,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.2539274696969697,
      "avg_duration_per_task_ms": 431832.7626262626,
      "avg_output_tokens_per_task": 12531.59090909091,
      "last_run_timestamp": "2026-10-17T10:59:55.039Z"
    },
    {
      "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": 0.7440801639344262,
      "accuracy_std_dev": 0.02729307539006798,
      "run_count": 11,
      "total_questions": 549,
      "avg_cost_per_task_usd": 0.011795925318761384,
      "avg_duration_per_task_ms": 451361.3533697632,
      "avg_output_tokens_per_task": 14179.424408014573,
      "last_run_timestamp": "2026-10-17T10:31:56.493Z"
    },
    {
      "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": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "Xiaomi",
      "accuracy": 0.743889,
      "accuracy_std_dev": 0.03872595709813596,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.02433945333333333,
      "avg_duration_per_task_ms": 401395.9033333333,
      "avg_output_tokens_per_task": 15168.051666666666,
      "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": "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": 0.7438888333333333,
      "accuracy_std_dev": 0.03327303342256143,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.027845846666666667,
      "avg_duration_per_task_ms": 428793.8,
      "avg_output_tokens_per_task": 14516.615,
      "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": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "Together",
      "accuracy": 0.7438479642058166,
      "accuracy_std_dev": 0.03102502627609039,
      "run_count": 9,
      "total_questions": 447,
      "avg_cost_per_task_usd": 0.06547314988814318,
      "avg_duration_per_task_ms": 990556.5682326622,
      "avg_output_tokens_per_task": 71441.88143176734,
      "last_run_timestamp": "2026-10-17T19:52:46.515Z"
    },
    {
      "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": 0.7438097142857142,
      "accuracy_std_dev": 0.03525096933505834,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.015975259999999998,
      "avg_duration_per_task_ms": 149636.74857142856,
      "avg_output_tokens_per_task": 15244.568571428572,
      "last_run_timestamp": "2026-10-17T10:31:13.895Z"
    },
    {
      "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": 0.7437710606060606,
      "accuracy_std_dev": 0.022759018292755954,
      "run_count": 10,
      "total_questions": 495,
      "avg_cost_per_task_usd": 0.024460747474747482,
      "avg_duration_per_task_ms": 280677.9696969697,
      "avg_output_tokens_per_task": 18749.616161616163,
      "last_run_timestamp": "2026-10-17T11:00:48.266Z"
    },
    {
      "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": "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": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Phala",
      "accuracy": 0.7433333333333333,
      "accuracy_std_dev": 0.019907322310033455,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.02349562666666667,
      "avg_duration_per_task_ms": 145712.99333333335,
      "avg_output_tokens_per_task": 16978.89,
      "last_run_timestamp": "2026-10-17T10:59:45.574Z"
    },
    {
      "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": "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": 0.7433332500000001,
      "accuracy_std_dev": 0.01452975873638307,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.1182949875,
      "avg_duration_per_task_ms": 350271.9625,
      "avg_output_tokens_per_task": 9768.27,
      "last_run_timestamp": "2026-10-17T20:28:21.183Z"
    },
    {
      "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": 0.7433035357142856,
      "accuracy_std_dev": 0.02587302115436699,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.03511485267857143,
      "avg_duration_per_task_ms": 607727.7254464285,
      "avg_output_tokens_per_task": 15955.859375,
      "last_run_timestamp": "2026-10-17T19:37:14.691Z"
    },
    {
      "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": 0.7431110666666667,
      "accuracy_std_dev": 0.017866552793666977,
      "run_count": 15,
      "total_questions": 750,
      "avg_cost_per_task_usd": 0.3618751893333333,
      "avg_duration_per_task_ms": 88230.65866666667,
      "avg_output_tokens_per_task": 7156.333333333333,
      "last_run_timestamp": "2026-10-17T10:59:45.574Z"
    },
    {
      "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": 0.742962888888889,
      "accuracy_std_dev": 0.028993307965208676,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.3506727244444445,
      "avg_duration_per_task_ms": 221762.51111111112,
      "avg_output_tokens_per_task": 10894.351111111111,
      "last_run_timestamp": "2026-10-17T10:58:41.105Z"
    },
    {
      "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": 0.7428571904761905,
      "accuracy_std_dev": 0.020426289140687226,
      "run_count": 21,
      "total_questions": 1050,
      "avg_cost_per_task_usd": 0.4526462142857143,
      "avg_duration_per_task_ms": 607353.6895238095,
      "avg_output_tokens_per_task": 9902.753333333334,
      "last_run_timestamp": "2026-10-18T01:21:52.177Z"
    },
    {
      "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": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "GMICloud",
      "accuracy": 0.7426667,
      "accuracy_std_dev": 0.031860777699390844,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.107968072,
      "avg_duration_per_task_ms": 317982.848,
      "avg_output_tokens_per_task": 18593.702,
      "last_run_timestamp": "2026-10-17T11:15:57.694Z"
    },
    {
      "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": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "Z.AI",
      "accuracy": 0.7422222222222222,
      "accuracy_std_dev": 0.05057690391819785,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.09490620666666666,
      "avg_duration_per_task_ms": 230369.54888888888,
      "avg_output_tokens_per_task": 11349.884444444444,
      "last_run_timestamp": "2026-10-17T20:40:39.123Z"
    },
    {
      "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": 0.7422221666666666,
      "accuracy_std_dev": 0.03541660697289089,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.026421163333333334,
      "avg_duration_per_task_ms": 160993.86333333334,
      "avg_output_tokens_per_task": 16769.46,
      "last_run_timestamp": "2026-10-17T10:59:46.342Z"
    },
    {
      "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": 0.7422221666666666,
      "accuracy_std_dev": 0.016178166968857207,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.05587016000000001,
      "avg_duration_per_task_ms": 223603.79666666666,
      "avg_output_tokens_per_task": 9185.596666666666,
      "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": "CoreWeave",
      "accuracy": 0.742222,
      "accuracy_std_dev": 0.020608136952831695,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.044708606666666664,
      "avg_duration_per_task_ms": 149318.53333333333,
      "avg_output_tokens_per_task": 38342.77333333333,
      "last_run_timestamp": "2026-10-10T11:04:02.145Z"
    },
    {
      "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-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Baidu",
      "accuracy": 0.742083375,
      "accuracy_std_dev": 0.02948325972877448,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0861245475,
      "avg_duration_per_task_ms": 179001.195,
      "avg_output_tokens_per_task": 10461.4425,
      "last_run_timestamp": "2026-10-17T19:46:06.956Z"
    },
    {
      "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": 0.7418181818181818,
      "accuracy_std_dev": 0.108770519591242,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.1487734618181818,
      "avg_duration_per_task_ms": 616954.0145454545,
      "avg_output_tokens_per_task": 27127.465454545454,
      "last_run_timestamp": "2026-10-17T11:01:24.247Z"
    },
    {
      "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": 0.7418154508928572,
      "accuracy_std_dev": 0.028897426433319922,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.09664293303571428,
      "avg_duration_per_task_ms": 357228.20089285716,
      "avg_output_tokens_per_task": 11297.408482142857,
      "last_run_timestamp": "2026-10-17T19:51:54.944Z"
    },
    {
      "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": 0.7417582051282051,
      "accuracy_std_dev": 0.022765247487641204,
      "run_count": 11,
      "total_questions": 546,
      "avg_cost_per_task_usd": 0.1424760476190476,
      "avg_duration_per_task_ms": 619826.9175824176,
      "avg_output_tokens_per_task": 26765.48534798535,
      "last_run_timestamp": "2026-10-17T19:26:00.646Z"
    },
    {
      "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": 0.7416668333333333,
      "accuracy_std_dev": 0.0247767924573021,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.04222387666666667,
      "avg_duration_per_task_ms": 129284.56,
      "avg_output_tokens_per_task": 26313.053333333333,
      "last_run_timestamp": "2026-10-17T10:27:15.231Z"
    },
    {
      "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": 0.7416668333333333,
      "accuracy_std_dev": 0.03963955422687743,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.00980467,
      "avg_duration_per_task_ms": 415763.42833333334,
      "avg_output_tokens_per_task": 14671.45,
      "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": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Alibaba",
      "accuracy": 0.741666625,
      "accuracy_std_dev": 0.037675458046112394,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.3525293925,
      "avg_duration_per_task_ms": 391694.48,
      "avg_output_tokens_per_task": 9580.7525,
      "last_run_timestamp": "2026-10-17T11:00:48.240Z"
    },
    {
      "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": 0.7414815555555555,
      "accuracy_std_dev": 0.03675450160520359,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.04532596444444444,
      "avg_duration_per_task_ms": 571751.6577777778,
      "avg_output_tokens_per_task": 9173.444444444445,
      "last_run_timestamp": "2026-10-17T20:41:08.909Z"
    },
    {
      "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": 0.7414814444444444,
      "accuracy_std_dev": 0.03043362969228142,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.5050336355555556,
      "avg_duration_per_task_ms": 191771.47333333333,
      "avg_output_tokens_per_task": 9950.68888888889,
      "last_run_timestamp": "2026-10-17T11:12:58.005Z"
    },
    {
      "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": 0.7414813333333333,
      "accuracy_std_dev": 0.021029672681977536,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.06887691555555556,
      "avg_duration_per_task_ms": 310202.34444444446,
      "avg_output_tokens_per_task": 12962.453333333333,
      "last_run_timestamp": "2026-10-17T19:46:38.851Z"
    },
    {
      "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": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "DeepInfra",
      "accuracy": 0.7411113333333332,
      "accuracy_std_dev": 0.00415742679497253,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.08487537333333334,
      "avg_duration_per_task_ms": 315752.68666666665,
      "avg_output_tokens_per_task": 9327.993333333334,
      "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-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "DigitalOcean",
      "accuracy": 0.740963843373494,
      "accuracy_std_dev": 0.0367474030332757,
      "run_count": 10,
      "total_questions": 498,
      "avg_cost_per_task_usd": 0.079212578313253,
      "avg_duration_per_task_ms": 799563.951807229,
      "avg_output_tokens_per_task": 14536.901606425703,
      "last_run_timestamp": "2026-10-17T21:16:39.089Z"
    },
    {
      "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": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "Novita",
      "accuracy": 0.7405555,
      "accuracy_std_dev": 0.028474594482040785,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.021673546666666665,
      "avg_duration_per_task_ms": 395217.175,
      "avg_output_tokens_per_task": 12192.15,
      "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": "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": "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": 0.7403200757575757,
      "accuracy_std_dev": 0.01686858005610948,
      "run_count": 8,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.018282113636363638,
      "avg_duration_per_task_ms": 87170.2398989899,
      "avg_output_tokens_per_task": 4071.1969696969695,
      "last_run_timestamp": "2026-10-17T11:12:55.113Z"
    },
    {
      "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": 0.7400003333333333,
      "accuracy_std_dev": 0.018856180831641253,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.019302796666666667,
      "avg_duration_per_task_ms": 499555.5366666667,
      "avg_output_tokens_per_task": 14674.223333333333,
      "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": "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-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Venice",
      "accuracy": 0.7400001428571428,
      "accuracy_std_dev": 0.031471771176398565,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.07991938000000001,
      "avg_duration_per_task_ms": 532201.7285714286,
      "avg_output_tokens_per_task": 12014.771428571428,
      "last_run_timestamp": "2026-10-17T23:14:09.649Z"
    },
    {
      "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": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "Chutes",
      "accuracy": 0.74,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.049757800000000005,
      "avg_duration_per_task_ms": 211927.58,
      "avg_output_tokens_per_task": 8527.66,
      "last_run_timestamp": "2026-07-25T08:07:26.285Z"
    },
    {
      "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": 0.74,
      "accuracy_std_dev": 0.01333300000000004,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.15468345,
      "avg_duration_per_task_ms": 99212.54,
      "avg_output_tokens_per_task": 8743.23,
      "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-5.6-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": "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": 0.74,
      "accuracy_std_dev": 0.01671431284857383,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.2243457,
      "avg_duration_per_task_ms": 261652.61428571428,
      "avg_output_tokens_per_task": 11882.217142857144,
      "last_run_timestamp": "2026-10-17T11:03:43.306Z"
    },
    {
      "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": 0.74,
      "accuracy_std_dev": 0.03858593583807446,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.42733946500000003,
      "avg_duration_per_task_ms": 38766.985,
      "avg_output_tokens_per_task": 9216.565,
      "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": "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": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.03319066,
      "avg_duration_per_task_ms": 77162.14,
      "avg_output_tokens_per_task": 4453.24,
      "last_run_timestamp": "2026-10-04T03:00:16.404Z"
    },
    {
      "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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7394871538461538,
      "accuracy_std_dev": 0.027606692440688974,
      "run_count": 13,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.056419910769230766,
      "avg_duration_per_task_ms": 240446.07846153845,
      "avg_output_tokens_per_task": 9604.89076923077,
      "last_run_timestamp": "2026-10-17T19:41:59.742Z"
    },
    {
      "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": 0.7394766996587031,
      "accuracy_std_dev": 0.028748585800363012,
      "run_count": 12,
      "total_questions": 586,
      "avg_cost_per_task_usd": 0.06203791467576791,
      "avg_duration_per_task_ms": 129086.0494880546,
      "avg_output_tokens_per_task": 8186.447098976109,
      "last_run_timestamp": "2026-10-17T19:18:24.840Z"
    },
    {
      "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": 0.7393334,
      "accuracy_std_dev": 0.049121084860169785,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.009324213999999999,
      "avg_duration_per_task_ms": 313360.034,
      "avg_output_tokens_per_task": 15426.052,
      "last_run_timestamp": "2026-10-17T11:16:23.436Z"
    },
    {
      "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": 0.7392550315186246,
      "accuracy_std_dev": 0.025013223548469583,
      "run_count": 7,
      "total_questions": 349,
      "avg_cost_per_task_usd": 0.23634143553008594,
      "avg_duration_per_task_ms": 371772.38968481374,
      "avg_output_tokens_per_task": 10117.719197707736,
      "last_run_timestamp": "2026-10-17T11:09:42.449Z"
    },
    {
      "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": 0.7390475714285715,
      "accuracy_std_dev": 0.03175881615667484,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.057812691428571435,
      "avg_duration_per_task_ms": 661457.1457142857,
      "avg_output_tokens_per_task": 14926.3,
      "last_run_timestamp": "2026-10-10T19:35:42.927Z"
    },
    {
      "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": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Minimax",
      "accuracy": 0.7388889166666667,
      "accuracy_std_dev": 0.0286530189318634,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.023847688333333335,
      "avg_duration_per_task_ms": 88794.73166666667,
      "avg_output_tokens_per_task": 7901.106666666667,
      "last_run_timestamp": "2026-10-17T19:36:03.091Z"
    },
    {
      "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": 0.7387772808022923,
      "accuracy_std_dev": 0.02413990130003665,
      "run_count": 7,
      "total_questions": 349,
      "avg_cost_per_task_usd": 0.05359102578796561,
      "avg_duration_per_task_ms": 822214.4985673352,
      "avg_output_tokens_per_task": 14206.361031518625,
      "last_run_timestamp": "2026-10-17T19:36:08.332Z"
    },
    {
      "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": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Baidu",
      "accuracy": 0.738333375,
      "accuracy_std_dev": 0.015898903703852527,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.07332558500000001,
      "avg_duration_per_task_ms": 262220.325,
      "avg_output_tokens_per_task": 13184.325,
      "last_run_timestamp": "2026-10-17T19:21:39.847Z"
    },
    {
      "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": 0.7383231627502177,
      "accuracy_std_dev": 0.027430622924368746,
      "run_count": 23,
      "total_questions": 1149,
      "avg_cost_per_task_usd": 0.12736605570060922,
      "avg_duration_per_task_ms": 169094.96257615316,
      "avg_output_tokens_per_task": 9298.362924281984,
      "last_run_timestamp": "2026-10-17T19:52:56.176Z"
    },
    {
      "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": 0.7382865983935742,
      "accuracy_std_dev": 0.027461257286584682,
      "run_count": 5,
      "total_questions": 249,
      "avg_cost_per_task_usd": 0.07498479919678715,
      "avg_duration_per_task_ms": 91080.29718875502,
      "avg_output_tokens_per_task": 4367.955823293173,
      "last_run_timestamp": "2026-09-26T08:11:26.495Z"
    },
    {
      "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": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "Novita",
      "accuracy": 0.7378469999999999,
      "accuracy_std_dev": 0.0017360000000000153,
      "run_count": 2,
      "total_questions": 96,
      "avg_cost_per_task_usd": 0.21588738541666666,
      "avg_duration_per_task_ms": 378388.65625,
      "avg_output_tokens_per_task": 25153.166666666668,
      "last_run_timestamp": "2026-10-17T19:37:05.922Z"
    },
    {
      "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": 0.7375,
      "accuracy_std_dev": 0.04795095874432546,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.17019316,
      "avg_duration_per_task_ms": 283667.8625,
      "avg_output_tokens_per_task": 9793.12,
      "last_run_timestamp": "2026-10-17T11:16:12.519Z"
    },
    {
      "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": 0.737499875,
      "accuracy_std_dev": 0.018837726259009493,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0148281175,
      "avg_duration_per_task_ms": 76477.91,
      "avg_output_tokens_per_task": 4476.9775,
      "last_run_timestamp": "2026-10-17T10:59:31.515Z"
    },
    {
      "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": 0.7372220833333334,
      "accuracy_std_dev": 0.023484995867781692,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.007983868333333335,
      "avg_duration_per_task_ms": 398396.11666666664,
      "avg_output_tokens_per_task": 14403.198333333334,
      "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-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Phala",
      "accuracy": 0.7370185075376885,
      "accuracy_std_dev": 0.05274392638908769,
      "run_count": 8,
      "total_questions": 398,
      "avg_cost_per_task_usd": 0.2985063944723619,
      "avg_duration_per_task_ms": 383061.77889447234,
      "avg_output_tokens_per_task": 11160.087939698493,
      "last_run_timestamp": "2026-10-17T11:17:31.758Z"
    },
    {
      "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": 0.7369696818181818,
      "accuracy_std_dev": 0.024677843497183632,
      "run_count": 22,
      "total_questions": 1100,
      "avg_cost_per_task_usd": 0.36901798272727276,
      "avg_duration_per_task_ms": 230204.26454545456,
      "avg_output_tokens_per_task": 10920.546363636364,
      "last_run_timestamp": "2026-10-17T11:09:53.773Z"
    },
    {
      "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": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Chutes",
      "accuracy": 0.736666625,
      "accuracy_std_dev": 0.03464100412422215,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.09860756999999999,
      "avg_duration_per_task_ms": 640483.4725,
      "avg_output_tokens_per_task": 9709.82,
      "last_run_timestamp": "2026-10-17T21:09:49.384Z"
    },
    {
      "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": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Modal",
      "accuracy": 0.7362964444444444,
      "accuracy_std_dev": 0.017808576717170822,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.38587963333333336,
      "avg_duration_per_task_ms": 178574.04,
      "avg_output_tokens_per_task": 10909.295555555556,
      "last_run_timestamp": "2026-10-10T11:25:32.854Z"
    },
    {
      "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": 0.7362962222222222,
      "accuracy_std_dev": 0.05087965035645451,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.047357815555555544,
      "avg_duration_per_task_ms": 286314.82666666666,
      "avg_output_tokens_per_task": 14790.351111111111,
      "last_run_timestamp": "2026-10-17T11:21:16.811Z"
    },
    {
      "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": 0.7361904285714286,
      "accuracy_std_dev": 0.022213041869632365,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.35529426857142854,
      "avg_duration_per_task_ms": 193600.25714285715,
      "avg_output_tokens_per_task": 8821.457142857143,
      "last_run_timestamp": "2026-10-17T10:59:54.343Z"
    },
    {
      "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": 0.7361110833333333,
      "accuracy_std_dev": 0.04367039035673625,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.082194435,
      "avg_duration_per_task_ms": 179037.31666666668,
      "avg_output_tokens_per_task": 8256.305,
      "last_run_timestamp": "2026-10-17T19:21:01.874Z"
    },
    {
      "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": 0.7360831871227365,
      "accuracy_std_dev": 0.029580267931342347,
      "run_count": 10,
      "total_questions": 497,
      "avg_cost_per_task_usd": 0.07848097384305834,
      "avg_duration_per_task_ms": 216753.75050301812,
      "avg_output_tokens_per_task": 8419.167002012073,
      "last_run_timestamp": "2026-10-17T19:14:56.925Z"
    },
    {
      "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": "Parasail",
      "accuracy": 0.7357142857142858,
      "accuracy_std_dev": 0.03689182323153652,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.07578944857142858,
      "avg_duration_per_task_ms": 210085.71714285715,
      "avg_output_tokens_per_task": 11277.097142857143,
      "last_run_timestamp": "2026-10-10T18:54:49.288Z"
    },
    {
      "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": 0.7355556666666666,
      "accuracy_std_dev": 0.036649745254533794,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.022963626666666664,
      "avg_duration_per_task_ms": 113662.22333333333,
      "avg_output_tokens_per_task": 7480.291666666667,
      "last_run_timestamp": "2026-10-17T19:17:31.925Z"
    },
    {
      "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": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "Alibaba",
      "accuracy": 0.7351189375,
      "accuracy_std_dev": 0.09797950654998103,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.07525125669642858,
      "avg_duration_per_task_ms": 194209.64732142858,
      "avg_output_tokens_per_task": 24190.323660714286,
      "last_run_timestamp": "2026-10-17T19:25:40.380Z"
    },
    {
      "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": 0.7350427912087912,
      "accuracy_std_dev": 0.06405752324575284,
      "run_count": 11,
      "total_questions": 546,
      "avg_cost_per_task_usd": 0.1534336978021978,
      "avg_duration_per_task_ms": 1367021.8919413919,
      "avg_output_tokens_per_task": 27730.919413919415,
      "last_run_timestamp": "2026-10-17T19:42:10.212Z"
    },
    {
      "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": 0.735000125,
      "accuracy_std_dev": 0.028819160776458658,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0161670325,
      "avg_duration_per_task_ms": 638661.92,
      "avg_output_tokens_per_task": 18640.1425,
      "last_run_timestamp": "2026-10-18T01:24:51.034Z"
    },
    {
      "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": 0.735000125,
      "accuracy_std_dev": 0.026405000054144565,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.29933707499999995,
      "avg_duration_per_task_ms": 500487.98,
      "avg_output_tokens_per_task": 9623.9275,
      "last_run_timestamp": "2026-10-17T21:38:40.884Z"
    },
    {
      "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": 0.735,
      "accuracy_std_dev": 0.011666999999999983,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.024598469999999997,
      "avg_duration_per_task_ms": 93010.71,
      "avg_output_tokens_per_task": 31378.94,
      "last_run_timestamp": "2026-10-17T19:09:13.040Z"
    },
    {
      "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": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Moonshot AI",
      "accuracy": 0.7347416579476861,
      "accuracy_std_dev": 0.025978725188315156,
      "run_count": 10,
      "total_questions": 497,
      "avg_cost_per_task_usd": 0.35989230985915494,
      "avg_duration_per_task_ms": 406164.88732394367,
      "avg_output_tokens_per_task": 9488.581488933602,
      "last_run_timestamp": "2026-10-17T10:59:45.574Z"
    },
    {
      "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": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "Z.AI",
      "accuracy": 0.7344445,
      "accuracy_std_dev": 0.028393141670422214,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.04939556333333334,
      "avg_duration_per_task_ms": 469523.68666666665,
      "avg_output_tokens_per_task": 9528.593333333334,
      "last_run_timestamp": "2026-10-17T20:03:15.965Z"
    },
    {
      "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": 0.7344348103448277,
      "accuracy_std_dev": 0.030951629540778514,
      "run_count": 14,
      "total_questions": 696,
      "avg_cost_per_task_usd": 0.1064283620689655,
      "avg_duration_per_task_ms": 100216.88362068965,
      "avg_output_tokens_per_task": 7768.466954022989,
      "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": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "DeepInfra",
      "accuracy": 0.7342658006993007,
      "accuracy_std_dev": 0.01683879230048547,
      "run_count": 6,
      "total_questions": 286,
      "avg_cost_per_task_usd": 0.14595530419580421,
      "avg_duration_per_task_ms": 709836.5559440559,
      "avg_output_tokens_per_task": 29021.49300699301,
      "last_run_timestamp": "2026-10-17T19:00:48.720Z"
    },
    {
      "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": 0.73416675,
      "accuracy_std_dev": 0.027975611900144373,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.05059770500000001,
      "avg_duration_per_task_ms": 367749.4825,
      "avg_output_tokens_per_task": 10213.9225,
      "last_run_timestamp": "2026-10-17T20:14:14.955Z"
    },
    {
      "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": 0.734166625,
      "accuracy_std_dev": 0.039077836323603884,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.12202924000000001,
      "avg_duration_per_task_ms": 675600.895,
      "avg_output_tokens_per_task": 33746.605,
      "last_run_timestamp": "2026-10-17T10:59:46.342Z"
    },
    {
      "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": 0.7341665,
      "accuracy_std_dev": 0.021650731321366502,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.065515015,
      "avg_duration_per_task_ms": 151485.83,
      "avg_output_tokens_per_task": 7332.795,
      "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": "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": 0.73413336,
      "accuracy_std_dev": 0.03348342146421718,
      "run_count": 25,
      "total_questions": 1250,
      "avg_cost_per_task_usd": 0.3556991112,
      "avg_duration_per_task_ms": 207103.4464,
      "avg_output_tokens_per_task": 10323.1728,
      "last_run_timestamp": "2026-10-17T11:02:34.524Z"
    },
    {
      "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": "auto-routing",
      "accuracy": 0.7338888333333333,
      "accuracy_std_dev": 0.02530674652746884,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.07665837666666667,
      "avg_duration_per_task_ms": 177081.09,
      "avg_output_tokens_per_task": 8668.4,
      "last_run_timestamp": "2026-10-17T19:13:29.274Z"
    },
    {
      "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": 0.7338531336302895,
      "accuracy_std_dev": 0.03071493333547167,
      "run_count": 9,
      "total_questions": 449,
      "avg_cost_per_task_usd": 0.06568974832962138,
      "avg_duration_per_task_ms": 161273.1714922049,
      "avg_output_tokens_per_task": 11527.013363028953,
      "last_run_timestamp": "2026-10-17T19:36:06.338Z"
    },
    {
      "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": "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": "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": "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": "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": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7331495477941177,
      "accuracy_std_dev": 0.04512078785249381,
      "run_count": 11,
      "total_questions": 544,
      "avg_cost_per_task_usd": 0.02227737316176471,
      "avg_duration_per_task_ms": 99731.28860294117,
      "avg_output_tokens_per_task": 7413.3125,
      "last_run_timestamp": "2026-10-17T19:12:23.569Z"
    },
    {
      "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": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Relace",
      "accuracy": 0.732857,
      "accuracy_std_dev": 0.02930024262210617,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.015632725714285714,
      "avg_duration_per_task_ms": 286314.63714285713,
      "avg_output_tokens_per_task": 22301.08,
      "last_run_timestamp": "2026-10-17T11:22:26.147Z"
    },
    {
      "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": "Venice",
      "accuracy": 0.7322221666666667,
      "accuracy_std_dev": 0.02705733510169758,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.04443540666666667,
      "avg_duration_per_task_ms": 142368.75,
      "avg_output_tokens_per_task": 22198.636666666665,
      "last_run_timestamp": "2026-10-17T11:01:48.208Z"
    },
    {
      "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": 0.7322221666666667,
      "accuracy_std_dev": 0.027866575847639113,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.03493169,
      "avg_duration_per_task_ms": 340890.4033333333,
      "avg_output_tokens_per_task": 8339.266666666666,
      "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": "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": "Together",
      "accuracy": 0.731515090909091,
      "accuracy_std_dev": 0.03470991263301791,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.3032066236363637,
      "avg_duration_per_task_ms": 332655.4072727273,
      "avg_output_tokens_per_task": 9387.027272727273,
      "last_run_timestamp": "2026-10-17T10:59:57.613Z"
    },
    {
      "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": 0.7313334,
      "accuracy_std_dev": 0.012666852633547132,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.243754672,
      "avg_duration_per_task_ms": 268035.924,
      "avg_output_tokens_per_task": 10219.814,
      "last_run_timestamp": "2026-10-17T11:01:22.455Z"
    },
    {
      "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": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "Novita",
      "accuracy": 0.7311112222222222,
      "accuracy_std_dev": 0.022443432322063868,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.05785997111111111,
      "avg_duration_per_task_ms": 325880.12666666665,
      "avg_output_tokens_per_task": 9968.124444444444,
      "last_run_timestamp": "2026-10-17T20:02:15.278Z"
    },
    {
      "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": 0.7311110000000001,
      "accuracy_std_dev": 0.026010406276463026,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.03343346333333334,
      "avg_duration_per_task_ms": 98916.58333333333,
      "avg_output_tokens_per_task": 23693.51,
      "last_run_timestamp": "2026-10-17T11:03:07.971Z"
    },
    {
      "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": 0.7311110000000001,
      "accuracy_std_dev": 0.02096437307274735,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.04746591,
      "avg_duration_per_task_ms": 142455.40666666668,
      "avg_output_tokens_per_task": 28668.966666666667,
      "last_run_timestamp": "2026-10-17T10:58:45.204Z"
    },
    {
      "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": "Parasail",
      "accuracy": 0.7305553333333333,
      "accuracy_std_dev": 0.020942233896018736,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.04416758666666667,
      "avg_duration_per_task_ms": 171467.14,
      "avg_output_tokens_per_task": 30432.296666666665,
      "last_run_timestamp": "2026-10-17T11:17:21.507Z"
    },
    {
      "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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7299999166666666,
      "accuracy_std_dev": 0.034047910145015584,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.09264119166666665,
      "avg_duration_per_task_ms": 141837.91666666666,
      "avg_output_tokens_per_task": 9263.338333333333,
      "last_run_timestamp": "2026-10-17T19:23:11.027Z"
    },
    {
      "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": 0.7296969090909091,
      "accuracy_std_dev": 0.03415505130871314,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.33621692363636363,
      "avg_duration_per_task_ms": 207110.29636363636,
      "avg_output_tokens_per_task": 10400.149090909092,
      "last_run_timestamp": "2026-10-17T10:58:55.551Z"
    },
    {
      "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": 0.729166625,
      "accuracy_std_dev": 0.026020732920776363,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.12688592750000002,
      "avg_duration_per_task_ms": 249207.73,
      "avg_output_tokens_per_task": 10044.14,
      "last_run_timestamp": "2026-10-17T21:29:38.906Z"
    },
    {
      "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": 0.728889,
      "accuracy_std_dev": 0.028631520229751458,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.05385624666666666,
      "avg_duration_per_task_ms": 289115.72444444447,
      "avg_output_tokens_per_task": 11182.26,
      "last_run_timestamp": "2026-10-17T19:41:41.353Z"
    },
    {
      "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": 0.7284644516853932,
      "accuracy_std_dev": 0.03760340361105365,
      "run_count": 9,
      "total_questions": 445,
      "avg_cost_per_task_usd": 0.09336272584269663,
      "avg_duration_per_task_ms": 298064.6898876405,
      "avg_output_tokens_per_task": 14598.31011235955,
      "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": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "BaseTen",
      "accuracy": 0.7284264213197968,
      "accuracy_std_dev": 0.023635771369896082,
      "run_count": 4,
      "total_questions": 197,
      "avg_cost_per_task_usd": 0.051129340101522834,
      "avg_duration_per_task_ms": 77511.28934010152,
      "avg_output_tokens_per_task": 19306.862944162436,
      "last_run_timestamp": "2026-10-17T10:21:09.018Z"
    },
    {
      "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": 0.7283333333333334,
      "accuracy_std_dev": 0.027369352961945523,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.08085183000000001,
      "avg_duration_per_task_ms": 227378.24666666667,
      "avg_output_tokens_per_task": 8744.926666666666,
      "last_run_timestamp": "2026-10-17T19:59:56.527Z"
    },
    {
      "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": 0.728333125,
      "accuracy_std_dev": 0.035862401001736835,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0596212475,
      "avg_duration_per_task_ms": 368463.805,
      "avg_output_tokens_per_task": 11783.2,
      "last_run_timestamp": "2026-10-17T20:00:29.267Z"
    },
    {
      "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": 0.7279999,
      "accuracy_std_dev": 0.029784221028759515,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.29079123999999995,
      "avg_duration_per_task_ms": 357975.664,
      "avg_output_tokens_per_task": 9410.784,
      "last_run_timestamp": "2026-10-17T10:36:28.747Z"
    },
    {
      "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": 0.7277775,
      "accuracy_std_dev": 0.028652944605805083,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.04601716333333333,
      "avg_duration_per_task_ms": 165703.69333333333,
      "avg_output_tokens_per_task": 8247.983333333334,
      "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": "Ambient",
      "accuracy": 0.7276191428571428,
      "accuracy_std_dev": 0.052906351183492316,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.07805452,
      "avg_duration_per_task_ms": 228092.80571428573,
      "avg_output_tokens_per_task": 8318.434285714286,
      "last_run_timestamp": "2026-09-19T08:12:24.315Z"
    },
    {
      "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": 0.727316106017192,
      "accuracy_std_dev": 0.028771894791938337,
      "run_count": 7,
      "total_questions": 349,
      "avg_cost_per_task_usd": 0.1317654670487106,
      "avg_duration_per_task_ms": 71027.62464183381,
      "avg_output_tokens_per_task": 10119.80229226361,
      "last_run_timestamp": "2026-10-10T19:19:07.184Z"
    },
    {
      "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": 0.72722225,
      "accuracy_std_dev": 0.050365918751547674,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.11962306833333335,
      "avg_duration_per_task_ms": 145630.83333333334,
      "avg_output_tokens_per_task": 9570.12,
      "last_run_timestamp": "2026-10-17T19:44:43.225Z"
    },
    {
      "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": 0.7272221666666667,
      "accuracy_std_dev": 0.024863716190576896,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.08010389666666666,
      "avg_duration_per_task_ms": 652312.8383333334,
      "avg_output_tokens_per_task": 8922.71,
      "last_run_timestamp": "2026-10-17T19:35:50.224Z"
    },
    {
      "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": 0.7272220833333334,
      "accuracy_std_dev": 0.025012427159508587,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.07207888500000001,
      "avg_duration_per_task_ms": 216694.285,
      "avg_output_tokens_per_task": 8508.088333333333,
      "last_run_timestamp": "2026-10-17T20:16:13.711Z"
    },
    {
      "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": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "StreamLake",
      "accuracy": 0.7266667142857143,
      "accuracy_std_dev": 0.023637648207372713,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.06641256285714285,
      "avg_duration_per_task_ms": 202688.7742857143,
      "avg_output_tokens_per_task": 9459.962857142857,
      "last_run_timestamp": "2026-10-17T19:21:20.002Z"
    },
    {
      "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": 0.7266665,
      "accuracy_std_dev": 0.00666650000000002,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.11245736999999999,
      "avg_duration_per_task_ms": 138229.98,
      "avg_output_tokens_per_task": 17958.65,
      "last_run_timestamp": "2026-10-17T10:25:34.919Z"
    },
    {
      "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": "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": "anthropic/claude-4.5-sonnet-20250929",
      "model_slug": "anthropic/claude-sonnet-4.5",
      "model_name": "Anthropic: Claude Sonnet 4.5",
      "provider_name": "Azure",
      "accuracy": 0.7260605454545455,
      "accuracy_std_dev": 0.03014507499459492,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.35443870000000005,
      "avg_duration_per_task_ms": 238811.5509090909,
      "avg_output_tokens_per_task": 10880.86,
      "last_run_timestamp": "2026-10-17T10:59:31.851Z"
    },
    {
      "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": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "Friendli",
      "accuracy": 0.7259258888888889,
      "accuracy_std_dev": 0.01817470468318503,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.10746231555555555,
      "avg_duration_per_task_ms": 332961.9111111111,
      "avg_output_tokens_per_task": 14617.324444444444,
      "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": "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": 0.7255891616161616,
      "accuracy_std_dev": 0.0865306861288959,
      "run_count": 8,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.009908151515151516,
      "avg_duration_per_task_ms": 192785.0606060606,
      "avg_output_tokens_per_task": 4205.436868686868,
      "last_run_timestamp": "2026-10-17T11:01:14.130Z"
    },
    {
      "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": 0.7250608065693431,
      "accuracy_std_dev": 0.027313728982127115,
      "run_count": 11,
      "total_questions": 548,
      "avg_cost_per_task_usd": 0.045819195255474454,
      "avg_duration_per_task_ms": 200398.98540145985,
      "avg_output_tokens_per_task": 9616.815693430657,
      "last_run_timestamp": "2026-10-17T19:49:20.976Z"
    },
    {
      "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": 0.725,
      "accuracy_std_dev": 0.025331238950157955,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.28661039000000005,
      "avg_duration_per_task_ms": 193369.4875,
      "avg_output_tokens_per_task": 11544.6525,
      "last_run_timestamp": "2026-10-17T11:22:12.479Z"
    },
    {
      "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": 0.7248484545454545,
      "accuracy_std_dev": 0.026831392056739796,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.028918963636363638,
      "avg_duration_per_task_ms": 207327.25454545455,
      "avg_output_tokens_per_task": 9728.478181818182,
      "last_run_timestamp": "2026-10-17T20:29:55.837Z"
    },
    {
      "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": "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": 0.7239058686868687,
      "accuracy_std_dev": 0.07830589921636386,
      "run_count": 12,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.044882882154882155,
      "avg_duration_per_task_ms": 272662.92087542085,
      "avg_output_tokens_per_task": 25044.632996632998,
      "last_run_timestamp": "2026-10-17T19:39:03.025Z"
    },
    {
      "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": 0.723680833976834,
      "accuracy_std_dev": 0.03403551347181265,
      "run_count": 26,
      "total_questions": 1295,
      "avg_cost_per_task_usd": 0.11716301235521236,
      "avg_duration_per_task_ms": 133519.80694980695,
      "avg_output_tokens_per_task": 8532.8,
      "last_run_timestamp": "2026-10-17T19:53:08.746Z"
    },
    {
      "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": "moonshotai/kimi-k2-thinking-20251106",
      "model_slug": "moonshotai/kimi-k2-thinking",
      "model_name": "MoonshotAI: Kimi K2 Thinking",
      "provider_name": "Novita",
      "accuracy": 0.7233335,
      "accuracy_std_dev": 0.023333500000000007,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.05323206,
      "avg_duration_per_task_ms": 276804.85,
      "avg_output_tokens_per_task": 8575.93,
      "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": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Novita",
      "accuracy": 0.72333325,
      "accuracy_std_dev": 0.01953163151627381,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.08062917333333333,
      "avg_duration_per_task_ms": 245562.14,
      "avg_output_tokens_per_task": 9203.731666666667,
      "last_run_timestamp": "2026-10-17T19:20:50.237Z"
    },
    {
      "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": 0.723137294117647,
      "accuracy_std_dev": 0.03345988135375875,
      "run_count": 17,
      "total_questions": 850,
      "avg_cost_per_task_usd": 0.07395268117647061,
      "avg_duration_per_task_ms": 120093.27058823529,
      "avg_output_tokens_per_task": 8763.595294117647,
      "last_run_timestamp": "2026-10-17T19:46:31.426Z"
    },
    {
      "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": 0.7229628888888889,
      "accuracy_std_dev": 0.015748534890899425,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.024130835555555553,
      "avg_duration_per_task_ms": 202737.1288888889,
      "avg_output_tokens_per_task": 9830.88,
      "last_run_timestamp": "2026-10-17T19:19:55.377Z"
    },
    {
      "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": 0.7224448316633266,
      "accuracy_std_dev": 0.04149011767939444,
      "run_count": 10,
      "total_questions": 499,
      "avg_cost_per_task_usd": 0.02709162324649299,
      "avg_duration_per_task_ms": 310028.67134268535,
      "avg_output_tokens_per_task": 6206.739478957916,
      "last_run_timestamp": "2026-10-17T11:01:17.062Z"
    },
    {
      "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": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "Novita",
      "accuracy": 0.7222221666666667,
      "accuracy_std_dev": 0.029228799430337363,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.05327505,
      "avg_duration_per_task_ms": 275570.87333333335,
      "avg_output_tokens_per_task": 8563.536666666667,
      "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": "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": 0.7218181818181818,
      "accuracy_std_dev": 0.026981663672064338,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.042815734545454544,
      "avg_duration_per_task_ms": 128618.97818181818,
      "avg_output_tokens_per_task": 8570.907272727272,
      "last_run_timestamp": "2026-10-17T19:19:33.079Z"
    },
    {
      "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": 0.7214815555555555,
      "accuracy_std_dev": 0.017149338413615003,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.05164317555555556,
      "avg_duration_per_task_ms": 391354.9733333333,
      "avg_output_tokens_per_task": 9429.368888888888,
      "last_run_timestamp": "2026-10-17T20:23:07.690Z"
    },
    {
      "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": 0.7213049198113207,
      "accuracy_std_dev": 0.02807943476656338,
      "run_count": 17,
      "total_questions": 848,
      "avg_cost_per_task_usd": 0.10996251533018866,
      "avg_duration_per_task_ms": 106559.05070754717,
      "avg_output_tokens_per_task": 9515.395047169812,
      "last_run_timestamp": "2026-10-17T19:37:10.352Z"
    },
    {
      "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": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "Novita",
      "accuracy": 0.7207207364864865,
      "accuracy_std_dev": 0.029283397598582478,
      "run_count": 6,
      "total_questions": 296,
      "avg_cost_per_task_usd": 0.0602562027027027,
      "avg_duration_per_task_ms": 500890.8614864865,
      "avg_output_tokens_per_task": 14217.908783783783,
      "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": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7205554166666667,
      "accuracy_std_dev": 0.06303779020484238,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.28826981500000004,
      "avg_duration_per_task_ms": 424234.8416666667,
      "avg_output_tokens_per_task": 10037.273333333333,
      "last_run_timestamp": "2026-10-17T11:17:27.404Z"
    },
    {
      "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": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "Minimax",
      "accuracy": 0.7203921764705883,
      "accuracy_std_dev": 0.017030604563326206,
      "run_count": 17,
      "total_questions": 850,
      "avg_cost_per_task_usd": 0.038154339999999995,
      "avg_duration_per_task_ms": 182427.35294117648,
      "avg_output_tokens_per_task": 9882.828235294117,
      "last_run_timestamp": "2026-10-17T19:25:39.114Z"
    },
    {
      "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": "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": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "Phala",
      "accuracy": 0.72,
      "accuracy_std_dev": 0.020000000000000018,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.06889947,
      "avg_duration_per_task_ms": 485555.13,
      "avg_output_tokens_per_task": 10139.72,
      "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-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-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "auto-routing",
      "accuracy": 0.7199998888888889,
      "accuracy_std_dev": 0.02413946106111294,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.06200596222222221,
      "avg_duration_per_task_ms": 305935.6088888889,
      "avg_output_tokens_per_task": 9573.94,
      "last_run_timestamp": "2026-10-17T19:52:16.244Z"
    },
    {
      "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": 0.71999975,
      "accuracy_std_dev": 0.03559033108426363,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.2422260125,
      "avg_duration_per_task_ms": 293804.47,
      "avg_output_tokens_per_task": 21452.32,
      "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": "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": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Fireworks",
      "accuracy": 0.7198066588628762,
      "accuracy_std_dev": 0.028976766258600285,
      "run_count": 18,
      "total_questions": 897,
      "avg_cost_per_task_usd": 0.4397325016722407,
      "avg_duration_per_task_ms": 240718.0635451505,
      "avg_output_tokens_per_task": 10095.885172798216,
      "last_run_timestamp": "2026-10-17T11:17:26.520Z"
    },
    {
      "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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "CoreWeave",
      "accuracy": 0.7192592222222223,
      "accuracy_std_dev": 0.03561726435242629,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.06394025333333334,
      "avg_duration_per_task_ms": 73154.98,
      "avg_output_tokens_per_task": 9365.82888888889,
      "last_run_timestamp": "2026-10-17T19:08:58.242Z"
    },
    {
      "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": 0.7188886666666667,
      "accuracy_std_dev": 0.010303924699949169,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.4212243666666666,
      "avg_duration_per_task_ms": 138973.84666666668,
      "avg_output_tokens_per_task": 11088.173333333334,
      "last_run_timestamp": "2026-10-17T19:18:39.432Z"
    },
    {
      "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": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "Venice",
      "accuracy": 0.7185185555555555,
      "accuracy_std_dev": 0.02347109362651624,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.07224364888888889,
      "avg_duration_per_task_ms": 300057.04,
      "avg_output_tokens_per_task": 10826.902222222223,
      "last_run_timestamp": "2026-10-17T22:09:39.682Z"
    },
    {
      "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": 0.7175757272727272,
      "accuracy_std_dev": 0.03858842754244266,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.10341143999999998,
      "avg_duration_per_task_ms": 109231.03818181818,
      "avg_output_tokens_per_task": 9277.165454545455,
      "last_run_timestamp": "2026-10-17T19:18:06.689Z"
    },
    {
      "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": "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": "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": 0.7169697272727272,
      "accuracy_std_dev": 0.03367855284920254,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.12441471090909093,
      "avg_duration_per_task_ms": 167003.52181818182,
      "avg_output_tokens_per_task": 11951.263636363636,
      "last_run_timestamp": "2026-10-17T19:47:56.038Z"
    },
    {
      "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": 0.7167631098265895,
      "accuracy_std_dev": 0.10367778899207734,
      "run_count": 7,
      "total_questions": 346,
      "avg_cost_per_task_usd": 0.07269199132947977,
      "avg_duration_per_task_ms": 149631.9450867052,
      "avg_output_tokens_per_task": 14661.630057803468,
      "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": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Relace",
      "accuracy": 0.716667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.015077799999999999,
      "avg_duration_per_task_ms": 476878.72,
      "avg_output_tokens_per_task": 26097.54,
      "last_run_timestamp": "2026-09-28T09:00:02.615Z"
    },
    {
      "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": 0.7166665833333333,
      "accuracy_std_dev": 0.0346943826179838,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.038253503333333334,
      "avg_duration_per_task_ms": 190560.97666666665,
      "avg_output_tokens_per_task": 9125.601666666667,
      "last_run_timestamp": "2026-10-17T20:01:54.778Z"
    },
    {
      "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": "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": "DeepInfra",
      "accuracy": 0.7161111666666666,
      "accuracy_std_dev": 0.030847173070999906,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.04424468333333333,
      "avg_duration_per_task_ms": 144050.02166666667,
      "avg_output_tokens_per_task": 8482.466666666667,
      "last_run_timestamp": "2026-10-17T19:16:32.484Z"
    },
    {
      "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": "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": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "CoreWeave",
      "accuracy": 0.7153846153846154,
      "accuracy_std_dev": 0.03837602219643538,
      "run_count": 13,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.08081019846153846,
      "avg_duration_per_task_ms": 94500.82769230769,
      "avg_output_tokens_per_task": 8578.88,
      "last_run_timestamp": "2026-10-17T19:41:19.491Z"
    },
    {
      "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": "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-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "StreamLake",
      "accuracy": 0.7144445,
      "accuracy_std_dev": 0.03515405426002334,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.13767769333333332,
      "avg_duration_per_task_ms": 330585.11666666664,
      "avg_output_tokens_per_task": 10602.806666666667,
      "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": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "Mara",
      "accuracy": 0.7139639594594595,
      "accuracy_std_dev": 0.015199555307822523,
      "run_count": 3,
      "total_questions": 148,
      "avg_cost_per_task_usd": 0.04553853378378378,
      "avg_duration_per_task_ms": 61304.33783783784,
      "avg_output_tokens_per_task": 8376.81081081081,
      "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": "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": "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": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "Phala",
      "accuracy": 0.7128572857142856,
      "accuracy_std_dev": 0.02897348072050064,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.20664069142857142,
      "avg_duration_per_task_ms": 1520570.0285714285,
      "avg_output_tokens_per_task": 13765.78,
      "last_run_timestamp": "2026-10-17T19:37:45.657Z"
    },
    {
      "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": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "OpenInference",
      "accuracy": 0.7121210774410774,
      "accuracy_std_dev": 0.067600587772757,
      "run_count": 6,
      "total_questions": 297,
      "avg_cost_per_task_usd": 0.05431976767676767,
      "avg_duration_per_task_ms": 1240970.0303030303,
      "avg_output_tokens_per_task": 60052.92929292929,
      "last_run_timestamp": "2026-10-17T11:01:17.149Z"
    },
    {
      "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": "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": "auto-routing",
      "accuracy": 0.711111,
      "accuracy_std_dev": 0.014401748867566209,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.02642942666666667,
      "avg_duration_per_task_ms": 207531.11555555556,
      "avg_output_tokens_per_task": 10062.235555555555,
      "last_run_timestamp": "2026-10-17T19:00:35.778Z"
    },
    {
      "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": "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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Parasail",
      "accuracy": 0.709697,
      "accuracy_std_dev": 0.029489058220419185,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.11235794000000002,
      "avg_duration_per_task_ms": 96345.32545454545,
      "avg_output_tokens_per_task": 9865.885454545454,
      "last_run_timestamp": "2026-10-17T19:23:11.802Z"
    },
    {
      "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": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "SambaNova",
      "accuracy": 0.709172389261745,
      "accuracy_std_dev": 0.06451706767635644,
      "run_count": 3,
      "total_questions": 149,
      "avg_cost_per_task_usd": 0.07478329530201343,
      "avg_duration_per_task_ms": 78607.22147651007,
      "avg_output_tokens_per_task": 13899.020134228187,
      "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": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "AtlasCloud",
      "accuracy": 0.709166875,
      "accuracy_std_dev": 0.054968456232728356,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.030966977499999996,
      "avg_duration_per_task_ms": 308720.575,
      "avg_output_tokens_per_task": 10036.525,
      "last_run_timestamp": "2026-10-17T19:40:45.559Z"
    },
    {
      "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": 0.7090909090909091,
      "accuracy_std_dev": 0.1551348146791596,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.058754954545454555,
      "avg_duration_per_task_ms": 149210.46545454545,
      "avg_output_tokens_per_task": 17057.438181818183,
      "last_run_timestamp": "2026-10-17T19:18:18.672Z"
    },
    {
      "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": 0.7088888888888889,
      "accuracy_std_dev": 0.057777619658757995,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.10292275111111111,
      "avg_duration_per_task_ms": 232970.33777777778,
      "avg_output_tokens_per_task": 10921.413333333334,
      "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/qwen3.6-27b-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-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "BaseTen",
      "accuracy": 0.7075098814229249,
      "accuracy_std_dev": 0.03261134857329563,
      "run_count": 31,
      "total_questions": 1518,
      "avg_cost_per_task_usd": 0.09765357707509884,
      "avg_duration_per_task_ms": 73613.54281949934,
      "avg_output_tokens_per_task": 7991.078392621871,
      "last_run_timestamp": "2026-10-17T19:21:40.007Z"
    },
    {
      "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": "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": "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": "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-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "DeepInfra",
      "accuracy": 0.7066664,
      "accuracy_std_dev": 0.01801227976242876,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.050093852,
      "avg_duration_per_task_ms": 393986.292,
      "avg_output_tokens_per_task": 8914.744,
      "last_run_timestamp": "2026-10-03T16:12:07.463Z"
    },
    {
      "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": 0.7061769766277128,
      "accuracy_std_dev": 0.033226512089578226,
      "run_count": 12,
      "total_questions": 599,
      "avg_cost_per_task_usd": 0.07507928213689483,
      "avg_duration_per_task_ms": 218057.23372287146,
      "avg_output_tokens_per_task": 8906.173622704508,
      "last_run_timestamp": "2026-10-17T19:12:23.653Z"
    },
    {
      "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": 0.7060416875,
      "accuracy_std_dev": 0.03260073424540993,
      "run_count": 16,
      "total_questions": 800,
      "avg_cost_per_task_usd": 0.12466288875,
      "avg_duration_per_task_ms": 129451.30375,
      "avg_output_tokens_per_task": 8877.84875,
      "last_run_timestamp": "2026-10-17T19:20:16.670Z"
    },
    {
      "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": 0.7059999333333333,
      "accuracy_std_dev": 0.024104784834735375,
      "run_count": 15,
      "total_questions": 750,
      "avg_cost_per_task_usd": 0.13286914133333336,
      "avg_duration_per_task_ms": 86225.15866666667,
      "avg_output_tokens_per_task": 10631.385333333334,
      "last_run_timestamp": "2026-10-17T19:38:35.557Z"
    },
    {
      "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": "deepseek/deepseek-v3.1-terminus",
      "model_slug": "deepseek/deepseek-v3.1-terminus",
      "model_name": "DeepSeek: DeepSeek V3.1 Terminus",
      "provider_name": "auto-routing",
      "accuracy": 0.7039998000000001,
      "accuracy_std_dev": 0.028782815862246673,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.059492744,
      "avg_duration_per_task_ms": 531413.344,
      "avg_output_tokens_per_task": 14170.264,
      "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": "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": 0.7037037777777777,
      "accuracy_std_dev": 0.0414682095178618,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.06722512,
      "avg_duration_per_task_ms": 338497.55333333334,
      "avg_output_tokens_per_task": 9866.23111111111,
      "last_run_timestamp": "2026-10-17T20:25:29.287Z"
    },
    {
      "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": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Nebius",
      "accuracy": 0.7028112570281124,
      "accuracy_std_dev": 0.05430403649085398,
      "run_count": 5,
      "total_questions": 249,
      "avg_cost_per_task_usd": 0.21449976305220883,
      "avg_duration_per_task_ms": 647680.6907630522,
      "avg_output_tokens_per_task": 10064.048192771084,
      "last_run_timestamp": "2026-10-17T20:39:24.044Z"
    },
    {
      "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": "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": "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": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "auto-routing",
      "accuracy": 0.700000125,
      "accuracy_std_dev": 0.028674427245533173,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.047772974999999995,
      "avg_duration_per_task_ms": 357321.1875,
      "avg_output_tokens_per_task": 9223.68,
      "last_run_timestamp": "2026-10-17T20:08:34.955Z"
    },
    {
      "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": "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": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "Venice",
      "accuracy": 0.6986607901785714,
      "accuracy_std_dev": 0.03366609083665969,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.045400854910714286,
      "avg_duration_per_task_ms": 512677.609375,
      "avg_output_tokens_per_task": 9135.091517857143,
      "last_run_timestamp": "2026-10-17T19:41:52.559Z"
    },
    {
      "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": 0.0438401425,
      "avg_duration_per_task_ms": 300334.5325,
      "avg_output_tokens_per_task": 10585.9875,
      "last_run_timestamp": "2026-10-17T21:09:43.089Z"
    },
    {
      "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": 0.6975756363636364,
      "accuracy_std_dev": 0.042571053337114,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.09368267636363638,
      "avg_duration_per_task_ms": 209598.28545454546,
      "avg_output_tokens_per_task": 8033.250909090909,
      "last_run_timestamp": "2026-10-17T19:56:11.216Z"
    },
    {
      "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": "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": 0.6963602643678161,
      "accuracy_std_dev": 0.1081019669259783,
      "run_count": 7,
      "total_questions": 348,
      "avg_cost_per_task_usd": 0.07385931609195404,
      "avg_duration_per_task_ms": 786132.5603448276,
      "avg_output_tokens_per_task": 45244.65804597701,
      "last_run_timestamp": "2026-10-17T19:13:10.419Z"
    },
    {
      "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": 0.6962752034383954,
      "accuracy_std_dev": 0.02843686742557259,
      "run_count": 7,
      "total_questions": 349,
      "avg_cost_per_task_usd": 0.019877143266475644,
      "avg_duration_per_task_ms": 515823.4040114613,
      "avg_output_tokens_per_task": 9119.95988538682,
      "last_run_timestamp": "2026-10-17T19:15:21.500Z"
    },
    {
      "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": "Mistral",
      "accuracy": 0.6959998500000001,
      "accuracy_std_dev": 0.02293954179637203,
      "run_count": 20,
      "total_questions": 1000,
      "avg_cost_per_task_usd": 0.10828757100000001,
      "avg_duration_per_task_ms": 77900.456,
      "avg_output_tokens_per_task": 9749.434,
      "last_run_timestamp": "2026-10-17T19:38:37.865Z"
    },
    {
      "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": "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": 0.6958144439461884,
      "accuracy_std_dev": 0.011732711186157182,
      "run_count": 9,
      "total_questions": 446,
      "avg_cost_per_task_usd": 0.02170840807174888,
      "avg_duration_per_task_ms": 570466.8318385651,
      "avg_output_tokens_per_task": 18922.206278026904,
      "last_run_timestamp": "2026-10-17T19:19:38.619Z"
    },
    {
      "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": 0.695,
      "accuracy_std_dev": 0.0016670000000000296,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.11442463,
      "avg_duration_per_task_ms": 208708.95,
      "avg_output_tokens_per_task": 8898.45,
      "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": "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": 0.69499975,
      "accuracy_std_dev": 0.041331999306681265,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.018806935000000004,
      "avg_duration_per_task_ms": 381666.575,
      "avg_output_tokens_per_task": 20477.295,
      "last_run_timestamp": "2026-10-17T20:55:29.683Z"
    },
    {
      "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": 0.6949152745762711,
      "accuracy_std_dev": 0.028368895357513555,
      "run_count": 6,
      "total_questions": 295,
      "avg_cost_per_task_usd": 0.01705773898305085,
      "avg_duration_per_task_ms": 275649.04745762714,
      "avg_output_tokens_per_task": 9029.440677966102,
      "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": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "Novita",
      "accuracy": 0.6933333333333334,
      "accuracy_std_dev": 0.054160228677918684,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.02421678666666667,
      "avg_duration_per_task_ms": 198756.01333333334,
      "avg_output_tokens_per_task": 9626.628888888888,
      "last_run_timestamp": "2026-10-17T19:18:29.012Z"
    },
    {
      "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": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "Venice",
      "accuracy": 0.6933333333333334,
      "accuracy_std_dev": 0.07159554093346564,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.07348721999999999,
      "avg_duration_per_task_ms": 284784.05555555556,
      "avg_output_tokens_per_task": 26224.762222222223,
      "last_run_timestamp": "2026-10-17T19:45:29.467Z"
    },
    {
      "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": "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": "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-4.7-flash-20260119",
      "model_slug": "z-ai/glm-4.7-flash",
      "model_name": "Z.ai: GLM 4.7 Flash",
      "provider_name": "Venice",
      "accuracy": 0.6923076113360324,
      "accuracy_std_dev": 0.004408577938519381,
      "run_count": 5,
      "total_questions": 247,
      "avg_cost_per_task_usd": 0.009435951417004049,
      "avg_duration_per_task_ms": 259361.54655870446,
      "avg_output_tokens_per_task": 9080.562753036438,
      "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": "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": "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": 0.6913265306122449,
      "accuracy_std_dev": 0.02634978866613541,
      "run_count": 4,
      "total_questions": 196,
      "avg_cost_per_task_usd": 0.024798336734693876,
      "avg_duration_per_task_ms": 305198.10204081633,
      "avg_output_tokens_per_task": 16421.35714285714,
      "last_run_timestamp": "2026-10-17T20:08:29.447Z"
    },
    {
      "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": "Minimax",
      "accuracy": 0.6905556666666667,
      "accuracy_std_dev": 0.03202522925323443,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.035373304999999994,
      "avg_duration_per_task_ms": 164008.30833333332,
      "avg_output_tokens_per_task": 10108.841666666667,
      "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": "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": "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": 0.69,
      "accuracy_std_dev": 0.003332999999999975,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01604401,
      "avg_duration_per_task_ms": 258586.65,
      "avg_output_tokens_per_task": 10109.57,
      "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": "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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Friendli",
      "accuracy": 0.688485,
      "accuracy_std_dev": 0.07169463814489388,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.11122534363636363,
      "avg_duration_per_task_ms": 72899.75090909092,
      "avg_output_tokens_per_task": 8778.474545454545,
      "last_run_timestamp": "2026-10-17T19:21:20.080Z"
    },
    {
      "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": 0.6872909230769231,
      "accuracy_std_dev": 0.021147758508094125,
      "run_count": 6,
      "total_questions": 299,
      "avg_cost_per_task_usd": 0.006207953177257525,
      "avg_duration_per_task_ms": 252164.59866220737,
      "avg_output_tokens_per_task": 9519.789297658863,
      "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": "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": "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": "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": 0.6866665000000001,
      "accuracy_std_dev": 0.01943636340085252,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.015455749999999999,
      "avg_duration_per_task_ms": 256063.75,
      "avg_output_tokens_per_task": 17914.69,
      "last_run_timestamp": "2026-10-17T20:59:34.683Z"
    },
    {
      "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": "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": "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": 0.6833333333333333,
      "accuracy_std_dev": 0.027148330554361683,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.007905433333333333,
      "avg_duration_per_task_ms": 275517.85333333333,
      "avg_output_tokens_per_task": 12243.556666666667,
      "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": "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": "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": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "AtlasCloud",
      "accuracy": 0.6814285714285714,
      "accuracy_std_dev": 0.07495729245159184,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.05172822285714287,
      "avg_duration_per_task_ms": 312143.8542857143,
      "avg_output_tokens_per_task": 8009.471428571429,
      "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": "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": "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": "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "DekaLLM",
      "accuracy": 0.6791665,
      "accuracy_std_dev": 0.017380712492587855,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.009063064999999999,
      "avg_duration_per_task_ms": 747112.36,
      "avg_output_tokens_per_task": 18370.14,
      "last_run_timestamp": "2026-10-17T19:40:56.720Z"
    },
    {
      "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": 0.6783334999999999,
      "accuracy_std_dev": 0.062070824243359295,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.02936051,
      "avg_duration_per_task_ms": 178882.05,
      "avg_output_tokens_per_task": 10616.835,
      "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": "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": "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": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "CoreWeave",
      "accuracy": 0.6761111666666667,
      "accuracy_std_dev": 0.020942326744471888,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.017369979999999997,
      "avg_duration_per_task_ms": 120886.62666666666,
      "avg_output_tokens_per_task": 21445.503333333334,
      "last_run_timestamp": "2026-10-17T10:59:56.016Z"
    },
    {
      "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": "auto-routing",
      "accuracy": 0.6755556666666667,
      "accuracy_std_dev": 0.020245505523174504,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.008486243333333334,
      "avg_duration_per_task_ms": 252988.28666666665,
      "avg_output_tokens_per_task": 9856.29,
      "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-k2.7-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": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "Parasail",
      "accuracy": 0.6734496104651163,
      "accuracy_std_dev": 0.08613577414329836,
      "run_count": 7,
      "total_questions": 344,
      "avg_cost_per_task_usd": 0.13526142151162793,
      "avg_duration_per_task_ms": 1917186.9389534884,
      "avg_output_tokens_per_task": 101904.42732558139,
      "last_run_timestamp": "2026-10-17T19:18:10.297Z"
    },
    {
      "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": "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": 0.6733331428571429,
      "accuracy_std_dev": 0.014253843816705011,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.015046291428571428,
      "avg_duration_per_task_ms": 230830.76285714286,
      "avg_output_tokens_per_task": 10594.968571428572,
      "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": "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": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Mancer 2",
      "accuracy": 0.672916875,
      "accuracy_std_dev": 0.0946110438485348,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.056751722500000004,
      "avg_duration_per_task_ms": 524628.53,
      "avg_output_tokens_per_task": 16401.87,
      "last_run_timestamp": "2026-10-10T19:25:46.224Z"
    },
    {
      "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": 0.6724726434426229,
      "accuracy_std_dev": 0.11469794368845504,
      "run_count": 10,
      "total_questions": 488,
      "avg_cost_per_task_usd": 0.347430487704918,
      "avg_duration_per_task_ms": 1805676.3360655739,
      "avg_output_tokens_per_task": 58259.237704918036,
      "last_run_timestamp": "2026-10-17T11:24:43.736Z"
    },
    {
      "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": "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": 0.671666625,
      "accuracy_std_dev": 0.0299072896180325,
      "run_count": 32,
      "total_questions": 1600,
      "avg_cost_per_task_usd": 0.12573722500000004,
      "avg_duration_per_task_ms": 121720.524375,
      "avg_output_tokens_per_task": 12750.396875,
      "last_run_timestamp": "2026-10-17T11:08:18.199Z"
    },
    {
      "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": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Friendli",
      "accuracy": 0.670000125,
      "accuracy_std_dev": 0.08762293110601456,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.09058643999999999,
      "avg_duration_per_task_ms": 103016.5525,
      "avg_output_tokens_per_task": 8462.2,
      "last_run_timestamp": "2026-10-17T19:21:30.530Z"
    },
    {
      "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": 0.6690908181818183,
      "accuracy_std_dev": 0.03244538308329071,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.11794495454545455,
      "avg_duration_per_task_ms": 125284.06909090909,
      "avg_output_tokens_per_task": 12716.46,
      "last_run_timestamp": "2026-10-17T10:30:47.278Z"
    },
    {
      "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": "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": 0.6666667246963563,
      "accuracy_std_dev": 0.034360494831273874,
      "run_count": 10,
      "total_questions": 494,
      "avg_cost_per_task_usd": 0.02177052024291498,
      "avg_duration_per_task_ms": 261725.5991902834,
      "avg_output_tokens_per_task": 16694.267206477733,
      "last_run_timestamp": "2026-10-17T19:37:21.849Z"
    },
    {
      "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": 0.6657576363636364,
      "accuracy_std_dev": 0.02534973407166497,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.11508544,
      "avg_duration_per_task_ms": 127141.62181818181,
      "avg_output_tokens_per_task": 12363.76909090909,
      "last_run_timestamp": "2026-10-17T10:30:02.901Z"
    },
    {
      "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": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Friendli",
      "accuracy": 0.6644445,
      "accuracy_std_dev": 0.0170691983091376,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.030060816666666667,
      "avg_duration_per_task_ms": 250207.05,
      "avg_output_tokens_per_task": 11154.95,
      "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": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "AtlasCloud",
      "accuracy": 0.664166625,
      "accuracy_std_dev": 0.11765935913362087,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.05179430000000001,
      "avg_duration_per_task_ms": 715447.9325,
      "avg_output_tokens_per_task": 23872.2825,
      "last_run_timestamp": "2026-10-17T19:52:00.769Z"
    },
    {
      "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": 0.6622221666666667,
      "accuracy_std_dev": 0.016629373093181295,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.026509213333333333,
      "avg_duration_per_task_ms": 393739.17,
      "avg_output_tokens_per_task": 10346.71,
      "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": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Tenstorrent",
      "accuracy": 0.66166675,
      "accuracy_std_dev": 0.21009922457421754,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.01252789,
      "avg_duration_per_task_ms": 392497.45,
      "avg_output_tokens_per_task": 10444.425,
      "last_run_timestamp": "2026-10-17T10:36:51.409Z"
    },
    {
      "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": 0.6616665,
      "accuracy_std_dev": 0.048333499999999974,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.006591860000000001,
      "avg_duration_per_task_ms": 431022.15,
      "avg_output_tokens_per_task": 20204.64,
      "last_run_timestamp": "2026-10-17T19:41:21.758Z"
    },
    {
      "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": 0.6600001666666666,
      "accuracy_std_dev": 0.026666791666894545,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.023134516666666664,
      "avg_duration_per_task_ms": 309314.74333333335,
      "avg_output_tokens_per_task": 10345.606666666667,
      "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": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "AtlasCloud",
      "accuracy": 0.6596296666666667,
      "accuracy_std_dev": 0.10722601669267481,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.09074199333333334,
      "avg_duration_per_task_ms": 219707.82,
      "avg_output_tokens_per_task": 22414.957777777778,
      "last_run_timestamp": "2026-10-17T19:35:52.505Z"
    },
    {
      "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": 0.6594443333333334,
      "accuracy_std_dev": 0.015081891688010765,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.02318081,
      "avg_duration_per_task_ms": 162431.44666666666,
      "avg_output_tokens_per_task": 17110.11,
      "last_run_timestamp": "2026-10-17T19:13:31.535Z"
    },
    {
      "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": 0.659393909090909,
      "accuracy_std_dev": 0.028349142244908185,
      "run_count": 33,
      "total_questions": 1650,
      "avg_cost_per_task_usd": 0.12236485999999999,
      "avg_duration_per_task_ms": 121638.10242424242,
      "avg_output_tokens_per_task": 12651.811515151516,
      "last_run_timestamp": "2026-10-17T11:12:55.032Z"
    },
    {
      "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": 0.6570882643678161,
      "accuracy_std_dev": 0.06117711592197373,
      "run_count": 7,
      "total_questions": 348,
      "avg_cost_per_task_usd": 0.024367669540229883,
      "avg_duration_per_task_ms": 460029.2816091954,
      "avg_output_tokens_per_task": 15624,
      "last_run_timestamp": "2026-10-17T19:35:41.485Z"
    },
    {
      "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Crusoe",
      "accuracy": 0.6555553333333334,
      "accuracy_std_dev": 0.04228062043326968,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.016284913333333335,
      "avg_duration_per_task_ms": 98025.37333333334,
      "avg_output_tokens_per_task": 20936.24,
      "last_run_timestamp": "2026-10-17T19:16:22.176Z"
    },
    {
      "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": 0.654166625,
      "accuracy_std_dev": 0.16981822491706944,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.13378775999999998,
      "avg_duration_per_task_ms": 366007.185,
      "avg_output_tokens_per_task": 8490.5075,
      "last_run_timestamp": "2026-10-17T20:38:51.263Z"
    },
    {
      "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": 0.6536609479768786,
      "accuracy_std_dev": 0.0348083909646944,
      "run_count": 7,
      "total_questions": 346,
      "avg_cost_per_task_usd": 0.013188358381502889,
      "avg_duration_per_task_ms": 991739.563583815,
      "avg_output_tokens_per_task": 22797.29479768786,
      "last_run_timestamp": "2026-10-17T11:54:26.628Z"
    },
    {
      "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": "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": "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": 0.6527314314381271,
      "accuracy_std_dev": 0.019242186339885847,
      "run_count": 6,
      "total_questions": 299,
      "avg_cost_per_task_usd": 0.01716779933110368,
      "avg_duration_per_task_ms": 303346.9264214047,
      "avg_output_tokens_per_task": 22796.782608695652,
      "last_run_timestamp": "2026-10-17T11:22:16.401Z"
    },
    {
      "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": 0.652196480620155,
      "accuracy_std_dev": 0.02424257709439149,
      "run_count": 13,
      "total_questions": 645,
      "avg_cost_per_task_usd": 0.010831201550387599,
      "avg_duration_per_task_ms": 324905.28992248065,
      "avg_output_tokens_per_task": 17153.43565891473,
      "last_run_timestamp": "2026-10-17T19:00:34.053Z"
    },
    {
      "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": 0.6521739130434783,
      "accuracy_std_dev": 0.026636806048627927,
      "run_count": 6,
      "total_questions": 299,
      "avg_cost_per_task_usd": 0.014278464882943145,
      "avg_duration_per_task_ms": 801415.0635451504,
      "avg_output_tokens_per_task": 24999.354515050167,
      "last_run_timestamp": "2026-10-17T19:51:56.033Z"
    },
    {
      "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": "anthropic/claude-4.5-haiku-20251001",
      "model_slug": "anthropic/claude-haiku-4.5",
      "model_name": "Anthropic: Claude Haiku 4.5",
      "provider_name": "Azure",
      "accuracy": 0.6515152727272728,
      "accuracy_std_dev": 0.024959225520075408,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.11694987454545454,
      "avg_duration_per_task_ms": 144270.65272727274,
      "avg_output_tokens_per_task": 12693.569090909092,
      "last_run_timestamp": "2026-10-17T11:02:55.214Z"
    },
    {
      "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": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "DeepInfra",
      "accuracy": 0.6495238571428571,
      "accuracy_std_dev": 0.04088017277510516,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.021405068571428574,
      "avg_duration_per_task_ms": 240437.21714285715,
      "avg_output_tokens_per_task": 27547.52857142857,
      "last_run_timestamp": "2026-10-17T10:59:26.562Z"
    },
    {
      "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": 0.6485186666666666,
      "accuracy_std_dev": 0.03216903997807692,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.03373100888888889,
      "avg_duration_per_task_ms": 73645.7311111111,
      "avg_output_tokens_per_task": 17406.14,
      "last_run_timestamp": "2026-10-17T19:37:10.435Z"
    },
    {
      "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "auto-routing",
      "accuracy": 0.64666675,
      "accuracy_std_dev": 0.032317934478204226,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.014356867499999997,
      "avg_duration_per_task_ms": 383641.005,
      "avg_output_tokens_per_task": 19373.5125,
      "last_run_timestamp": "2026-10-17T19:56:57.382Z"
    },
    {
      "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": 0.6465773571428571,
      "accuracy_std_dev": 0.12243509635005365,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.04724405580357143,
      "avg_duration_per_task_ms": 2135419.910714286,
      "avg_output_tokens_per_task": 77401.015625,
      "last_run_timestamp": "2026-10-17T11:23:30.347Z"
    },
    {
      "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": 0.6464647272727273,
      "accuracy_std_dev": 0.040612499999999996,
      "run_count": 2,
      "total_questions": 99,
      "avg_cost_per_task_usd": 0.03484631313131314,
      "avg_duration_per_task_ms": 278588.8080808081,
      "avg_output_tokens_per_task": 18286.49494949495,
      "last_run_timestamp": "2026-10-17T19:18:50.561Z"
    },
    {
      "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": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Novita",
      "accuracy": 0.6455555000000001,
      "accuracy_std_dev": 0.022249880012785092,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.024322700000000003,
      "avg_duration_per_task_ms": 151214.55,
      "avg_output_tokens_per_task": 10419.556666666667,
      "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": 0.6455555000000001,
      "accuracy_std_dev": 0.03206863203479896,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.02137825,
      "avg_duration_per_task_ms": 227028.51,
      "avg_output_tokens_per_task": 10295.053333333333,
      "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": "Mara",
      "accuracy": 0.644208,
      "accuracy_std_dev": 0.03904293739234052,
      "run_count": 3,
      "total_questions": 141,
      "avg_cost_per_task_usd": 0.024240567375886524,
      "avg_duration_per_task_ms": 70321.02836879433,
      "avg_output_tokens_per_task": 14782.985815602837,
      "last_run_timestamp": "2026-10-10T18:42:24.416Z"
    },
    {
      "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "DeepInfra",
      "accuracy": 0.6409490933014353,
      "accuracy_std_dev": 0.040131205785428226,
      "run_count": 17,
      "total_questions": 836,
      "avg_cost_per_task_usd": 0.020747283492822968,
      "avg_duration_per_task_ms": 318678.3744019139,
      "avg_output_tokens_per_task": 17826.28947368421,
      "last_run_timestamp": "2026-10-17T20:17:34.516Z"
    },
    {
      "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": 0.64083325,
      "accuracy_std_dev": 0.027525485666696242,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.02212257,
      "avg_duration_per_task_ms": 261130.045,
      "avg_output_tokens_per_task": 10462.685,
      "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": "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": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "Venice",
      "accuracy": 0.6396396317567568,
      "accuracy_std_dev": 0.06834614418790695,
      "run_count": 12,
      "total_questions": 592,
      "avg_cost_per_task_usd": 0.024531244932432434,
      "avg_duration_per_task_ms": 263890.6858108108,
      "avg_output_tokens_per_task": 25052.733108108107,
      "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": "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": "moonshotai/kimi-k2-0905",
      "model_slug": "moonshotai/kimi-k2-0905",
      "model_name": "MoonshotAI: Kimi K2 0905",
      "provider_name": "Novita",
      "accuracy": 0.6366665,
      "accuracy_std_dev": 0.03666649999999999,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.09132923,
      "avg_duration_per_task_ms": 148180.52,
      "avg_output_tokens_per_task": 3934.54,
      "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": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "GMICloud",
      "accuracy": 0.6333331666666667,
      "accuracy_std_dev": 0.10029589925551403,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.017528723333333333,
      "avg_duration_per_task_ms": 260405.91333333333,
      "avg_output_tokens_per_task": 9076.7,
      "last_run_timestamp": "2026-10-17T19:37:16.888Z"
    },
    {
      "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": 0.633333,
      "accuracy_std_dev": 0.020000000000000018,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.024781250000000005,
      "avg_duration_per_task_ms": 171636.65,
      "avg_output_tokens_per_task": 11166.2,
      "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Venice",
      "accuracy": 0.633333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.009703119999999999,
      "avg_duration_per_task_ms": 1596682.66,
      "avg_output_tokens_per_task": 19652.38,
      "last_run_timestamp": "2026-10-17T21:47:41.282Z"
    },
    {
      "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": 0.63,
      "accuracy_std_dev": 0.016666999999999987,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.02616634,
      "avg_duration_per_task_ms": 174261.43,
      "avg_output_tokens_per_task": 11763.02,
      "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": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "CoreWeave",
      "accuracy": 0.63,
      "accuracy_std_dev": 0.023333000000000048,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.07672618,
      "avg_duration_per_task_ms": 231628.06,
      "avg_output_tokens_per_task": 10580.58,
      "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": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "AtlasCloud",
      "accuracy": 0.6299998333333334,
      "accuracy_std_dev": 0.019531617294501993,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.02836308333333333,
      "avg_duration_per_task_ms": 208192.55666666667,
      "avg_output_tokens_per_task": 10017.453333333333,
      "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": "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": "SambaNova",
      "accuracy": 0.6239699483146067,
      "accuracy_std_dev": 0.026459500316288177,
      "run_count": 9,
      "total_questions": 445,
      "avg_cost_per_task_usd": 0.04252493033707865,
      "avg_duration_per_task_ms": 84689.76629213484,
      "avg_output_tokens_per_task": 23543.703370786516,
      "last_run_timestamp": "2026-10-17T19:09:20.411Z"
    },
    {
      "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": 0.6233332500000001,
      "accuracy_std_dev": 0.038872984049819245,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.09536726749999998,
      "avg_duration_per_task_ms": 55598.6275,
      "avg_output_tokens_per_task": 3623.22,
      "last_run_timestamp": "2026-10-17T19:10:42.003Z"
    },
    {
      "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Cerebras",
      "accuracy": 0.619999875,
      "accuracy_std_dev": 0.024036858272440176,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.06868683249999999,
      "avg_duration_per_task_ms": 32416.59,
      "avg_output_tokens_per_task": 21460.735,
      "last_run_timestamp": "2026-10-17T19:10:22.459Z"
    },
    {
      "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": 0.6196237419354839,
      "accuracy_std_dev": 0.011182496025485556,
      "run_count": 5,
      "total_questions": 248,
      "avg_cost_per_task_usd": 0.1178863306451613,
      "avg_duration_per_task_ms": 650732.7620967742,
      "avg_output_tokens_per_task": 30809.334677419356,
      "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-m2.5-20260211",
      "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": "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": 0.6006711006711409,
      "accuracy_std_dev": 0.09595263420158,
      "run_count": 6,
      "total_questions": 298,
      "avg_cost_per_task_usd": 0.03432377852348993,
      "avg_duration_per_task_ms": 2205358.8926174496,
      "avg_output_tokens_per_task": 69060.20469798657,
      "last_run_timestamp": "2026-10-17T20:16:20.049Z"
    },
    {
      "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": 0.59958325,
      "accuracy_std_dev": 0.026793203689322042,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.047474890000000006,
      "avg_duration_per_task_ms": 92900.3,
      "avg_output_tokens_per_task": 4620.69,
      "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": 0.5943487528735633,
      "accuracy_std_dev": 0.021321805852608424,
      "run_count": 7,
      "total_questions": 348,
      "avg_cost_per_task_usd": 0.019723531609195398,
      "avg_duration_per_task_ms": 98994.94827586207,
      "avg_output_tokens_per_task": 4290.752873563219,
      "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": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "AtlasCloud",
      "accuracy": 0.5935779504587155,
      "accuracy_std_dev": 0.242524807897738,
      "run_count": 11,
      "total_questions": 545,
      "avg_cost_per_task_usd": 0.017899249541284405,
      "avg_duration_per_task_ms": 294924.47155963304,
      "avg_output_tokens_per_task": 8919.59633027523,
      "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": "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": "moonshotai/kimi-k2-0905",
      "model_slug": "moonshotai/kimi-k2-0905",
      "model_name": "MoonshotAI: Kimi K2 0905",
      "provider_name": "auto-routing",
      "accuracy": 0.5918367959183675,
      "accuracy_std_dev": 0.0052774999999999905,
      "run_count": 2,
      "total_questions": 98,
      "avg_cost_per_task_usd": 0.09532348979591837,
      "avg_duration_per_task_ms": 107161.72448979592,
      "avg_output_tokens_per_task": 3646.7755102040815,
      "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": "Ionstream",
      "accuracy": 0.5902778229166666,
      "accuracy_std_dev": 0.015776500000000027,
      "run_count": 2,
      "total_questions": 96,
      "avg_cost_per_task_usd": 0.01728564583333333,
      "avg_duration_per_task_ms": 131526.78125,
      "avg_output_tokens_per_task": 3837.7916666666665,
      "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": "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": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "AkashML",
      "accuracy": 0.58166675,
      "accuracy_std_dev": 0.11415155982809651,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.05917653,
      "avg_duration_per_task_ms": 650135.44,
      "avg_output_tokens_per_task": 38980.17,
      "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": "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": "Alibaba",
      "accuracy": 0.5762463812316716,
      "accuracy_std_dev": 0.04222508395322788,
      "run_count": 7,
      "total_questions": 341,
      "avg_cost_per_task_usd": 0.07726211143695015,
      "avg_duration_per_task_ms": 93864.66862170088,
      "avg_output_tokens_per_task": 5954.222873900293,
      "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": "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": 0.5733334999999999,
      "accuracy_std_dev": 0.08212467486845838,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0333021775,
      "avg_duration_per_task_ms": 408902.2175,
      "avg_output_tokens_per_task": 17562.9,
      "last_run_timestamp": "2026-10-17T20:21:35.847Z"
    },
    {
      "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Novita",
      "accuracy": 0.5687678137535817,
      "accuracy_std_dev": 0.03326857759790391,
      "run_count": 7,
      "total_questions": 349,
      "avg_cost_per_task_usd": 0.011248378223495703,
      "avg_duration_per_task_ms": 234594.50429799428,
      "avg_output_tokens_per_task": 18173.54441260745,
      "last_run_timestamp": "2026-10-17T19:50:06.036Z"
    },
    {
      "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": 0.5590475714285714,
      "accuracy_std_dev": 0.10244195347729805,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.03983742857142857,
      "avg_duration_per_task_ms": 100656.9,
      "avg_output_tokens_per_task": 4373.865714285715,
      "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": "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": 0.5438595101214574,
      "accuracy_std_dev": 0.02876780459472013,
      "run_count": 5,
      "total_questions": 247,
      "avg_cost_per_task_usd": 0.002696097165991903,
      "avg_duration_per_task_ms": 431895.5668016194,
      "avg_output_tokens_per_task": 6592.368421052632,
      "last_run_timestamp": "2026-10-17T19:51:20.926Z"
    },
    {
      "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": 0.5387596395348836,
      "accuracy_std_dev": 0.016016421054272853,
      "run_count": 7,
      "total_questions": 344,
      "avg_cost_per_task_usd": 0.021050491279069765,
      "avg_duration_per_task_ms": 1421665.261627907,
      "avg_output_tokens_per_task": 95835.8953488372,
      "last_run_timestamp": "2026-10-17T19:40:18.671Z"
    },
    {
      "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": 0.5360134020100502,
      "accuracy_std_dev": 0.08239086668587728,
      "run_count": 8,
      "total_questions": 398,
      "avg_cost_per_task_usd": 0.0516702864321608,
      "avg_duration_per_task_ms": 428067,
      "avg_output_tokens_per_task": 8053.70351758794,
      "last_run_timestamp": "2026-10-17T19:13:31.456Z"
    },
    {
      "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": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "SiliconFlow",
      "accuracy": 0.5282739330357142,
      "accuracy_std_dev": 0.019775495122330913,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.06609532812499999,
      "avg_duration_per_task_ms": 286684.00223214284,
      "avg_output_tokens_per_task": 4048.9084821428573,
      "last_run_timestamp": "2026-10-17T19:35:54.698Z"
    },
    {
      "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": 0.5271991666666667,
      "accuracy_std_dev": 0.10370817001031737,
      "run_count": 6,
      "total_questions": 288,
      "avg_cost_per_task_usd": 0.062181392361111115,
      "avg_duration_per_task_ms": 408554.8958333333,
      "avg_output_tokens_per_task": 26661.76388888889,
      "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-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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "CoreWeave",
      "accuracy": 0.5188173145161291,
      "accuracy_std_dev": 0.2566565496323832,
      "run_count": 5,
      "total_questions": 248,
      "avg_cost_per_task_usd": 0.007278975806451614,
      "avg_duration_per_task_ms": 482880.30241935485,
      "avg_output_tokens_per_task": 15769.91935483871,
      "last_run_timestamp": "2026-10-10T18:54:19.893Z"
    },
    {
      "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": "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-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Together",
      "accuracy": 0.5099998888888889,
      "accuracy_std_dev": 0.03456977556718343,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.02963195333333334,
      "avg_duration_per_task_ms": 260966.97555555555,
      "avg_output_tokens_per_task": 20350.56,
      "last_run_timestamp": "2026-10-17T19:36:01.995Z"
    },
    {
      "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": "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": "qwen/qwen3-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "CoreWeave",
      "accuracy": 0.49666650000000007,
      "accuracy_std_dev": 0.023333500000000007,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.23109727999999996,
      "avg_duration_per_task_ms": 80168.37,
      "avg_output_tokens_per_task": 2746.03,
      "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": "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": 0.48833333333333334,
      "accuracy_std_dev": 0.018831396165859726,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.03323320666666667,
      "avg_duration_per_task_ms": 665027.6466666666,
      "avg_output_tokens_per_task": 61694.416666666664,
      "last_run_timestamp": "2026-10-03T16:04:32.703Z"
    },
    {
      "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-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": "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": 0.47333357142857135,
      "accuracy_std_dev": 0.03837997162902674,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.02437984,
      "avg_duration_per_task_ms": 61492.5,
      "avg_output_tokens_per_task": 3778.7314285714288,
      "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": "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": "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": 0.4707204932432433,
      "accuracy_std_dev": 0.04373885161863152,
      "run_count": 6,
      "total_questions": 296,
      "avg_cost_per_task_usd": 0.028365344594594593,
      "avg_duration_per_task_ms": 2449502.6824324327,
      "avg_output_tokens_per_task": 141559.9391891892,
      "last_run_timestamp": "2026-10-17T19:53:16.929Z"
    },
    {
      "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": 0.47047614285714284,
      "accuracy_std_dev": 0.0261516387382106,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.025336362857142858,
      "avg_duration_per_task_ms": 119394.00571428571,
      "avg_output_tokens_per_task": 3796.2971428571427,
      "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-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "auto-routing",
      "accuracy": 0.467619,
      "accuracy_std_dev": 0.035173696625258384,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.08029666571428572,
      "avg_duration_per_task_ms": 101903.78,
      "avg_output_tokens_per_task": 2752.694285714286,
      "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": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Groq",
      "accuracy": 0.4672696511627907,
      "accuracy_std_dev": 0.037956496396985506,
      "run_count": 8,
      "total_questions": 387,
      "avg_cost_per_task_usd": 0.0519110826873385,
      "avg_duration_per_task_ms": 199894.31782945737,
      "avg_output_tokens_per_task": 89805.21447028423,
      "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": "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": "qwen/qwen3-32b-04-28",
      "model_slug": "qwen/qwen3-32b",
      "model_name": "Qwen: Qwen3 32B",
      "provider_name": "auto-routing",
      "accuracy": 0.4666666666666667,
      "accuracy_std_dev": 0.019626040903747127,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.020897343333333335,
      "avg_duration_per_task_ms": 623575.84,
      "avg_output_tokens_per_task": 20727.386666666665,
      "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": "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-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": "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": 0.46416674999999996,
      "accuracy_std_dev": 0.03386045095590872,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.022192442499999996,
      "avg_duration_per_task_ms": 225702.4075,
      "avg_output_tokens_per_task": 3591.99,
      "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-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-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": "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": 0.45666666666666667,
      "accuracy_std_dev": 0.03880952218778845,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.053214168888888885,
      "avg_duration_per_task_ms": 167547.09333333332,
      "avg_output_tokens_per_task": 3746.711111111111,
      "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": 0.4563758489932886,
      "accuracy_std_dev": 0.0483687203813925,
      "run_count": 6,
      "total_questions": 298,
      "avg_cost_per_task_usd": 0.015332644295302013,
      "avg_duration_per_task_ms": 126207.14429530202,
      "avg_output_tokens_per_task": 3312.3657718120803,
      "last_run_timestamp": "2026-10-17T19:19:30.676Z"
    },
    {
      "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": 0.453333375,
      "accuracy_std_dev": 0.04533822277597982,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0537951275,
      "avg_duration_per_task_ms": 125879.785,
      "avg_output_tokens_per_task": 3538.1725,
      "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": 0.45183054913294796,
      "accuracy_std_dev": 0.04493510200731899,
      "run_count": 7,
      "total_questions": 346,
      "avg_cost_per_task_usd": 0.05361997109826589,
      "avg_duration_per_task_ms": 73759.52312138729,
      "avg_output_tokens_per_task": 3369.815028901734,
      "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/qwen3-32b-04-28",
      "model_slug": "qwen/qwen3-32b",
      "model_name": "Qwen: Qwen3 32B",
      "provider_name": "DeepInfra",
      "accuracy": 0.4516665,
      "accuracy_std_dev": 0.02954922760270845,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.021153063333333333,
      "avg_duration_per_task_ms": 604099.4133333333,
      "avg_output_tokens_per_task": 21039.02666666667,
      "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": "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-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/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Venice",
      "accuracy": 0.44708325000000004,
      "accuracy_std_dev": 0.03075165420245715,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.040560375,
      "avg_duration_per_task_ms": 277235.58,
      "avg_output_tokens_per_task": 3614.425,
      "last_run_timestamp": "2026-10-10T19:42:35.202Z"
    },
    {
      "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": "qwen/qwen3-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "Novita",
      "accuracy": 0.4428572857142857,
      "accuracy_std_dev": 0.06825330771202708,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.17141447999999998,
      "avg_duration_per_task_ms": 117506.95428571428,
      "avg_output_tokens_per_task": 3352.2,
      "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": "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-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": 0.43999985714285716,
      "accuracy_std_dev": 0.02225407399895624,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.036167245714285715,
      "avg_duration_per_task_ms": 123578.75714285714,
      "avg_output_tokens_per_task": 3292.997142857143,
      "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": "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": "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": 0.4383965367088608,
      "accuracy_std_dev": 0.03960836597472182,
      "run_count": 8,
      "total_questions": 395,
      "avg_cost_per_task_usd": 0.05240046329113924,
      "avg_duration_per_task_ms": 422067.2658227848,
      "avg_output_tokens_per_task": 20533.427848101266,
      "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": "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": "qwen/qwen3-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "Venice",
      "accuracy": 0.4362416845637584,
      "accuracy_std_dev": 0.024054442567272728,
      "run_count": 6,
      "total_questions": 298,
      "avg_cost_per_task_usd": 0.016195271812080535,
      "avg_duration_per_task_ms": 219855.4865771812,
      "avg_output_tokens_per_task": 2778.412751677852,
      "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "BaseTen",
      "accuracy": 0.4358108783783784,
      "accuracy_std_dev": 0.3084304226783661,
      "run_count": 6,
      "total_questions": 296,
      "avg_cost_per_task_usd": 0.015068885135135136,
      "avg_duration_per_task_ms": 91458.4831081081,
      "avg_output_tokens_per_task": 15700.83445945946,
      "last_run_timestamp": "2026-10-10T18:32:50.715Z"
    },
    {
      "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": "deepseek/deepseek-chat-v3-0324",
      "model_slug": "deepseek/deepseek-chat-v3-0324",
      "model_name": "DeepSeek: DeepSeek V3 0324",
      "provider_name": "DeepInfra",
      "accuracy": 0.4337431393442623,
      "accuracy_std_dev": 0.021677717946315293,
      "run_count": 5,
      "total_questions": 244,
      "avg_cost_per_task_usd": 0.032824139344262294,
      "avg_duration_per_task_ms": 478154.40573770495,
      "avg_output_tokens_per_task": 4694.745901639344,
      "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-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "Alibaba",
      "accuracy": 0.43333342857142854,
      "accuracy_std_dev": 0.04507497105888349,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.4237074657142857,
      "avg_duration_per_task_ms": 109389.9,
      "avg_output_tokens_per_task": 3189.074285714286,
      "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-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "DeepInfra",
      "accuracy": 0.4313334,
      "accuracy_std_dev": 0.02362666666798345,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.027534839999999994,
      "avg_duration_per_task_ms": 80491.464,
      "avg_output_tokens_per_task": 2779.172,
      "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-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/qwen3-14b-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/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Novita",
      "accuracy": 0.4238095714285714,
      "accuracy_std_dev": 0.013384183649337146,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.024649345714285717,
      "avg_duration_per_task_ms": 219498.08857142858,
      "avg_output_tokens_per_task": 3556.6628571428573,
      "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/qwen3-30b-a3b-04-28",
      "model_slug": "qwen/qwen3-30b-a3b",
      "model_name": "Qwen: Qwen3 30B A3B",
      "provider_name": "Alibaba",
      "accuracy": 0.423333,
      "accuracy_std_dev": 0.010000000000000009,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.03482589,
      "avg_duration_per_task_ms": 291563.16,
      "avg_output_tokens_per_task": 21648.1,
      "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": "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": 0.4222225,
      "accuracy_std_dev": 0.04331906317758191,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.05459885,
      "avg_duration_per_task_ms": 144935.76666666666,
      "avg_output_tokens_per_task": 3474.9266666666667,
      "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": 0.42196971363636365,
      "accuracy_std_dev": 0.02414578127403073,
      "run_count": 9,
      "total_questions": 440,
      "avg_cost_per_task_usd": 0.03133452727272727,
      "avg_duration_per_task_ms": 644947.5272727272,
      "avg_output_tokens_per_task": 6264.340909090909,
      "last_run_timestamp": "2026-10-17T19:21:15.015Z"
    },
    {
      "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": "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": 0.42,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.02779026,
      "avg_duration_per_task_ms": 140935.14,
      "avg_output_tokens_per_task": 10384.88,
      "last_run_timestamp": "2026-10-17T11:21:49.304Z"
    },
    {
      "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": 0.41901883467741935,
      "accuracy_std_dev": 0.027582004591581083,
      "run_count": 10,
      "total_questions": 496,
      "avg_cost_per_task_usd": 0.02511988508064516,
      "avg_duration_per_task_ms": 390541.47782258067,
      "avg_output_tokens_per_task": 3601.5826612903224,
      "last_run_timestamp": "2026-10-17T19:23:43.385Z"
    },
    {
      "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": "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": "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": 0.40897435897435896,
      "accuracy_std_dev": 0.021673182250407416,
      "run_count": 8,
      "total_questions": 390,
      "avg_cost_per_task_usd": 0.011223169230769232,
      "avg_duration_per_task_ms": 108184.48717948717,
      "avg_output_tokens_per_task": 3251.0871794871796,
      "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-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.408060476070529,
      "accuracy_std_dev": 0.1308846142804411,
      "run_count": 16,
      "total_questions": 794,
      "avg_cost_per_task_usd": 0.039315931989924426,
      "avg_duration_per_task_ms": 629857.338790932,
      "avg_output_tokens_per_task": 160922.7934508816,
      "last_run_timestamp": "2026-10-17T19:35:54.288Z"
    },
    {
      "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": "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": 0.40333342857142856,
      "accuracy_std_dev": 0.021894458241150196,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.06258026857142857,
      "avg_duration_per_task_ms": 99625.90285714286,
      "avg_output_tokens_per_task": 2910.902857142857,
      "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-vl-235b-a22b-instruct",
      "model_slug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_name": "Qwen: Qwen3 VL 235B A22B Instruct",
      "provider_name": "Novita",
      "accuracy": 0.4019045714285715,
      "accuracy_std_dev": 0.02611661099134934,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.07661166857142858,
      "avg_duration_per_task_ms": 119529.87428571428,
      "avg_output_tokens_per_task": 3108.5485714285714,
      "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": "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": "Venice",
      "accuracy": 0.4,
      "accuracy_std_dev": 0.03126941809500138,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.06669086285714285,
      "avg_duration_per_task_ms": 316867.28285714285,
      "avg_output_tokens_per_task": 12448.365714285714,
      "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": 0.39809542857142854,
      "accuracy_std_dev": 0.024355591348993847,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.027166705714285717,
      "avg_duration_per_task_ms": 311043.09428571427,
      "avg_output_tokens_per_task": 4197.714285714285,
      "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": "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": 0.39809514285714287,
      "accuracy_std_dev": 0.03431195953278671,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.06120218857142856,
      "avg_duration_per_task_ms": 337486.8342857143,
      "avg_output_tokens_per_task": 8320.111428571428,
      "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": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Chutes",
      "accuracy": 0.3980540203045685,
      "accuracy_std_dev": 0.06137496179721317,
      "run_count": 8,
      "total_questions": 394,
      "avg_cost_per_task_usd": 0.004494365482233502,
      "avg_duration_per_task_ms": 319107.7182741117,
      "avg_output_tokens_per_task": 4432.3045685279185,
      "last_run_timestamp": "2026-10-17T19:41:17.469Z"
    },
    {
      "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": 0.3975043048128342,
      "accuracy_std_dev": 0.017113317027610396,
      "run_count": 4,
      "total_questions": 187,
      "avg_cost_per_task_usd": 0.008529352941176469,
      "avg_duration_per_task_ms": 72525.10695187165,
      "avg_output_tokens_per_task": 1231.951871657754,
      "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/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "auto-routing",
      "accuracy": 0.3970371111111111,
      "accuracy_std_dev": 0.12102898971407583,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.023592877777777778,
      "avg_duration_per_task_ms": 195708.14444444445,
      "avg_output_tokens_per_task": 3363.0533333333333,
      "last_run_timestamp": "2026-10-17T19:58:01.087Z"
    },
    {
      "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": "Parasail",
      "accuracy": 0.396667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.0528661,
      "avg_duration_per_task_ms": 135118.22,
      "avg_output_tokens_per_task": 7825.32,
      "last_run_timestamp": "2026-07-25T08:07:19.515Z"
    },
    {
      "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": 0.3961905714285714,
      "accuracy_std_dev": 0.05390861756411,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.04983603428571428,
      "avg_duration_per_task_ms": 88538.82571428572,
      "avg_output_tokens_per_task": 3736.18,
      "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": "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": "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": 0.3942857142857143,
      "accuracy_std_dev": 0.02060319637278425,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.025239111428571426,
      "avg_duration_per_task_ms": 231992.80571428573,
      "avg_output_tokens_per_task": 4356.64,
      "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": "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": 0.39351850505050506,
      "accuracy_std_dev": 0.030148094697001596,
      "run_count": 8,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.05853486616161616,
      "avg_duration_per_task_ms": 386334.952020202,
      "avg_output_tokens_per_task": 3836.3636363636365,
      "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": "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-vl-235b-a22b-instruct",
      "model_slug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_name": "Qwen: Qwen3 VL 235B A22B Instruct",
      "provider_name": "Parasail",
      "accuracy": 0.39047600000000005,
      "accuracy_std_dev": 0.017036713725699232,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.03704184285714286,
      "avg_duration_per_task_ms": 191510.86285714287,
      "avg_output_tokens_per_task": 3720.674285714286,
      "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/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "CoreWeave",
      "accuracy": 0.39,
      "accuracy_std_dev": 0.014782396941407474,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.028387520000000006,
      "avg_duration_per_task_ms": 78569.82333333333,
      "avg_output_tokens_per_task": 4258.516666666666,
      "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": "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": 0.3877778333333333,
      "accuracy_std_dev": 0.00895805727481629,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.014963976666666667,
      "avg_duration_per_task_ms": 92139.43333333333,
      "avg_output_tokens_per_task": 608.4966666666667,
      "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/qwen3-30b-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": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Parasail",
      "accuracy": 0.38583349999999994,
      "accuracy_std_dev": 0.06396070625391811,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.006338907499999999,
      "avg_duration_per_task_ms": 153451.2725,
      "avg_output_tokens_per_task": 3929.855,
      "last_run_timestamp": "2026-10-10T19:09:43.617Z"
    },
    {
      "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": 0.372222,
      "accuracy_std_dev": 0.019019243018585152,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.03316533666666667,
      "avg_duration_per_task_ms": 160394.92,
      "avg_output_tokens_per_task": 4455.986666666667,
      "last_run_timestamp": "2026-10-03T16:07:02.084Z"
    },
    {
      "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": 0.37148306580829754,
      "accuracy_std_dev": 0.022649811310652534,
      "run_count": 14,
      "total_questions": 699,
      "avg_cost_per_task_usd": 0.01332730329041488,
      "avg_duration_per_task_ms": 37995.6452074392,
      "avg_output_tokens_per_task": 713.623748211731,
      "last_run_timestamp": "2026-10-17T19:12:18.209Z"
    },
    {
      "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": 0.3695237142857143,
      "accuracy_std_dev": 0.029996155461822036,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.05283062,
      "avg_duration_per_task_ms": 74136.42571428571,
      "avg_output_tokens_per_task": 4382.251428571429,
      "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/qwen3-235b-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/qwen3-30b-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": "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": 0.35925922222222223,
      "accuracy_std_dev": 0.032270942308928206,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.08270178888888888,
      "avg_duration_per_task_ms": 25491.448888888888,
      "avg_output_tokens_per_task": 960.1,
      "last_run_timestamp": "2026-10-17T18:14:44.847Z"
    },
    {
      "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": "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": "qwen/qwen3-vl-8b-instruct",
      "model_slug": "qwen/qwen3-vl-8b-instruct",
      "model_name": "Qwen: Qwen3 VL 8B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.348333,
      "accuracy_std_dev": 0.015000000000000013,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.042842359999999996,
      "avg_duration_per_task_ms": 116081.37,
      "avg_output_tokens_per_task": 4431.82,
      "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": "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": 0,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.02900084,
      "avg_duration_per_task_ms": 82898.03,
      "avg_output_tokens_per_task": 3525.09,
      "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-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 VL 30B A3B Instruct",
      "provider_name": "Alibaba",
      "accuracy": 0.32800009999999996,
      "accuracy_std_dev": 0.024367614210053474,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.08013712,
      "avg_duration_per_task_ms": 148773.196,
      "avg_output_tokens_per_task": 8016.85,
      "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": "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": 0.03824817264573991,
      "avg_duration_per_task_ms": 96365.55156950673,
      "avg_output_tokens_per_task": 1277.7242152466367,
      "last_run_timestamp": "2026-10-17T18:24:53.657Z"
    },
    {
      "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": 0.3093607534246575,
      "accuracy_std_dev": 0.03377945099423942,
      "run_count": 6,
      "total_questions": 292,
      "avg_cost_per_task_usd": 0.014691332191780824,
      "avg_duration_per_task_ms": 145559.72602739726,
      "avg_output_tokens_per_task": 1356.4246575342465,
      "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Nebius",
      "accuracy": 0.30916662500000003,
      "accuracy_std_dev": 0.23192285519970726,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0994847125,
      "avg_duration_per_task_ms": 229836.265,
      "avg_output_tokens_per_task": 28937.29,
      "last_run_timestamp": "2026-10-17T19:13:29.194Z"
    },
    {
      "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": 0.29925922222222223,
      "accuracy_std_dev": 0.03737628995552537,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.17081659111111108,
      "avg_duration_per_task_ms": 141383.24444444446,
      "avg_output_tokens_per_task": 1902.9044444444444,
      "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-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "CoreWeave",
      "accuracy": 0.28901757803468203,
      "accuracy_std_dev": 0.24326737555175115,
      "run_count": 7,
      "total_questions": 346,
      "avg_cost_per_task_usd": 0.01199693063583815,
      "avg_duration_per_task_ms": 1302513.62716763,
      "avg_output_tokens_per_task": 80546.70520231214,
      "last_run_timestamp": "2026-10-17T19:19:57.292Z"
    },
    {
      "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/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "Venice",
      "accuracy": 0.2683333333333333,
      "accuracy_std_dev": 0.3228473030880423,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.016053078333333335,
      "avg_duration_per_task_ms": 49421.746666666666,
      "avg_output_tokens_per_task": 8722.081666666667,
      "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": "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": "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": 0.22095242857142855,
      "accuracy_std_dev": 0.06555531427048489,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.012673305714285711,
      "avg_duration_per_task_ms": 20818.634285714284,
      "avg_output_tokens_per_task": 1052.7457142857143,
      "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": "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": "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": 0.19666649999999997,
      "accuracy_std_dev": 0.023804971471312466,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.00425299,
      "avg_duration_per_task_ms": 15962.445,
      "avg_output_tokens_per_task": 1579.045,
      "last_run_timestamp": "2026-10-03T16:07:44.178Z"
    },
    {
      "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": "deepseek/deepseek-v3.1-terminus",
      "model_slug": "deepseek/deepseek-v3.1-terminus",
      "model_name": "DeepSeek: DeepSeek V3.1 Terminus",
      "provider_name": "AtlasCloud",
      "accuracy": 0.18666649999999999,
      "accuracy_std_dev": 0.13029887012397048,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.01464795,
      "avg_duration_per_task_ms": 65215.473333333335,
      "avg_output_tokens_per_task": 1173.1066666666666,
      "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": "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"
    }
  ]
}
