{
  "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": "2026-10-31",
  "retrieved_at": "2026-10-31T01:45:40.293Z",
  "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": 2230,
  "rows": [
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-4-argon-20261013",
      "model_slug": "google/gemini-4-argon",
      "model_name": "Google: Gemini 4 Argon",
      "provider_name": "Google",
      "accuracy": 0.964646,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.09468125757575757,
      "avg_duration_per_task_ms": 101980.28282828283,
      "avg_output_tokens_per_task": 8439.414141414141,
      "last_run_timestamp": "2026-10-14T15:00:03.588Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-4-argon",
      "model_slug": "google/gemini-4-argon",
      "model_name": "Gemini 4 Argon",
      "provider_name": "auto-routing",
      "accuracy": 0.959596,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.029260161616161615,
      "avg_duration_per_task_ms": 40252.82323232323,
      "avg_output_tokens_per_task": 2870.0252525252527,
      "last_run_timestamp": "2026-10-23T07:38:37.725Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.8-flash-20260902",
      "model_slug": "google/gemini-3.8-flash",
      "model_name": "Google: Gemini 3.8 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.95637,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 191,
      "avg_cost_per_task_usd": 0.06822796858638744,
      "avg_duration_per_task_ms": 132197.8062827225,
      "avg_output_tokens_per_task": 18049.068062827224,
      "last_run_timestamp": "2026-10-01T08:31:33.481Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-astra-pro-20260903",
      "model_slug": "openai/gpt-6-astra-pro",
      "model_name": "OpenAI: GPT-6 Astra Pro",
      "provider_name": "auto-routing",
      "accuracy": 0.954545,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.332555398989899,
      "avg_duration_per_task_ms": 75446.5505050505,
      "avg_output_tokens_per_task": 4530.388888888889,
      "last_run_timestamp": "2026-10-01T08:27:35.414Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.8-flash-20260902",
      "model_slug": "google/gemini-3.8-flash",
      "model_name": "Google: Gemini 3.8 Flash",
      "provider_name": "Google AI Studio",
      "accuracy": 0.952862,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.07423223737373738,
      "avg_duration_per_task_ms": 86748.28787878787,
      "avg_output_tokens_per_task": 19646.409090909092,
      "last_run_timestamp": "2026-10-01T08:29:11.347Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.1-pro-preview-20260219",
      "model_slug": "google/gemini-3.1-pro-preview",
      "model_name": "Google: Gemini 3.1 Pro Preview",
      "provider_name": "Google",
      "accuracy": 0.95202,
      "accuracy_std_dev": 0.0008420000000000094,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.20364760606060606,
      "avg_duration_per_task_ms": 135706.18686868687,
      "avg_output_tokens_per_task": 16847.588383838385,
      "last_run_timestamp": "2026-10-01T19:14:26.157Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-astra-pro-20260903",
      "model_slug": "openai/gpt-6-astra-pro",
      "model_name": "OpenAI: GPT-6 Astra Pro",
      "provider_name": "OpenAI",
      "accuracy": 0.9511785,
      "accuracy_std_dev": 0.0016835000000000044,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.33620659090909094,
      "avg_duration_per_task_ms": 70544.52272727272,
      "avg_output_tokens_per_task": 4588.830808080808,
      "last_run_timestamp": "2026-10-02T23:23:01.262Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.9511779999999999,
      "accuracy_std_dev": 0.003367000000000009,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.20247136363636364,
      "avg_duration_per_task_ms": 137578.30555555556,
      "avg_output_tokens_per_task": 16748.56818181818,
      "last_run_timestamp": "2026-10-01T19:12:42.155Z"
    },
    {
      "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-3.8-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": 0,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.021484267676767678,
      "avg_duration_per_task_ms": 55623.856060606064,
      "avg_output_tokens_per_task": 1890.328282828283,
      "last_run_timestamp": "2026-10-30T00:03:33.414Z"
    },
    {
      "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": "openai/gpt-6-astra-20260903",
      "model_slug": "openai/gpt-6-astra",
      "model_name": "OpenAI: GPT-6 Astra",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.9461276666666666,
      "accuracy_std_dev": 0.001374572014200141,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.1181814292929293,
      "avg_duration_per_task_ms": 36670.56565656565,
      "avg_output_tokens_per_task": 2059.811447811448,
      "last_run_timestamp": "2026-10-30T00:02:55.917Z"
    },
    {
      "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": "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": 0.941077,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.15125092424242426,
      "avg_duration_per_task_ms": 97603.77272727272,
      "avg_output_tokens_per_task": 11559.585858585859,
      "last_run_timestamp": "2026-10-01T16:10:33.208Z"
    },
    {
      "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": 0.939394,
      "accuracy_std_dev": 0.0033669999999999534,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.31252995454545457,
      "avg_duration_per_task_ms": 150832.55050505052,
      "avg_output_tokens_per_task": 10288.712121212122,
      "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-5.6-sol-pro-20260709",
      "model_slug": "openai/gpt-5.6-sol-pro",
      "model_name": "OpenAI: GPT-5.6 Sol Pro",
      "provider_name": "Azure",
      "accuracy": 0.939394,
      "accuracy_std_dev": 0.0033669999999999534,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.32760574494949496,
      "avg_duration_per_task_ms": 78455.45202020202,
      "avg_output_tokens_per_task": 11977.780303030304,
      "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": 0.9393935,
      "accuracy_std_dev": 0.005050499999999958,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.15067959090909092,
      "avg_duration_per_task_ms": 95544.60353535354,
      "avg_output_tokens_per_task": 11515.699494949495,
      "last_run_timestamp": "2026-10-02T23:21:48.863Z"
    },
    {
      "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": "OpenAI",
      "accuracy": 0.9360266666666668,
      "accuracy_std_dev": 0.0036369299629708854,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.3141816919191919,
      "avg_duration_per_task_ms": 142474.1195286195,
      "avg_output_tokens_per_task": 10342.412457912458,
      "last_run_timestamp": "2026-10-02T23:25:41.614Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "Io Net",
      "accuracy": 0.9353121872146118,
      "accuracy_std_dev": 0.0026674999999999893,
      "run_count": 2,
      "total_questions": 219,
      "avg_cost_per_task_usd": 0.014858730593607304,
      "avg_duration_per_task_ms": 395134.2511415525,
      "avg_output_tokens_per_task": 29583.762557077625,
      "last_run_timestamp": "2026-10-24T08:19:41.879Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.8-flash",
      "model_slug": "google/gemini-3.8-flash",
      "model_name": "Gemini 3.8 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.9343430000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.023175040404040403,
      "avg_duration_per_task_ms": 44381.02525252525,
      "avg_output_tokens_per_task": 6131.126262626262,
      "last_run_timestamp": "2026-10-23T07:37:43.257Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.5-20260423",
      "model_slug": "openai/gpt-5.5",
      "model_name": "OpenAI: GPT-5.5",
      "provider_name": "Azure",
      "accuracy": 0.9336215714285715,
      "accuracy_std_dev": 0.008810277924222438,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.3281882835497835,
      "avg_duration_per_task_ms": 164668.6948051948,
      "avg_output_tokens_per_task": 10091.254689754689,
      "last_run_timestamp": "2026-10-01T19:14:07.472Z"
    },
    {
      "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": "google/gemini-3.6-flash-20260721",
      "model_slug": "google/gemini-3.6-flash",
      "model_name": "Google: Gemini 3.6 Flash",
      "provider_name": "Google",
      "accuracy": 0.9309762,
      "accuracy_std_dev": 0.005106374267520954,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.057293524242424246,
      "avg_duration_per_task_ms": 66775.28888888888,
      "avg_output_tokens_per_task": 11798.8,
      "last_run_timestamp": "2026-10-01T19:12:42.153Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-sonnet-5.5-20260928",
      "model_slug": "anthropic/claude-sonnet-5.5",
      "model_name": "Anthropic: Claude Sonnet 5.5",
      "provider_name": "Google",
      "accuracy": 0.9308640666666669,
      "accuracy_std_dev": 0.007021551143602255,
      "run_count": 15,
      "total_questions": 2970,
      "avg_cost_per_task_usd": 0.026242806060606066,
      "avg_duration_per_task_ms": 18771.915488215487,
      "avg_output_tokens_per_task": 2231.500336700337,
      "last_run_timestamp": "2026-10-24T08:30:19.120Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-sonnet-5.5-20260928",
      "model_slug": "anthropic/claude-sonnet-5.5",
      "model_name": "Anthropic: Claude Sonnet 5.5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.9306396,
      "accuracy_std_dev": 0.006683344480123684,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.025279440404040406,
      "avg_duration_per_task_ms": 20334.968686868688,
      "avg_output_tokens_per_task": 2296.2020202020203,
      "last_run_timestamp": "2026-10-24T08:11:43.558Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-sonnet-5.5-20260928",
      "model_slug": "anthropic/claude-sonnet-5.5",
      "model_name": "Anthropic: Claude Sonnet 5.5",
      "provider_name": "Anthropic",
      "accuracy": 0.9303028,
      "accuracy_std_dev": 0.00377921650081072,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.02488049898989899,
      "avg_duration_per_task_ms": 19525.145454545454,
      "avg_output_tokens_per_task": 2247.960606060606,
      "last_run_timestamp": "2026-10-24T08:10:23.014Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.9301345,
      "accuracy_std_dev": 0.002525500000000014,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.044492378787878796,
      "avg_duration_per_task_ms": 69170.04040404041,
      "avg_output_tokens_per_task": 11715.780303030304,
      "last_run_timestamp": "2026-10-01T19:12:42.162Z"
    },
    {
      "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": 0.928451,
      "accuracy_std_dev": 0.0025249999999999995,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.3476244823232323,
      "avg_duration_per_task_ms": 112460.52777777778,
      "avg_output_tokens_per_task": 10405.60101010101,
      "last_run_timestamp": "2026-10-01T19:13:06.559Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "x-ai/grok-4.6-20260810",
      "model_slug": "x-ai/grok-4.6",
      "model_name": "SpaceXAI: Grok 4.6",
      "provider_name": "auto-routing",
      "accuracy": 0.928451,
      "accuracy_std_dev": 0.004209000000000018,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.12326839393939394,
      "avg_duration_per_task_ms": 361758.7878787879,
      "avg_output_tokens_per_task": 20322.222222222223,
      "last_run_timestamp": "2026-10-01T16:09:08.142Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Parasail",
      "accuracy": 0.9280388242811503,
      "accuracy_std_dev": 0.010739874350082303,
      "run_count": 12,
      "total_questions": 2191,
      "avg_cost_per_task_usd": 0.1293467288909174,
      "avg_duration_per_task_ms": 132537.31081697854,
      "avg_output_tokens_per_task": 8542.681424007302,
      "last_run_timestamp": "2026-10-24T08:42:17.989Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.9276095000000001,
      "accuracy_std_dev": 0.005050500000000013,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.04436604545454545,
      "avg_duration_per_task_ms": 61670.28282828283,
      "avg_output_tokens_per_task": 11683.290404040405,
      "last_run_timestamp": "2026-10-01T19:11:04.522Z"
    },
    {
      "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": 0.9276091428571428,
      "accuracy_std_dev": 0.005010400338398178,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.10054259307359308,
      "avg_duration_per_task_ms": 55310.95093795094,
      "avg_output_tokens_per_task": 3496.1031746031745,
      "last_run_timestamp": "2026-10-01T19:42:48.407Z"
    },
    {
      "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": 0.9271917641601563,
      "accuracy_std_dev": 0.007748903788582068,
      "run_count": 33,
      "total_questions": 6144,
      "avg_cost_per_task_usd": 0.17010933951822915,
      "avg_duration_per_task_ms": 162569.57161458334,
      "avg_output_tokens_per_task": 9313.541341145834,
      "last_run_timestamp": "2026-10-24T08:50:28.834Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-sonnet-5.5-20260928",
      "model_slug": "anthropic/claude-sonnet-5.5",
      "model_name": "Anthropic: Claude Sonnet 5.5",
      "provider_name": "auto-routing",
      "accuracy": 0.9271882499999999,
      "accuracy_std_dev": 0.006560703749408305,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.025096623737373737,
      "avg_duration_per_task_ms": 19787.91414141414,
      "avg_output_tokens_per_task": 2272.02904040404,
      "last_run_timestamp": "2026-10-24T08:11:40.974Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-sol-20260709",
      "model_slug": "openai/gpt-5.6-sol",
      "model_name": "OpenAI: GPT-5.6 Sol",
      "provider_name": "OpenAI",
      "accuracy": 0.9264870000000001,
      "accuracy_std_dev": 0.007815979699734826,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.03737200673400674,
      "avg_duration_per_task_ms": 60770.02693602694,
      "avg_output_tokens_per_task": 3580.875420875421,
      "last_run_timestamp": "2026-10-02T23:20:36.509Z"
    },
    {
      "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": 0.9261662857142858,
      "accuracy_std_dev": 0.0030420094728079157,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.026267576479076477,
      "avg_duration_per_task_ms": 19507.747474747473,
      "avg_output_tokens_per_task": 2314.836940836941,
      "last_run_timestamp": "2026-10-29T00:04:34.815Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Decart",
      "accuracy": 0.9259260000000001,
      "accuracy_std_dev": 0.004610436378478734,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.10788488484848485,
      "avg_duration_per_task_ms": 187765.80707070706,
      "avg_output_tokens_per_task": 9499.632323232323,
      "last_run_timestamp": "2026-10-24T08:37:09.424Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Relace",
      "accuracy": 0.925795996178344,
      "accuracy_std_dev": 0.007145577022492671,
      "run_count": 8,
      "total_questions": 1570,
      "avg_cost_per_task_usd": 0.1095707592356688,
      "avg_duration_per_task_ms": 200872.2719745223,
      "avg_output_tokens_per_task": 9447.89617834395,
      "last_run_timestamp": "2026-10-24T08:14:36.241Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": "Darkbloom",
      "accuracy": 0.925214,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 78,
      "avg_cost_per_task_usd": 0.0012695128205128204,
      "avg_duration_per_task_ms": 237789.78205128206,
      "avg_output_tokens_per_task": 6844.423076923077,
      "last_run_timestamp": "2026-10-30T01:38:27.650Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "DeepInfra",
      "accuracy": 0.9247310806451613,
      "accuracy_std_dev": 0.013119500000000006,
      "run_count": 2,
      "total_questions": 372,
      "avg_cost_per_task_usd": 0.005701161290322581,
      "avg_duration_per_task_ms": 222245.45967741936,
      "avg_output_tokens_per_task": 19159.182795698925,
      "last_run_timestamp": "2026-10-24T08:13:25.170Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "x-ai/grok-4.6-20260810",
      "model_slug": "x-ai/grok-4.6",
      "model_name": "SpaceXAI: Grok 4.6",
      "provider_name": "xAI",
      "accuracy": 0.9246375695652174,
      "accuracy_std_dev": 0.006863146440316182,
      "run_count": 7,
      "total_questions": 1380,
      "avg_cost_per_task_usd": 0.12292555942028985,
      "avg_duration_per_task_ms": 340480.1579710145,
      "avg_output_tokens_per_task": 19662.684057971015,
      "last_run_timestamp": "2026-10-01T16:09:31.184Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-sol-20260922",
      "model_slug": "openai/gpt-6-sol",
      "model_name": "OpenAI: GPT-6 Sol",
      "provider_name": "Azure",
      "accuracy": 0.924599700084246,
      "accuracy_std_dev": 0.006819818440072705,
      "run_count": 6,
      "total_questions": 1187,
      "avg_cost_per_task_usd": 0.02751449789385004,
      "avg_duration_per_task_ms": 24957.139848357205,
      "avg_output_tokens_per_task": 2420.7127211457455,
      "last_run_timestamp": "2026-10-01T08:24:52.927Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.5-flash-20260519",
      "model_slug": "google/gemini-3.5-flash",
      "model_name": "Google: Gemini 3.5 Flash",
      "provider_name": "Google",
      "accuracy": 0.9245788000000001,
      "accuracy_std_dev": 0.005259575777569916,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.1499325777777778,
      "avg_duration_per_task_ms": 85124.36363636363,
      "avg_output_tokens_per_task": 15611.38383838384,
      "last_run_timestamp": "2026-10-01T19:13:26.156Z"
    },
    {
      "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": 0.9242420000000001,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.03863523484848485,
      "avg_duration_per_task_ms": 63161.229797979795,
      "avg_output_tokens_per_task": 3708.2954545454545,
      "last_run_timestamp": "2026-10-01T19:12:28.573Z"
    },
    {
      "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": "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": 0.9234005000000001,
      "accuracy_std_dev": 0.000841499999999995,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.1436656515151515,
      "avg_duration_per_task_ms": 89621.648989899,
      "avg_output_tokens_per_task": 15839.800505050505,
      "last_run_timestamp": "2026-10-01T19:18:40.706Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "InferenceNet",
      "accuracy": 0.9230042575935954,
      "accuracy_std_dev": 0.028834365810508387,
      "run_count": 22,
      "total_questions": 4247,
      "avg_cost_per_task_usd": 0.12647735083588418,
      "avg_duration_per_task_ms": 173368.8584883447,
      "avg_output_tokens_per_task": 9831.030138921591,
      "last_run_timestamp": "2026-10-24T08:45:15.682Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.9225589999999999,
      "accuracy_std_dev": 0.003367000000000009,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.14588472727272728,
      "avg_duration_per_task_ms": 63933.833333333336,
      "avg_output_tokens_per_task": 16085.368686868687,
      "last_run_timestamp": "2026-10-01T19:12:44.398Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Makora",
      "accuracy": 0.9220161548643283,
      "accuracy_std_dev": 0.015489389982329555,
      "run_count": 9,
      "total_questions": 1511,
      "avg_cost_per_task_usd": 0.1115360423560556,
      "avg_duration_per_task_ms": 205913.44606221045,
      "avg_output_tokens_per_task": 8295.92322964924,
      "last_run_timestamp": "2026-10-24T08:11:11.433Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.9219976666666666,
      "accuracy_std_dev": 0.0028614651647170097,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.08309612457912459,
      "avg_duration_per_task_ms": 49364.952861952865,
      "avg_output_tokens_per_task": 3748.658249158249,
      "last_run_timestamp": "2026-10-30T00:00:06.279Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Sail Research",
      "accuracy": 0.9214363999999999,
      "accuracy_std_dev": 0.007595474019001951,
      "run_count": 15,
      "total_questions": 2970,
      "avg_cost_per_task_usd": 0.13534853400673405,
      "avg_duration_per_task_ms": 180051.0377104377,
      "avg_output_tokens_per_task": 11618.102693602694,
      "last_run_timestamp": "2026-10-24T09:13:18.915Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Modal",
      "accuracy": 0.9214090472319009,
      "accuracy_std_dev": 0.0068651349363197224,
      "run_count": 14,
      "total_questions": 2583,
      "avg_cost_per_task_usd": 0.14661487727448702,
      "avg_duration_per_task_ms": 125761.86372435153,
      "avg_output_tokens_per_task": 9628.765001935733,
      "last_run_timestamp": "2026-10-24T08:48:19.650Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.92087525,
      "accuracy_std_dev": 0.004610545920766882,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.14864516792929292,
      "avg_duration_per_task_ms": 124980.99747474748,
      "avg_output_tokens_per_task": 9671.253787878788,
      "last_run_timestamp": "2026-10-24T08:58:50.031Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-sol-20260922",
      "model_slug": "openai/gpt-6-sol",
      "model_name": "OpenAI: GPT-6 Sol",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.9208750000000001,
      "accuracy_std_dev": 0.004761657064510223,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.026268675084175085,
      "avg_duration_per_task_ms": 25466.75420875421,
      "avg_output_tokens_per_task": 2308.39898989899,
      "last_run_timestamp": "2026-10-30T00:00:05.373Z"
    },
    {
      "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": "ParetoAlpha",
      "accuracy": 0.9203848188976379,
      "accuracy_std_dev": 0.004243859645122607,
      "run_count": 4,
      "total_questions": 762,
      "avg_cost_per_task_usd": 0.15456283070866142,
      "avg_duration_per_task_ms": 390118.6771653543,
      "avg_output_tokens_per_task": 10091.72965879265,
      "last_run_timestamp": "2026-10-30T06:35:32.515Z"
    },
    {
      "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": 0.9203701999999999,
      "accuracy_std_dev": 0.006736210266908262,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.17098447676767675,
      "avg_duration_per_task_ms": 297723.4101010101,
      "avg_output_tokens_per_task": 9720.665151515152,
      "last_run_timestamp": "2026-10-28T03:40:38.193Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "DigitalOcean",
      "accuracy": 0.9200335,
      "accuracy_std_dev": 0.004208500000000004,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.021749656565656566,
      "avg_duration_per_task_ms": 75214.94191919192,
      "avg_output_tokens_per_task": 20570.64898989899,
      "last_run_timestamp": "2026-10-24T08:29:16.810Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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-6.1-terra-20261026",
      "model_slug": "openai/gpt-6.1-terra",
      "model_name": "OpenAI: GPT-6.1 Terra",
      "provider_name": "Azure",
      "accuracy": 0.9191919999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.01357689898989899,
      "avg_duration_per_task_ms": 18304.29292929293,
      "avg_output_tokens_per_task": 2104.2979797979797,
      "last_run_timestamp": "2026-10-28T00:00:04.630Z"
    },
    {
      "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": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "BaseTen",
      "accuracy": 0.9182648918918918,
      "accuracy_std_dev": 0.012947711305692437,
      "run_count": 12,
      "total_questions": 2294,
      "avg_cost_per_task_usd": 0.1290118766346992,
      "avg_duration_per_task_ms": 117697.42414995641,
      "avg_output_tokens_per_task": 8571.878378378378,
      "last_run_timestamp": "2026-10-24T08:44:47.181Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "auto-routing",
      "accuracy": 0.9181606468965516,
      "accuracy_std_dev": 0.010622287016200703,
      "run_count": 11,
      "total_questions": 2175,
      "avg_cost_per_task_usd": 0.13311453471264367,
      "avg_duration_per_task_ms": 278187.9034482759,
      "avg_output_tokens_per_task": 10469.161379310344,
      "last_run_timestamp": "2026-10-24T08:16:43.073Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "Venice",
      "accuracy": 0.9180774433042584,
      "accuracy_std_dev": 0.01435885589070384,
      "run_count": 10,
      "total_questions": 1949,
      "avg_cost_per_task_usd": 0.24670780708055412,
      "avg_duration_per_task_ms": 414365.66136480245,
      "avg_output_tokens_per_task": 36895.286813750645,
      "last_run_timestamp": "2026-10-24T08:18:32.510Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-sonnet-5.5-20260928",
      "model_slug": "anthropic/claude-sonnet-5.5",
      "model_name": "Anthropic: Claude Sonnet 5.5",
      "provider_name": "Azure",
      "accuracy": 0.9171716,
      "accuracy_std_dev": 0.0037495390703391955,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.02583652727272727,
      "avg_duration_per_task_ms": 19190.16717171717,
      "avg_output_tokens_per_task": 2225.3469696969696,
      "last_run_timestamp": "2026-10-24T08:28:10.076Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Moonshot AI",
      "accuracy": 0.9165345411718132,
      "accuracy_std_dev": 0.00728827618680496,
      "run_count": 13,
      "total_questions": 2526,
      "avg_cost_per_task_usd": 0.14153307442596988,
      "avg_duration_per_task_ms": 283509.3226444972,
      "avg_output_tokens_per_task": 9267.701108471892,
      "last_run_timestamp": "2026-10-24T08:44:42.687Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Cohere",
      "accuracy": 0.915825,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.014131747474747476,
      "avg_duration_per_task_ms": 59960.4595959596,
      "avg_output_tokens_per_task": 21225.651515151516,
      "last_run_timestamp": "2026-10-28T21:00:06.478Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "SiliconFlow",
      "accuracy": 0.9158245,
      "accuracy_std_dev": 0.011784500000000031,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.026372479797979794,
      "avg_duration_per_task_ms": 193752.02525252526,
      "avg_output_tokens_per_task": 21866.449494949495,
      "last_run_timestamp": "2026-10-24T08:46:05.254Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Baidu",
      "accuracy": 0.9152636666666667,
      "accuracy_std_dev": 0.006963814615568124,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.09072644781144781,
      "avg_duration_per_task_ms": 657647.413580247,
      "avg_output_tokens_per_task": 40463.242985409655,
      "last_run_timestamp": "2026-10-24T08:13:28.067Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "AkashML",
      "accuracy": 0.9150769353846153,
      "accuracy_std_dev": 0.01273628821599266,
      "run_count": 12,
      "total_questions": 1625,
      "avg_cost_per_task_usd": 0.11317312553846154,
      "avg_duration_per_task_ms": 182427.93907692307,
      "avg_output_tokens_per_task": 7798.748307692308,
      "last_run_timestamp": "2026-10-30T00:27:14.164Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Baidu",
      "accuracy": 0.9147832881355932,
      "accuracy_std_dev": 0.01218950000000002,
      "run_count": 2,
      "total_questions": 354,
      "avg_cost_per_task_usd": 0.02172540395480226,
      "avg_duration_per_task_ms": 124806.83333333333,
      "avg_output_tokens_per_task": 19165.833333333332,
      "last_run_timestamp": "2026-10-24T08:44:43.907Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Wafer",
      "accuracy": 0.9144541630847027,
      "accuracy_std_dev": 0.007898223803049383,
      "run_count": 12,
      "total_questions": 2373,
      "avg_cost_per_task_usd": 0.09740494732406238,
      "avg_duration_per_task_ms": 296054.32490518334,
      "avg_output_tokens_per_task": 25118.074589127686,
      "last_run_timestamp": "2026-10-29T22:45:19.967Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Mancer 2",
      "accuracy": 0.9142155073529412,
      "accuracy_std_dev": 0.012753000000000014,
      "run_count": 2,
      "total_questions": 272,
      "avg_cost_per_task_usd": 0.014779014705882354,
      "avg_duration_per_task_ms": 301053.6323529412,
      "avg_output_tokens_per_task": 14660.216911764706,
      "last_run_timestamp": "2026-10-24T08:32:55.335Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6.1-terra",
      "model_slug": "openai/gpt-6.1-terra",
      "model_name": "GPT-6.1-Terra",
      "provider_name": "auto-routing",
      "accuracy": 0.9141410000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0075055656565656565,
      "avg_duration_per_task_ms": 21128.91414141414,
      "avg_output_tokens_per_task": 1395.4343434343434,
      "last_run_timestamp": "2026-10-27T03:32:35.022Z"
    },
    {
      "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": 0.91386075,
      "accuracy_std_dev": 0.013535011243222777,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.034369089225589226,
      "avg_duration_per_task_ms": 228439.68476430976,
      "avg_output_tokens_per_task": 28471.782828282827,
      "last_run_timestamp": "2026-10-24T16:21:24.997Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Morph",
      "accuracy": 0.913200685767098,
      "accuracy_std_dev": 0.02916766954216023,
      "run_count": 22,
      "total_questions": 4328,
      "avg_cost_per_task_usd": 0.1485330272643253,
      "avg_duration_per_task_ms": 403035.12731053605,
      "avg_output_tokens_per_task": 9145.585720887246,
      "last_run_timestamp": "2026-10-24T08:17:56.389Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "AtlasCloud",
      "accuracy": 0.9131792857142857,
      "accuracy_std_dev": 0.008577350568532835,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.15955438744588746,
      "avg_duration_per_task_ms": 565778.9992784993,
      "avg_output_tokens_per_task": 39975.08946608946,
      "last_run_timestamp": "2026-10-24T08:28:29.525Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Alibaba",
      "accuracy": 0.913019,
      "accuracy_std_dev": 0.006496204173720757,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.027868947811447813,
      "avg_duration_per_task_ms": 190559.73569023568,
      "avg_output_tokens_per_task": 21553.755892255893,
      "last_run_timestamp": "2026-10-28T03:31:51.626Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "GMICloud",
      "accuracy": 0.9118966666666666,
      "accuracy_std_dev": 0.0028614651647170097,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.020073099326599324,
      "avg_duration_per_task_ms": 146602.46801346802,
      "avg_output_tokens_per_task": 22818.382154882154,
      "last_run_timestamp": "2026-10-27T00:36:02.541Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Chutes",
      "accuracy": 0.9118396713596139,
      "accuracy_std_dev": 0.015391401769750214,
      "run_count": 13,
      "total_questions": 2486,
      "avg_cost_per_task_usd": 0.1705156729686243,
      "avg_duration_per_task_ms": 504972.3576025744,
      "avg_output_tokens_per_task": 11645.380128720837,
      "last_run_timestamp": "2026-10-24T08:25:10.474Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Decart",
      "accuracy": 0.911616,
      "accuracy_std_dev": 0.0025249999999999995,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.007887881313131314,
      "avg_duration_per_task_ms": 185118.85353535353,
      "avg_output_tokens_per_task": 21770.37626262626,
      "last_run_timestamp": "2026-10-24T08:29:45.567Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Alibaba",
      "accuracy": 0.9115664705882354,
      "accuracy_std_dev": 0.014309678991415931,
      "run_count": 17,
      "total_questions": 3366,
      "avg_cost_per_task_usd": 0.13998793790849673,
      "avg_duration_per_task_ms": 602578.7964943553,
      "avg_output_tokens_per_task": 39991.1972667855,
      "last_run_timestamp": "2026-10-28T04:18:14.838Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "DekaLLM",
      "accuracy": 0.911111096,
      "accuracy_std_dev": 0.009498497417355022,
      "run_count": 4,
      "total_questions": 750,
      "avg_cost_per_task_usd": 0.136844476,
      "avg_duration_per_task_ms": 321154.29333333333,
      "avg_output_tokens_per_task": 35843.412,
      "last_run_timestamp": "2026-10-24T08:11:11.441Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6.1-terra-20261026",
      "model_slug": "openai/gpt-6.1-terra",
      "model_name": "OpenAI: GPT-6.1 Terra",
      "provider_name": "OpenAI",
      "accuracy": 0.910774,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.011462444444444444,
      "avg_duration_per_task_ms": 27557.444444444445,
      "avg_output_tokens_per_task": 2133.969696969697,
      "last_run_timestamp": "2026-10-28T00:00:04.598Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "AtlasCloud",
      "accuracy": 0.9099324999999999,
      "accuracy_std_dev": 0.004208500000000004,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.02111154797979798,
      "avg_duration_per_task_ms": 145444.96212121213,
      "avg_output_tokens_per_task": 21297.967171717173,
      "last_run_timestamp": "2026-10-24T08:04:44.510Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Fireworks",
      "accuracy": 0.9097220343749999,
      "accuracy_std_dev": 0.007467330848435752,
      "run_count": 5,
      "total_questions": 960,
      "avg_cost_per_task_usd": 0.023086179166666665,
      "avg_duration_per_task_ms": 81909.69166666667,
      "avg_output_tokens_per_task": 20192.428125,
      "last_run_timestamp": "2026-10-28T00:00:04.630Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Nebius",
      "accuracy": 0.909652,
      "accuracy_std_dev": 0.0028613344206273234,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.025303424242424242,
      "avg_duration_per_task_ms": 91659.5925925926,
      "avg_output_tokens_per_task": 20911.968013468013,
      "last_run_timestamp": "2026-10-24T08:24:39.071Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-luna-pro-20260709",
      "model_slug": "openai/gpt-5.6-luna-pro",
      "model_name": "OpenAI: GPT-5.6 Luna Pro",
      "provider_name": "Azure",
      "accuracy": 0.9095115,
      "accuracy_std_dev": 0.002759842885020819,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.0416229191919192,
      "avg_duration_per_task_ms": 138307.48484848486,
      "avg_output_tokens_per_task": 27969.503787878788,
      "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-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Particle AI",
      "accuracy": 0.9090910000000001,
      "accuracy_std_dev": 0.006734000000000018,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.017200570707070706,
      "avg_duration_per_task_ms": 67652.35606060606,
      "avg_output_tokens_per_task": 17022.85606060606,
      "last_run_timestamp": "2026-10-28T22:43:29.733Z"
    },
    {
      "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": "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": 0.9090908333333333,
      "accuracy_std_dev": 0.008189992133017543,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.03826549410774411,
      "avg_duration_per_task_ms": 134585.28703703705,
      "avg_output_tokens_per_task": 28750.484848484848,
      "last_run_timestamp": "2026-10-27T00:36:05.210Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "CoreWeave",
      "accuracy": 0.908733465037911,
      "accuracy_std_dev": 0.006850022329140406,
      "run_count": 12,
      "total_questions": 2374,
      "avg_cost_per_task_usd": 0.030451491996630154,
      "avg_duration_per_task_ms": 408677.97219882056,
      "avg_output_tokens_per_task": 31497.24094355518,
      "last_run_timestamp": "2026-10-24T16:30:24.394Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Phala",
      "accuracy": 0.9084910475789473,
      "accuracy_std_dev": 0.04945617807940908,
      "run_count": 12,
      "total_questions": 2375,
      "avg_cost_per_task_usd": 0.14244542484210523,
      "avg_duration_per_task_ms": 331138.315368421,
      "avg_output_tokens_per_task": 10400.368421052632,
      "last_run_timestamp": "2026-10-24T08:48:54.139Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Together",
      "accuracy": 0.9080194545454545,
      "accuracy_std_dev": 0.015874287061234574,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.035897662534435264,
      "avg_duration_per_task_ms": 311230.74242424243,
      "avg_output_tokens_per_task": 29833.370982552802,
      "last_run_timestamp": "2026-10-24T16:24:24.501Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "DeepInfra",
      "accuracy": 0.9077097088435375,
      "accuracy_std_dev": 0.01602150486845725,
      "run_count": 4,
      "total_questions": 735,
      "avg_cost_per_task_usd": 0.12614392789115647,
      "avg_duration_per_task_ms": 1486736.511564626,
      "avg_output_tokens_per_task": 50215.49115646259,
      "last_run_timestamp": "2026-10-24T08:43:13.447Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "InferenceNet",
      "accuracy": 0.9074072500000001,
      "accuracy_std_dev": 0.008828586845441347,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.009289530303030303,
      "avg_duration_per_task_ms": 172567.80050505052,
      "avg_output_tokens_per_task": 18481.776515151516,
      "last_run_timestamp": "2026-10-24T08:44:46.341Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Venice",
      "accuracy": 0.907407,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.025121388888888888,
      "avg_duration_per_task_ms": 92618.0808080808,
      "avg_output_tokens_per_task": 20743.772727272728,
      "last_run_timestamp": "2026-10-29T00:00:06.218Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Ionstream",
      "accuracy": 0.907407,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.018446676767676768,
      "avg_duration_per_task_ms": 140631.38888888888,
      "avg_output_tokens_per_task": 20189.31818181818,
      "last_run_timestamp": "2026-10-24T08:54:33.388Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Relace",
      "accuracy": 0.907407,
      "accuracy_std_dev": 0.003367000000000009,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.012070199494949497,
      "avg_duration_per_task_ms": 98657.74747474748,
      "avg_output_tokens_per_task": 21841.54797979798,
      "last_run_timestamp": "2026-10-24T08:21:56.264Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "auto-routing",
      "accuracy": 0.907407,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.010421979797979798,
      "avg_duration_per_task_ms": 107076.06565656565,
      "avg_output_tokens_per_task": 18400.97474747475,
      "last_run_timestamp": "2026-10-24T08:14:51.110Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "GMICloud",
      "accuracy": 0.9070186923076924,
      "accuracy_std_dev": 0.024866682548789026,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.03558504700854701,
      "avg_duration_per_task_ms": 231138.5062160062,
      "avg_output_tokens_per_task": 29778.67948717949,
      "last_run_timestamp": "2026-10-27T00:51:48.120Z"
    },
    {
      "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": 0.9067937377877597,
      "accuracy_std_dev": 0.03179893182614709,
      "run_count": 9,
      "total_questions": 1781,
      "avg_cost_per_task_usd": 0.01609888826501965,
      "avg_duration_per_task_ms": 58211.32790567097,
      "avg_output_tokens_per_task": 3297.93037619315,
      "last_run_timestamp": "2026-10-16T18:32:13.741Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "SiliconFlow",
      "accuracy": 0.9067057499999999,
      "accuracy_std_dev": 0.02181467158101399,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.15476828703703704,
      "avg_duration_per_task_ms": 601020.0946969697,
      "avg_output_tokens_per_task": 39038.01388888889,
      "last_run_timestamp": "2026-10-24T08:50:20.872Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Phala",
      "accuracy": 0.9066696814159291,
      "accuracy_std_dev": 0.02221528527057198,
      "run_count": 11,
      "total_questions": 2034,
      "avg_cost_per_task_usd": 0.150266023107178,
      "avg_duration_per_task_ms": 558942.4459193707,
      "avg_output_tokens_per_task": 37986.727138643066,
      "last_run_timestamp": "2026-10-24T09:27:08.802Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Alibaba",
      "accuracy": 0.9065654999999999,
      "accuracy_std_dev": 0.00811767264984244,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.03409049431818181,
      "avg_duration_per_task_ms": 218247.69949494948,
      "avg_output_tokens_per_task": 27597.452651515152,
      "last_run_timestamp": "2026-10-28T03:27:34.225Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Sail Research",
      "accuracy": 0.9065654999999999,
      "accuracy_std_dev": 0.008117672649842422,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.0029748573232323233,
      "avg_duration_per_task_ms": 218585.40025252526,
      "avg_output_tokens_per_task": 19647.71212121212,
      "last_run_timestamp": "2026-10-24T08:58:07.435Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Novita",
      "accuracy": 0.9064251666666665,
      "accuracy_std_dev": 0.020964442642377977,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.03777529545454546,
      "avg_duration_per_task_ms": 260085.33417508416,
      "avg_output_tokens_per_task": 31644.78872053872,
      "last_run_timestamp": "2026-10-24T16:42:35.935Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "BaseTen",
      "accuracy": 0.9063117514792899,
      "accuracy_std_dev": 0.0007095000000000296,
      "run_count": 2,
      "total_questions": 338,
      "avg_cost_per_task_usd": 0.0210026775147929,
      "avg_duration_per_task_ms": 63992.53550295858,
      "avg_output_tokens_per_task": 17351.381656804733,
      "last_run_timestamp": "2026-10-24T08:13:28.071Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Baidu",
      "accuracy": 0.9060604000000001,
      "accuracy_std_dev": 0.0032644069966840917,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.02883772424242424,
      "avg_duration_per_task_ms": 166818.5292929293,
      "avg_output_tokens_per_task": 28846.466666666667,
      "last_run_timestamp": "2026-10-24T08:33:28.371Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Wafer",
      "accuracy": 0.9060041666666666,
      "accuracy_std_dev": 0.005181136857764807,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.008250873737373737,
      "avg_duration_per_task_ms": 82060.31565656565,
      "avg_output_tokens_per_task": 16356.271043771043,
      "last_run_timestamp": "2026-10-29T22:25:59.214Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "StreamLake",
      "accuracy": 0.9059828461538462,
      "accuracy_std_dev": 0.02626261808627545,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.04094654972804972,
      "avg_duration_per_task_ms": 283710.56487956486,
      "avg_output_tokens_per_task": 31884.02874902875,
      "last_run_timestamp": "2026-10-24T16:17:39.031Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "StreamLake",
      "accuracy": 0.9058946753684209,
      "accuracy_std_dev": 0.029872349099497063,
      "run_count": 12,
      "total_questions": 2375,
      "avg_cost_per_task_usd": 0.10017258273684211,
      "avg_duration_per_task_ms": 566875.4896842105,
      "avg_output_tokens_per_task": 39376.88463157895,
      "last_run_timestamp": "2026-10-24T08:54:07.383Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "StreamLake",
      "accuracy": 0.905724,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.026723479797979795,
      "avg_duration_per_task_ms": 119393.66666666667,
      "avg_output_tokens_per_task": 22072.065656565657,
      "last_run_timestamp": "2026-10-26T03:00:10.427Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta/muse-spark-open-20261015",
      "model_slug": "meta/muse-spark-open",
      "model_name": "Meta: Muse Spark Open",
      "provider_name": "auto-routing",
      "accuracy": 0.905724,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.06900820707070707,
      "avg_duration_per_task_ms": 101485.72222222222,
      "avg_output_tokens_per_task": 16020.974747474747,
      "last_run_timestamp": "2026-10-24T08:27:04.527Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Relace",
      "accuracy": 0.9057143142857144,
      "accuracy_std_dev": 0.001963571554591265,
      "run_count": 4,
      "total_questions": 700,
      "avg_cost_per_task_usd": 0.18660921714285714,
      "avg_duration_per_task_ms": 435582.91714285716,
      "avg_output_tokens_per_task": 41735.86142857143,
      "last_run_timestamp": "2026-10-24T08:46:32.813Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta/muse-spark-open-20261015",
      "model_slug": "meta/muse-spark-open",
      "model_name": "Meta: Muse Spark Open",
      "provider_name": "Modal",
      "accuracy": 0.9056120000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 196,
      "avg_cost_per_task_usd": 0.06647105102040816,
      "avg_duration_per_task_ms": 64791.01530612245,
      "avg_output_tokens_per_task": 15327.612244897959,
      "last_run_timestamp": "2026-10-24T08:14:36.238Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "InferenceNet",
      "accuracy": 0.905376,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 155,
      "avg_cost_per_task_usd": 0.09219292258064517,
      "avg_duration_per_task_ms": 892432.9419354838,
      "avg_output_tokens_per_task": 36790.02580645162,
      "last_run_timestamp": "2026-10-02T06:44:44.722Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Together",
      "accuracy": 0.905323592920354,
      "accuracy_std_dev": 0.03949844256599099,
      "run_count": 12,
      "total_questions": 2373,
      "avg_cost_per_task_usd": 0.1419002410450906,
      "avg_duration_per_task_ms": 207365.2397808681,
      "avg_output_tokens_per_task": 9535.341761483354,
      "last_run_timestamp": "2026-10-24T08:46:33.301Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Makora",
      "accuracy": 0.9053030000000001,
      "accuracy_std_dev": 0.004187691070267732,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.026182512626262627,
      "avg_duration_per_task_ms": 138194.00252525252,
      "avg_output_tokens_per_task": 21714.06186868687,
      "last_run_timestamp": "2026-10-24T08:02:23.005Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Morph",
      "accuracy": 0.9051626666666666,
      "accuracy_std_dev": 0.0055553002519115975,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.009504934343434343,
      "avg_duration_per_task_ms": 118639.06902356903,
      "avg_output_tokens_per_task": 16467.956228956227,
      "last_run_timestamp": "2026-10-24T08:45:01.104Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-flash-20260826",
      "model_slug": "qwen/qwen3.8-flash",
      "model_name": "Qwen: Qwen3.8 Flash",
      "provider_name": "Alibaba",
      "accuracy": 0.9049724513274336,
      "accuracy_std_dev": 0.008906991898924113,
      "run_count": 4,
      "total_questions": 791,
      "avg_cost_per_task_usd": 0.020146505689001264,
      "avg_duration_per_task_ms": 644404.02528445,
      "avg_output_tokens_per_task": 42703.32616940582,
      "last_run_timestamp": "2026-10-28T03:59:43.480Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Modal",
      "accuracy": 0.904882,
      "accuracy_std_dev": 0.005892000000000008,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.022279229797979798,
      "avg_duration_per_task_ms": 122775.77777777778,
      "avg_output_tokens_per_task": 22121.87626262626,
      "last_run_timestamp": "2026-10-24T08:33:42.016Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "SiliconFlow",
      "accuracy": 0.9044612500000001,
      "accuracy_std_dev": 0.009172855468037184,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.03501683964646465,
      "avg_duration_per_task_ms": 199099.8143939394,
      "avg_output_tokens_per_task": 29016.732954545456,
      "last_run_timestamp": "2026-10-24T08:11:11.436Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "StreamLake",
      "accuracy": 0.904461125,
      "accuracy_std_dev": 0.008737580020198659,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.02118971275252525,
      "avg_duration_per_task_ms": 207950.51199494948,
      "avg_output_tokens_per_task": 28934.199494949495,
      "last_run_timestamp": "2026-10-24T08:11:11.332Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3-flash-preview-20251217",
      "model_slug": "google/gemini-3-flash-preview",
      "model_name": "Google: Gemini 3 Flash Preview",
      "provider_name": "Google",
      "accuracy": 0.9040405,
      "accuracy_std_dev": 0.011784499999999976,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.13174327272727274,
      "avg_duration_per_task_ms": 249918.9898989899,
      "avg_output_tokens_per_task": 43791.378787878784,
      "last_run_timestamp": "2026-10-01T19:18:35.588Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3.1-pro",
      "model_slug": "moonshotai/kimi-k3.1-pro",
      "model_name": "Kimi K3.1 Pro",
      "provider_name": "auto-routing",
      "accuracy": 0.9040400000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.021051848484848486,
      "avg_duration_per_task_ms": 58149.92929292929,
      "avg_output_tokens_per_task": 2035.7424242424242,
      "last_run_timestamp": "2026-10-26T19:26:14.323Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "Novita",
      "accuracy": 0.9040400000000001,
      "accuracy_std_dev": 0.004761657064510192,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.03280091414141414,
      "avg_duration_per_task_ms": 814621.6686868687,
      "avg_output_tokens_per_task": 37346.743434343436,
      "last_run_timestamp": "2026-10-24T08:23:31.992Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta/muse-spark-open-20261015",
      "model_slug": "meta/muse-spark-open",
      "model_name": "Meta: Muse Spark Open",
      "provider_name": "Makora",
      "accuracy": 0.9039929892328398,
      "accuracy_std_dev": 0.013615822071674532,
      "run_count": 4,
      "total_questions": 743,
      "avg_cost_per_task_usd": 0.061841156123822336,
      "avg_duration_per_task_ms": 74303.45625841184,
      "avg_output_tokens_per_task": 14413.555854643339,
      "last_run_timestamp": "2026-10-24T08:25:42.800Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Alibaba",
      "accuracy": 0.9036307954401521,
      "accuracy_std_dev": 0.011636066678482626,
      "run_count": 16,
      "total_questions": 3158,
      "avg_cost_per_task_usd": 0.027261241925269155,
      "avg_duration_per_task_ms": 186560.3065231159,
      "avg_output_tokens_per_task": 25898.152311589613,
      "last_run_timestamp": "2026-10-28T03:34:20.229Z"
    },
    {
      "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": 0.9036194166666667,
      "accuracy_std_dev": 0.01817259761499133,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.034217859006734005,
      "avg_duration_per_task_ms": 319774.7689393939,
      "avg_output_tokens_per_task": 30123.76936026936,
      "last_run_timestamp": "2026-10-24T16:12:49.069Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.90361925,
      "accuracy_std_dev": 0.005630970448111022,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.026822416666666668,
      "avg_duration_per_task_ms": 92441.06818181818,
      "avg_output_tokens_per_task": 22174.864898989897,
      "last_run_timestamp": "2026-10-29T23:05:04.190Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Venice",
      "accuracy": 0.9036160645842126,
      "accuracy_std_dev": 0.011349068793175144,
      "run_count": 12,
      "total_questions": 2369,
      "avg_cost_per_task_usd": 0.028033349514563106,
      "avg_duration_per_task_ms": 292905.91599831154,
      "avg_output_tokens_per_task": 23207.665681722247,
      "last_run_timestamp": "2026-10-24T16:03:25.397Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "x-ai/grok-4.6-20260810",
      "model_slug": "x-ai/grok-4.6",
      "model_name": "SpaceXAI: Grok 4.6",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.903508802631579,
      "accuracy_std_dev": 0.0984941779678147,
      "run_count": 3,
      "total_questions": 456,
      "avg_cost_per_task_usd": 0.11285829166666667,
      "avg_duration_per_task_ms": 423954.7916666667,
      "avg_output_tokens_per_task": 16846.184210526317,
      "last_run_timestamp": "2026-10-01T16:04:59.818Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "DeepInfra",
      "accuracy": 0.9033163440134907,
      "accuracy_std_dev": 0.008355089002931883,
      "run_count": 9,
      "total_questions": 1779,
      "avg_cost_per_task_usd": 0.015168129286115795,
      "avg_duration_per_task_ms": 392884.92523889826,
      "avg_output_tokens_per_task": 28918.40921866217,
      "last_run_timestamp": "2026-10-24T08:18:49.230Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.4-20260305",
      "model_slug": "openai/gpt-5.4",
      "model_name": "OpenAI: GPT-5.4",
      "provider_name": "OpenAI",
      "accuracy": 0.9031985,
      "accuracy_std_dev": 0.004208500000000004,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.14226843434343436,
      "avg_duration_per_task_ms": 149341.0782828283,
      "avg_output_tokens_per_task": 9356.287878787878,
      "last_run_timestamp": "2026-10-01T19:14:55.028Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "DeepInfra",
      "accuracy": 0.9031626720031235,
      "accuracy_std_dev": 0.019975453712141735,
      "run_count": 13,
      "total_questions": 2561,
      "avg_cost_per_task_usd": 0.11988065286997265,
      "avg_duration_per_task_ms": 343221.44474814524,
      "avg_output_tokens_per_task": 8205.648574775478,
      "last_run_timestamp": "2026-10-24T08:49:03.637Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Novita",
      "accuracy": 0.9026110979153505,
      "accuracy_std_dev": 0.006660054560587301,
      "run_count": 8,
      "total_questions": 1583,
      "avg_cost_per_task_usd": 0.028997156664560955,
      "avg_duration_per_task_ms": 168876.1749842072,
      "avg_output_tokens_per_task": 28895.867340492736,
      "last_run_timestamp": "2026-10-24T08:51:57.001Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "DeepSeek",
      "accuracy": 0.902357,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.007275146464646465,
      "avg_duration_per_task_ms": 217667.1313131313,
      "avg_output_tokens_per_task": 25683.540404040403,
      "last_run_timestamp": "2026-08-08T09:30:52.456Z"
    },
    {
      "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": 0.902357,
      "accuracy_std_dev": 0.002749143987983661,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.06939250673400674,
      "avg_duration_per_task_ms": 257971.41245791246,
      "avg_output_tokens_per_task": 16102.877104377105,
      "last_run_timestamp": "2026-10-30T00:00:05.460Z"
    },
    {
      "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": 0.902357,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.1437863712121212,
      "avg_duration_per_task_ms": 149585.63636363635,
      "avg_output_tokens_per_task": 9456.126262626263,
      "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-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Mistral",
      "accuracy": 0.9023569999999999,
      "accuracy_std_dev": 0.006734000000000018,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.02541817676767677,
      "avg_duration_per_task_ms": 99436.77525252526,
      "avg_output_tokens_per_task": 21016.33585858586,
      "last_run_timestamp": "2026-10-30T16:01:57.510Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-opus-5.5-20260921",
      "model_slug": "anthropic/claude-opus-5.5",
      "model_name": "Anthropic: Claude Opus 5.5",
      "provider_name": "auto-routing",
      "accuracy": 0.9023568,
      "accuracy_std_dev": 0.005733794394639539,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.05410575151515151,
      "avg_duration_per_task_ms": 31371.09595959596,
      "avg_output_tokens_per_task": 2496.787878787879,
      "last_run_timestamp": "2026-10-24T08:04:44.407Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-opus-5.5-20260921",
      "model_slug": "anthropic/claude-opus-5.5",
      "model_name": "Anthropic: Claude Opus 5.5",
      "provider_name": "Anthropic",
      "accuracy": 0.9023566666666667,
      "accuracy_std_dev": 0.003636621356271357,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.053918992424242425,
      "avg_duration_per_task_ms": 26868.823232323233,
      "avg_output_tokens_per_task": 2480.0361952861954,
      "last_run_timestamp": "2026-10-24T08:14:51.096Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "reflection/beam-1.0-501b-a23b-20261016",
      "model_slug": "reflection/beam-1.0-501b-a23b",
      "model_name": "Reflection: Beam 1.0 501B A23B",
      "provider_name": "Nebius",
      "accuracy": 0.9017956666666668,
      "accuracy_std_dev": 0.003459479954495376,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.014097868686868686,
      "avg_duration_per_task_ms": 101233.50336700337,
      "avg_output_tokens_per_task": 13920.223905723906,
      "last_run_timestamp": "2026-10-24T08:29:16.809Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.8-opus-20260528",
      "model_slug": "anthropic/claude-opus-4.8",
      "model_name": "Anthropic: Claude Opus 4.8",
      "provider_name": "Azure",
      "accuracy": 0.9017254583333333,
      "accuracy_std_dev": 0.0058683477301193655,
      "run_count": 24,
      "total_questions": 4752,
      "avg_cost_per_task_usd": 0.2056731323653199,
      "avg_duration_per_task_ms": 97891.09090909091,
      "avg_output_tokens_per_task": 7837.397516835017,
      "last_run_timestamp": "2026-10-24T08:44:41.805Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "auto-routing",
      "accuracy": 0.9016089931034483,
      "accuracy_std_dev": 0.017035826057784066,
      "run_count": 11,
      "total_questions": 2175,
      "avg_cost_per_task_usd": 0.10899811586206896,
      "avg_duration_per_task_ms": 561470.3995402299,
      "avg_output_tokens_per_task": 37112.96183908046,
      "last_run_timestamp": "2026-10-24T08:01:41.138Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-1017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "Deepseek V4.1 Flash 1017",
      "provider_name": "auto-routing",
      "accuracy": 0.901515,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.006598151515151515,
      "avg_duration_per_task_ms": 72990.44444444444,
      "avg_output_tokens_per_task": 11623.267676767677,
      "last_run_timestamp": "2026-10-23T06:09:54.243Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Together",
      "accuracy": 0.901515,
      "accuracy_std_dev": 0.0025249999999999995,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.022678441919191918,
      "avg_duration_per_task_ms": 49322.41161616162,
      "avg_output_tokens_per_task": 18730.81818181818,
      "last_run_timestamp": "2026-10-24T08:41:23.349Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.4-20260305",
      "model_slug": "openai/gpt-5.4",
      "model_name": "OpenAI: GPT-5.4",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.901515,
      "accuracy_std_dev": 0.0025249999999999995,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.16319313383838385,
      "avg_duration_per_task_ms": 88415.74242424243,
      "avg_output_tokens_per_task": 9762.232323232323,
      "last_run_timestamp": "2026-10-01T19:12:25.243Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Cloudflare",
      "accuracy": 0.9009352959183673,
      "accuracy_std_dev": 0.0002650000000000152,
      "run_count": 2,
      "total_questions": 392,
      "avg_cost_per_task_usd": 0.013316757653061225,
      "avg_duration_per_task_ms": 199818.51020408163,
      "avg_output_tokens_per_task": 22103.426020408162,
      "last_run_timestamp": "2026-10-24T08:16:43.075Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.8-opus-20260528",
      "model_slug": "anthropic/claude-opus-4.8",
      "model_name": "Anthropic: Claude Opus 4.8",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.9007668055555554,
      "accuracy_std_dev": 0.005776883838096259,
      "run_count": 36,
      "total_questions": 7128,
      "avg_cost_per_task_usd": 0.21669425,
      "avg_duration_per_task_ms": 93977.99873737374,
      "avg_output_tokens_per_task": 7906.665684624018,
      "last_run_timestamp": "2026-10-24T08:14:03.138Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Google",
      "accuracy": 0.9006734999999999,
      "accuracy_std_dev": 0.005050500000000013,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.020602151515151518,
      "avg_duration_per_task_ms": 398490.0909090909,
      "avg_output_tokens_per_task": 16966.28787878788,
      "last_run_timestamp": "2026-10-24T08:44:49.435Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta/muse-spark-open-20261015",
      "model_slug": "meta/muse-spark-open",
      "model_name": "Meta: Muse Spark Open",
      "provider_name": "Together",
      "accuracy": 0.900673,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.07405119191919192,
      "avg_duration_per_task_ms": 94171.00505050505,
      "avg_output_tokens_per_task": 17264.035353535353,
      "last_run_timestamp": "2026-10-24T08:14:51.112Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "GMICloud",
      "accuracy": 0.900673,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.11481898484848484,
      "avg_duration_per_task_ms": 1609821.202020202,
      "avg_output_tokens_per_task": 45736.868686868685,
      "last_run_timestamp": "2026-10-27T09:00:04.213Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Parasail",
      "accuracy": 0.9002912013585639,
      "accuracy_std_dev": 0.008613175564538058,
      "run_count": 12,
      "total_questions": 2061,
      "avg_cost_per_task_usd": 0.024704503153808833,
      "avg_duration_per_task_ms": 166852.90538573507,
      "avg_output_tokens_per_task": 20435.77389616691,
      "last_run_timestamp": "2026-10-24T16:26:43.118Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "Xiaomi",
      "accuracy": 0.8999997999999999,
      "accuracy_std_dev": 0.0027351667883330276,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.03378601313131313,
      "avg_duration_per_task_ms": 836485.5171717171,
      "avg_output_tokens_per_task": 38467.88484848485,
      "last_run_timestamp": "2026-10-24T08:37:44.264Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Baidu",
      "accuracy": 0.8999869065420562,
      "accuracy_std_dev": 0.0055303889812583175,
      "run_count": 13,
      "total_questions": 2568,
      "avg_cost_per_task_usd": 0.012645620327102805,
      "avg_duration_per_task_ms": 176614.65965732088,
      "avg_output_tokens_per_task": 26316.880841121496,
      "last_run_timestamp": "2026-10-24T08:45:10.619Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Crusoe",
      "accuracy": 0.899832,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.00798121212121212,
      "avg_duration_per_task_ms": 376079.0606060606,
      "avg_output_tokens_per_task": 28057.621212121212,
      "last_run_timestamp": "2026-10-26T03:00:10.512Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-opus-5.5-20260921",
      "model_slug": "anthropic/claude-opus-5.5",
      "model_name": "Anthropic: Claude Opus 5.5",
      "provider_name": "Google",
      "accuracy": 0.8998315555555554,
      "accuracy_std_dev": 0.004532988273904737,
      "run_count": 18,
      "total_questions": 3564,
      "avg_cost_per_task_usd": 0.05740673849607183,
      "avg_duration_per_task_ms": 25170.2278338945,
      "avg_output_tokens_per_task": 2482.2177328843995,
      "last_run_timestamp": "2026-10-24T08:41:23.935Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Parasail",
      "accuracy": 0.8998315,
      "accuracy_std_dev": 0.007575499999999957,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.02710091414141414,
      "avg_duration_per_task_ms": 156920.601010101,
      "avg_output_tokens_per_task": 22433.31818181818,
      "last_run_timestamp": "2026-10-24T08:34:01.458Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "Cohere",
      "accuracy": 0.8998315,
      "accuracy_std_dev": 0.0025254999999999583,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.11557131818181818,
      "avg_duration_per_task_ms": 1030735.8762626263,
      "avg_output_tokens_per_task": 45994.38383838384,
      "last_run_timestamp": "2026-10-24T08:14:03.133Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "auto-routing",
      "accuracy": 0.8995506666666668,
      "accuracy_std_dev": 0.005722799567422267,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.11900113804713805,
      "avg_duration_per_task_ms": 846691.8215488215,
      "avg_output_tokens_per_task": 47399.26936026936,
      "last_run_timestamp": "2026-10-24T08:02:41.727Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "Novita",
      "accuracy": 0.8994372877637131,
      "accuracy_std_dev": 0.01977995405637288,
      "run_count": 6,
      "total_questions": 1185,
      "avg_cost_per_task_usd": 0.03544181097046414,
      "avg_duration_per_task_ms": 152645.21687763714,
      "avg_output_tokens_per_task": 26597.288607594935,
      "last_run_timestamp": "2026-10-03T09:28:39.969Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-opus-5.5-20260921",
      "model_slug": "anthropic/claude-opus-5.5",
      "model_name": "Anthropic: Claude Opus 5.5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.8993638888888889,
      "accuracy_std_dev": 0.004257353623619148,
      "run_count": 18,
      "total_questions": 3564,
      "avg_cost_per_task_usd": 0.058306676767676764,
      "avg_duration_per_task_ms": 37477.73063973064,
      "avg_output_tokens_per_task": 2520.0830527497196,
      "last_run_timestamp": "2026-10-24T08:44:59.329Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "NextBit",
      "accuracy": 0.8993638888888887,
      "accuracy_std_dev": 0.004029367042778795,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.030411617283950615,
      "avg_duration_per_task_ms": 237553.6509539843,
      "avg_output_tokens_per_task": 26633.1290684624,
      "last_run_timestamp": "2026-10-03T10:19:34.320Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-fable-5.1-20260831",
      "model_slug": "anthropic/claude-fable-5.1",
      "model_name": "Anthropic: Claude Fable 5.1",
      "provider_name": "auto-routing",
      "accuracy": 0.8993264,
      "accuracy_std_dev": 0.006511664598242117,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.13052546969696968,
      "avg_duration_per_task_ms": 30874.333333333332,
      "avg_output_tokens_per_task": 2403.130303030303,
      "last_run_timestamp": "2026-10-01T08:24:30.121Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Novita",
      "accuracy": 0.8992834083438684,
      "accuracy_std_dev": 0.01959300954191881,
      "run_count": 12,
      "total_questions": 2373,
      "avg_cost_per_task_usd": 0.12703601896333752,
      "avg_duration_per_task_ms": 516266.79730299197,
      "avg_output_tokens_per_task": 38334.854193004634,
      "last_run_timestamp": "2026-10-24T08:37:37.787Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Cloudflare",
      "accuracy": 0.8991558025316456,
      "accuracy_std_dev": 0.04033796787732645,
      "run_count": 6,
      "total_questions": 1185,
      "avg_cost_per_task_usd": 0.09046152742616034,
      "avg_duration_per_task_ms": 742340.6286919832,
      "avg_output_tokens_per_task": 35318.2,
      "last_run_timestamp": "2026-10-24T16:21:36.007Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-opus-5.5-20260921",
      "model_slug": "anthropic/claude-opus-5.5",
      "model_name": "Anthropic: Claude Opus 5.5",
      "provider_name": "Azure",
      "accuracy": 0.89913,
      "accuracy_std_dev": 0.005890503218458193,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.05769051388888888,
      "avg_duration_per_task_ms": 26119.287037037036,
      "avg_output_tokens_per_task": 2529.9574915824915,
      "last_run_timestamp": "2026-10-24T08:24:00.716Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta/muse-spark-open-20261015",
      "model_slug": "meta/muse-spark-open",
      "model_name": "Meta: Muse Spark Open",
      "provider_name": "BaseTen",
      "accuracy": 0.8989899999999998,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.07316585858585858,
      "avg_duration_per_task_ms": 97804.9696969697,
      "avg_output_tokens_per_task": 16993.328282828283,
      "last_run_timestamp": "2026-10-24T08:12:37.563Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.8-opus-20260528",
      "model_slug": "anthropic/claude-opus-4.8",
      "model_name": "Anthropic: Claude Opus 4.8",
      "provider_name": "Google",
      "accuracy": 0.8989897837837839,
      "accuracy_std_dev": 0.006188632325591686,
      "run_count": 37,
      "total_questions": 7326,
      "avg_cost_per_task_usd": 0.22242387441987438,
      "avg_duration_per_task_ms": 94205.91345891346,
      "avg_output_tokens_per_task": 8109.393939393939,
      "last_run_timestamp": "2026-10-24T08:20:25.133Z"
    },
    {
      "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": 0.8986389583333334,
      "accuracy_std_dev": 0.007699096453909172,
      "run_count": 24,
      "total_questions": 4752,
      "avg_cost_per_task_usd": 0.12657329903198655,
      "avg_duration_per_task_ms": 58006.606060606064,
      "avg_output_tokens_per_task": 4727.50441919192,
      "last_run_timestamp": "2026-10-24T08:33:28.370Z"
    },
    {
      "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": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "ModelRun",
      "accuracy": 0.8984635624999999,
      "accuracy_std_dev": 0.01886644038355125,
      "run_count": 16,
      "total_questions": 3168,
      "avg_cost_per_task_usd": 0.07989067045454545,
      "avg_duration_per_task_ms": 218000.37657828283,
      "avg_output_tokens_per_task": 30348.85164141414,
      "last_run_timestamp": "2026-10-28T05:05:35.330Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.7-opus-20260416",
      "model_slug": "anthropic/claude-opus-4.7",
      "model_name": "Anthropic: Claude Opus 4.7",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.8983774545454545,
      "accuracy_std_dev": 0.004829955304962313,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.2299064136822773,
      "avg_duration_per_task_ms": 110399.28466483012,
      "avg_output_tokens_per_task": 8983.414600550965,
      "last_run_timestamp": "2026-10-24T08:27:26.668Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8983162,
      "accuracy_std_dev": 0.008080975000580047,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.1253350707070707,
      "avg_duration_per_task_ms": 994044.4858585858,
      "avg_output_tokens_per_task": 49876.97171717172,
      "last_run_timestamp": "2026-10-24T08:12:02.438Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "DigitalOcean",
      "accuracy": 0.8981654060721064,
      "accuracy_std_dev": 0.006690047440041041,
      "run_count": 8,
      "total_questions": 1581,
      "avg_cost_per_task_usd": 0.01891549778621126,
      "avg_duration_per_task_ms": 414943.30676786846,
      "avg_output_tokens_per_task": 22047.87349778621,
      "last_run_timestamp": "2026-10-24T09:25:31.426Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": "openai/gpt-5.4-20260305",
      "model_slug": "openai/gpt-5.4",
      "model_name": "OpenAI: GPT-5.4",
      "provider_name": "Azure",
      "accuracy": 0.8977871428571428,
      "accuracy_std_dev": 0.005599042607665047,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.15678494083694083,
      "avg_duration_per_task_ms": 138524.9898989899,
      "avg_output_tokens_per_task": 9631.916305916306,
      "last_run_timestamp": "2026-10-01T19:18:24.536Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.8-opus-20260528",
      "model_slug": "anthropic/claude-opus-4.8",
      "model_name": "Anthropic: Claude Opus 4.8",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.8977654545454545,
      "accuracy_std_dev": 0.00783374636212555,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.19846787741046834,
      "avg_duration_per_task_ms": 95325.20982552801,
      "avg_output_tokens_per_task": 7728.792011019284,
      "last_run_timestamp": "2026-10-24T08:17:56.390Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "DeepInfra",
      "accuracy": 0.8975652307692306,
      "accuracy_std_dev": 0.01420604709353269,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.004205098290598292,
      "avg_duration_per_task_ms": 359286.8714063714,
      "avg_output_tokens_per_task": 23062.278943278943,
      "last_run_timestamp": "2026-10-24T08:41:47.029Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-luna-pro-20260709",
      "model_slug": "openai/gpt-5.6-luna-pro",
      "model_name": "OpenAI: GPT-5.6 Luna Pro",
      "provider_name": "OpenAI",
      "accuracy": 0.8973063333333333,
      "accuracy_std_dev": 0.004955863351178606,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.03703221212121212,
      "avg_duration_per_task_ms": 137593.69696969696,
      "avg_output_tokens_per_task": 26012.949494949495,
      "last_run_timestamp": "2026-10-02T23:26:10.109Z"
    },
    {
      "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": 0.8973062499999999,
      "accuracy_std_dev": 0.01842900724458137,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.030114116582491583,
      "avg_duration_per_task_ms": 249874.79587542088,
      "avg_output_tokens_per_task": 27231.26473063973,
      "last_run_timestamp": "2026-10-24T08:56:13.509Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Cloudflare",
      "accuracy": 0.8972558314475225,
      "accuracy_std_dev": 0.033281996803894424,
      "run_count": 13,
      "total_questions": 2563,
      "avg_cost_per_task_usd": 0.1455034553257901,
      "avg_duration_per_task_ms": 811067.1357783847,
      "avg_output_tokens_per_task": 36442.67655091689,
      "last_run_timestamp": "2026-10-24T08:25:27.617Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "GMICloud",
      "accuracy": 0.8971767692307694,
      "accuracy_std_dev": 0.016248030982026696,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.019287930069930074,
      "avg_duration_per_task_ms": 202055.53224553223,
      "avg_output_tokens_per_task": 25397.89898989899,
      "last_run_timestamp": "2026-10-27T00:43:30.287Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.8-opus-20260528",
      "model_slug": "anthropic/claude-opus-4.8",
      "model_name": "Anthropic: Claude Opus 4.8",
      "provider_name": "auto-routing",
      "accuracy": 0.8971659166666668,
      "accuracy_std_dev": 0.011261007344359652,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.19972315109427613,
      "avg_duration_per_task_ms": 95894.5180976431,
      "avg_output_tokens_per_task": 7770.697390572391,
      "last_run_timestamp": "2026-10-24T08:28:56.790Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Wafer",
      "accuracy": 0.8970973821999613,
      "accuracy_std_dev": 0.12179084265551246,
      "run_count": 28,
      "total_questions": 5191,
      "avg_cost_per_task_usd": 0.1641544681178963,
      "avg_duration_per_task_ms": 204815.23540743595,
      "avg_output_tokens_per_task": 10801.00500866885,
      "last_run_timestamp": "2026-10-29T22:49:50.464Z"
    },
    {
      "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": 0.8965847857142858,
      "accuracy_std_dev": 0.006695321321410832,
      "run_count": 14,
      "total_questions": 2772,
      "avg_cost_per_task_usd": 0.20728678535353537,
      "avg_duration_per_task_ms": 100509.16955266955,
      "avg_output_tokens_per_task": 8077.520562770563,
      "last_run_timestamp": "2026-10-24T08:30:35.509Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "reflection/beam-1.0-501b-a23b-20261016",
      "model_slug": "reflection/beam-1.0-501b-a23b",
      "model_name": "Reflection: Beam 1.0 501B A23B",
      "provider_name": "CoreWeave",
      "accuracy": 0.8964644999999999,
      "accuracy_std_dev": 0.0025254999999999583,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.027913444444444446,
      "avg_duration_per_task_ms": 54299.99747474748,
      "avg_output_tokens_per_task": 13871.598484848484,
      "last_run_timestamp": "2026-10-24T08:12:49.568Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Venice",
      "accuracy": 0.8962877525832377,
      "accuracy_std_dev": 0.04356588502231736,
      "run_count": 9,
      "total_questions": 1742,
      "avg_cost_per_task_usd": 0.04917616016073478,
      "avg_duration_per_task_ms": 485084.33180252585,
      "avg_output_tokens_per_task": 15776.056831228472,
      "last_run_timestamp": "2026-10-24T16:26:47.562Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta/muse-spark-open-20261015",
      "model_slug": "meta/muse-spark-open",
      "model_name": "Meta: Muse Spark Open",
      "provider_name": "Nebius",
      "accuracy": 0.8960434999999999,
      "accuracy_std_dev": 0.007996499999999962,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.06795408585858587,
      "avg_duration_per_task_ms": 278542.00757575757,
      "avg_output_tokens_per_task": 15808.588383838383,
      "last_run_timestamp": "2026-10-24T08:15:15.302Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "DeepInfra",
      "accuracy": 0.8959899189640768,
      "accuracy_std_dev": 0.02896400676270464,
      "run_count": 8,
      "total_questions": 1197,
      "avg_cost_per_task_usd": 0.024112938178780283,
      "avg_duration_per_task_ms": 783093.3592314118,
      "avg_output_tokens_per_task": 22333.22807017544,
      "last_run_timestamp": "2026-10-17T16:03:42.308Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.7-opus-20260416",
      "model_slug": "anthropic/claude-opus-4.7",
      "model_name": "Anthropic: Claude Opus 4.7",
      "provider_name": "Google",
      "accuracy": 0.8958047837837839,
      "accuracy_std_dev": 0.007786697038158158,
      "run_count": 37,
      "total_questions": 7326,
      "avg_cost_per_task_usd": 0.20536014018564022,
      "avg_duration_per_task_ms": 70550.94512694512,
      "avg_output_tokens_per_task": 7473.251979251979,
      "last_run_timestamp": "2026-10-24T08:43:08.078Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "OpenInference",
      "accuracy": 0.895623,
      "accuracy_std_dev": 0.010100999999999971,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.014240813131313132,
      "avg_duration_per_task_ms": 199874.40656565657,
      "avg_output_tokens_per_task": 21243.522727272728,
      "last_run_timestamp": "2026-10-24T08:54:21.561Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.895623,
      "accuracy_std_dev": 0.0033669999999999534,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.03841230555555556,
      "avg_duration_per_task_ms": 144059.0782828283,
      "avg_output_tokens_per_task": 27039.489898989897,
      "last_run_timestamp": "2026-10-01T16:12:34.647Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Crusoe",
      "accuracy": 0.8956225,
      "accuracy_std_dev": 0.005050499999999958,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.026159166666666667,
      "avg_duration_per_task_ms": 177167.38888888888,
      "avg_output_tokens_per_task": 21649.300505050505,
      "last_run_timestamp": "2026-10-24T08:22:00.564Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-opus-5.5-20260921",
      "model_slug": "anthropic/claude-opus-5.5",
      "model_name": "Anthropic: Claude Opus 5.5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.8953421666666667,
      "accuracy_std_dev": 0.0047035083153133565,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.05463887542087542,
      "avg_duration_per_task_ms": 31241.656565656565,
      "avg_output_tokens_per_task": 2527.686026936027,
      "last_run_timestamp": "2026-10-24T08:08:54.167Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "GMICloud",
      "accuracy": 0.8952485555555553,
      "accuracy_std_dev": 0.018860205708263727,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.02531761784511784,
      "avg_duration_per_task_ms": 163044.16722783388,
      "avg_output_tokens_per_task": 28644.837822671157,
      "last_run_timestamp": "2026-10-27T00:41:39.750Z"
    },
    {
      "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": 0.8952178253164556,
      "accuracy_std_dev": 0.009821681674220657,
      "run_count": 12,
      "total_questions": 2370,
      "avg_cost_per_task_usd": 0.01650155105485232,
      "avg_duration_per_task_ms": 347394.32742616034,
      "avg_output_tokens_per_task": 28453.445147679326,
      "last_run_timestamp": "2026-10-24T08:49:27.630Z"
    },
    {
      "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": 0.8952138875490746,
      "accuracy_std_dev": 0.03543317040246446,
      "run_count": 19,
      "total_questions": 3566,
      "avg_cost_per_task_usd": 0.3690802299495232,
      "avg_duration_per_task_ms": 494589.1385305665,
      "avg_output_tokens_per_task": 57144.12535053281,
      "last_run_timestamp": "2026-10-29T22:54:33.936Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Cohere",
      "accuracy": 0.8950615,
      "accuracy_std_dev": 0.01045325149973282,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.008697955387205388,
      "avg_duration_per_task_ms": 170461.9739057239,
      "avg_output_tokens_per_task": 30584.97053872054,
      "last_run_timestamp": "2026-10-24T08:08:54.161Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-opus-5-20260723",
      "model_slug": "anthropic/claude-opus-5",
      "model_name": "Anthropic: Claude Opus 5",
      "provider_name": "Anthropic",
      "accuracy": 0.8949211666666667,
      "accuracy_std_dev": 0.007698231721563628,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.12088044486531986,
      "avg_duration_per_task_ms": 58308.99410774411,
      "avg_output_tokens_per_task": 4615.2697811447815,
      "last_run_timestamp": "2026-10-24T08:16:38.969Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "reflection/beam-1.0-501b-a23b-20261016",
      "model_slug": "reflection/beam-1.0-501b-a23b",
      "model_name": "Reflection: Beam 1.0 501B A23B",
      "provider_name": "BaseTen",
      "accuracy": 0.8947810000000002,
      "accuracy_std_dev": 0.0025249999999999995,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.03846228535353535,
      "avg_duration_per_task_ms": 31568.290404040403,
      "avg_output_tokens_per_task": 13846.992424242424,
      "last_run_timestamp": "2026-10-24T08:14:03.134Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "StreamLake",
      "accuracy": 0.894781,
      "accuracy_std_dev": 0.007575999999999972,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.036287184343434344,
      "avg_duration_per_task_ms": 737979.1590909091,
      "avg_output_tokens_per_task": 37423.542929292926,
      "last_run_timestamp": "2026-10-24T08:21:38.159Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8946608571428573,
      "accuracy_std_dev": 0.021330086138923062,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.025403829725829725,
      "avg_duration_per_task_ms": 208889.86363636365,
      "avg_output_tokens_per_task": 27754.472582972583,
      "last_run_timestamp": "2026-10-24T08:01:41.237Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.8946406666666666,
      "accuracy_std_dev": 0.009509966424172523,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.23073020496632995,
      "avg_duration_per_task_ms": 110250.86616161616,
      "avg_output_tokens_per_task": 9005.217171717171,
      "last_run_timestamp": "2026-10-24T08:44:41.392Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "BaseTen",
      "accuracy": 0.8945726512842839,
      "accuracy_std_dev": 0.020487988184645325,
      "run_count": 12,
      "total_questions": 2297,
      "avg_cost_per_task_usd": 0.15815831606443187,
      "avg_duration_per_task_ms": 350172.93774488464,
      "avg_output_tokens_per_task": 39679.15150195908,
      "last_run_timestamp": "2026-10-26T03:00:10.340Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "ModelRun",
      "accuracy": 0.8943087542213883,
      "accuracy_std_dev": 0.03942204850926279,
      "run_count": 3,
      "total_questions": 533,
      "avg_cost_per_task_usd": 0.019450917448405252,
      "avg_duration_per_task_ms": 177927.2420262664,
      "avg_output_tokens_per_task": 28174.943714821762,
      "last_run_timestamp": "2026-10-28T05:00:14.622Z"
    },
    {
      "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": 0.8940067465753425,
      "accuracy_std_dev": 0.01621889745889719,
      "run_count": 15,
      "total_questions": 2920,
      "avg_cost_per_task_usd": 0.014883987671232876,
      "avg_duration_per_task_ms": 203624.06232876712,
      "avg_output_tokens_per_task": 20864.14212328767,
      "last_run_timestamp": "2026-10-29T22:27:58.466Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "Makora",
      "accuracy": 0.8934709999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 194,
      "avg_cost_per_task_usd": 0.11686864948453608,
      "avg_duration_per_task_ms": 605687.7474226804,
      "avg_output_tokens_per_task": 46496.701030927834,
      "last_run_timestamp": "2026-10-13T03:00:07.016Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Phala",
      "accuracy": 0.8932483811659193,
      "accuracy_std_dev": 0.017336749323389345,
      "run_count": 7,
      "total_questions": 1338,
      "avg_cost_per_task_usd": 0.03355175710014948,
      "avg_duration_per_task_ms": 240320.3542600897,
      "avg_output_tokens_per_task": 32642.40284005979,
      "last_run_timestamp": "2026-10-24T08:41:23.835Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.8930975,
      "accuracy_std_dev": 0.010942499999999966,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.04334855050505051,
      "avg_duration_per_task_ms": 40520.42929292929,
      "avg_output_tokens_per_task": 3151.881313131313,
      "last_run_timestamp": "2026-10-01T19:11:43.444Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "Tencent",
      "accuracy": 0.8930975,
      "accuracy_std_dev": 0.009560420192125432,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.11553578787878788,
      "avg_duration_per_task_ms": 625203.1957070708,
      "avg_output_tokens_per_task": 46042.48863636364,
      "last_run_timestamp": "2026-10-24T08:01:41.143Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Fireworks",
      "accuracy": 0.8928570861678005,
      "accuracy_std_dev": 0.015340394005370288,
      "run_count": 5,
      "total_questions": 882,
      "avg_cost_per_task_usd": 0.12683532766439912,
      "avg_duration_per_task_ms": 285532.1031746032,
      "avg_output_tokens_per_task": 28366.77097505669,
      "last_run_timestamp": "2026-09-19T09:14:09.704Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "GMICloud",
      "accuracy": 0.8924427777777778,
      "accuracy_std_dev": 0.010300002662111597,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.005114002244668912,
      "avg_duration_per_task_ms": 223535.9657687991,
      "avg_output_tokens_per_task": 28038.338383838385,
      "last_run_timestamp": "2026-10-27T01:02:43.701Z"
    },
    {
      "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": 0.8922560000000002,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0070467222222222224,
      "avg_duration_per_task_ms": 212650.5101010101,
      "avg_output_tokens_per_task": 24920.08080808081,
      "last_run_timestamp": "2026-08-08T10:06:09.471Z"
    },
    {
      "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": 0.8922560000000002,
      "accuracy_std_dev": 0.010100999999999971,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.14220060606060608,
      "avg_duration_per_task_ms": 188559.2904040404,
      "avg_output_tokens_per_task": 47276.156565656565,
      "last_run_timestamp": "2026-10-01T19:12:30.168Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "Novita",
      "accuracy": 0.8922557999999999,
      "accuracy_std_dev": 0.007299435934371887,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.11377392424242423,
      "avg_duration_per_task_ms": 848578.0848484848,
      "avg_output_tokens_per_task": 45335.65555555555,
      "last_run_timestamp": "2026-10-24T08:01:41.133Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.7-opus-20260416",
      "model_slug": "anthropic/claude-opus-4.7",
      "model_name": "Anthropic: Claude Opus 4.7",
      "provider_name": "Anthropic",
      "accuracy": 0.8920152857142859,
      "accuracy_std_dev": 0.011608927534005704,
      "run_count": 14,
      "total_questions": 2772,
      "avg_cost_per_task_usd": 0.20755713311688312,
      "avg_duration_per_task_ms": 99648.18037518038,
      "avg_output_tokens_per_task": 8074.249278499278,
      "last_run_timestamp": "2026-10-24T08:05:46.408Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.8919751666666667,
      "accuracy_std_dev": 0.008661714291768222,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.024034642255892257,
      "avg_duration_per_task_ms": 245030.94696969696,
      "avg_output_tokens_per_task": 30463.32239057239,
      "last_run_timestamp": "2026-10-24T08:01:41.135Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-terra-20260709",
      "model_slug": "openai/gpt-5.6-terra",
      "model_name": "OpenAI: GPT-5.6 Terra",
      "provider_name": "Azure",
      "accuracy": 0.8917748571428571,
      "accuracy_std_dev": 0.009721593908534165,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.04607931240981241,
      "avg_duration_per_task_ms": 43786.40764790765,
      "avg_output_tokens_per_task": 3453.090909090909,
      "last_run_timestamp": "2026-10-01T19:13:06.514Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Sail Research",
      "accuracy": 0.8916709261744967,
      "accuracy_std_dev": 0.019973452544815567,
      "run_count": 10,
      "total_questions": 1937,
      "avg_cost_per_task_usd": 0.13092793236964376,
      "avg_duration_per_task_ms": 553822.8373773877,
      "avg_output_tokens_per_task": 42781.386164171396,
      "last_run_timestamp": "2026-10-24T09:11:22.395Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.7-opus-20260416",
      "model_slug": "anthropic/claude-opus-4.7",
      "model_name": "Anthropic: Claude Opus 4.7",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.8916244166666666,
      "accuracy_std_dev": 0.008865984904475566,
      "run_count": 24,
      "total_questions": 4752,
      "avg_cost_per_task_usd": 0.24456361553030298,
      "avg_duration_per_task_ms": 85715.98253367003,
      "avg_output_tokens_per_task": 9091.301978114478,
      "last_run_timestamp": "2026-10-24T08:14:36.267Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "DeepInfra",
      "accuracy": 0.8915753430656934,
      "accuracy_std_dev": 0.004887565515332395,
      "run_count": 6,
      "total_questions": 1096,
      "avg_cost_per_task_usd": 0.03103391879562044,
      "avg_duration_per_task_ms": 795798.2390510949,
      "avg_output_tokens_per_task": 36089.42244525548,
      "last_run_timestamp": "2026-10-24T08:07:59.343Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Morph",
      "accuracy": 0.8915342857142857,
      "accuracy_std_dev": 0.010905197538189691,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.011785490620490622,
      "avg_duration_per_task_ms": 284433.1883116883,
      "avg_output_tokens_per_task": 21375.483405483406,
      "last_run_timestamp": "2026-10-24T08:45:15.768Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Parasail",
      "accuracy": 0.8914141,
      "accuracy_std_dev": 0.016939939187907378,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.1456811595959596,
      "avg_duration_per_task_ms": 619546.0101010101,
      "avg_output_tokens_per_task": 36709.93131313131,
      "last_run_timestamp": "2026-10-17T08:18:00.876Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Novita",
      "accuracy": 0.891414,
      "accuracy_std_dev": 0.01094299999999998,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.007142083333333334,
      "avg_duration_per_task_ms": 125847.88383838384,
      "avg_output_tokens_per_task": 11714.222222222223,
      "last_run_timestamp": "2026-10-24T08:16:38.970Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "reflection/beam-1.0-501b-a23b-20261016",
      "model_slug": "reflection/beam-1.0-501b-a23b",
      "model_name": "Reflection: Beam 1.0 501B A23B",
      "provider_name": "DigitalOcean",
      "accuracy": 0.891414,
      "accuracy_std_dev": 0.0025249999999999995,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.016427454545454544,
      "avg_duration_per_task_ms": 96439.17424242424,
      "avg_output_tokens_per_task": 13618.247474747475,
      "last_run_timestamp": "2026-10-24T08:14:51.086Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Crusoe",
      "accuracy": 0.8912280000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 190,
      "avg_cost_per_task_usd": 0.14632182105263158,
      "avg_duration_per_task_ms": 676754.2315789474,
      "avg_output_tokens_per_task": 41671.152631578945,
      "last_run_timestamp": "2026-10-26T03:00:09.830Z"
    },
    {
      "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": "SiliconFlow",
      "accuracy": 0.8907934361904761,
      "accuracy_std_dev": 0.020177999596946557,
      "run_count": 8,
      "total_questions": 1575,
      "avg_cost_per_task_usd": 0.18863606476190478,
      "avg_duration_per_task_ms": 1541820.2831746032,
      "avg_output_tokens_per_task": 55365.57968253968,
      "last_run_timestamp": "2026-10-24T16:20:08.208Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Parasail",
      "accuracy": 0.8906711574504008,
      "accuracy_std_dev": 0.010298496956136629,
      "run_count": 12,
      "total_questions": 2369,
      "avg_cost_per_task_usd": 0.007867603630223723,
      "avg_duration_per_task_ms": 337616.2794428029,
      "avg_output_tokens_per_task": 27764.332207682568,
      "last_run_timestamp": "2026-10-24T08:50:11.572Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "GMICloud",
      "accuracy": 0.8905723333333333,
      "accuracy_std_dev": 0.0076535129334327255,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.007299473063973064,
      "avg_duration_per_task_ms": 131190.53535353535,
      "avg_output_tokens_per_task": 12389.143097643098,
      "last_run_timestamp": "2026-10-27T00:37:23.267Z"
    },
    {
      "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": 0.8901177599344533,
      "accuracy_std_dev": 0.009279724700032873,
      "run_count": 37,
      "total_questions": 7323,
      "avg_cost_per_task_usd": 0.1335219325413082,
      "avg_duration_per_task_ms": 59352.6749965861,
      "avg_output_tokens_per_task": 4769.18080021849,
      "last_run_timestamp": "2026-10-24T08:16:43.073Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-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-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "GMICloud",
      "accuracy": 0.8899192124789208,
      "accuracy_std_dev": 0.02293354363329505,
      "run_count": 9,
      "total_questions": 1779,
      "avg_cost_per_task_usd": 0.06400783642495786,
      "avg_duration_per_task_ms": 535683.2737492974,
      "avg_output_tokens_per_task": 35982.14896008994,
      "last_run_timestamp": "2026-10-27T02:49:26.609Z"
    },
    {
      "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": "openai/gpt-5.6-terra-20260709",
      "model_slug": "openai/gpt-5.6-terra",
      "model_name": "OpenAI: GPT-5.6 Terra",
      "provider_name": "OpenAI",
      "accuracy": 0.8888890000000002,
      "accuracy_std_dev": 0.002749143987983661,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0431288872053872,
      "avg_duration_per_task_ms": 52519.67676767677,
      "avg_output_tokens_per_task": 3463.8030303030305,
      "last_run_timestamp": "2026-10-02T23:21:06.556Z"
    },
    {
      "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": 0.888889,
      "accuracy_std_dev": 0.006733999999999962,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.1979060606060606,
      "avg_duration_per_task_ms": 95702.44696969698,
      "avg_output_tokens_per_task": 7691.454545454545,
      "last_run_timestamp": "2026-10-01T19:11:18.427Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Reka",
      "accuracy": 0.8887654163757536,
      "accuracy_std_dev": 0.008192069777191544,
      "run_count": 16,
      "total_questions": 3151,
      "avg_cost_per_task_usd": 0.01559116534433513,
      "avg_duration_per_task_ms": 197619.02634084417,
      "avg_output_tokens_per_task": 26956.708981275784,
      "last_run_timestamp": "2026-10-24T08:24:10.667Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Ionstream",
      "accuracy": 0.8886482857142856,
      "accuracy_std_dev": 0.013807459233682208,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.1021348318903319,
      "avg_duration_per_task_ms": 444900.012987013,
      "avg_output_tokens_per_task": 35591.61255411255,
      "last_run_timestamp": "2026-10-24T08:42:40.868Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Relace",
      "accuracy": 0.8885464319465846,
      "accuracy_std_dev": 0.029017071552622252,
      "run_count": 10,
      "total_questions": 1947,
      "avg_cost_per_task_usd": 0.009370765279917823,
      "avg_duration_per_task_ms": 312934.7123780175,
      "avg_output_tokens_per_task": 26808.334874165383,
      "last_run_timestamp": "2026-10-24T08:13:41.086Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.7-opus-20260416",
      "model_slug": "anthropic/claude-opus-4.7",
      "model_name": "Anthropic: Claude Opus 4.7",
      "provider_name": "Azure",
      "accuracy": 0.888468,
      "accuracy_std_dev": 0.008504715878068274,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.20666023274410775,
      "avg_duration_per_task_ms": 100029.38846801347,
      "avg_output_tokens_per_task": 8038.521464646465,
      "last_run_timestamp": "2026-10-24T08:44:29.191Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Novita",
      "accuracy": 0.888467875,
      "accuracy_std_dev": 0.02891189332021296,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.10564124999999999,
      "avg_duration_per_task_ms": 446348.9172979798,
      "avg_output_tokens_per_task": 36610.40214646464,
      "last_run_timestamp": "2026-10-24T16:12:09.455Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Phala",
      "accuracy": 0.8880469999999999,
      "accuracy_std_dev": 0.01094299999999998,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.02377390909090909,
      "avg_duration_per_task_ms": 131245.66919191918,
      "avg_output_tokens_per_task": 19627.237373737375,
      "last_run_timestamp": "2026-10-24T08:41:24.320Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Fireworks",
      "accuracy": 0.8880229138638227,
      "accuracy_std_dev": 0.007263570665748784,
      "run_count": 7,
      "total_questions": 1219,
      "avg_cost_per_task_usd": 0.01426234372436423,
      "avg_duration_per_task_ms": 205415.2715340443,
      "avg_output_tokens_per_task": 24591.009844134536,
      "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-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "GMICloud",
      "accuracy": 0.8878528461538461,
      "accuracy_std_dev": 0.03528588596427789,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.11375538578088577,
      "avg_duration_per_task_ms": 500007.0135975136,
      "avg_output_tokens_per_task": 36427.52797202797,
      "last_run_timestamp": "2026-10-27T01:11:27.731Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "NextBit",
      "accuracy": 0.887537924012158,
      "accuracy_std_dev": 0.020250680116973837,
      "run_count": 5,
      "total_questions": 987,
      "avg_cost_per_task_usd": 0.12033997061803446,
      "avg_duration_per_task_ms": 580755.4113475177,
      "avg_output_tokens_per_task": 34357.5663627153,
      "last_run_timestamp": "2026-10-03T16:44:56.199Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-fable-5.1-20260831",
      "model_slug": "anthropic/claude-fable-5.1",
      "model_name": "Anthropic: Claude Fable 5.1",
      "provider_name": "Azure",
      "accuracy": 0.8872051666666666,
      "accuracy_std_dev": 0.024761493125527695,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.1434989882154882,
      "avg_duration_per_task_ms": 42121.643939393936,
      "avg_output_tokens_per_task": 2660.465488215488,
      "last_run_timestamp": "2026-10-01T08:24:34.535Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "AtlasCloud",
      "accuracy": 0.887205,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.11313772222222222,
      "avg_duration_per_task_ms": 834184.0151515151,
      "avg_output_tokens_per_task": 45085.207070707074,
      "last_run_timestamp": "2026-10-27T03:00:05.728Z"
    },
    {
      "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": 0.887042338150289,
      "accuracy_std_dev": 0.010371066166434932,
      "run_count": 7,
      "total_questions": 1384,
      "avg_cost_per_task_usd": 0.010374203757225434,
      "avg_duration_per_task_ms": 78284.89884393064,
      "avg_output_tokens_per_task": 7947.398121387283,
      "last_run_timestamp": "2026-10-01T19:12:42.994Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Phala",
      "accuracy": 0.8869462307692308,
      "accuracy_std_dev": 0.01086717272801863,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.025130035353535356,
      "avg_duration_per_task_ms": 337121.5742035742,
      "avg_output_tokens_per_task": 26419.156565656565,
      "last_run_timestamp": "2026-10-24T08:11:11.359Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Wafer",
      "accuracy": 0.8869174161290322,
      "accuracy_std_dev": 0.06211553768895107,
      "run_count": 20,
      "total_questions": 3720,
      "avg_cost_per_task_usd": 0.00878794623655914,
      "avg_duration_per_task_ms": 150880.02284946237,
      "avg_output_tokens_per_task": 24534.02741935484,
      "last_run_timestamp": "2026-10-29T22:36:27.977Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "auto-routing",
      "accuracy": 0.8867541405460061,
      "accuracy_std_dev": 0.003543823280018365,
      "run_count": 5,
      "total_questions": 989,
      "avg_cost_per_task_usd": 0.03347630637007078,
      "avg_duration_per_task_ms": 767457.7462082912,
      "avg_output_tokens_per_task": 38389.49544994944,
      "last_run_timestamp": "2026-10-24T08:26:00.565Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Inceptron",
      "accuracy": 0.8867506138022562,
      "accuracy_std_dev": 0.020605395430793746,
      "run_count": 8,
      "total_questions": 1507,
      "avg_cost_per_task_usd": 0.1588602813536828,
      "avg_duration_per_task_ms": 705608.4233576643,
      "avg_output_tokens_per_task": 51029.565361645655,
      "last_run_timestamp": "2026-10-24T16:27:11.137Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Inceptron",
      "accuracy": 0.8867197848605577,
      "accuracy_std_dev": 0.007952474956944445,
      "run_count": 13,
      "total_questions": 2510,
      "avg_cost_per_task_usd": 0.09700779960159363,
      "avg_duration_per_task_ms": 461798.00398406375,
      "avg_output_tokens_per_task": 28989.141832669324,
      "last_run_timestamp": "2026-10-24T16:06:30.904Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Fireworks",
      "accuracy": 0.8867051248554914,
      "accuracy_std_dev": 0.021208571037200977,
      "run_count": 5,
      "total_questions": 865,
      "avg_cost_per_task_usd": 0.16137210289017342,
      "avg_duration_per_task_ms": 430828.78150289017,
      "avg_output_tokens_per_task": 40500.787283237,
      "last_run_timestamp": "2026-09-19T08:53:57.162Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8866872307692307,
      "accuracy_std_dev": 0.01902814900716969,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.027383982128982128,
      "avg_duration_per_task_ms": 210380.49145299144,
      "avg_output_tokens_per_task": 25687.588966588966,
      "last_run_timestamp": "2026-10-24T08:19:41.879Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "amazon/nova-micro-v1",
      "model_slug": "amazon/nova-micro-v1",
      "model_name": "Amazon: Nova Micro 1.0",
      "provider_name": "auto-routing",
      "accuracy": 0.8863635000000001,
      "accuracy_std_dev": 0.002525500000000014,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.22157512626262627,
      "avg_duration_per_task_ms": 106335.13383838384,
      "avg_output_tokens_per_task": 8638.217171717171,
      "last_run_timestamp": "2026-10-01T19:10:42.213Z"
    },
    {
      "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": 0.8861492509064539,
      "accuracy_std_dev": 0.01420488521313591,
      "run_count": 7,
      "total_questions": 1379,
      "avg_cost_per_task_usd": 0.12877551051486583,
      "avg_duration_per_task_ms": 1724556.7200870195,
      "avg_output_tokens_per_task": 32019.171863669326,
      "last_run_timestamp": "2026-10-24T16:28:09.530Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "InferenceNet",
      "accuracy": 0.8859760038834952,
      "accuracy_std_dev": 0.023393307207604786,
      "run_count": 8,
      "total_questions": 1545,
      "avg_cost_per_task_usd": 0.02020688673139159,
      "avg_duration_per_task_ms": 264412.80194174754,
      "avg_output_tokens_per_task": 30725.21488673139,
      "last_run_timestamp": "2026-10-24T09:10:08.243Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Novita",
      "accuracy": 0.885662,
      "accuracy_std_dev": 0.014374902133000174,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.11714179755892257,
      "avg_duration_per_task_ms": 572173.5332491583,
      "avg_output_tokens_per_task": 30582.406986531987,
      "last_run_timestamp": "2026-10-24T16:30:32.666Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Sail Research",
      "accuracy": 0.8856471629778674,
      "accuracy_std_dev": 0.009013068317726206,
      "run_count": 8,
      "total_questions": 1491,
      "avg_cost_per_task_usd": 0.010023977867203217,
      "avg_duration_per_task_ms": 270739.96177062375,
      "avg_output_tokens_per_task": 28220.264922870556,
      "last_run_timestamp": "2026-10-24T09:01:43.485Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "sakana/fugu-max",
      "model_slug": "sakana/fugu-max",
      "model_name": "Fugu Max",
      "provider_name": "auto-routing",
      "accuracy": 0.885522,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.07555469191919191,
      "avg_duration_per_task_ms": 225091.94949494948,
      "avg_output_tokens_per_task": 12279.550505050505,
      "last_run_timestamp": "2026-09-15T01:07:47.957Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Minimax",
      "accuracy": 0.885522,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.007081343434343435,
      "avg_duration_per_task_ms": 121564.28282828283,
      "avg_output_tokens_per_task": 11603.333333333334,
      "last_run_timestamp": "2026-10-24T08:10:23.026Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": "Nebius",
      "accuracy": 0.885522,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.05586563636363637,
      "avg_duration_per_task_ms": 471872.54545454547,
      "avg_output_tokens_per_task": 61805.67676767677,
      "last_run_timestamp": "2026-10-27T00:00:04.640Z"
    },
    {
      "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-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "DekaLLM",
      "accuracy": 0.8850689656218401,
      "accuracy_std_dev": 0.014411932672615443,
      "run_count": 5,
      "total_questions": 989,
      "avg_cost_per_task_usd": 0.026149528816986855,
      "avg_duration_per_task_ms": 355797.47320525785,
      "avg_output_tokens_per_task": 26995.41152679474,
      "last_run_timestamp": "2026-10-10T08:22:45.418Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Moonshot AI",
      "accuracy": 0.884890375,
      "accuracy_std_dev": 0.021787089743340537,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.22976492803030302,
      "avg_duration_per_task_ms": 949653.6900252525,
      "avg_output_tokens_per_task": 57582.06123737374,
      "last_run_timestamp": "2026-10-24T16:07:10.471Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "reflection/beam-1.0-501b-a23b-20261016",
      "model_slug": "reflection/beam-1.0-501b-a23b",
      "model_name": "Reflection: Beam 1.0 501B A23B",
      "provider_name": "Together",
      "accuracy": 0.8846800000000001,
      "accuracy_std_dev": 0.0025249999999999995,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.01685389393939394,
      "avg_duration_per_task_ms": 60063.59848484849,
      "avg_output_tokens_per_task": 13866.29292929293,
      "last_run_timestamp": "2026-10-24T08:11:43.555Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.88468,
      "accuracy_std_dev": 0.0008420000000000094,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.1319599772727273,
      "avg_duration_per_task_ms": 249775.19949494948,
      "avg_output_tokens_per_task": 43863.606060606064,
      "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": "PrimeIntellect",
      "accuracy": 0.8845502920892495,
      "accuracy_std_dev": 0.01776525517520085,
      "run_count": 5,
      "total_questions": 986,
      "avg_cost_per_task_usd": 0.040126028397565926,
      "avg_duration_per_task_ms": 169902.08620689655,
      "avg_output_tokens_per_task": 33271.34381338742,
      "last_run_timestamp": "2026-10-29T23:12:01.788Z"
    },
    {
      "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": 0.8843169242648614,
      "accuracy_std_dev": 0.01377803394592024,
      "run_count": 24,
      "total_questions": 4727,
      "avg_cost_per_task_usd": 0.19024729342077426,
      "avg_duration_per_task_ms": 371375.61032367253,
      "avg_output_tokens_per_task": 31453.42183202877,
      "last_run_timestamp": "2026-10-24T16:13:46.475Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-terra-20260709",
      "model_slug": "openai/gpt-5.6-terra",
      "model_name": "OpenAI: GPT-5.6 Terra",
      "provider_name": "auto-routing",
      "accuracy": 0.883838,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.04134767424242424,
      "avg_duration_per_task_ms": 54936.305555555555,
      "avg_output_tokens_per_task": 3316.282828282828,
      "last_run_timestamp": "2026-10-01T19:13:07.806Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-luna-pro-20260922",
      "model_slug": "openai/gpt-6-luna-pro",
      "model_name": "OpenAI: GPT-6 Luna Pro",
      "provider_name": "auto-routing",
      "accuracy": 0.883838,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.008301368686868687,
      "avg_duration_per_task_ms": 72985.27777777778,
      "avg_output_tokens_per_task": 13098.969696969696,
      "last_run_timestamp": "2026-10-01T08:24:38.949Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "reflection/beam-1.0-501b-a23b-20261016",
      "model_slug": "reflection/beam-1.0-501b-a23b",
      "model_name": "Reflection: Beam 1.0 501B A23B",
      "provider_name": "auto-routing",
      "accuracy": 0.883838,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.014282979797979798,
      "avg_duration_per_task_ms": 55075.11111111111,
      "avg_output_tokens_per_task": 13355.585858585859,
      "last_run_timestamp": "2026-10-24T08:02:32.162Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Relace",
      "accuracy": 0.8838218675352877,
      "accuracy_std_dev": 0.0032754605783004037,
      "run_count": 5,
      "total_questions": 921,
      "avg_cost_per_task_usd": 0.10965064820846905,
      "avg_duration_per_task_ms": 399019.9750271444,
      "avg_output_tokens_per_task": 36828.9565689468,
      "last_run_timestamp": "2026-10-24T16:03:25.471Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Nebius",
      "accuracy": 0.88341725,
      "accuracy_std_dev": 0.003834202809907171,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.03130394823232323,
      "avg_duration_per_task_ms": 201934.6616161616,
      "avg_output_tokens_per_task": 25861.328282828283,
      "last_run_timestamp": "2026-10-24T08:53:53.931Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "DeepInfra",
      "accuracy": 0.8833333472727273,
      "accuracy_std_dev": 0.012158692807021237,
      "run_count": 7,
      "total_questions": 1100,
      "avg_cost_per_task_usd": 0.007285390000000001,
      "avg_duration_per_task_ms": 942880.7490909091,
      "avg_output_tokens_per_task": 28917.175454545453,
      "last_run_timestamp": "2026-10-17T08:19:36.266Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Together",
      "accuracy": 0.8828761428571429,
      "accuracy_std_dev": 0.004395219722398795,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.029912125541125538,
      "avg_duration_per_task_ms": 82923.05122655122,
      "avg_output_tokens_per_task": 24775.42784992785,
      "last_run_timestamp": "2026-10-24T08:49:27.543Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Decart",
      "accuracy": 0.8828537731629394,
      "accuracy_std_dev": 0.008622317150279293,
      "run_count": 5,
      "total_questions": 939,
      "avg_cost_per_task_usd": 0.006383692225772098,
      "avg_duration_per_task_ms": 468696.3525026624,
      "avg_output_tokens_per_task": 35072.061767838124,
      "last_run_timestamp": "2026-10-24T08:44:41.546Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-opus-5-20260723",
      "model_slug": "anthropic/claude-opus-5",
      "model_name": "Anthropic: Claude Opus 5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.882503491800857,
      "accuracy_std_dev": 0.009500768601346247,
      "run_count": 35,
      "total_questions": 6769,
      "avg_cost_per_task_usd": 0.14009028970305806,
      "avg_duration_per_task_ms": 63158.250849460775,
      "avg_output_tokens_per_task": 5048.065888609839,
      "last_run_timestamp": "2026-10-24T08:22:55.673Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Alibaba",
      "accuracy": 0.882356599609375,
      "accuracy_std_dev": 0.02350065555877767,
      "run_count": 13,
      "total_questions": 2560,
      "avg_cost_per_task_usd": 0.0909012125,
      "avg_duration_per_task_ms": 446817.64765625,
      "avg_output_tokens_per_task": 32851.274609375,
      "last_run_timestamp": "2026-10-28T04:23:20.591Z"
    },
    {
      "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": "ModelRun",
      "accuracy": 0.8821550000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.026309151515151515,
      "avg_duration_per_task_ms": 202247.18686868687,
      "avg_output_tokens_per_task": 39119.5505050505,
      "last_run_timestamp": "2026-10-29T03:00:03.996Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Phala",
      "accuracy": 0.8821550000000001,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.129083601010101,
      "avg_duration_per_task_ms": 388751.6590909091,
      "avg_output_tokens_per_task": 32112.176767676767,
      "last_run_timestamp": "2026-10-17T16:07:32.047Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "SiliconFlow",
      "accuracy": 0.88215475,
      "accuracy_std_dev": 0.018821912135527027,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.008138106691919191,
      "avg_duration_per_task_ms": 380823.72411616164,
      "avg_output_tokens_per_task": 28681.33712121212,
      "last_run_timestamp": "2026-10-24T16:29:33.430Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "StreamLake",
      "accuracy": 0.882034357142857,
      "accuracy_std_dev": 0.011301592736090304,
      "run_count": 14,
      "total_questions": 2772,
      "avg_cost_per_task_usd": 0.006003468614718615,
      "avg_duration_per_task_ms": 401285.20815295813,
      "avg_output_tokens_per_task": 29705.93326118326,
      "last_run_timestamp": "2026-10-24T08:46:33.521Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "CoreWeave",
      "accuracy": 0.8818485454545456,
      "accuracy_std_dev": 0.01327175411002725,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.18638605096418728,
      "avg_duration_per_task_ms": 321116.0505050505,
      "avg_output_tokens_per_task": 46783.020661157025,
      "last_run_timestamp": "2026-10-24T08:02:27.414Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "Parasail",
      "accuracy": 0.8817956985294118,
      "accuracy_std_dev": 0.012032407447571536,
      "run_count": 8,
      "total_questions": 1496,
      "avg_cost_per_task_usd": 0.06691362366310159,
      "avg_duration_per_task_ms": 318081.7332887701,
      "avg_output_tokens_per_task": 18489.88435828877,
      "last_run_timestamp": "2026-10-24T16:08:29.567Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "CoreWeave",
      "accuracy": 0.8816240000000001,
      "accuracy_std_dev": 0.004701000000000011,
      "run_count": 2,
      "total_questions": 390,
      "avg_cost_per_task_usd": 0.020361825641025642,
      "avg_duration_per_task_ms": 90976.65641025642,
      "avg_output_tokens_per_task": 40519.597435897434,
      "last_run_timestamp": "2026-10-24T08:02:32.160Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.8814814000000001,
      "accuracy_std_dev": 0.011789297597397391,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.007892147474747474,
      "avg_duration_per_task_ms": 236163.43737373737,
      "avg_output_tokens_per_task": 27867.079797979797,
      "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-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Inceptron",
      "accuracy": 0.8814552816901408,
      "accuracy_std_dev": 0.00965776820360461,
      "run_count": 13,
      "total_questions": 2556,
      "avg_cost_per_task_usd": 0.00679829303599374,
      "avg_duration_per_task_ms": 375447.8978873239,
      "avg_output_tokens_per_task": 19618.11854460094,
      "last_run_timestamp": "2026-10-24T08:46:28.764Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "BaseTen",
      "accuracy": 0.881313,
      "accuracy_std_dev": 0.012626000000000026,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.022825318181818183,
      "avg_duration_per_task_ms": 147723.58080808082,
      "avg_output_tokens_per_task": 45335.5101010101,
      "last_run_timestamp": "2026-10-24T08:12:49.570Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Novita",
      "accuracy": 0.8813128749999999,
      "accuracy_std_dev": 0.02861962867700026,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.19908026893939393,
      "avg_duration_per_task_ms": 958354.9299242424,
      "avg_output_tokens_per_task": 58904.72348484849,
      "last_run_timestamp": "2026-10-24T16:17:57.472Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "DeepInfra",
      "accuracy": 0.8811927142857143,
      "accuracy_std_dev": 0.007888829783113527,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.015891973304473304,
      "avg_duration_per_task_ms": 279407.10173160175,
      "avg_output_tokens_per_task": 24276.554834054834,
      "last_run_timestamp": "2026-10-08T22:18:51.642Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "Together",
      "accuracy": 0.8808292124352332,
      "accuracy_std_dev": 0.0020954999999999724,
      "run_count": 2,
      "total_questions": 386,
      "avg_cost_per_task_usd": 0.03105037564766839,
      "avg_duration_per_task_ms": 282019.7538860104,
      "avg_output_tokens_per_task": 61860.47668393782,
      "last_run_timestamp": "2026-10-24T08:07:42.017Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "DeepInfra",
      "accuracy": 0.8807385965665235,
      "accuracy_std_dev": 0.04941426514861386,
      "run_count": 13,
      "total_questions": 2563,
      "avg_cost_per_task_usd": 0.0748979442060086,
      "avg_duration_per_task_ms": 557063.2532188841,
      "avg_output_tokens_per_task": 39438.98985563793,
      "last_run_timestamp": "2026-10-24T16:40:43.534Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "DigitalOcean",
      "accuracy": 0.8805647777777778,
      "accuracy_std_dev": 0.0331443107562395,
      "run_count": 18,
      "total_questions": 3564,
      "avg_cost_per_task_usd": 0.14604359764309763,
      "avg_duration_per_task_ms": 370818.19837261504,
      "avg_output_tokens_per_task": 10967.793209876543,
      "last_run_timestamp": "2026-10-24T08:22:00.565Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-luna-20260922",
      "model_slug": "openai/gpt-6-luna",
      "model_name": "OpenAI: GPT-6 Luna",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.880471,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0032475000000000004,
      "avg_duration_per_task_ms": 58949.292929292926,
      "avg_output_tokens_per_task": 5746.025252525253,
      "last_run_timestamp": "2026-09-23T06:09:34.299Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "auto-routing",
      "accuracy": 0.8801906666666667,
      "accuracy_std_dev": 0.011226983784119793,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.0227881835016835,
      "avg_duration_per_task_ms": 234700.68434343435,
      "avg_output_tokens_per_task": 25264.190235690236,
      "last_run_timestamp": "2026-10-03T09:30:18.263Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Modal",
      "accuracy": 0.8800047366136035,
      "accuracy_std_dev": 0.007207181987717666,
      "run_count": 7,
      "total_questions": 1382,
      "avg_cost_per_task_usd": 0.03244589001447178,
      "avg_duration_per_task_ms": 148073.17149059335,
      "avg_output_tokens_per_task": 26851.936324167873,
      "last_run_timestamp": "2026-10-24T08:26:33.257Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "StreamLake",
      "accuracy": 0.879965,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 193,
      "avg_cost_per_task_usd": 0.006318481865284975,
      "avg_duration_per_task_ms": 107877.19170984456,
      "avg_output_tokens_per_task": 10411.435233160622,
      "last_run_timestamp": "2026-10-27T15:00:04.487Z"
    },
    {
      "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": "anthropic/claude-opus-5-20260723",
      "model_slug": "anthropic/claude-opus-5",
      "model_name": "Anthropic: Claude Opus 5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.8794892499999999,
      "accuracy_std_dev": 0.0183834671245887,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.12937238047138044,
      "avg_duration_per_task_ms": 63385.43771043771,
      "avg_output_tokens_per_task": 4942.099747474747,
      "last_run_timestamp": "2026-10-24T08:27:56.033Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "OpenInference",
      "accuracy": 0.879386,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 76,
      "avg_cost_per_task_usd": 0.03120878947368421,
      "avg_duration_per_task_ms": 274583.63157894736,
      "avg_output_tokens_per_task": 9741.526315789473,
      "last_run_timestamp": "2026-10-09T00:00:06.206Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "DigitalOcean",
      "accuracy": 0.879370525058275,
      "accuracy_std_dev": 0.02189493848211963,
      "run_count": 9,
      "total_questions": 1716,
      "avg_cost_per_task_usd": 0.09564452447552448,
      "avg_duration_per_task_ms": 551582.8269230769,
      "avg_output_tokens_per_task": 28335.815268065267,
      "last_run_timestamp": "2026-10-24T08:13:46.723Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Relace",
      "accuracy": 0.8792376607706888,
      "accuracy_std_dev": 0.006337813294770935,
      "run_count": 8,
      "total_questions": 1583,
      "avg_cost_per_task_usd": 0.016976666456096023,
      "avg_duration_per_task_ms": 306778.0650663297,
      "avg_output_tokens_per_task": 30476.99052432091,
      "last_run_timestamp": "2026-10-24T08:43:17.831Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Venice",
      "accuracy": 0.8790282857142857,
      "accuracy_std_dev": 0.019161523885971404,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.03542920851370852,
      "avg_duration_per_task_ms": 140156.7987012987,
      "avg_output_tokens_per_task": 27224.489898989897,
      "last_run_timestamp": "2026-10-24T08:28:56.672Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.8788318459807073,
      "accuracy_std_dev": 0.05618657197923134,
      "run_count": 16,
      "total_questions": 3110,
      "avg_cost_per_task_usd": 0.010130650482315113,
      "avg_duration_per_task_ms": 433648.34083601285,
      "avg_output_tokens_per_task": 25902.576527331188,
      "last_run_timestamp": "2026-10-29T22:45:55.902Z"
    },
    {
      "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": 0.8787880000000001,
      "accuracy_std_dev": 0.007273551310513108,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.009556249158249158,
      "avg_duration_per_task_ms": 52070.99494949495,
      "avg_output_tokens_per_task": 7331.292929292929,
      "last_run_timestamp": "2026-10-30T00:00:06.145Z"
    },
    {
      "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": "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": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Cloudflare",
      "accuracy": 0.8787668916494136,
      "accuracy_std_dev": 0.03281417517628624,
      "run_count": 8,
      "total_questions": 1449,
      "avg_cost_per_task_usd": 0.2005868881987578,
      "avg_duration_per_task_ms": 1332066.3747412008,
      "avg_output_tokens_per_task": 50003.33195307109,
      "last_run_timestamp": "2026-10-24T16:26:57.307Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Parasail",
      "accuracy": 0.8786261545293074,
      "accuracy_std_dev": 0.0068412882173039926,
      "run_count": 6,
      "total_questions": 1126,
      "avg_cost_per_task_usd": 0.11316016607460035,
      "avg_duration_per_task_ms": 476858.84191829484,
      "avg_output_tokens_per_task": 32318.380994671403,
      "last_run_timestamp": "2026-10-17T16:07:16.258Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-opus-5-20260723",
      "model_slug": "anthropic/claude-opus-5",
      "model_name": "Anthropic: Claude Opus 5",
      "provider_name": "auto-routing",
      "accuracy": 0.87780575,
      "accuracy_std_dev": 0.016897481104812654,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.1286623606902357,
      "avg_duration_per_task_ms": 62852.628367003366,
      "avg_output_tokens_per_task": 4923.856481481482,
      "last_run_timestamp": "2026-10-24T08:15:51.136Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8776513109987357,
      "accuracy_std_dev": 0.007333672175410435,
      "run_count": 12,
      "total_questions": 2373,
      "avg_cost_per_task_usd": 0.11767066118836916,
      "avg_duration_per_task_ms": 752365.9431099873,
      "avg_output_tokens_per_task": 30843.225874420565,
      "last_run_timestamp": "2026-10-24T16:07:46.989Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Mancer 2",
      "accuracy": 0.8775252499999999,
      "accuracy_std_dev": 0.03188403682128148,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.01210295664983165,
      "avg_duration_per_task_ms": 348579.569023569,
      "avg_output_tokens_per_task": 22092.35732323232,
      "last_run_timestamp": "2026-10-24T08:45:03.893Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Ionstream",
      "accuracy": 0.8774407999999999,
      "accuracy_std_dev": 0.00954717737134911,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.03921924343434343,
      "avg_duration_per_task_ms": 242470.04242424242,
      "avg_output_tokens_per_task": 37354.009090909094,
      "last_run_timestamp": "2026-10-21T11:23:15.101Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Parasail",
      "accuracy": 0.8773447142857143,
      "accuracy_std_dev": 0.012707857195050472,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.03401366883116883,
      "avg_duration_per_task_ms": 154330.79004329004,
      "avg_output_tokens_per_task": 28188.37950937951,
      "last_run_timestamp": "2026-10-24T09:09:44.617Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Reka",
      "accuracy": 0.8772070529695024,
      "accuracy_std_dev": 0.03893659390197737,
      "run_count": 4,
      "total_questions": 623,
      "avg_cost_per_task_usd": 0.09425666452648475,
      "avg_duration_per_task_ms": 461375.2568218299,
      "avg_output_tokens_per_task": 23922.329052969504,
      "last_run_timestamp": "2026-09-30T11:37:23.219Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-luna-20260922",
      "model_slug": "openai/gpt-6-luna",
      "model_name": "OpenAI: GPT-6 Luna",
      "provider_name": "Azure",
      "accuracy": 0.8771043333333335,
      "accuracy_std_dev": 0.007896528133018712,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.003143374579124579,
      "avg_duration_per_task_ms": 35450.33585858586,
      "avg_output_tokens_per_task": 5733.669191919192,
      "last_run_timestamp": "2026-10-01T08:25:15.234Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "GMICloud",
      "accuracy": 0.877104,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.07509321212121212,
      "avg_duration_per_task_ms": 263175.32828282827,
      "avg_output_tokens_per_task": 20738.459595959597,
      "last_run_timestamp": "2026-08-29T19:12:43.653Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Parasail",
      "accuracy": 0.8770437482014388,
      "accuracy_std_dev": 0.016843957465790968,
      "run_count": 8,
      "total_questions": 1529,
      "avg_cost_per_task_usd": 0.1855886553302812,
      "avg_duration_per_task_ms": 777107.6200130804,
      "avg_output_tokens_per_task": 53473.10595160235,
      "last_run_timestamp": "2026-10-24T16:28:49.711Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "SiliconFlow",
      "accuracy": 0.876999,
      "accuracy_std_dev": 0.06406231714705611,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.1307801066919192,
      "avg_duration_per_task_ms": 1601351.4103535353,
      "avg_output_tokens_per_task": 55537.77272727273,
      "last_run_timestamp": "2026-10-24T16:02:47.376Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "StreamLake",
      "accuracy": 0.8769172222222222,
      "accuracy_std_dev": 0.01657946057812884,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.02028404994388328,
      "avg_duration_per_task_ms": 725618.4191919192,
      "avg_output_tokens_per_task": 35595.71436588103,
      "last_run_timestamp": "2026-10-24T16:27:17.186Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "ModelRun",
      "accuracy": 0.876893965909091,
      "accuracy_std_dev": 0.02910034944176192,
      "run_count": 7,
      "total_questions": 1232,
      "avg_cost_per_task_usd": 0.09717019561688313,
      "avg_duration_per_task_ms": 193376.24837662338,
      "avg_output_tokens_per_task": 26307.616071428572,
      "last_run_timestamp": "2026-09-26T09:19:13.012Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Crusoe",
      "accuracy": 0.8767675999999999,
      "accuracy_std_dev": 0.014101374310328764,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.042260786868686874,
      "avg_duration_per_task_ms": 210147.5181818182,
      "avg_output_tokens_per_task": 35072.08888888889,
      "last_run_timestamp": "2026-10-24T08:50:45.433Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "AkashML",
      "accuracy": 0.8765966933797908,
      "accuracy_std_dev": 0.020189124617118887,
      "run_count": 6,
      "total_questions": 1148,
      "avg_cost_per_task_usd": 0.004125730836236933,
      "avg_duration_per_task_ms": 453615.28397212544,
      "avg_output_tokens_per_task": 16054.564459930314,
      "last_run_timestamp": "2026-09-05T08:24:32.148Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-fable-5.1-20260831",
      "model_slug": "anthropic/claude-fable-5.1",
      "model_name": "Anthropic: Claude Fable 5.1",
      "provider_name": "Google",
      "accuracy": 0.8765431666666668,
      "accuracy_std_dev": 0.03217293038884639,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.11818556060606061,
      "avg_duration_per_task_ms": 28260.9031986532,
      "avg_output_tokens_per_task": 2161.995791245791,
      "last_run_timestamp": "2026-10-01T08:25:56.890Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Parasail",
      "accuracy": 0.8765007667238421,
      "accuracy_std_dev": 0.009087011145341243,
      "run_count": 3,
      "total_questions": 583,
      "avg_cost_per_task_usd": 0.09344886277873071,
      "avg_duration_per_task_ms": 568169.3550600343,
      "avg_output_tokens_per_task": 27396.61578044597,
      "last_run_timestamp": "2026-08-22T09:51:24.721Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Reka",
      "accuracy": 0.8763533692307693,
      "accuracy_std_dev": 0.006821595577445394,
      "run_count": 3,
      "total_questions": 585,
      "avg_cost_per_task_usd": 0.027450420512820508,
      "avg_duration_per_task_ms": 188376.252991453,
      "avg_output_tokens_per_task": 30391.492307692308,
      "last_run_timestamp": "2026-10-23T06:17:23.826Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.8762625,
      "accuracy_std_dev": 0.005892500000000023,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.009382047979797978,
      "avg_duration_per_task_ms": 83118.84595959596,
      "avg_output_tokens_per_task": 7689.015151515152,
      "last_run_timestamp": "2026-10-01T19:14:22.294Z"
    },
    {
      "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": 0.8762115941845765,
      "accuracy_std_dev": 0.007231120010715849,
      "run_count": 8,
      "total_questions": 1582,
      "avg_cost_per_task_usd": 0.12366235651074589,
      "avg_duration_per_task_ms": 903569.9494310999,
      "avg_output_tokens_per_task": 43400.72819216182,
      "last_run_timestamp": "2026-10-24T16:19:29.598Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Z.AI",
      "accuracy": 0.8760746752655538,
      "accuracy_std_dev": 0.011470076619186123,
      "run_count": 10,
      "total_questions": 1977,
      "avg_cost_per_task_usd": 0.009704542235710673,
      "avg_duration_per_task_ms": 429066.8811330298,
      "avg_output_tokens_per_task": 22619.601416287303,
      "last_run_timestamp": "2026-10-24T08:41:28.565Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "CoreWeave",
      "accuracy": 0.8760220000000001,
      "accuracy_std_dev": 0.014151103773204405,
      "run_count": 14,
      "total_questions": 2772,
      "avg_cost_per_task_usd": 0.11097613275613274,
      "avg_duration_per_task_ms": 457617.5707070707,
      "avg_output_tokens_per_task": 31565.57683982684,
      "last_run_timestamp": "2026-10-24T16:02:47.387Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Cloudflare",
      "accuracy": 0.8754205,
      "accuracy_std_dev": 0.0016835000000000044,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.04156198232323233,
      "avg_duration_per_task_ms": 309570.8888888889,
      "avg_output_tokens_per_task": 39159.9494949495,
      "last_run_timestamp": "2026-10-17T08:26:18.265Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "DeepInfra",
      "accuracy": 0.8754205,
      "accuracy_std_dev": 0.0016835000000000044,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.01960839393939394,
      "avg_duration_per_task_ms": 110709.8308080808,
      "avg_output_tokens_per_task": 43264.59848484849,
      "last_run_timestamp": "2026-10-24T08:12:37.561Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "DigitalOcean",
      "accuracy": 0.8752913076923076,
      "accuracy_std_dev": 0.04445532892411749,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.09291413209013209,
      "avg_duration_per_task_ms": 618152.5388500389,
      "avg_output_tokens_per_task": 41235.80536130536,
      "last_run_timestamp": "2026-10-24T16:04:29.298Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "Phala",
      "accuracy": 0.8749998749999999,
      "accuracy_std_dev": 0.005171981690742456,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.0810850505050505,
      "avg_duration_per_task_ms": 406962.2190656566,
      "avg_output_tokens_per_task": 23056.515151515152,
      "last_run_timestamp": "2026-10-24T16:39:19.451Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Venice",
      "accuracy": 0.8749675384615386,
      "accuracy_std_dev": 0.013017813792677657,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.006246133255633255,
      "avg_duration_per_task_ms": 265528.69425019424,
      "avg_output_tokens_per_task": 18684.931235431235,
      "last_run_timestamp": "2026-10-24T08:20:25.161Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Decart",
      "accuracy": 0.8749014196217493,
      "accuracy_std_dev": 0.02259480542425625,
      "run_count": 5,
      "total_questions": 846,
      "avg_cost_per_task_usd": 0.12702199408983453,
      "avg_duration_per_task_ms": 548158.9716312056,
      "avg_output_tokens_per_task": 48997.07092198582,
      "last_run_timestamp": "2026-10-03T17:01:24.374Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Baidu",
      "accuracy": 0.8747894999999999,
      "accuracy_std_dev": 0.02495521020748971,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.1433024273989899,
      "avg_duration_per_task_ms": 926726.0309343435,
      "avg_output_tokens_per_task": 61552.77146464647,
      "last_run_timestamp": "2026-10-24T16:07:10.480Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Reka",
      "accuracy": 0.8746087667870037,
      "accuracy_std_dev": 0.007592482877484693,
      "run_count": 7,
      "total_questions": 1385,
      "avg_cost_per_task_usd": 0.2179415407942238,
      "avg_duration_per_task_ms": 1239989.4288808664,
      "avg_output_tokens_per_task": 48485.83321299639,
      "last_run_timestamp": "2026-10-24T16:18:35.554Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Reka",
      "accuracy": 0.8745790000000001,
      "accuracy_std_dev": 0.04079651270022963,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.010708270202020202,
      "avg_duration_per_task_ms": 125365.15404040404,
      "avg_output_tokens_per_task": 15791.75,
      "last_run_timestamp": "2026-10-24T08:20:32.397Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Together",
      "accuracy": 0.874579,
      "accuracy_std_dev": 0.005892000000000008,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.018065838383838382,
      "avg_duration_per_task_ms": 168823.5202020202,
      "avg_output_tokens_per_task": 29810.62373737374,
      "last_run_timestamp": "2026-10-24T08:41:24.133Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Makora",
      "accuracy": 0.8745317312013828,
      "accuracy_std_dev": 0.011186512454241005,
      "run_count": 6,
      "total_questions": 1157,
      "avg_cost_per_task_usd": 0.0043274416594641305,
      "avg_duration_per_task_ms": 185028.52031114954,
      "avg_output_tokens_per_task": 21870.51080380294,
      "last_run_timestamp": "2026-09-26T08:53:59.525Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Makora",
      "accuracy": 0.8740843414084506,
      "accuracy_std_dev": 0.004005737471597214,
      "run_count": 9,
      "total_questions": 1775,
      "avg_cost_per_task_usd": 0.044288470985915485,
      "avg_duration_per_task_ms": 203522.5785915493,
      "avg_output_tokens_per_task": 38212.04394366197,
      "last_run_timestamp": "2026-10-24T08:44:12.949Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Together",
      "accuracy": 0.8740044607508533,
      "accuracy_std_dev": 0.007605951280996289,
      "run_count": 3,
      "total_questions": 586,
      "avg_cost_per_task_usd": 0.10401818771331058,
      "avg_duration_per_task_ms": 459122.319112628,
      "avg_output_tokens_per_task": 30022.856655290103,
      "last_run_timestamp": "2026-08-22T16:49:15.579Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "GMICloud",
      "accuracy": 0.8739801189873418,
      "accuracy_std_dev": 0.018911107403381266,
      "run_count": 6,
      "total_questions": 1185,
      "avg_cost_per_task_usd": 0.0366216776371308,
      "avg_duration_per_task_ms": 691274.4362869199,
      "avg_output_tokens_per_task": 42052.38227848101,
      "last_run_timestamp": "2026-10-27T01:33:21.133Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Together",
      "accuracy": 0.8739276971507354,
      "accuracy_std_dev": 0.014730569376258348,
      "run_count": 11,
      "total_questions": 2176,
      "avg_cost_per_task_usd": 0.09131274172794118,
      "avg_duration_per_task_ms": 205017.0390625,
      "avg_output_tokens_per_task": 23097.8125,
      "last_run_timestamp": "2026-10-24T08:48:36.491Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "CoreWeave",
      "accuracy": 0.8737371428571428,
      "accuracy_std_dev": 0.01647031431245381,
      "run_count": 14,
      "total_questions": 2772,
      "avg_cost_per_task_usd": 0.008272517676767677,
      "avg_duration_per_task_ms": 219776.6417748918,
      "avg_output_tokens_per_task": 29274.76370851371,
      "last_run_timestamp": "2026-10-24T09:11:12.594Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "SambaNova",
      "accuracy": 0.8737370000000001,
      "accuracy_std_dev": 0.006734000000000018,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.03531422474747475,
      "avg_duration_per_task_ms": 145858.30808080808,
      "avg_output_tokens_per_task": 29116.540404040403,
      "last_run_timestamp": "2026-10-24T08:59:28.837Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "auto-routing",
      "accuracy": 0.8737370000000001,
      "accuracy_std_dev": 0.006734000000000018,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.023037462121212122,
      "avg_duration_per_task_ms": 506854.46717171714,
      "avg_output_tokens_per_task": 48027.71464646464,
      "last_run_timestamp": "2026-10-24T08:01:41.049Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-luna-20260922",
      "model_slug": "openai/gpt-6-luna",
      "model_name": "OpenAI: GPT-6 Luna",
      "provider_name": "auto-routing",
      "accuracy": 0.873737,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0029526111111111116,
      "avg_duration_per_task_ms": 49722.71717171717,
      "avg_output_tokens_per_task": 5746.707070707071,
      "last_run_timestamp": "2026-10-01T08:23:37.659Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.7-plus-20260602",
      "model_slug": "qwen/qwen3.7-plus",
      "model_name": "Qwen: Qwen3.7 Plus",
      "provider_name": "Alibaba",
      "accuracy": 0.873737,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.045554454545454544,
      "avg_duration_per_task_ms": 605480.6616161616,
      "avg_output_tokens_per_task": 35385.22727272727,
      "last_run_timestamp": "2026-10-01T08:40:06.225Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Venice",
      "accuracy": 0.873716393364929,
      "accuracy_std_dev": 0.03410588644533702,
      "run_count": 9,
      "total_questions": 1688,
      "avg_cost_per_task_usd": 0.012903949052132702,
      "avg_duration_per_task_ms": 360645.17417061614,
      "avg_output_tokens_per_task": 25970.889218009477,
      "last_run_timestamp": "2026-10-24T08:43:07.915Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Cloudflare",
      "accuracy": 0.8736468057110863,
      "accuracy_std_dev": 0.03036521292219109,
      "run_count": 10,
      "total_questions": 1786,
      "avg_cost_per_task_usd": 0.025540696528555434,
      "avg_duration_per_task_ms": 360273.59238521836,
      "avg_output_tokens_per_task": 24510.389137737962,
      "last_run_timestamp": "2026-10-24T16:24:04.274Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Venice",
      "accuracy": 0.8733163333333334,
      "accuracy_std_dev": 0.030039827993663497,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.17326605218855218,
      "avg_duration_per_task_ms": 357944.43055555556,
      "avg_output_tokens_per_task": 39522.67929292929,
      "last_run_timestamp": "2026-10-24T16:27:20.826Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Cloudflare",
      "accuracy": 0.8732174942820837,
      "accuracy_std_dev": 0.03602354128311638,
      "run_count": 12,
      "total_questions": 2361,
      "avg_cost_per_task_usd": 0.10224943286742906,
      "avg_duration_per_task_ms": 640310.1185938162,
      "avg_output_tokens_per_task": 25410.69970351546,
      "last_run_timestamp": "2026-10-24T16:26:55.135Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "CoreWeave",
      "accuracy": 0.8722641250000001,
      "accuracy_std_dev": 0.019820282720218067,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.19987492803030302,
      "avg_duration_per_task_ms": 567054.6963383838,
      "avg_output_tokens_per_task": 58481.14962121212,
      "last_run_timestamp": "2026-10-24T16:09:40.215Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Together",
      "accuracy": 0.8721539397539244,
      "accuracy_std_dev": 0.008970575607073516,
      "run_count": 12,
      "total_questions": 2357,
      "avg_cost_per_task_usd": 0.008344311412812899,
      "avg_duration_per_task_ms": 339433.18073822657,
      "avg_output_tokens_per_task": 29494.21552821383,
      "last_run_timestamp": "2026-10-17T08:23:51.334Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "AtlasCloud",
      "accuracy": 0.872053625,
      "accuracy_std_dev": 0.02370397511989021,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.1190935284090909,
      "avg_duration_per_task_ms": 606301.1369949495,
      "avg_output_tokens_per_task": 35231.555555555555,
      "last_run_timestamp": "2026-10-24T16:26:49.397Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Alibaba",
      "accuracy": 0.8720047896405919,
      "accuracy_std_dev": 0.013430501280667074,
      "run_count": 10,
      "total_questions": 1892,
      "avg_cost_per_task_usd": 0.00839126691331924,
      "avg_duration_per_task_ms": 234781.10200845666,
      "avg_output_tokens_per_task": 29482.704545454544,
      "last_run_timestamp": "2026-10-28T03:34:12.001Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Morph",
      "accuracy": 0.871875,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 160,
      "avg_cost_per_task_usd": 0.04507655,
      "avg_duration_per_task_ms": 1504372.61875,
      "avg_output_tokens_per_task": 18693.5375,
      "last_run_timestamp": "2026-08-27T15:00:05.074Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.8717731666666667,
      "accuracy_std_dev": 0.004601771648567064,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.00424808627946128,
      "avg_duration_per_task_ms": 39417.8244949495,
      "avg_output_tokens_per_task": 7731.690235690236,
      "last_run_timestamp": "2026-10-24T08:30:49.056Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Wafer",
      "accuracy": 0.871717,
      "accuracy_std_dev": 0.007924857828377761,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.007889449494949494,
      "avg_duration_per_task_ms": 615342.6777777778,
      "avg_output_tokens_per_task": 44673.57070707071,
      "last_run_timestamp": "2026-10-29T22:53:02.505Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "DeepInfra",
      "accuracy": 0.8714859264426126,
      "accuracy_std_dev": 0.00752103815972237,
      "run_count": 8,
      "total_questions": 1577,
      "avg_cost_per_task_usd": 0.06945589727330374,
      "avg_duration_per_task_ms": 276858.71528218134,
      "avg_output_tokens_per_task": 11308.629676601142,
      "last_run_timestamp": "2026-10-10T08:22:14.215Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "auto-routing",
      "accuracy": 0.8713054444444444,
      "accuracy_std_dev": 0.011226478265552096,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.08585465039281706,
      "avg_duration_per_task_ms": 191685.55331088664,
      "avg_output_tokens_per_task": 13992.872615039281,
      "last_run_timestamp": "2026-10-24T08:29:16.805Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "DigitalOcean",
      "accuracy": 0.871212,
      "accuracy_std_dev": 0.0008420000000000094,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.017874631313131314,
      "avg_duration_per_task_ms": 86992.41414141415,
      "avg_output_tokens_per_task": 29675.651515151516,
      "last_run_timestamp": "2026-10-24T08:40:56.206Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "OpenInference",
      "accuracy": 0.8710915714285715,
      "accuracy_std_dev": 0.014505946572326232,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.021487838383838383,
      "avg_duration_per_task_ms": 272335.1349206349,
      "avg_output_tokens_per_task": 35144.1277056277,
      "last_run_timestamp": "2026-10-24T08:13:46.727Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "thinkingmachines/inkling-small-20260730",
      "model_slug": "thinkingmachines/inkling-small",
      "model_name": "Thinking Machines: Inkling Small",
      "provider_name": "Together",
      "accuracy": 0.8710116425661915,
      "accuracy_std_dev": 0.01025225770257459,
      "run_count": 5,
      "total_questions": 982,
      "avg_cost_per_task_usd": 0.025059682281059068,
      "avg_duration_per_task_ms": 328330.6405295316,
      "avg_output_tokens_per_task": 20642.19348268839,
      "last_run_timestamp": "2026-09-12T08:54:29.103Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Fireworks",
      "accuracy": 0.8706863196512894,
      "accuracy_std_dev": 0.010805462536728977,
      "run_count": 14,
      "total_questions": 2753,
      "avg_cost_per_task_usd": 0.047070880130766435,
      "avg_duration_per_task_ms": 208861.94733018524,
      "avg_output_tokens_per_task": 36299.43697784236,
      "last_run_timestamp": "2026-10-24T09:06:03.113Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "Alibaba",
      "accuracy": 0.870484655811342,
      "accuracy_std_dev": 0.008268565007542176,
      "run_count": 9,
      "total_questions": 1781,
      "avg_cost_per_task_usd": 0.0494731560920831,
      "avg_duration_per_task_ms": 288367.13307130826,
      "avg_output_tokens_per_task": 21014.321167883212,
      "last_run_timestamp": "2026-10-28T03:32:25.744Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Parasail",
      "accuracy": 0.8703702499999999,
      "accuracy_std_dev": 0.005050541672682263,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.00817759595959596,
      "avg_duration_per_task_ms": 373579.7626262626,
      "avg_output_tokens_per_task": 28859.406565656565,
      "last_run_timestamp": "2026-10-24T16:12:07.782Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "thinkingmachines/inkling-small-20260730",
      "model_slug": "thinkingmachines/inkling-small",
      "model_name": "Thinking Machines: Inkling Small",
      "provider_name": "BaseTen",
      "accuracy": 0.8703701666666666,
      "accuracy_std_dev": 0.006941288747215641,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.024583196127946128,
      "avg_duration_per_task_ms": 150981.7457912458,
      "avg_output_tokens_per_task": 20283.477272727272,
      "last_run_timestamp": "2026-09-19T08:49:51.735Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8701915282758622,
      "accuracy_std_dev": 0.09957007568085358,
      "run_count": 11,
      "total_questions": 2175,
      "avg_cost_per_task_usd": 0.027101170114942532,
      "avg_duration_per_task_ms": 202993.58068965518,
      "avg_output_tokens_per_task": 23323.951724137933,
      "last_run_timestamp": "2026-10-24T16:28:20.445Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Ionstream",
      "accuracy": 0.8699495,
      "accuracy_std_dev": 0.002104500000000009,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.00502325,
      "avg_duration_per_task_ms": 389434.8611111111,
      "avg_output_tokens_per_task": 17944.10101010101,
      "last_run_timestamp": "2026-08-08T10:05:07.443Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Baidu",
      "accuracy": 0.86994925,
      "accuracy_std_dev": 0.017018176890827654,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.03481110921717172,
      "avg_duration_per_task_ms": 505625.7146464646,
      "avg_output_tokens_per_task": 35174.42866161616,
      "last_run_timestamp": "2026-10-24T16:26:47.554Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "Together",
      "accuracy": 0.869809111111111,
      "accuracy_std_dev": 0.008765689330115383,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.07596271436588103,
      "avg_duration_per_task_ms": 102049.09147025814,
      "avg_output_tokens_per_task": 12345.802469135802,
      "last_run_timestamp": "2026-10-24T08:49:10.928Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "BaseTen",
      "accuracy": 0.8697483287395923,
      "accuracy_std_dev": 0.025647694972356895,
      "run_count": 18,
      "total_questions": 3483,
      "avg_cost_per_task_usd": 0.03920849038185472,
      "avg_duration_per_task_ms": 78569.80017226528,
      "avg_output_tokens_per_task": 25460.256675279932,
      "last_run_timestamp": "2026-10-24T08:27:04.531Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "CoreWeave",
      "accuracy": 0.8695285000000001,
      "accuracy_std_dev": 0.002525500000000014,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.10538721464646464,
      "avg_duration_per_task_ms": 581981.0101010101,
      "avg_output_tokens_per_task": 34695.7398989899,
      "last_run_timestamp": "2026-08-08T09:03:54.762Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.1-20251113",
      "model_slug": "openai/gpt-5.1",
      "model_name": "OpenAI: GPT-5.1",
      "provider_name": "OpenAI",
      "accuracy": 0.8695284999999999,
      "accuracy_std_dev": 0.005892500000000023,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.2202375707070707,
      "avg_duration_per_task_ms": 264875.98484848486,
      "avg_output_tokens_per_task": 21926.78787878788,
      "last_run_timestamp": "2026-10-01T16:12:49.000Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "DeepInfra",
      "accuracy": 0.8694541060070672,
      "accuracy_std_dev": 0.017661177225418892,
      "run_count": 9,
      "total_questions": 1698,
      "avg_cost_per_task_usd": 0.07418639340400471,
      "avg_duration_per_task_ms": 396957.8833922261,
      "avg_output_tokens_per_task": 24048.35217903416,
      "last_run_timestamp": "2026-10-24T08:28:10.076Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Sail Research",
      "accuracy": 0.8692809647058825,
      "accuracy_std_dev": 0.009104176033557385,
      "run_count": 4,
      "total_questions": 765,
      "avg_cost_per_task_usd": 0.0047775359477124175,
      "avg_duration_per_task_ms": 482494.045751634,
      "avg_output_tokens_per_task": 26353.58431372549,
      "last_run_timestamp": "2026-08-22T16:48:48.690Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "Anthropic",
      "accuracy": 0.8691077499999998,
      "accuracy_std_dev": 0.005877315899924065,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.003982401515151515,
      "avg_duration_per_task_ms": 37577.292929292926,
      "avg_output_tokens_per_task": 7724.75,
      "last_run_timestamp": "2026-10-24T08:16:38.969Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-5-fable-20260609",
      "model_slug": "anthropic/claude-fable-5",
      "model_name": "Anthropic: Claude Fable 5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.8689927272727271,
      "accuracy_std_dev": 0.03917244891523285,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.17977295500459137,
      "avg_duration_per_task_ms": 45320.05463728191,
      "avg_output_tokens_per_task": 3392.5229568411387,
      "last_run_timestamp": "2026-10-24T08:48:02.790Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "DeepInfra",
      "accuracy": 0.8689616459016394,
      "accuracy_std_dev": 0.006730575179500894,
      "run_count": 9,
      "total_questions": 1525,
      "avg_cost_per_task_usd": 0.05556747672131148,
      "avg_duration_per_task_ms": 433024.3068852459,
      "avg_output_tokens_per_task": 18421.848524590165,
      "last_run_timestamp": "2026-10-24T16:13:46.398Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.868687,
      "accuracy_std_dev": 0.006299080210633947,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.004035915404040404,
      "avg_duration_per_task_ms": 38787.454545454544,
      "avg_output_tokens_per_task": 7838.214646464647,
      "last_run_timestamp": "2026-10-24T08:07:26.994Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "Azure",
      "accuracy": 0.8686867500000001,
      "accuracy_std_dev": 0.005583575528950976,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.004182061868686869,
      "avg_duration_per_task_ms": 36016.292929292926,
      "avg_output_tokens_per_task": 7729.580176767677,
      "last_run_timestamp": "2026-10-24T08:19:41.880Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "auto-routing",
      "accuracy": 0.8686867500000001,
      "accuracy_std_dev": 0.016015464836760137,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.038159512626262625,
      "avg_duration_per_task_ms": 691231.6849747475,
      "avg_output_tokens_per_task": 34768.117424242424,
      "last_run_timestamp": "2026-10-24T16:16:05.009Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.1-20251113",
      "model_slug": "openai/gpt-5.1",
      "model_name": "OpenAI: GPT-5.1",
      "provider_name": "Azure",
      "accuracy": 0.8686867500000001,
      "accuracy_std_dev": 0.009143975622643609,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.2541581641414141,
      "avg_duration_per_task_ms": 251149.7840909091,
      "avg_output_tokens_per_task": 24116.972222222223,
      "last_run_timestamp": "2026-10-01T16:14:35.237Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "auto-routing",
      "accuracy": 0.8686866666666666,
      "accuracy_std_dev": 0.004955863351178644,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.004029786195286195,
      "avg_duration_per_task_ms": 38652.87037037037,
      "avg_output_tokens_per_task": 7824.787878787879,
      "last_run_timestamp": "2026-10-24T08:01:41.043Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Nebius",
      "accuracy": 0.8686865,
      "accuracy_std_dev": 0.0016835000000000044,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.01774992424242424,
      "avg_duration_per_task_ms": 402650.36616161617,
      "avg_output_tokens_per_task": 29383.83585858586,
      "last_run_timestamp": "2026-10-24T08:28:10.100Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "Chutes",
      "accuracy": 0.8686865,
      "accuracy_std_dev": 0.0016835000000000044,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.06319552525252525,
      "avg_duration_per_task_ms": 266969.2752525252,
      "avg_output_tokens_per_task": 20992.954545454544,
      "last_run_timestamp": "2026-08-15T15:55:02.752Z"
    },
    {
      "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": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Z.AI",
      "accuracy": 0.8682876231003039,
      "accuracy_std_dev": 0.028807543557374,
      "run_count": 10,
      "total_questions": 1974,
      "avg_cost_per_task_usd": 0.10277313323201621,
      "avg_duration_per_task_ms": 355978.4974670719,
      "avg_output_tokens_per_task": 23116.26595744681,
      "last_run_timestamp": "2026-10-24T08:57:08.885Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "Modal",
      "accuracy": 0.8682058571428571,
      "accuracy_std_dev": 0.007673379706651383,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.07688844300144301,
      "avg_duration_per_task_ms": 83582.13492063493,
      "avg_output_tokens_per_task": 12574.14357864358,
      "last_run_timestamp": "2026-10-10T08:21:46.324Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "Novita",
      "accuracy": 0.8676766000000001,
      "accuracy_std_dev": 0.004205309600968765,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.07033428585858588,
      "avg_duration_per_task_ms": 232782.28383838382,
      "avg_output_tokens_per_task": 11454.874242424243,
      "last_run_timestamp": "2026-10-24T08:34:00.054Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "thinkingmachines/inkling-small-20260730",
      "model_slug": "thinkingmachines/inkling-small",
      "model_name": "Thinking Machines: Inkling Small",
      "provider_name": "DeepInfra",
      "accuracy": 0.867634625,
      "accuracy_std_dev": 0.007431114989311851,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.025694838383838382,
      "avg_duration_per_task_ms": 248942.36553030304,
      "avg_output_tokens_per_task": 21167.008207070707,
      "last_run_timestamp": "2026-10-08T22:14:16.087Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "Google",
      "accuracy": 0.8675645,
      "accuracy_std_dev": 0.008228295808772726,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.004289666245791245,
      "avg_duration_per_task_ms": 35945.74074074074,
      "avg_output_tokens_per_task": 7815.2925084175085,
      "last_run_timestamp": "2026-10-24T08:43:41.154Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-luna-20260709",
      "model_slug": "openai/gpt-5.6-luna",
      "model_name": "OpenAI: GPT-5.6 Luna",
      "provider_name": "OpenAI",
      "accuracy": 0.8675643333333334,
      "accuracy_std_dev": 0.0028614651647170092,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.009448984848484849,
      "avg_duration_per_task_ms": 79377.02525252526,
      "avg_output_tokens_per_task": 7744.79797979798,
      "last_run_timestamp": "2026-10-02T23:19:10.742Z"
    },
    {
      "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": 0.8674239999999999,
      "accuracy_std_dev": 0.03557408983609655,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.18567804966329968,
      "avg_duration_per_task_ms": 45738.964225589225,
      "avg_output_tokens_per_task": 3505.928872053872,
      "last_run_timestamp": "2026-10-24T08:45:03.898Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.867003,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.013525050505050505,
      "avg_duration_per_task_ms": 181228.28787878787,
      "avg_output_tokens_per_task": 22673.38888888889,
      "last_run_timestamp": "2026-10-24T08:02:32.162Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.1-20251113",
      "model_slug": "openai/gpt-5.1",
      "model_name": "OpenAI: GPT-5.1",
      "provider_name": "auto-routing",
      "accuracy": 0.867003,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.22265377272727274,
      "avg_duration_per_task_ms": 265427.9646464646,
      "avg_output_tokens_per_task": 22169.171717171717,
      "last_run_timestamp": "2026-10-01T16:12:54.241Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "auto-routing",
      "accuracy": 0.8666971818181819,
      "accuracy_std_dev": 0.02628554379335525,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.007686401744719926,
      "avg_duration_per_task_ms": 446263.08815426996,
      "avg_output_tokens_per_task": 25752.568411386594,
      "last_run_timestamp": "2026-10-24T08:15:15.304Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Phala",
      "accuracy": 0.8664308674130818,
      "accuracy_std_dev": 0.027802409922633914,
      "run_count": 9,
      "total_questions": 1697,
      "avg_cost_per_task_usd": 0.011504347672362995,
      "avg_duration_per_task_ms": 435596.48084855627,
      "avg_output_tokens_per_task": 25872.670005892753,
      "last_run_timestamp": "2026-10-24T08:11:43.556Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-5-fable-20260609",
      "model_slug": "anthropic/claude-fable-5",
      "model_name": "Anthropic: Claude Fable 5",
      "provider_name": "auto-routing",
      "accuracy": 0.8663018333333332,
      "accuracy_std_dev": 0.04118241672492712,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.1777705147306397,
      "avg_duration_per_task_ms": 43633.450336700334,
      "avg_output_tokens_per_task": 3347.0820707070707,
      "last_run_timestamp": "2026-10-24T08:27:26.672Z"
    },
    {
      "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": "anthropic/claude-sonnet-5-20260630",
      "model_slug": "anthropic/claude-sonnet-5",
      "model_name": "Anthropic: Claude Sonnet 5",
      "provider_name": "Azure",
      "accuracy": 0.8654601250000001,
      "accuracy_std_dev": 0.02784951724493697,
      "run_count": 24,
      "total_questions": 4752,
      "avg_cost_per_task_usd": 0.15895532175925928,
      "avg_duration_per_task_ms": 153299.00252525252,
      "avg_output_tokens_per_task": 15366.623316498317,
      "last_run_timestamp": "2026-10-24T08:20:16.287Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "DeepInfra",
      "accuracy": 0.8653352747021724,
      "accuracy_std_dev": 0.017080671231962035,
      "run_count": 8,
      "total_questions": 1427,
      "avg_cost_per_task_usd": 0.005114180098107919,
      "avg_duration_per_task_ms": 702127.0336370007,
      "avg_output_tokens_per_task": 28117.81079187106,
      "last_run_timestamp": "2026-10-24T16:15:47.409Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Relace",
      "accuracy": 0.86532,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.01495640909090909,
      "avg_duration_per_task_ms": 120441.0101010101,
      "avg_output_tokens_per_task": 29795.70707070707,
      "last_run_timestamp": "2026-10-24T15:00:02.228Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "NextBit",
      "accuracy": 0.8653195,
      "accuracy_std_dev": 0.0016835000000000044,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.008587179292929294,
      "avg_duration_per_task_ms": 240899.30303030304,
      "avg_output_tokens_per_task": 30387.573232323233,
      "last_run_timestamp": "2026-08-29T18:38:15.907Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Wafer",
      "accuracy": 0.8652359694129255,
      "accuracy_std_dev": 0.07811822893653826,
      "run_count": 21,
      "total_questions": 4054,
      "avg_cost_per_task_usd": 0.22709830858411445,
      "avg_duration_per_task_ms": 501409.10236803157,
      "avg_output_tokens_per_task": 63065.26467686236,
      "last_run_timestamp": "2026-10-29T22:46:44.614Z"
    },
    {
      "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": 0.8650137272727274,
      "accuracy_std_dev": 0.051263649171514956,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.1117407479338843,
      "avg_duration_per_task_ms": 477041.2093663912,
      "avg_output_tokens_per_task": 28184.2681359045,
      "last_run_timestamp": "2026-10-27T01:15:58.793Z"
    },
    {
      "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": 0.864899,
      "accuracy_std_dev": 0.017388735477314038,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.07374217613636364,
      "avg_duration_per_task_ms": 286692.6893939394,
      "avg_output_tokens_per_task": 20985.801767676767,
      "last_run_timestamp": "2026-10-24T16:08:41.781Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Cohere",
      "accuracy": 0.8644780000000001,
      "accuracy_std_dev": 0.01094299999999998,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.05228046464646464,
      "avg_duration_per_task_ms": 195135.57323232322,
      "avg_output_tokens_per_task": 43383.04797979798,
      "last_run_timestamp": "2026-10-24T08:15:15.306Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.6-opus-20260205",
      "model_slug": "anthropic/claude-opus-4.6",
      "model_name": "Anthropic: Claude Opus 4.6",
      "provider_name": "auto-routing",
      "accuracy": 0.86433775,
      "accuracy_std_dev": 0.014059447115285145,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.7497476914983165,
      "avg_duration_per_task_ms": 499048.3042929293,
      "avg_output_tokens_per_task": 29826.423821548822,
      "last_run_timestamp": "2026-10-24T08:11:56.068Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-sonnet-5-20260630",
      "model_slug": "anthropic/claude-sonnet-5",
      "model_name": "Anthropic: Claude Sonnet 5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.8638838235294116,
      "accuracy_std_dev": 0.02774133966979138,
      "run_count": 34,
      "total_questions": 6732,
      "avg_cost_per_task_usd": 0.17359224925727867,
      "avg_duration_per_task_ms": 159152.2174688057,
      "avg_output_tokens_per_task": 16169.014557338087,
      "last_run_timestamp": "2026-10-24T08:48:23.815Z"
    },
    {
      "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": 0.8636430722945668,
      "accuracy_std_dev": 0.030671339204155882,
      "run_count": 12,
      "total_questions": 2227,
      "avg_cost_per_task_usd": 0.02386835383924562,
      "avg_duration_per_task_ms": 460447.3304894477,
      "avg_output_tokens_per_task": 22718.12303547373,
      "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-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": "anthropic/claude-4.5-opus-20251124",
      "model_slug": "anthropic/claude-opus-4.5",
      "model_name": "Anthropic: Claude Opus 4.5",
      "provider_name": "auto-routing",
      "accuracy": 0.8634960000000002,
      "accuracy_std_dev": 0.01008821268444846,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.8421511931818182,
      "avg_duration_per_task_ms": 412180.98316498316,
      "avg_output_tokens_per_task": 33494.958333333336,
      "last_run_timestamp": "2026-10-24T08:26:00.557Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Venice",
      "accuracy": 0.8634435114503817,
      "accuracy_std_dev": 0.01366835982067709,
      "run_count": 8,
      "total_questions": 1572,
      "avg_cost_per_task_usd": 0.13531842366412217,
      "avg_duration_per_task_ms": 289599.17048346053,
      "avg_output_tokens_per_task": 28034.88358778626,
      "last_run_timestamp": "2026-10-17T08:23:25.121Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "Fireworks",
      "accuracy": 0.86321525,
      "accuracy_std_dev": 0.010127171825712258,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.022006194444444443,
      "avg_duration_per_task_ms": 265798.13636363635,
      "avg_output_tokens_per_task": 33108.22601010101,
      "last_run_timestamp": "2026-09-19T08:51:58.825Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "ModelRun",
      "accuracy": 0.8630749999999998,
      "accuracy_std_dev": 0.0034590473640392213,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.018059974747474748,
      "avg_duration_per_task_ms": 144168.21043771043,
      "avg_output_tokens_per_task": 29911.44107744108,
      "last_run_timestamp": "2026-10-28T04:56:05.805Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Relace",
      "accuracy": 0.862963,
      "accuracy_std_dev": 0.009134508372101892,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.06279948181818182,
      "avg_duration_per_task_ms": 441139.7515151515,
      "avg_output_tokens_per_task": 56049.472727272725,
      "last_run_timestamp": "2026-10-24T16:14:38.647Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.6-opus-20260205",
      "model_slug": "anthropic/claude-opus-4.6",
      "model_name": "Anthropic: Claude Opus 4.6",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.8629348333333335,
      "accuracy_std_dev": 0.008541693195080755,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.8005278598484848,
      "avg_duration_per_task_ms": 459239.8367003367,
      "avg_output_tokens_per_task": 31848.97138047138,
      "last_run_timestamp": "2026-10-24T08:27:05.617Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8626261,
      "accuracy_std_dev": 0.011472416183611877,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.07024428484848486,
      "avg_duration_per_task_ms": 231811.16010101011,
      "avg_output_tokens_per_task": 11437.808585858586,
      "last_run_timestamp": "2026-10-24T08:31:24.553Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Baidu",
      "accuracy": 0.8625841249999999,
      "accuracy_std_dev": 0.006783236993454885,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.004199931818181818,
      "avg_duration_per_task_ms": 250234.78724747474,
      "avg_output_tokens_per_task": 29101.075757575756,
      "last_run_timestamp": "2026-10-24T16:02:06.935Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Phala",
      "accuracy": 0.8625343605313093,
      "accuracy_std_dev": 0.061262981528279976,
      "run_count": 8,
      "total_questions": 1581,
      "avg_cost_per_task_usd": 0.2509589614168248,
      "avg_duration_per_task_ms": 1179158.6533839342,
      "avg_output_tokens_per_task": 55775.307400379505,
      "last_run_timestamp": "2026-10-24T16:03:06.207Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "auto-routing",
      "accuracy": 0.8619528181818182,
      "accuracy_std_dev": 0.017198547374273336,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.09933105968778695,
      "avg_duration_per_task_ms": 470423.74885215797,
      "avg_output_tokens_per_task": 26261.255280073463,
      "last_run_timestamp": "2026-10-24T16:12:09.459Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-5-fable-20260609",
      "model_slug": "anthropic/claude-fable-5",
      "model_name": "Anthropic: Claude Fable 5",
      "provider_name": "Azure",
      "accuracy": 0.86195275,
      "accuracy_std_dev": 0.04035143977217541,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.181786273989899,
      "avg_duration_per_task_ms": 45601.15025252525,
      "avg_output_tokens_per_task": 3426.050925925926,
      "last_run_timestamp": "2026-10-24T08:08:40.733Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Azure",
      "accuracy": 0.8619527499999999,
      "accuracy_std_dev": 0.02717181928832702,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.13016119191919193,
      "avg_duration_per_task_ms": 482692.7720959596,
      "avg_output_tokens_per_task": 34118.55934343435,
      "last_run_timestamp": "2026-10-24T16:05:51.240Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "Novita",
      "accuracy": 0.861837910657307,
      "accuracy_std_dev": 0.01299998475419779,
      "run_count": 8,
      "total_questions": 1567,
      "avg_cost_per_task_usd": 0.07543853095086152,
      "avg_duration_per_task_ms": 279737.1965539247,
      "avg_output_tokens_per_task": 20830.50095724314,
      "last_run_timestamp": "2026-10-24T16:12:49.070Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "CoreWeave",
      "accuracy": 0.861708258064516,
      "accuracy_std_dev": 0.012137624210372552,
      "run_count": 6,
      "total_questions": 1116,
      "avg_cost_per_task_usd": 0.02452295071684588,
      "avg_duration_per_task_ms": 145394.43727598566,
      "avg_output_tokens_per_task": 37540.51523297491,
      "last_run_timestamp": "2026-10-24T08:12:02.536Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.8614937272727271,
      "accuracy_std_dev": 0.007701093359102963,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.06505497428833792,
      "avg_duration_per_task_ms": 215627.76170798898,
      "avg_output_tokens_per_task": 10574.208907254362,
      "last_run_timestamp": "2026-10-28T03:28:22.479Z"
    },
    {
      "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": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "StreamLake",
      "accuracy": 0.8613914444444443,
      "accuracy_std_dev": 0.017477429375633225,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.09228736588103255,
      "avg_duration_per_task_ms": 186627.3569023569,
      "avg_output_tokens_per_task": 25504.406285072953,
      "last_run_timestamp": "2026-10-24T16:20:59.111Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.5-opus-20251124",
      "model_slug": "anthropic/claude-opus-4.5",
      "model_name": "Anthropic: Claude Opus 4.5",
      "provider_name": "Anthropic",
      "accuracy": 0.8613514285714288,
      "accuracy_std_dev": 0.011656715536635543,
      "run_count": 14,
      "total_questions": 2772,
      "avg_cost_per_task_usd": 0.842712674963925,
      "avg_duration_per_task_ms": 456401.95995670994,
      "avg_output_tokens_per_task": 33529.13383838384,
      "last_run_timestamp": "2026-10-24T08:15:51.139Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Morph",
      "accuracy": 0.8612792,
      "accuracy_std_dev": 0.03136197207702347,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.005120404040404041,
      "avg_duration_per_task_ms": 530889.794949495,
      "avg_output_tokens_per_task": 17542.636363636364,
      "last_run_timestamp": "2026-09-05T09:22:54.724Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Novita",
      "accuracy": 0.861111,
      "accuracy_std_dev": 0.04797999999999997,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.021780800505050504,
      "avg_duration_per_task_ms": 138470.81818181818,
      "avg_output_tokens_per_task": 20189.02525252525,
      "last_run_timestamp": "2026-10-24T08:12:37.562Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.5-opus-20251124",
      "model_slug": "anthropic/claude-opus-4.5",
      "model_name": "Anthropic: Claude Opus 4.5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.8609707916666667,
      "accuracy_std_dev": 0.007924941871391254,
      "run_count": 24,
      "total_questions": 4752,
      "avg_cost_per_task_usd": 0.8830173162878786,
      "avg_duration_per_task_ms": 414862.5126262626,
      "avg_output_tokens_per_task": 33448.23737373737,
      "last_run_timestamp": "2026-10-24T08:44:41.418Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "DekaLLM",
      "accuracy": 0.8608923543307087,
      "accuracy_std_dev": 0.017439499999999997,
      "run_count": 2,
      "total_questions": 254,
      "avg_cost_per_task_usd": 0.003837814960629921,
      "avg_duration_per_task_ms": 298144.1771653543,
      "avg_output_tokens_per_task": 11523.807086614173,
      "last_run_timestamp": "2026-10-03T17:04:09.464Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "InferenceNet",
      "accuracy": 0.8605965053212455,
      "accuracy_std_dev": 0.03960090537740342,
      "run_count": 13,
      "total_questions": 2537,
      "avg_cost_per_task_usd": 0.01657229877808435,
      "avg_duration_per_task_ms": 463843.5068979109,
      "avg_output_tokens_per_task": 41792.180134016555,
      "last_run_timestamp": "2026-10-24T08:58:59.550Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "DigitalOcean",
      "accuracy": 0.8602692499999999,
      "accuracy_std_dev": 0.07328871721500406,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.0069893973063973065,
      "avg_duration_per_task_ms": 603844.6805555555,
      "avg_output_tokens_per_task": 28814.18181818182,
      "last_run_timestamp": "2026-10-24T08:23:31.990Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "reflection/beam-1.0-501b-a23b-20261016",
      "model_slug": "reflection/beam-1.0-501b-a23b",
      "model_name": "Reflection: Beam 1.0 501B A23B",
      "provider_name": "Io Net",
      "accuracy": 0.860269,
      "accuracy_std_dev": 0.010100999999999971,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.00987439393939394,
      "avg_duration_per_task_ms": 74831.28787878787,
      "avg_output_tokens_per_task": 9997.058080808081,
      "last_run_timestamp": "2026-10-24T08:14:36.264Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "auto-routing",
      "accuracy": 0.8601809488313328,
      "accuracy_std_dev": 0.03964118133468496,
      "run_count": 8,
      "total_questions": 1583,
      "avg_cost_per_task_usd": 0.16196978711307647,
      "avg_duration_per_task_ms": 990576.0770688566,
      "avg_output_tokens_per_task": 55234.00947567909,
      "last_run_timestamp": "2026-10-24T16:06:09.255Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.5-opus-20251124",
      "model_slug": "anthropic/claude-opus-4.5",
      "model_name": "Anthropic: Claude Opus 4.5",
      "provider_name": "Google",
      "accuracy": 0.8599885833333334,
      "accuracy_std_dev": 0.009617771098842076,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.8533179797979797,
      "avg_duration_per_task_ms": 482512.8122895623,
      "avg_output_tokens_per_task": 33941.89983164983,
      "last_run_timestamp": "2026-10-24T08:14:51.084Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-sonnet-5-20260630",
      "model_slug": "anthropic/claude-sonnet-5",
      "model_name": "Anthropic: Claude Sonnet 5",
      "provider_name": "Anthropic",
      "accuracy": 0.8598806153846154,
      "accuracy_std_dev": 0.030036815288878832,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.15524810489510485,
      "avg_duration_per_task_ms": 152128.8811188811,
      "avg_output_tokens_per_task": 15307.565656565657,
      "last_run_timestamp": "2026-10-24T08:12:37.564Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "InferenceNet",
      "accuracy": 0.8597419723629663,
      "accuracy_std_dev": 0.016505636989518208,
      "run_count": 11,
      "total_questions": 2171,
      "avg_cost_per_task_usd": 0.08549684339014277,
      "avg_duration_per_task_ms": 179079.73422385997,
      "avg_output_tokens_per_task": 24385.910640257945,
      "last_run_timestamp": "2026-10-24T08:41:25.897Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "Io Net",
      "accuracy": 0.8596746149425287,
      "accuracy_std_dev": 0.000831499999999985,
      "run_count": 2,
      "total_questions": 348,
      "avg_cost_per_task_usd": 0.0899417959770115,
      "avg_duration_per_task_ms": 1024745.4281609196,
      "avg_output_tokens_per_task": 35825.26724137931,
      "last_run_timestamp": "2026-10-27T13:57:40.548Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Inceptron",
      "accuracy": 0.8596613057736721,
      "accuracy_std_dev": 0.0311431999096363,
      "run_count": 11,
      "total_questions": 2165,
      "avg_cost_per_task_usd": 0.1269020115473441,
      "avg_duration_per_task_ms": 528998.0508083141,
      "avg_output_tokens_per_task": 43596.739953810626,
      "last_run_timestamp": "2026-10-24T16:12:49.068Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.6-opus-20260205",
      "model_slug": "anthropic/claude-opus-4.6",
      "model_name": "Anthropic: Claude Opus 4.6",
      "provider_name": "Azure",
      "accuracy": 0.8594274166666667,
      "accuracy_std_dev": 0.018601260147538098,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.7585617697811449,
      "avg_duration_per_task_ms": 508230.9494949495,
      "avg_output_tokens_per_task": 30213.497474747473,
      "last_run_timestamp": "2026-10-24T08:10:23.014Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Fireworks",
      "accuracy": 0.8590258723037101,
      "accuracy_std_dev": 0.009219054418533492,
      "run_count": 21,
      "total_questions": 3477,
      "avg_cost_per_task_usd": 0.1720266922634455,
      "avg_duration_per_task_ms": 256422.9841817659,
      "avg_output_tokens_per_task": 31690.788323267185,
      "last_run_timestamp": "2026-09-26T08:54:22.466Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "DigitalOcean",
      "accuracy": 0.8586340031847132,
      "accuracy_std_dev": 0.05344060455646062,
      "run_count": 10,
      "total_questions": 1884,
      "avg_cost_per_task_usd": 0.0115320483014862,
      "avg_duration_per_task_ms": 468303.98089171975,
      "avg_output_tokens_per_task": 23774.996815286624,
      "last_run_timestamp": "2026-10-24T08:32:58.238Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.858586,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.08651329797979797,
      "avg_duration_per_task_ms": 499241.5707070707,
      "avg_output_tokens_per_task": 35858.23737373737,
      "last_run_timestamp": "2026-10-04T03:06:10.317Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Chutes",
      "accuracy": 0.85858575,
      "accuracy_std_dev": 0.021526546714173627,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.1476876622474748,
      "avg_duration_per_task_ms": 1253876.004419192,
      "avg_output_tokens_per_task": 52218.26641414141,
      "last_run_timestamp": "2026-10-24T16:26:15.939Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "InferenceNet",
      "accuracy": 0.8585855,
      "accuracy_std_dev": 0.012710146271778291,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.014605150252525251,
      "avg_duration_per_task_ms": 477497.9494949495,
      "avg_output_tokens_per_task": 26944.89898989899,
      "last_run_timestamp": "2026-10-24T08:15:32.562Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Cloudflare",
      "accuracy": 0.8585232040276647,
      "accuracy_std_dev": 0.03867598457396728,
      "run_count": 29,
      "total_questions": 4916,
      "avg_cost_per_task_usd": 0.01779470951993491,
      "avg_duration_per_task_ms": 399750.05593978846,
      "avg_output_tokens_per_task": 18974.740846216435,
      "last_run_timestamp": "2026-10-24T08:49:56.512Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "SiliconFlow",
      "accuracy": 0.858164875,
      "accuracy_std_dev": 0.016185868044667075,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.1437290511363636,
      "avg_duration_per_task_ms": 908116.1098484849,
      "avg_output_tokens_per_task": 45706.13320707071,
      "last_run_timestamp": "2026-10-24T16:14:06.777Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "BaseTen",
      "accuracy": 0.8579636833065611,
      "accuracy_std_dev": 0.020319706431789875,
      "run_count": 41,
      "total_questions": 8093,
      "avg_cost_per_task_usd": 0.28457569974051644,
      "avg_duration_per_task_ms": 335733.34103546274,
      "avg_output_tokens_per_task": 51175.235635734585,
      "last_run_timestamp": "2026-10-24T16:37:01.548Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.8577439999999998,
      "accuracy_std_dev": 0.019433210349810964,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.02084834090909091,
      "avg_duration_per_task_ms": 460409.3712121212,
      "avg_output_tokens_per_task": 41456.85101010101,
      "last_run_timestamp": "2026-10-29T23:33:29.301Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.6-opus-20260205",
      "model_slug": "anthropic/claude-opus-4.6",
      "model_name": "Anthropic: Claude Opus 4.6",
      "provider_name": "Anthropic",
      "accuracy": 0.8577439285714286,
      "accuracy_std_dev": 0.016868073763619186,
      "run_count": 14,
      "total_questions": 2772,
      "avg_cost_per_task_usd": 0.7558037463924961,
      "avg_duration_per_task_ms": 505406.40187590185,
      "avg_output_tokens_per_task": 30123.084054834057,
      "last_run_timestamp": "2026-10-24T08:30:49.044Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.6-opus-20260205",
      "model_slug": "anthropic/claude-opus-4.6",
      "model_name": "Anthropic: Claude Opus 4.6",
      "provider_name": "Google",
      "accuracy": 0.8576738333333334,
      "accuracy_std_dev": 0.01763649745004817,
      "run_count": 24,
      "total_questions": 4752,
      "avg_cost_per_task_usd": 0.8048634583333334,
      "avg_duration_per_task_ms": 484140.92781986535,
      "avg_output_tokens_per_task": 30507.85164141414,
      "last_run_timestamp": "2026-10-24T08:15:15.299Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "DigitalOcean",
      "accuracy": 0.857663467955239,
      "accuracy_std_dev": 0.05434631352575813,
      "run_count": 10,
      "total_questions": 1966,
      "avg_cost_per_task_usd": 0.0781413972533062,
      "avg_duration_per_task_ms": 509131.62817904376,
      "avg_output_tokens_per_task": 21110.80518819939,
      "last_run_timestamp": "2026-10-24T08:57:31.359Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "Minimax",
      "accuracy": 0.8573231875,
      "accuracy_std_dev": 0.007907506704223743,
      "run_count": 16,
      "total_questions": 3168,
      "avg_cost_per_task_usd": 0.06346724873737374,
      "avg_duration_per_task_ms": 591294.5839646464,
      "avg_output_tokens_per_task": 35019.75441919192,
      "last_run_timestamp": "2026-10-24T16:03:59.167Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.5-opus-20251124",
      "model_slug": "anthropic/claude-opus-4.5",
      "model_name": "Anthropic: Claude Opus 4.5",
      "provider_name": "Azure",
      "accuracy": 0.8573231666666665,
      "accuracy_std_dev": 0.0170736054571246,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.8399580702861953,
      "avg_duration_per_task_ms": 458034.78324915824,
      "avg_output_tokens_per_task": 33422.68813131313,
      "last_run_timestamp": "2026-10-24T08:13:46.758Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Together",
      "accuracy": 0.8571427108013937,
      "accuracy_std_dev": 0.01011902539224449,
      "run_count": 3,
      "total_questions": 574,
      "avg_cost_per_task_usd": 0.11599589721254357,
      "avg_duration_per_task_ms": 363022.99128919863,
      "avg_output_tokens_per_task": 29391.742160278747,
      "last_run_timestamp": "2026-08-22T10:19:36.962Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.5-opus-20251124",
      "model_slug": "anthropic/claude-opus-4.5",
      "model_name": "Anthropic: Claude Opus 4.5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.8570552727272727,
      "accuracy_std_dev": 0.02678680556845074,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.8339512213039484,
      "avg_duration_per_task_ms": 436211.7764003673,
      "avg_output_tokens_per_task": 33251.8347107438,
      "last_run_timestamp": "2026-10-24T08:08:52.220Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "BaseTen",
      "accuracy": 0.8569032243034418,
      "accuracy_std_dev": 0.036552330331727854,
      "run_count": 22,
      "total_questions": 4271,
      "avg_cost_per_task_usd": 0.0041591236244439235,
      "avg_duration_per_task_ms": 82317.1749004917,
      "avg_output_tokens_per_task": 15615.828377429174,
      "last_run_timestamp": "2026-10-24T08:41:46.943Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Tenstorrent",
      "accuracy": 0.8569025,
      "accuracy_std_dev": 0.0016835000000000044,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.017741116161616163,
      "avg_duration_per_task_ms": 206976.69949494948,
      "avg_output_tokens_per_task": 29405.780303030304,
      "last_run_timestamp": "2026-10-30T18:32:25.440Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Wafer",
      "accuracy": 0.8569025,
      "accuracy_std_dev": 0.005050499999999958,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.018898722222222223,
      "avg_duration_per_task_ms": 164757.63383838383,
      "avg_output_tokens_per_task": 31337.035353535353,
      "last_run_timestamp": "2026-10-24T08:17:56.393Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "CoreWeave",
      "accuracy": 0.8569022500000001,
      "accuracy_std_dev": 0.013202173483464743,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.007818540404040404,
      "avg_duration_per_task_ms": 506894.8308080808,
      "avg_output_tokens_per_task": 28009.079545454544,
      "last_run_timestamp": "2026-08-29T18:49:16.416Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5-2025-08-07",
      "model_slug": "openai/gpt-5",
      "model_name": "OpenAI: GPT-5",
      "provider_name": "auto-routing",
      "accuracy": 0.8569019999999999,
      "accuracy_std_dev": 0.0033669999999999534,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.20989276262626264,
      "avg_duration_per_task_ms": 263850.72222222225,
      "avg_output_tokens_per_task": 20893.070707070707,
      "last_run_timestamp": "2026-10-01T16:31:36.833Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Near AI",
      "accuracy": 0.8568129407236336,
      "accuracy_std_dev": 0.01596270457778066,
      "run_count": 8,
      "total_questions": 1299,
      "avg_cost_per_task_usd": 0.004350760585065434,
      "avg_duration_per_task_ms": 410967.9938414165,
      "avg_output_tokens_per_task": 9402.565819861431,
      "last_run_timestamp": "2026-10-24T08:24:10.666Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-5-fable-20260609",
      "model_slug": "anthropic/claude-fable-5",
      "model_name": "Anthropic: Claude Fable 5",
      "provider_name": "Google",
      "accuracy": 0.8563635600000002,
      "accuracy_std_dev": 0.050198181182254005,
      "run_count": 25,
      "total_questions": 4950,
      "avg_cost_per_task_usd": 0.18820920787878787,
      "avg_duration_per_task_ms": 41314.649292929294,
      "avg_output_tokens_per_task": 3389.456767676768,
      "last_run_timestamp": "2026-10-24T08:44:43.918Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Crusoe",
      "accuracy": 0.856164284931507,
      "accuracy_std_dev": 0.05295168533540149,
      "run_count": 8,
      "total_questions": 1460,
      "avg_cost_per_task_usd": 0.18168324041095887,
      "avg_duration_per_task_ms": 1626314.2294520547,
      "avg_output_tokens_per_task": 52899.55684931507,
      "last_run_timestamp": "2026-10-24T16:16:05.013Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Azure",
      "accuracy": 0.8558499999999999,
      "accuracy_std_dev": 0.03281670770735542,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.015880464015151512,
      "avg_duration_per_task_ms": 344363.2039141414,
      "avg_output_tokens_per_task": 28437.20138888889,
      "last_run_timestamp": "2026-10-24T16:21:33.540Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Inceptron",
      "accuracy": 0.8558169770378875,
      "accuracy_std_dev": 0.018638102580532955,
      "run_count": 9,
      "total_questions": 1742,
      "avg_cost_per_task_usd": 0.08209278185993112,
      "avg_duration_per_task_ms": 337904.9672789897,
      "avg_output_tokens_per_task": 28132.247990815154,
      "last_run_timestamp": "2026-10-24T08:19:59.729Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Fireworks",
      "accuracy": 0.855721,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 67,
      "avg_cost_per_task_usd": 0.015361298507462687,
      "avg_duration_per_task_ms": 180200.07462686568,
      "avg_output_tokens_per_task": 25482.86567164179,
      "last_run_timestamp": "2026-10-28T00:00:04.637Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "auto-routing",
      "accuracy": 0.8554993333333334,
      "accuracy_std_dev": 0.009394238245979377,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.005048751683501684,
      "avg_duration_per_task_ms": 532489.6868686868,
      "avg_output_tokens_per_task": 30189.478956228955,
      "last_run_timestamp": "2026-10-24T16:02:08.817Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Baidu",
      "accuracy": 0.85530295,
      "accuracy_std_dev": 0.01065364351513132,
      "run_count": 20,
      "total_questions": 3960,
      "avg_cost_per_task_usd": 0.14301361767676768,
      "avg_duration_per_task_ms": 429371.0404040404,
      "avg_output_tokens_per_task": 43957.30757575758,
      "last_run_timestamp": "2026-10-24T16:26:47.544Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8552187499999999,
      "accuracy_std_dev": 0.022460588392504314,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.04356151957070707,
      "avg_duration_per_task_ms": 653045.7601010101,
      "avg_output_tokens_per_task": 36314.97159090909,
      "last_run_timestamp": "2026-10-24T16:04:33.484Z"
    },
    {
      "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": "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": 0.85479775,
      "accuracy_std_dev": 0.009017138608644082,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.09609784722222224,
      "avg_duration_per_task_ms": 435031.48232323234,
      "avg_output_tokens_per_task": 30609.608585858587,
      "last_run_timestamp": "2026-08-29T11:19:17.684Z"
    },
    {
      "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": 0.8541006015180266,
      "accuracy_std_dev": 0.015925939750529645,
      "run_count": 8,
      "total_questions": 1581,
      "avg_cost_per_task_usd": 0.037091079696394684,
      "avg_duration_per_task_ms": 591845.9070208728,
      "avg_output_tokens_per_task": 34537.16951296648,
      "last_run_timestamp": "2026-10-24T16:36:57.539Z"
    },
    {
      "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": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Fireworks",
      "accuracy": 0.8533329920000001,
      "accuracy_std_dev": 0.013089000000000017,
      "run_count": 2,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.005868832,
      "avg_duration_per_task_ms": 205553.388,
      "avg_output_tokens_per_task": 21941.544,
      "last_run_timestamp": "2026-10-25T21:00:03.046Z"
    },
    {
      "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": 0.8532693421052633,
      "accuracy_std_dev": 0.035890582042967266,
      "run_count": 38,
      "total_questions": 7524,
      "avg_cost_per_task_usd": 0.1652858625730994,
      "avg_duration_per_task_ms": 151312.44696969696,
      "avg_output_tokens_per_task": 15290.7336523126,
      "last_run_timestamp": "2026-10-24T08:32:13.238Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-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-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "GMICloud",
      "accuracy": 0.8530043052097741,
      "accuracy_std_dev": 0.0200463394003836,
      "run_count": 11,
      "total_questions": 2169,
      "avg_cost_per_task_usd": 0.09074984416781928,
      "avg_duration_per_task_ms": 309963.8911940987,
      "avg_output_tokens_per_task": 26287.04610419548,
      "last_run_timestamp": "2026-10-27T01:51:15.089Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Baidu",
      "accuracy": 0.8528355426944971,
      "accuracy_std_dev": 0.013503895836017096,
      "run_count": 8,
      "total_questions": 1581,
      "avg_cost_per_task_usd": 0.0925430898165718,
      "avg_duration_per_task_ms": 296081.5515496521,
      "avg_output_tokens_per_task": 29809.33523086654,
      "last_run_timestamp": "2026-10-24T08:21:56.261Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Minimax",
      "accuracy": 0.852693625,
      "accuracy_std_dev": 0.006627993756362076,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.05533727904040404,
      "avg_duration_per_task_ms": 425236.9147727273,
      "avg_output_tokens_per_task": 28304.02904040404,
      "last_run_timestamp": "2026-08-29T19:17:08.288Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "DekaLLM",
      "accuracy": 0.8524375810276679,
      "accuracy_std_dev": 0.011217700923094702,
      "run_count": 4,
      "total_questions": 759,
      "avg_cost_per_task_usd": 0.07750415546772069,
      "avg_duration_per_task_ms": 263910.441370224,
      "avg_output_tokens_per_task": 26225.246376811596,
      "last_run_timestamp": "2026-10-24T08:40:56.292Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "auto-routing",
      "accuracy": 0.8524129999999999,
      "accuracy_std_dev": 0.012672927610549278,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.1049078827160494,
      "avg_duration_per_task_ms": 476715.49887766555,
      "avg_output_tokens_per_task": 30743.156565656565,
      "last_run_timestamp": "2026-10-24T08:15:15.299Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Venice",
      "accuracy": 0.8521062280219781,
      "accuracy_std_dev": 0.006055500000000047,
      "run_count": 2,
      "total_questions": 364,
      "avg_cost_per_task_usd": 0.005501717032967033,
      "avg_duration_per_task_ms": 539158.467032967,
      "avg_output_tokens_per_task": 15491.851648351649,
      "last_run_timestamp": "2026-10-17T08:17:58.236Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Parasail",
      "accuracy": 0.851852,
      "accuracy_std_dev": 0.006733999999999962,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.01747031565656566,
      "avg_duration_per_task_ms": 122982.43434343435,
      "avg_output_tokens_per_task": 28903.234848484848,
      "last_run_timestamp": "2026-10-24T08:15:51.061Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Ionstream",
      "accuracy": 0.851852,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.16585097979797978,
      "avg_duration_per_task_ms": 389107.60101010103,
      "avg_output_tokens_per_task": 37536.631313131315,
      "last_run_timestamp": "2026-08-04T03:12:10.622Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.6-opus-20260205",
      "model_slug": "anthropic/claude-opus-4.6",
      "model_name": "Anthropic: Claude Opus 4.6",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.8518518181818182,
      "accuracy_std_dev": 0.03385326940520644,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.7072226216712579,
      "avg_duration_per_task_ms": 481789.99357208447,
      "avg_output_tokens_per_task": 28288.861799816346,
      "last_run_timestamp": "2026-10-24T08:43:42.744Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5-2025-08-07",
      "model_slug": "openai/gpt-5",
      "model_name": "OpenAI: GPT-5",
      "provider_name": "Azure",
      "accuracy": 0.8518517499999999,
      "accuracy_std_dev": 0.0020620621201845155,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.21063902146464644,
      "avg_duration_per_task_ms": 251655.29545454544,
      "avg_output_tokens_per_task": 19971.159090909092,
      "last_run_timestamp": "2026-10-01T16:27:58.636Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "auto-routing",
      "accuracy": 0.8516414999999999,
      "accuracy_std_dev": 0.028797790015034144,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.07571042234848485,
      "avg_duration_per_task_ms": 512797.04924242425,
      "avg_output_tokens_per_task": 22901.14393939394,
      "last_run_timestamp": "2026-10-24T16:15:14.403Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Ambient",
      "accuracy": 0.8515965972461275,
      "accuracy_std_dev": 0.006985364039578179,
      "run_count": 9,
      "total_questions": 1743,
      "avg_cost_per_task_usd": 0.09721534480780264,
      "avg_duration_per_task_ms": 672514.3052208836,
      "avg_output_tokens_per_task": 40113.75157773953,
      "last_run_timestamp": "2026-10-24T16:04:55.275Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "Moonshot AI",
      "accuracy": 0.8510099999999999,
      "accuracy_std_dev": 0.037427264881366905,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.12598769823232323,
      "avg_duration_per_task_ms": 640494.7992424242,
      "avg_output_tokens_per_task": 42382.61111111111,
      "last_run_timestamp": "2026-08-29T18:25:54.650Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8509267650397275,
      "accuracy_std_dev": 0.029137382315888,
      "run_count": 9,
      "total_questions": 1762,
      "avg_cost_per_task_usd": 0.14751040351872874,
      "avg_duration_per_task_ms": 406196.22701475595,
      "avg_output_tokens_per_task": 33481.99659477866,
      "last_run_timestamp": "2026-10-24T08:12:02.530Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "BaseTen",
      "accuracy": 0.8508249284795883,
      "accuracy_std_dev": 0.01638111560444773,
      "run_count": 33,
      "total_questions": 6222,
      "avg_cost_per_task_usd": 0.1628150337512054,
      "avg_duration_per_task_ms": 188242.73111539698,
      "avg_output_tokens_per_task": 29107.008357441337,
      "last_run_timestamp": "2026-10-24T09:28:03.268Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Mancer 2",
      "accuracy": 0.850799625,
      "accuracy_std_dev": 0.01687570508851038,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.013269784090909089,
      "avg_duration_per_task_ms": 808248.1130050505,
      "avg_output_tokens_per_task": 28025.626893939392,
      "last_run_timestamp": "2026-10-24T16:04:55.172Z"
    },
    {
      "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": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Crusoe",
      "accuracy": 0.8504104417831005,
      "accuracy_std_dev": 0.011757939662675208,
      "run_count": 9,
      "total_questions": 1503,
      "avg_cost_per_task_usd": 0.006495053226879575,
      "avg_duration_per_task_ms": 421761.22222222225,
      "avg_output_tokens_per_task": 16002.619427811045,
      "last_run_timestamp": "2026-10-24T16:11:33.499Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8504087142857142,
      "accuracy_std_dev": 0.010549160084565759,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.012687737373737374,
      "avg_duration_per_task_ms": 687002.3088023087,
      "avg_output_tokens_per_task": 18864.385281385283,
      "last_run_timestamp": "2026-10-17T16:05:11.655Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Venice",
      "accuracy": 0.8502384166666668,
      "accuracy_std_dev": 0.009974367911955898,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.09800930387205385,
      "avg_duration_per_task_ms": 567891.4086700337,
      "avg_output_tokens_per_task": 30514.193602693602,
      "last_run_timestamp": "2026-10-24T16:13:46.473Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5-2025-08-07",
      "model_slug": "openai/gpt-5",
      "model_name": "OpenAI: GPT-5",
      "provider_name": "OpenAI",
      "accuracy": 0.8501685000000001,
      "accuracy_std_dev": 0.0016835000000000044,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.20853091414141414,
      "avg_duration_per_task_ms": 280266.345959596,
      "avg_output_tokens_per_task": 20755.358585858587,
      "last_run_timestamp": "2026-10-01T16:20:44.419Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "Parasail",
      "accuracy": 0.8501683333333333,
      "accuracy_std_dev": 0.009523284669109108,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.03896970159932659,
      "avg_duration_per_task_ms": 390855.180976431,
      "avg_output_tokens_per_task": 39310.545454545456,
      "last_run_timestamp": "2026-10-24T16:23:26.712Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "GMICloud",
      "accuracy": 0.8501682222222223,
      "accuracy_std_dev": 0.0192438426734243,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.03105917845117845,
      "avg_duration_per_task_ms": 604384.6245791246,
      "avg_output_tokens_per_task": 33888.867003367006,
      "last_run_timestamp": "2026-10-27T03:42:35.337Z"
    },
    {
      "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": 0.8501303479061206,
      "accuracy_std_dev": 0.03129572966618359,
      "run_count": 11,
      "total_questions": 2173,
      "avg_cost_per_task_usd": 0.008647949838932353,
      "avg_duration_per_task_ms": 492602.7482742752,
      "avg_output_tokens_per_task": 30274.529682466637,
      "last_run_timestamp": "2026-10-27T01:57:27.248Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Inceptron",
      "accuracy": 0.8499550449236298,
      "accuracy_std_dev": 0.04367521121615224,
      "run_count": 7,
      "total_questions": 1113,
      "avg_cost_per_task_usd": 0.023873053908355795,
      "avg_duration_per_task_ms": 467262.93441150046,
      "avg_output_tokens_per_task": 37562.04132973944,
      "last_run_timestamp": "2026-10-24T08:33:50.644Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "AkashML",
      "accuracy": 0.8497263442622951,
      "accuracy_std_dev": 0.01704,
      "run_count": 2,
      "total_questions": 366,
      "avg_cost_per_task_usd": 0.06129340163934426,
      "avg_duration_per_task_ms": 530265.6393442623,
      "avg_output_tokens_per_task": 25778.054644808744,
      "last_run_timestamp": "2026-08-15T08:47:51.369Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Decart",
      "accuracy": 0.8497037279483036,
      "accuracy_std_dev": 0.02211279185187559,
      "run_count": 16,
      "total_questions": 3095,
      "avg_cost_per_task_usd": 0.2026981760904685,
      "avg_duration_per_task_ms": 259750.08206785138,
      "avg_output_tokens_per_task": 43394.09111470113,
      "last_run_timestamp": "2026-10-24T16:29:51.093Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "Phala",
      "accuracy": 0.8492062857142857,
      "accuracy_std_dev": 0.009075999999999973,
      "run_count": 2,
      "total_questions": 336,
      "avg_cost_per_task_usd": 0.1515307232142857,
      "avg_duration_per_task_ms": 1515335.5773809524,
      "avg_output_tokens_per_task": 50423.354166666664,
      "last_run_timestamp": "2026-08-29T18:25:52.753Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Sail Research",
      "accuracy": 0.8489659064484613,
      "accuracy_std_dev": 0.09543669529109948,
      "run_count": 21,
      "total_questions": 4094,
      "avg_cost_per_task_usd": 0.00791240718124084,
      "avg_duration_per_task_ms": 580031.3243771372,
      "avg_output_tokens_per_task": 27584.65388373229,
      "last_run_timestamp": "2026-10-24T08:58:59.998Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "auto-routing",
      "accuracy": 0.8486952499999999,
      "accuracy_std_dev": 0.02356047809229474,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.04730151515151515,
      "avg_duration_per_task_ms": 620661.3125,
      "avg_output_tokens_per_task": 38343.770202020205,
      "last_run_timestamp": "2026-10-24T16:27:04.555Z"
    },
    {
      "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": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Phala",
      "accuracy": 0.8482398753606462,
      "accuracy_std_dev": 0.025739734220236125,
      "run_count": 9,
      "total_questions": 1733,
      "avg_cost_per_task_usd": 0.10720182342758222,
      "avg_duration_per_task_ms": 497095.3819965378,
      "avg_output_tokens_per_task": 34429.2602423543,
      "last_run_timestamp": "2026-10-24T08:11:11.432Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8480638333333334,
      "accuracy_std_dev": 0.01645355840354568,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.052020993265993264,
      "avg_duration_per_task_ms": 270683.26851851854,
      "avg_output_tokens_per_task": 45678.64351851852,
      "last_run_timestamp": "2026-10-24T16:18:16.998Z"
    },
    {
      "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": 0.8480328171206225,
      "accuracy_std_dev": 0.009316525307514355,
      "run_count": 12,
      "total_questions": 2313,
      "avg_cost_per_task_usd": 0.06691969217466492,
      "avg_duration_per_task_ms": 195295.10635538262,
      "avg_output_tokens_per_task": 20814.787721573713,
      "last_run_timestamp": "2026-10-24T16:04:55.266Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "ModelRun",
      "accuracy": 0.8479167041666666,
      "accuracy_std_dev": 0.00828816902819917,
      "run_count": 5,
      "total_questions": 960,
      "avg_cost_per_task_usd": 0.16138468541666665,
      "avg_duration_per_task_ms": 277988.090625,
      "avg_output_tokens_per_task": 24274.81875,
      "last_run_timestamp": "2026-10-28T05:08:57.249Z"
    },
    {
      "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": 0.847853375,
      "accuracy_std_dev": 0.033941681060819205,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.010883301767676767,
      "avg_duration_per_task_ms": 412406.98232323234,
      "avg_output_tokens_per_task": 38550.31060606061,
      "last_run_timestamp": "2026-10-24T16:02:06.909Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "auto-routing",
      "accuracy": 0.84771325,
      "accuracy_std_dev": 0.017907204896004843,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.12780570496632998,
      "avg_duration_per_task_ms": 482349.20328282827,
      "avg_output_tokens_per_task": 47672.79882154882,
      "last_run_timestamp": "2026-10-24T16:02:47.284Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8476430833333334,
      "accuracy_std_dev": 0.017202829546610125,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.12729099452861953,
      "avg_duration_per_task_ms": 386377.8813131313,
      "avg_output_tokens_per_task": 38419.52356902357,
      "last_run_timestamp": "2026-10-24T16:18:17.002Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "DeepInfra",
      "accuracy": 0.84722225,
      "accuracy_std_dev": 0.04063783312613384,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.09673938762626262,
      "avg_duration_per_task_ms": 1384478.7525252525,
      "avg_output_tokens_per_task": 42908.3345959596,
      "last_run_timestamp": "2026-08-29T18:36:10.818Z"
    },
    {
      "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": 0.8466609166666665,
      "accuracy_std_dev": 0.04565389387456148,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.16722773358585863,
      "avg_duration_per_task_ms": 165303.9515993266,
      "avg_output_tokens_per_task": 16503.33207070707,
      "last_run_timestamp": "2026-10-24T08:13:41.058Z"
    },
    {
      "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": 0.8464799745547075,
      "accuracy_std_dev": 0.010187806019943643,
      "run_count": 8,
      "total_questions": 1572,
      "avg_cost_per_task_usd": 0.007149598600508905,
      "avg_duration_per_task_ms": 580512.5,
      "avg_output_tokens_per_task": 17687.171119592876,
      "last_run_timestamp": "2026-10-24T16:26:58.608Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Novita",
      "accuracy": 0.8463488417849898,
      "accuracy_std_dev": 0.018032907906380492,
      "run_count": 10,
      "total_questions": 1972,
      "avg_cost_per_task_usd": 0.010114639452332656,
      "avg_duration_per_task_ms": 449682.4756592292,
      "avg_output_tokens_per_task": 32532.80730223124,
      "last_run_timestamp": "2026-10-24T08:45:13.862Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8462835489864864,
      "accuracy_std_dev": 0.018064668045116323,
      "run_count": 9,
      "total_questions": 1776,
      "avg_cost_per_task_usd": 0.06724871058558558,
      "avg_duration_per_task_ms": 247566.1204954955,
      "avg_output_tokens_per_task": 19807.05518018018,
      "last_run_timestamp": "2026-10-24T08:41:57.620Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "Venice",
      "accuracy": 0.84624,
      "accuracy_std_dev": 0.02002973311521316,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.043265813131313134,
      "avg_duration_per_task_ms": 121558.87037037036,
      "avg_output_tokens_per_task": 34452.29797979798,
      "last_run_timestamp": "2026-10-24T08:13:28.067Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "StreamLake",
      "accuracy": 0.8458658888888888,
      "accuracy_std_dev": 0.043929875719895144,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.003920089225589225,
      "avg_duration_per_task_ms": 538039.7317620651,
      "avg_output_tokens_per_task": 39775.50785634119,
      "last_run_timestamp": "2026-10-24T16:14:40.610Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Together",
      "accuracy": 0.8456602006079027,
      "accuracy_std_dev": 0.016336342615469363,
      "run_count": 10,
      "total_questions": 1974,
      "avg_cost_per_task_usd": 0.020697631205673757,
      "avg_duration_per_task_ms": 324636.22441742657,
      "avg_output_tokens_per_task": 41178.20871327254,
      "last_run_timestamp": "2026-10-24T08:56:29.712Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Decart",
      "accuracy": 0.8455742847457628,
      "accuracy_std_dev": 0.019722621271080296,
      "run_count": 9,
      "total_questions": 1770,
      "avg_cost_per_task_usd": 0.10396145480225989,
      "avg_duration_per_task_ms": 199057.26101694914,
      "avg_output_tokens_per_task": 30600.066666666666,
      "last_run_timestamp": "2026-10-24T08:23:31.990Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8455386666666667,
      "accuracy_std_dev": 0.01527369648193331,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.020861409090909087,
      "avg_duration_per_task_ms": 420940.80723905726,
      "avg_output_tokens_per_task": 41492.801346801345,
      "last_run_timestamp": "2026-10-24T08:13:28.065Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Mistral",
      "accuracy": 0.8453048148148149,
      "accuracy_std_dev": 0.010350814475153139,
      "run_count": 27,
      "total_questions": 5346,
      "avg_cost_per_task_usd": 0.1469017957351291,
      "avg_duration_per_task_ms": 178191.52618780397,
      "avg_output_tokens_per_task": 32277.341937897494,
      "last_run_timestamp": "2026-10-30T16:28:33.613Z"
    },
    {
      "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": 0.845217870886076,
      "accuracy_std_dev": 0.011235203598016757,
      "run_count": 12,
      "total_questions": 2370,
      "avg_cost_per_task_usd": 0.018473258227848097,
      "avg_duration_per_task_ms": 138420.12109704642,
      "avg_output_tokens_per_task": 18520.817721518986,
      "last_run_timestamp": "2026-10-28T04:49:31.929Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-sonnet-5-20260630",
      "model_slug": "anthropic/claude-sonnet-5",
      "model_name": "Anthropic: Claude Sonnet 5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.8449648181818183,
      "accuracy_std_dev": 0.05078720577757961,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.16265579063360883,
      "avg_duration_per_task_ms": 162118.34848484848,
      "avg_output_tokens_per_task": 16092.63269054178,
      "last_run_timestamp": "2026-10-24T08:18:32.507Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Fireworks",
      "accuracy": 0.844913105281815,
      "accuracy_std_dev": 0.019516578349245098,
      "run_count": 15,
      "total_questions": 2821,
      "avg_cost_per_task_usd": 0.13749859695143568,
      "avg_duration_per_task_ms": 278207.19319390284,
      "avg_output_tokens_per_task": 28367.055654023396,
      "last_run_timestamp": "2026-10-24T08:44:43.305Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.8448274942528734,
      "accuracy_std_dev": 0.009144082018821893,
      "run_count": 11,
      "total_questions": 1914,
      "avg_cost_per_task_usd": 0.12633844566353186,
      "avg_duration_per_task_ms": 154602.84848484848,
      "avg_output_tokens_per_task": 31467.47910135841,
      "last_run_timestamp": "2026-10-29T23:21:54.785Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "StreamLake",
      "accuracy": 0.8448157507598784,
      "accuracy_std_dev": 0.02762329063670728,
      "run_count": 10,
      "total_questions": 1974,
      "avg_cost_per_task_usd": 0.013124052684903749,
      "avg_duration_per_task_ms": 664201.0739614995,
      "avg_output_tokens_per_task": 40826.44782168186,
      "last_run_timestamp": "2026-10-24T08:41:23.787Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Alibaba",
      "accuracy": 0.8445204193548387,
      "accuracy_std_dev": 0.018404855556345665,
      "run_count": 31,
      "total_questions": 6138,
      "avg_cost_per_task_usd": 0.2421819868035191,
      "avg_duration_per_task_ms": 500404.2896709026,
      "avg_output_tokens_per_task": 47667.435158031934,
      "last_run_timestamp": "2026-10-28T04:01:19.334Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "StreamLake",
      "accuracy": 0.8443693333333332,
      "accuracy_std_dev": 0.01866431008338881,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.1411051307519641,
      "avg_duration_per_task_ms": 894934.3092031425,
      "avg_output_tokens_per_task": 55451.94893378227,
      "last_run_timestamp": "2026-10-24T16:03:25.470Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.844276,
      "accuracy_std_dev": 0.009259000000000017,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.036391522727272724,
      "avg_duration_per_task_ms": 53072.40404040404,
      "avg_output_tokens_per_task": 14468.015151515152,
      "last_run_timestamp": "2026-10-01T19:12:26.032Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.8441825555555555,
      "accuracy_std_dev": 0.02302992819070732,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.014299253647586982,
      "avg_duration_per_task_ms": 527298.7811447812,
      "avg_output_tokens_per_task": 38122.5847362514,
      "last_run_timestamp": "2026-10-24T08:42:05.088Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Relace",
      "accuracy": 0.8440474857142857,
      "accuracy_std_dev": 0.0037822104436426945,
      "run_count": 5,
      "total_questions": 980,
      "avg_cost_per_task_usd": 0.11554378775510205,
      "avg_duration_per_task_ms": 243142.18265306123,
      "avg_output_tokens_per_task": 31758.279591836734,
      "last_run_timestamp": "2026-10-24T08:42:05.876Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-122b-a10b-20260224",
      "model_slug": "qwen/qwen3.5-122b-a10b",
      "model_name": "Qwen: Qwen3.5-122B-A10B",
      "provider_name": "DeepInfra",
      "accuracy": 0.8440043168316831,
      "accuracy_std_dev": 0.006860781750965057,
      "run_count": 8,
      "total_questions": 1515,
      "avg_cost_per_task_usd": 0.05332531353135314,
      "avg_duration_per_task_ms": 266581.80264026404,
      "avg_output_tokens_per_task": 22128.052145214522,
      "last_run_timestamp": "2026-09-26T09:50:39.842Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Crusoe",
      "accuracy": 0.843855125,
      "accuracy_std_dev": 0.007974456116838023,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.1439108863636364,
      "avg_duration_per_task_ms": 148669.27525252526,
      "avg_output_tokens_per_task": 32529.41856060606,
      "last_run_timestamp": "2026-10-24T08:10:23.024Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Nebius",
      "accuracy": 0.8438239857868021,
      "accuracy_std_dev": 0.012941643289783599,
      "run_count": 5,
      "total_questions": 985,
      "avg_cost_per_task_usd": 0.272832538071066,
      "avg_duration_per_task_ms": 511204.8578680203,
      "avg_output_tokens_per_task": 61801.45989847716,
      "last_run_timestamp": "2026-10-24T16:14:04.272Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "Z.AI",
      "accuracy": 0.8435453974862529,
      "accuracy_std_dev": 0.029302991838613988,
      "run_count": 7,
      "total_questions": 1273,
      "avg_cost_per_task_usd": 0.09004028358208956,
      "avg_duration_per_task_ms": 1451205.2136684996,
      "avg_output_tokens_per_task": 41372.553024351924,
      "last_run_timestamp": "2026-10-24T16:07:10.473Z"
    },
    {
      "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": 0.8434345000000001,
      "accuracy_std_dev": 0.005050500000000013,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.020032184343434345,
      "avg_duration_per_task_ms": 751978.8939393939,
      "avg_output_tokens_per_task": 23503.446969696968,
      "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-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "Novita",
      "accuracy": 0.8434342499999999,
      "accuracy_std_dev": 0.01044584279450443,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.08852940214646465,
      "avg_duration_per_task_ms": 1063581.5290404041,
      "avg_output_tokens_per_task": 44778.66351010101,
      "last_run_timestamp": "2026-10-24T16:33:50.860Z"
    },
    {
      "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": 0.8434174840336134,
      "accuracy_std_dev": 0.0058964576706039785,
      "run_count": 7,
      "total_questions": 1190,
      "avg_cost_per_task_usd": 0.07875789159663865,
      "avg_duration_per_task_ms": 236163.5705882353,
      "avg_output_tokens_per_task": 19324.579831932773,
      "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-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Fireworks",
      "accuracy": 0.8427474515050167,
      "accuracy_std_dev": 0.014940270952109593,
      "run_count": 14,
      "total_questions": 2691,
      "avg_cost_per_task_usd": 0.021943730955035302,
      "avg_duration_per_task_ms": 258715.1181716834,
      "avg_output_tokens_per_task": 37901.94314381271,
      "last_run_timestamp": "2026-10-24T08:49:23.400Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "thinkingmachines/inkling-20260715",
      "model_slug": "thinkingmachines/inkling",
      "model_name": "Thinking Machines: Inkling",
      "provider_name": "Together",
      "accuracy": 0.8425437868421053,
      "accuracy_std_dev": 0.00573439306339873,
      "run_count": 6,
      "total_questions": 1140,
      "avg_cost_per_task_usd": 0.08756267807017544,
      "avg_duration_per_task_ms": 282086.36140350875,
      "avg_output_tokens_per_task": 21480.616666666665,
      "last_run_timestamp": "2026-10-01T16:06:12.801Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-122b-a10b-20260224",
      "model_slug": "qwen/qwen3.5-122b-a10b",
      "model_name": "Qwen: Qwen3.5-122B-A10B",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8424523333333334,
      "accuracy_std_dev": 0.03339142982446578,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.05276241708754208,
      "avg_duration_per_task_ms": 208964.68855218854,
      "avg_output_tokens_per_task": 21882.910353535353,
      "last_run_timestamp": "2026-10-24T16:02:47.280Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Novita",
      "accuracy": 0.8423346063291139,
      "accuracy_std_dev": 0.054296645253950386,
      "run_count": 12,
      "total_questions": 2370,
      "avg_cost_per_task_usd": 0.07867724514767932,
      "avg_duration_per_task_ms": 564357.3624472574,
      "avg_output_tokens_per_task": 43556.894092827,
      "last_run_timestamp": "2026-10-24T16:27:11.126Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "SambaNova",
      "accuracy": 0.842279922077922,
      "accuracy_std_dev": 0.020944075957997636,
      "run_count": 8,
      "total_questions": 1155,
      "avg_cost_per_task_usd": 0.017810632034632037,
      "avg_duration_per_task_ms": 95543.98354978354,
      "avg_output_tokens_per_task": 15283.030303030304,
      "last_run_timestamp": "2026-10-24T16:27:17.205Z"
    },
    {
      "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": "openai/gpt-5.4-mini-20260317",
      "model_slug": "openai/gpt-5.4-mini",
      "model_name": "OpenAI: GPT-5.4 Mini",
      "provider_name": "OpenAI",
      "accuracy": 0.841751,
      "accuracy_std_dev": 0.004761657064510223,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.06159810101010101,
      "avg_duration_per_task_ms": 79946.62794612795,
      "avg_output_tokens_per_task": 13559.319865319865,
      "last_run_timestamp": "2026-10-02T23:24:30.140Z"
    },
    {
      "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": 0.8417508571428571,
      "accuracy_std_dev": 0.04451412312458842,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.09816640909090911,
      "avg_duration_per_task_ms": 1173545.2987012987,
      "avg_output_tokens_per_task": 39422.10028860029,
      "last_run_timestamp": "2026-10-24T16:02:28.490Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Sail Research",
      "accuracy": 0.8415046100309946,
      "accuracy_std_dev": 0.012823272459877284,
      "run_count": 18,
      "total_questions": 3549,
      "avg_cost_per_task_usd": 0.11274710115525499,
      "avg_duration_per_task_ms": 322682.1735700197,
      "avg_output_tokens_per_task": 34551.41730064807,
      "last_run_timestamp": "2026-10-24T08:41:42.263Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Together",
      "accuracy": 0.84141817433047,
      "accuracy_std_dev": 0.011968232237469333,
      "run_count": 10,
      "total_questions": 1979,
      "avg_cost_per_task_usd": 0.14348533804951996,
      "avg_duration_per_task_ms": 205361.71248105104,
      "avg_output_tokens_per_task": 32378.852956038405,
      "last_run_timestamp": "2026-10-24T08:12:37.564Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Together",
      "accuracy": 0.8411750606060605,
      "accuracy_std_dev": 0.008569864104844968,
      "run_count": 7,
      "total_questions": 1254,
      "avg_cost_per_task_usd": 0.01576657097288676,
      "avg_duration_per_task_ms": 669466.4673046252,
      "avg_output_tokens_per_task": 16726.208931419456,
      "last_run_timestamp": "2026-08-29T18:29:51.235Z"
    },
    {
      "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": 0.8409089999999999,
      "accuracy_std_dev": 0.010943000000000036,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.03708839646464646,
      "avg_duration_per_task_ms": 47515.843434343435,
      "avg_output_tokens_per_task": 14746.045454545454,
      "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": "Google",
      "accuracy": 0.840488,
      "accuracy_std_dev": 0.00819350402758187,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.03994661426767677,
      "avg_duration_per_task_ms": 52435.7619949495,
      "avg_output_tokens_per_task": 14777.814393939394,
      "last_run_timestamp": "2026-10-01T19:13:57.987Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Makora",
      "accuracy": 0.8398882325043376,
      "accuracy_std_dev": 0.015516849481773015,
      "run_count": 10,
      "total_questions": 1729,
      "avg_cost_per_task_usd": 0.14290184152689414,
      "avg_duration_per_task_ms": 273124.987854251,
      "avg_output_tokens_per_task": 33153.25390399075,
      "last_run_timestamp": "2026-10-24T08:53:46.038Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Reka",
      "accuracy": 0.839837363066202,
      "accuracy_std_dev": 0.014816504049501316,
      "run_count": 15,
      "total_questions": 2870,
      "avg_cost_per_task_usd": 0.07680571289198605,
      "avg_duration_per_task_ms": 389476.49477351917,
      "avg_output_tokens_per_task": 22208.951219512193,
      "last_run_timestamp": "2026-10-24T08:10:23.015Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "OpenInference",
      "accuracy": 0.8398090817995909,
      "accuracy_std_dev": 0.059804520823169324,
      "run_count": 3,
      "total_questions": 489,
      "avg_cost_per_task_usd": 0.019161192229038854,
      "avg_duration_per_task_ms": 228340.47034764825,
      "avg_output_tokens_per_task": 18072.97955010225,
      "last_run_timestamp": "2026-10-17T16:05:00.316Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Io Net",
      "accuracy": 0.8397870000000001,
      "accuracy_std_dev": 0.06305207943174174,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.013501392255892255,
      "avg_duration_per_task_ms": 269245.1767676768,
      "avg_output_tokens_per_task": 24293.242424242424,
      "last_run_timestamp": "2026-10-01T17:44:47.331Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Venice",
      "accuracy": 0.839646375,
      "accuracy_std_dev": 0.01029199634348825,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.006103154040404041,
      "avg_duration_per_task_ms": 431518.17424242425,
      "avg_output_tokens_per_task": 25837.31818181818,
      "last_run_timestamp": "2026-10-24T16:03:25.398Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Cloudflare",
      "accuracy": 0.8396166498586238,
      "accuracy_std_dev": 0.03493831082667691,
      "run_count": 11,
      "total_questions": 2122,
      "avg_cost_per_task_usd": 0.029913213949104622,
      "avg_duration_per_task_ms": 288209.38784165884,
      "avg_output_tokens_per_task": 9239.83930254477,
      "last_run_timestamp": "2026-10-24T08:08:54.186Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "StreamLake",
      "accuracy": 0.8391320000000001,
      "accuracy_std_dev": 0.025876072714210524,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.15949684006734002,
      "avg_duration_per_task_ms": 973112.1127946128,
      "avg_output_tokens_per_task": 52412.06341189674,
      "last_run_timestamp": "2026-10-24T16:05:13.521Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Together",
      "accuracy": 0.8389520859671303,
      "accuracy_std_dev": 0.014418680686632029,
      "run_count": 12,
      "total_questions": 2373,
      "avg_cost_per_task_usd": 0.22912267003792672,
      "avg_duration_per_task_ms": 279595.97260851244,
      "avg_output_tokens_per_task": 51927.226717235564,
      "last_run_timestamp": "2026-10-24T16:04:55.201Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "DigitalOcean",
      "accuracy": 0.8389448333333334,
      "accuracy_std_dev": 0.03599784248820413,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.004737409090909091,
      "avg_duration_per_task_ms": 1034495.3484848485,
      "avg_output_tokens_per_task": 27462.325757575756,
      "last_run_timestamp": "2026-10-24T16:19:29.603Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Mistral",
      "accuracy": 0.8385887456445994,
      "accuracy_std_dev": 0.01177027910963325,
      "run_count": 29,
      "total_questions": 5740,
      "avg_cost_per_task_usd": 0.2625816120209059,
      "avg_duration_per_task_ms": 276081.5918118467,
      "avg_output_tokens_per_task": 57083.71358885017,
      "last_run_timestamp": "2026-10-30T16:10:59.813Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Novita",
      "accuracy": 0.838522452928095,
      "accuracy_std_dev": 0.010501612316687394,
      "run_count": 8,
      "total_questions": 1349,
      "avg_cost_per_task_usd": 0.006855791697553743,
      "avg_duration_per_task_ms": 622066.7916975537,
      "avg_output_tokens_per_task": 16928.25722757598,
      "last_run_timestamp": "2026-10-24T16:21:33.540Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Modal",
      "accuracy": 0.8383968759493671,
      "accuracy_std_dev": 0.006763999999999992,
      "run_count": 2,
      "total_questions": 395,
      "avg_cost_per_task_usd": 0.033454222784810125,
      "avg_duration_per_task_ms": 81357.47848101266,
      "avg_output_tokens_per_task": 14534.177215189873,
      "last_run_timestamp": "2026-10-24T09:12:44.274Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "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": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "Venice",
      "accuracy": 0.8377524999999999,
      "accuracy_std_dev": 0.05290935711667643,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.09474105176767675,
      "avg_duration_per_task_ms": 401286.8162878788,
      "avg_output_tokens_per_task": 21287.625,
      "last_run_timestamp": "2026-10-24T16:06:30.903Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Novita",
      "accuracy": 0.8377103,
      "accuracy_std_dev": 0.023183391999662194,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.10627496717171718,
      "avg_duration_per_task_ms": 386419.75505050505,
      "avg_output_tokens_per_task": 35856.45909090909,
      "last_run_timestamp": "2026-10-24T08:25:29.458Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "StreamLake",
      "accuracy": 0.837692958038147,
      "accuracy_std_dev": 0.020571358270420545,
      "run_count": 10,
      "total_questions": 1835,
      "avg_cost_per_task_usd": 0.07314428610354223,
      "avg_duration_per_task_ms": 380364.9722070845,
      "avg_output_tokens_per_task": 24128.798910081743,
      "last_run_timestamp": "2026-10-24T16:08:29.374Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8376859309031556,
      "accuracy_std_dev": 0.027367300407603226,
      "run_count": 10,
      "total_questions": 1838,
      "avg_cost_per_task_usd": 0.011182803590859632,
      "avg_duration_per_task_ms": 390045.14526659413,
      "avg_output_tokens_per_task": 22148.7992383025,
      "last_run_timestamp": "2026-10-24T08:39:40.816Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8375420000000001,
      "accuracy_std_dev": 0.005646742468007568,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.02810060606060606,
      "avg_duration_per_task_ms": 530216.4090909091,
      "avg_output_tokens_per_task": 23231.272727272728,
      "last_run_timestamp": "2026-08-29T18:16:39.840Z"
    },
    {
      "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": 0.837542,
      "accuracy_std_dev": 0.005892000000000008,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.06186689898989899,
      "avg_duration_per_task_ms": 75089.61363636363,
      "avg_output_tokens_per_task": 13620.507575757576,
      "last_run_timestamp": "2026-10-01T19:12:44.401Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Nebius",
      "accuracy": 0.8375419999999999,
      "accuracy_std_dev": 0.012922002863333565,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.15197074579124578,
      "avg_duration_per_task_ms": 226641.9739057239,
      "avg_output_tokens_per_task": 34291.69781144781,
      "last_run_timestamp": "2026-10-24T08:43:12.480Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Chutes",
      "accuracy": 0.8372897388724037,
      "accuracy_std_dev": 0.028847852416095617,
      "run_count": 12,
      "total_questions": 2359,
      "avg_cost_per_task_usd": 0.06617999915218314,
      "avg_duration_per_task_ms": 1053779.2886816447,
      "avg_output_tokens_per_task": 33011.97880457821,
      "last_run_timestamp": "2026-10-24T16:02:06.932Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8371210833333332,
      "accuracy_std_dev": 0.012487611784340083,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.0845222638888889,
      "avg_duration_per_task_ms": 727103.742003367,
      "avg_output_tokens_per_task": 48945.89772727273,
      "last_run_timestamp": "2026-10-24T16:11:13.852Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Modal",
      "accuracy": 0.8368686,
      "accuracy_std_dev": 0.01362896250783603,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.1397023909090909,
      "avg_duration_per_task_ms": 182914.46161616163,
      "avg_output_tokens_per_task": 31522.08787878788,
      "last_run_timestamp": "2026-10-24T08:24:39.068Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.8367000000000001,
      "accuracy_std_dev": 0.006734000000000018,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.18573769191919193,
      "avg_duration_per_task_ms": 281534.297979798,
      "avg_output_tokens_per_task": 34913.7095959596,
      "last_run_timestamp": "2026-10-24T08:56:15.847Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "DeepInfra",
      "accuracy": 0.8366667090476191,
      "accuracy_std_dev": 0.01978861195373632,
      "run_count": 12,
      "total_questions": 2100,
      "avg_cost_per_task_usd": 0.09574838047619047,
      "avg_duration_per_task_ms": 613293.1757142857,
      "avg_output_tokens_per_task": 29845.292857142857,
      "last_run_timestamp": "2026-10-24T16:10:16.903Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "AtlasCloud",
      "accuracy": 0.836653454183267,
      "accuracy_std_dev": 0.007833999999999952,
      "run_count": 2,
      "total_questions": 251,
      "avg_cost_per_task_usd": 0.06977503187250997,
      "avg_duration_per_task_ms": 237256.35059760956,
      "avg_output_tokens_per_task": 17511.450199203187,
      "last_run_timestamp": "2026-08-15T08:47:41.523Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Nebius",
      "accuracy": 0.836139,
      "accuracy_std_dev": 0.05678065343641852,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.005011018518518518,
      "avg_duration_per_task_ms": 214818.47643097644,
      "avg_output_tokens_per_task": 17529.722222222223,
      "last_run_timestamp": "2026-09-26T11:06:41.988Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "Reka",
      "accuracy": 0.836139,
      "accuracy_std_dev": 0.0007933738084913017,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.05002690909090909,
      "avg_duration_per_task_ms": 537584.335016835,
      "avg_output_tokens_per_task": 49932.06734006734,
      "last_run_timestamp": "2026-09-30T11:44:08.104Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Reka",
      "accuracy": 0.8360894826189674,
      "accuracy_std_dev": 0.027487401181389745,
      "run_count": 15,
      "total_questions": 2963,
      "avg_cost_per_task_usd": 0.02392725244684441,
      "avg_duration_per_task_ms": 312624.6992912589,
      "avg_output_tokens_per_task": 17783.592642591968,
      "last_run_timestamp": "2026-10-24T08:22:23.704Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Z.AI",
      "accuracy": 0.83599875,
      "accuracy_std_dev": 0.06733100699544627,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.20365952861952863,
      "avg_duration_per_task_ms": 537295.401936027,
      "avg_output_tokens_per_task": 46280.556818181816,
      "last_run_timestamp": "2026-10-24T16:48:27.857Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "inclusionai/ling-3.0-flash-vl-20260910",
      "model_slug": "inclusionai/ling-3.0-flash-vl",
      "model_name": "inclusionAI: Ling 3.0 Flash VL",
      "provider_name": "Novita",
      "accuracy": 0.8358585,
      "accuracy_std_dev": 0.01430950000000003,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.0011451691919191918,
      "avg_duration_per_task_ms": 116299.33333333333,
      "avg_output_tokens_per_task": 18355.219696969696,
      "last_run_timestamp": "2026-10-01T08:23:48.135Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.8354376249999999,
      "accuracy_std_dev": 0.020682897348639886,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.09533192992424241,
      "avg_duration_per_task_ms": 1095682.7007575757,
      "avg_output_tokens_per_task": 43690.37752525252,
      "last_run_timestamp": "2026-10-24T16:12:09.457Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Mara",
      "accuracy": 0.835384616923077,
      "accuracy_std_dev": 0.032642666915557006,
      "run_count": 5,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.09615046153846153,
      "avg_duration_per_task_ms": 334242.9476923077,
      "avg_output_tokens_per_task": 39819.60307692308,
      "last_run_timestamp": "2026-10-24T16:40:24.303Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Relace",
      "accuracy": 0.8353175277777779,
      "accuracy_std_dev": 0.01836981616741691,
      "run_count": 9,
      "total_questions": 1764,
      "avg_cost_per_task_usd": 0.014329623015873016,
      "avg_duration_per_task_ms": 445894.7193877551,
      "avg_output_tokens_per_task": 39122.69671201814,
      "last_run_timestamp": "2026-10-24T08:17:53.301Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "inclusionai/ling-3.0-flash-vl-20260910",
      "model_slug": "inclusionai/ling-3.0-flash-vl",
      "model_name": "inclusionAI: Ling 3.0 Flash VL",
      "provider_name": "DeepInfra",
      "accuracy": 0.8351849999999998,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 180,
      "avg_cost_per_task_usd": 0.0030483999999999997,
      "avg_duration_per_task_ms": 288802.22222222225,
      "avg_output_tokens_per_task": 16703.5,
      "last_run_timestamp": "2026-10-01T08:43:11.415Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "AkashML",
      "accuracy": 0.8349678663276426,
      "accuracy_std_dev": 0.012269176788711802,
      "run_count": 17,
      "total_questions": 2753,
      "avg_cost_per_task_usd": 0.11747415800944423,
      "avg_duration_per_task_ms": 293233.16164184525,
      "avg_output_tokens_per_task": 27740.3185615692,
      "last_run_timestamp": "2026-10-28T12:31:58.209Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "SiliconFlow",
      "accuracy": 0.834789191724138,
      "accuracy_std_dev": 0.03295145245372225,
      "run_count": 11,
      "total_questions": 2175,
      "avg_cost_per_task_usd": 0.10863842022988504,
      "avg_duration_per_task_ms": 1052937.216091954,
      "avg_output_tokens_per_task": 33937.42436781609,
      "last_run_timestamp": "2026-10-24T16:17:57.470Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Alibaba",
      "accuracy": 0.8344555333333333,
      "accuracy_std_dev": 0.011296009695856743,
      "run_count": 15,
      "total_questions": 2970,
      "avg_cost_per_task_usd": 0.24910045925925925,
      "avg_duration_per_task_ms": 387517.9037037037,
      "avg_output_tokens_per_task": 40180.79898989899,
      "last_run_timestamp": "2026-10-28T04:11:19.238Z"
    },
    {
      "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": 0.8343080263157894,
      "accuracy_std_dev": 0.018459495220003813,
      "run_count": 16,
      "total_questions": 3078,
      "avg_cost_per_task_usd": 0.260129997725796,
      "avg_duration_per_task_ms": 393015.818388564,
      "avg_output_tokens_per_task": 45642.69135802469,
      "last_run_timestamp": "2026-10-24T16:03:06.205Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Phala",
      "accuracy": 0.8341984152249134,
      "accuracy_std_dev": 0.029002188381032377,
      "run_count": 3,
      "total_questions": 578,
      "avg_cost_per_task_usd": 0.005492470588235294,
      "avg_duration_per_task_ms": 956260.6712802768,
      "avg_output_tokens_per_task": 15404.778546712803,
      "last_run_timestamp": "2026-08-29T18:08:44.008Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "DekaLLM",
      "accuracy": 0.8340433944004868,
      "accuracy_std_dev": 0.008747099804372552,
      "run_count": 9,
      "total_questions": 1643,
      "avg_cost_per_task_usd": 0.04404786731588557,
      "avg_duration_per_task_ms": 94664.89896530737,
      "avg_output_tokens_per_task": 15175.068167985393,
      "last_run_timestamp": "2026-10-24T08:10:23.016Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "thinkingmachines/inkling-20260715",
      "model_slug": "thinkingmachines/inkling",
      "model_name": "Thinking Machines: Inkling",
      "provider_name": "auto-routing",
      "accuracy": 0.8340364430379746,
      "accuracy_std_dev": 0.01069546220626094,
      "run_count": 6,
      "total_questions": 1185,
      "avg_cost_per_task_usd": 0.0954248565400844,
      "avg_duration_per_task_ms": 293770.05400843883,
      "avg_output_tokens_per_task": 23403.932489451476,
      "last_run_timestamp": "2026-10-01T16:11:05.832Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8338943333333333,
      "accuracy_std_dev": 0.011690806112307073,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.03773407239057239,
      "avg_duration_per_task_ms": 276396.1228956229,
      "avg_output_tokens_per_task": 28290.345117845118,
      "last_run_timestamp": "2026-09-12T09:56:44.384Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Parasail",
      "accuracy": 0.8336574603112842,
      "accuracy_std_dev": 0.021426026057470033,
      "run_count": 13,
      "total_questions": 2570,
      "avg_cost_per_task_usd": 0.024801823735408558,
      "avg_duration_per_task_ms": 393313.6081712062,
      "avg_output_tokens_per_task": 46122.12840466926,
      "last_run_timestamp": "2026-10-24T08:48:13.246Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-122b-a10b-20260224",
      "model_slug": "qwen/qwen3.5-122b-a10b",
      "model_name": "Qwen: Qwen3.5-122B-A10B",
      "provider_name": "auto-routing",
      "accuracy": 0.8334735000000002,
      "accuracy_std_dev": 0.03349024054482143,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.05413035648148148,
      "avg_duration_per_task_ms": 251545.94865319866,
      "avg_output_tokens_per_task": 23002.68813131313,
      "last_run_timestamp": "2026-10-24T16:06:30.902Z"
    },
    {
      "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": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Alibaba",
      "accuracy": 0.833193,
      "accuracy_std_dev": 0.011323617332813778,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.03877796464646465,
      "avg_duration_per_task_ms": 229512.37962962964,
      "avg_output_tokens_per_task": 14499.016414141413,
      "last_run_timestamp": "2026-10-28T03:33:49.253Z"
    },
    {
      "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-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Parasail",
      "accuracy": 0.8327628225898459,
      "accuracy_std_dev": 0.014159242972544667,
      "run_count": 9,
      "total_questions": 1753,
      "avg_cost_per_task_usd": 0.15039315516257845,
      "avg_duration_per_task_ms": 277652.9520821449,
      "avg_output_tokens_per_task": 33995.16029663434,
      "last_run_timestamp": "2026-10-24T08:12:02.633Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Parasail",
      "accuracy": 0.832178375487224,
      "accuracy_std_dev": 0.011051436981414083,
      "run_count": 12,
      "total_questions": 2309,
      "avg_cost_per_task_usd": 0.2290666569943699,
      "avg_duration_per_task_ms": 296861.6738847986,
      "avg_output_tokens_per_task": 51892.169770463406,
      "last_run_timestamp": "2026-10-24T16:31:42.371Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "StreamLake",
      "accuracy": 0.8320383846153845,
      "accuracy_std_dev": 0.008064443483478893,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.07931912276612277,
      "avg_duration_per_task_ms": 701901.0648795649,
      "avg_output_tokens_per_task": 57457.6013986014,
      "last_run_timestamp": "2026-10-24T16:24:24.498Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.4-mini-20260317",
      "model_slug": "openai/gpt-5.4-mini",
      "model_name": "OpenAI: GPT-5.4 Mini",
      "provider_name": "Azure",
      "accuracy": 0.8319864,
      "accuracy_std_dev": 0.01402072559605957,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.06441977171717173,
      "avg_duration_per_task_ms": 117114.99292929293,
      "avg_output_tokens_per_task": 13401.907070707071,
      "last_run_timestamp": "2026-10-01T19:14:05.661Z"
    },
    {
      "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": 0.8318702952978058,
      "accuracy_std_dev": 0.01302850195570758,
      "run_count": 9,
      "total_questions": 1595,
      "avg_cost_per_task_usd": 0.21279191097178685,
      "avg_duration_per_task_ms": 458880.45830721,
      "avg_output_tokens_per_task": 48250.53855799373,
      "last_run_timestamp": "2026-10-28T11:00:59.554Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Crusoe",
      "accuracy": 0.83176317025572,
      "accuracy_std_dev": 0.014297695836545667,
      "run_count": 8,
      "total_questions": 1486,
      "avg_cost_per_task_usd": 0.023221829744279946,
      "avg_duration_per_task_ms": 299260.8001345895,
      "avg_output_tokens_per_task": 46221.84320323015,
      "last_run_timestamp": "2026-10-24T08:22:23.703Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "CoreWeave",
      "accuracy": 0.8316682557442558,
      "accuracy_std_dev": 0.01526659275642514,
      "run_count": 13,
      "total_questions": 2002,
      "avg_cost_per_task_usd": 0.039858755744255736,
      "avg_duration_per_task_ms": 286266.92507492506,
      "avg_output_tokens_per_task": 31778.31018981019,
      "last_run_timestamp": "2026-10-24T16:05:40.120Z"
    },
    {
      "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": 0.83143925,
      "accuracy_std_dev": 0.015980774330034836,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.1378152821969697,
      "avg_duration_per_task_ms": 293606.83648989897,
      "avg_output_tokens_per_task": 56806.145833333336,
      "last_run_timestamp": "2026-10-24T16:04:33.470Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "DeepInfra",
      "accuracy": 0.8313810512445096,
      "accuracy_std_dev": 0.012942947508356888,
      "run_count": 7,
      "total_questions": 1366,
      "avg_cost_per_task_usd": 0.027591674963396776,
      "avg_duration_per_task_ms": 601596.977306003,
      "avg_output_tokens_per_task": 14650.56149341142,
      "last_run_timestamp": "2026-10-24T08:48:04.402Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Io Net",
      "accuracy": 0.8312629347826086,
      "accuracy_std_dev": 0.044536486139877485,
      "run_count": 4,
      "total_questions": 644,
      "avg_cost_per_task_usd": 0.005694647515527951,
      "avg_duration_per_task_ms": 270388.4906832298,
      "avg_output_tokens_per_task": 13521.194099378881,
      "last_run_timestamp": "2026-08-22T09:56:05.550Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Cerebras",
      "accuracy": 0.83122875,
      "accuracy_std_dev": 0.007839878231675557,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.022249369949494953,
      "avg_duration_per_task_ms": 11972.136363636364,
      "avg_output_tokens_per_task": 14384.287878787878,
      "last_run_timestamp": "2026-10-24T08:21:05.266Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Parasail",
      "accuracy": 0.8311627944186045,
      "accuracy_std_dev": 0.006786821620057476,
      "run_count": 11,
      "total_questions": 2150,
      "avg_cost_per_task_usd": 0.03491511023255814,
      "avg_duration_per_task_ms": 201279.6446511628,
      "avg_output_tokens_per_task": 14261.299069767441,
      "last_run_timestamp": "2026-10-24T08:41:23.420Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "AkashML",
      "accuracy": 0.8311506707700367,
      "accuracy_std_dev": 0.02069707210422527,
      "run_count": 20,
      "total_questions": 3818,
      "avg_cost_per_task_usd": 0.04166823703509692,
      "avg_duration_per_task_ms": 443524.7218438973,
      "avg_output_tokens_per_task": 45446.65243583028,
      "last_run_timestamp": "2026-10-30T00:41:07.469Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "BaseTen",
      "accuracy": 0.8310956896816264,
      "accuracy_std_dev": 0.010867420886961873,
      "run_count": 16,
      "total_questions": 2607,
      "avg_cost_per_task_usd": 0.020108517453011125,
      "avg_duration_per_task_ms": 239467.13693901035,
      "avg_output_tokens_per_task": 40022.99923283468,
      "last_run_timestamp": "2026-10-24T08:13:46.749Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Wafer",
      "accuracy": 0.8309013333333333,
      "accuracy_std_dev": 0.004566576227814942,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.048388571829405154,
      "avg_duration_per_task_ms": 267954.62850729516,
      "avg_output_tokens_per_task": 15225.407968574635,
      "last_run_timestamp": "2026-10-29T22:38:23.385Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Chutes",
      "accuracy": 0.830874032193159,
      "accuracy_std_dev": 0.029241810305408605,
      "run_count": 8,
      "total_questions": 1491,
      "avg_cost_per_task_usd": 0.00674413145539906,
      "avg_duration_per_task_ms": 717174.3158953722,
      "avg_output_tokens_per_task": 18114.10865191147,
      "last_run_timestamp": "2026-10-24T16:05:36.255Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8306676666666667,
      "accuracy_std_dev": 0.02403995285052688,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.13645198905723904,
      "avg_duration_per_task_ms": 551462.0761784512,
      "avg_output_tokens_per_task": 43445.10816498317,
      "last_run_timestamp": "2026-10-24T16:02:47.376Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Parasail",
      "accuracy": 0.8302375046439628,
      "accuracy_std_dev": 0.02221590376784662,
      "run_count": 3,
      "total_questions": 323,
      "avg_cost_per_task_usd": 0.10860701238390094,
      "avg_duration_per_task_ms": 289034.8173374613,
      "avg_output_tokens_per_task": 24728.442724458204,
      "last_run_timestamp": "2026-08-22T17:29:49.714Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "Venice",
      "accuracy": 0.8301606153846154,
      "accuracy_std_dev": 0.025057780433164603,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.038417987567987574,
      "avg_duration_per_task_ms": 349233.05205905205,
      "avg_output_tokens_per_task": 39432.61577311577,
      "last_run_timestamp": "2026-10-24T16:01:48.224Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Friendli",
      "accuracy": 0.8299001138748543,
      "accuracy_std_dev": 0.01880026569622147,
      "run_count": 13,
      "total_questions": 2573,
      "avg_cost_per_task_usd": 0.2475731441896619,
      "avg_duration_per_task_ms": 394929.43801010493,
      "avg_output_tokens_per_task": 56101.092499028375,
      "last_run_timestamp": "2026-10-28T11:16:13.363Z"
    },
    {
      "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": 0.8295452499999999,
      "accuracy_std_dev": 0.02345987124192076,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.07824036237373737,
      "avg_duration_per_task_ms": 700469.6212121212,
      "avg_output_tokens_per_task": 42273.732323232325,
      "last_run_timestamp": "2026-10-04T00:07:08.217Z"
    },
    {
      "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": 0.8293709188811189,
      "accuracy_std_dev": 0.009670510440897125,
      "run_count": 4,
      "total_questions": 715,
      "avg_cost_per_task_usd": 0.025556813986013987,
      "avg_duration_per_task_ms": 118281.40559440559,
      "avg_output_tokens_per_task": 20315.50909090909,
      "last_run_timestamp": "2026-08-29T18:29:13.946Z"
    },
    {
      "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": 0.8288160468661773,
      "accuracy_std_dev": 0.05013602361575955,
      "run_count": 9,
      "total_questions": 1771,
      "avg_cost_per_task_usd": 0.1908068712591756,
      "avg_duration_per_task_ms": 608860.6832298137,
      "avg_output_tokens_per_task": 45412.08865047996,
      "last_run_timestamp": "2026-10-28T04:00:18.005Z"
    },
    {
      "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": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Alibaba",
      "accuracy": 0.8284027318022578,
      "accuracy_std_dev": 0.02855983228366381,
      "run_count": 13,
      "total_questions": 2569,
      "avg_cost_per_task_usd": 0.08278847567146748,
      "avg_duration_per_task_ms": 354761.10003892565,
      "avg_output_tokens_per_task": 20569.06267029973,
      "last_run_timestamp": "2026-10-28T03:29:22.918Z"
    },
    {
      "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-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Friendli",
      "accuracy": 0.8279138147973608,
      "accuracy_std_dev": 0.02296828124369033,
      "run_count": 11,
      "total_questions": 2122,
      "avg_cost_per_task_usd": 0.021867254476908574,
      "avg_duration_per_task_ms": 225918.76531573987,
      "avg_output_tokens_per_task": 43510.79641847314,
      "last_run_timestamp": "2026-10-28T10:43:17.083Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Cloudflare",
      "accuracy": 0.8276692169459963,
      "accuracy_std_dev": 0.017541679518523267,
      "run_count": 11,
      "total_questions": 2148,
      "avg_cost_per_task_usd": 0.14409020577281192,
      "avg_duration_per_task_ms": 292895.9767225326,
      "avg_output_tokens_per_task": 32531.135009310987,
      "last_run_timestamp": "2026-10-24T08:12:02.536Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "SiliconFlow",
      "accuracy": 0.827651625,
      "accuracy_std_dev": 0.00789359865551672,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.04827323547979799,
      "avg_duration_per_task_ms": 460754.56123737374,
      "avg_output_tokens_per_task": 26739.357954545456,
      "last_run_timestamp": "2026-10-24T16:31:34.472Z"
    },
    {
      "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": 0.827440875,
      "accuracy_std_dev": 0.011355911736156446,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.008075719065656566,
      "avg_duration_per_task_ms": 508001.31565656565,
      "avg_output_tokens_per_task": 20787.5827020202,
      "last_run_timestamp": "2026-10-24T16:12:11.567Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-122b-a10b-20260224",
      "model_slug": "qwen/qwen3.5-122b-a10b",
      "model_name": "Qwen: Qwen3.5-122B-A10B",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8273709166666667,
      "accuracy_std_dev": 0.056014389335030584,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.051560792508417504,
      "avg_duration_per_task_ms": 531412.6561447812,
      "avg_output_tokens_per_task": 24685.848905723906,
      "last_run_timestamp": "2026-10-24T16:11:29.497Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Modal",
      "accuracy": 0.8273257868177137,
      "accuracy_std_dev": 0.013461365335284555,
      "run_count": 10,
      "total_questions": 1942,
      "avg_cost_per_task_usd": 0.0351326035015448,
      "avg_duration_per_task_ms": 424285.9001029866,
      "avg_output_tokens_per_task": 48592.330072090626,
      "last_run_timestamp": "2026-10-17T08:18:01.015Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Morph",
      "accuracy": 0.8272187207100591,
      "accuracy_std_dev": 0.0496186451996,
      "run_count": 13,
      "total_questions": 2535,
      "avg_cost_per_task_usd": 0.10137412702169626,
      "avg_duration_per_task_ms": 538485.309270217,
      "avg_output_tokens_per_task": 29237.781459566075,
      "last_run_timestamp": "2026-10-24T08:19:59.641Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "Phala",
      "accuracy": 0.8268798333333333,
      "accuracy_std_dev": 0.014166079426064071,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.06329198569023568,
      "avg_duration_per_task_ms": 528541.8636363636,
      "avg_output_tokens_per_task": 49829.92003367003,
      "last_run_timestamp": "2026-10-24T16:07:10.471Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "Google",
      "accuracy": 0.8268397142857142,
      "accuracy_std_dev": 0.013924120672049498,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.09914931746031745,
      "avg_duration_per_task_ms": 361736.2164502164,
      "avg_output_tokens_per_task": 45492.343434343435,
      "last_run_timestamp": "2026-10-17T16:06:16.399Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Ionstream",
      "accuracy": 0.8267677,
      "accuracy_std_dev": 0.010999569410208776,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.03662119747474748,
      "avg_duration_per_task_ms": 290125.5409090909,
      "avg_output_tokens_per_task": 15502.71515151515,
      "last_run_timestamp": "2026-10-24T08:13:28.065Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Novita",
      "accuracy": 0.82659925,
      "accuracy_std_dev": 0.019009280711996968,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.02540684343434343,
      "avg_duration_per_task_ms": 237042.44823232322,
      "avg_output_tokens_per_task": 21223.366161616163,
      "last_run_timestamp": "2026-08-29T18:12:43.943Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Crusoe",
      "accuracy": 0.8264543613323679,
      "accuracy_std_dev": 0.018188360323715465,
      "run_count": 7,
      "total_questions": 1381,
      "avg_cost_per_task_usd": 0.20616153874004342,
      "avg_duration_per_task_ms": 315117.1817523534,
      "avg_output_tokens_per_task": 46875.9225199131,
      "last_run_timestamp": "2026-09-12T08:53:53.147Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Novita",
      "accuracy": 0.8264309000000002,
      "accuracy_std_dev": 0.009654922303674958,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.04312889949494949,
      "avg_duration_per_task_ms": 229184.7,
      "avg_output_tokens_per_task": 14260.256565656566,
      "last_run_timestamp": "2026-10-24T08:44:44.600Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Friendli",
      "accuracy": 0.8262931818181818,
      "accuracy_std_dev": 0.009811627071241754,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.14266403719008267,
      "avg_duration_per_task_ms": 193077.25711662075,
      "avg_output_tokens_per_task": 35369.469696969696,
      "last_run_timestamp": "2026-10-28T11:04:39.923Z"
    },
    {
      "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": 0.8259678749999999,
      "accuracy_std_dev": 0.013440054505446589,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.19472516477272722,
      "avg_duration_per_task_ms": 963909.0416666666,
      "avg_output_tokens_per_task": 51831.05113636364,
      "last_run_timestamp": "2026-10-24T16:12:49.071Z"
    },
    {
      "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": 0.8258736540920717,
      "accuracy_std_dev": 0.01098913364419598,
      "run_count": 8,
      "total_questions": 1564,
      "avg_cost_per_task_usd": 0.00797793094629156,
      "avg_duration_per_task_ms": 338576.25767263427,
      "avg_output_tokens_per_task": 19235.87084398977,
      "last_run_timestamp": "2026-10-24T16:01:08.902Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Cohere",
      "accuracy": 0.8257575,
      "accuracy_std_dev": 0.017676499999999984,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.02807809090909091,
      "avg_duration_per_task_ms": 230919.77777777778,
      "avg_output_tokens_per_task": 55916.378787878784,
      "last_run_timestamp": "2026-10-24T08:15:10.265Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-2.5-pro",
      "model_slug": "google/gemini-2.5-pro",
      "model_name": "Google: Gemini 2.5 Pro",
      "provider_name": "auto-routing",
      "accuracy": 0.8257574999999999,
      "accuracy_std_dev": 0.015993499999999994,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.2609003181818182,
      "avg_duration_per_task_ms": 231557.84848484848,
      "avg_output_tokens_per_task": 25997.717171717173,
      "last_run_timestamp": "2026-10-01T19:12:42.150Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "Darkbloom",
      "accuracy": 0.8257141371428571,
      "accuracy_std_dev": 0.01176326794163367,
      "run_count": 3,
      "total_questions": 525,
      "avg_cost_per_task_usd": 0.010625175238095239,
      "avg_duration_per_task_ms": 1617186.0876190476,
      "avg_output_tokens_per_task": 30239.152380952382,
      "last_run_timestamp": "2026-10-30T05:26:04.868Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Morph",
      "accuracy": 0.8253999912733042,
      "accuracy_std_dev": 0.022060427402287506,
      "run_count": 13,
      "total_questions": 2521,
      "avg_cost_per_task_usd": 0.027058457754859183,
      "avg_duration_per_task_ms": 873701.1269337564,
      "avg_output_tokens_per_task": 48650.64220547402,
      "last_run_timestamp": "2026-10-24T08:26:00.559Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "CoreWeave",
      "accuracy": 0.8253968696428573,
      "accuracy_std_dev": 0.014772865576229071,
      "run_count": 9,
      "total_questions": 1680,
      "avg_cost_per_task_usd": 0.024977942261904765,
      "avg_duration_per_task_ms": 272190.5416666667,
      "avg_output_tokens_per_task": 49745.9630952381,
      "last_run_timestamp": "2026-10-24T08:27:40.063Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "auto-routing",
      "accuracy": 0.8252562970922881,
      "accuracy_std_dev": 0.043641090936944585,
      "run_count": 12,
      "total_questions": 2373,
      "avg_cost_per_task_usd": 0.08621131184155077,
      "avg_duration_per_task_ms": 759710.0198061526,
      "avg_output_tokens_per_task": 31352.24863042562,
      "last_run_timestamp": "2026-10-24T16:13:03.130Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "OpenInference",
      "accuracy": 0.8246474711042312,
      "accuracy_std_dev": 0.04610697461436829,
      "run_count": 10,
      "total_questions": 1938,
      "avg_cost_per_task_usd": 0.019926908668730647,
      "avg_duration_per_task_ms": 787644.0030959753,
      "avg_output_tokens_per_task": 45616.08255933953,
      "last_run_timestamp": "2026-10-24T08:22:23.706Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "auto-routing",
      "accuracy": 0.8244948000000002,
      "accuracy_std_dev": 0.0601779527199123,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.028505976262626263,
      "avg_duration_per_task_ms": 700158.9757575757,
      "avg_output_tokens_per_task": 29700.529292929292,
      "last_run_timestamp": "2026-10-17T08:18:04.911Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-2.5-pro",
      "model_slug": "google/gemini-2.5-pro",
      "model_name": "Google: Gemini 2.5 Pro",
      "provider_name": "Google",
      "accuracy": 0.8243546666666667,
      "accuracy_std_dev": 0.03111308863734931,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.25956171212121215,
      "avg_duration_per_task_ms": 233681.0909090909,
      "avg_output_tokens_per_task": 25863.887205387204,
      "last_run_timestamp": "2026-10-01T19:17:25.275Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "CoreWeave",
      "accuracy": 0.8240257172774869,
      "accuracy_std_dev": 0.000479814547507715,
      "run_count": 3,
      "total_questions": 573,
      "avg_cost_per_task_usd": 0.1285148184991274,
      "avg_duration_per_task_ms": 295843.96684118675,
      "avg_output_tokens_per_task": 31965.242582897034,
      "last_run_timestamp": "2026-10-24T08:50:12.983Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "Xiaomi",
      "accuracy": 0.8239975454545455,
      "accuracy_std_dev": 0.022862830634275574,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.010785916437098255,
      "avg_duration_per_task_ms": 717852.0280073462,
      "avg_output_tokens_per_task": 38282.08126721763,
      "last_run_timestamp": "2026-10-17T08:02:36.531Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Venice",
      "accuracy": 0.8239540000000001,
      "accuracy_std_dev": 0.009474601205327877,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.14372687085137084,
      "avg_duration_per_task_ms": 420021.29581529583,
      "avg_output_tokens_per_task": 28768.17316017316,
      "last_run_timestamp": "2026-10-24T08:48:18.601Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.1-terminus",
      "model_slug": "deepseek/deepseek-v3.1-terminus",
      "model_name": "DeepSeek: DeepSeek V3.1 Terminus",
      "provider_name": "Novita",
      "accuracy": 0.8236529999999999,
      "accuracy_std_dev": 0.0034448118235979193,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.018703443181818182,
      "avg_duration_per_task_ms": 518705.38636363635,
      "avg_output_tokens_per_task": 18559.679292929293,
      "last_run_timestamp": "2026-08-29T18:35:57.732Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "AkashML",
      "accuracy": 0.823635741091314,
      "accuracy_std_dev": 0.01667549518358386,
      "run_count": 19,
      "total_questions": 3592,
      "avg_cost_per_task_usd": 0.03049394097995545,
      "avg_duration_per_task_ms": 266457.14532293985,
      "avg_output_tokens_per_task": 14043.167873051225,
      "last_run_timestamp": "2026-10-30T00:29:53.146Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "Novita",
      "accuracy": 0.823232125,
      "accuracy_std_dev": 0.02934086116850314,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.09752844696969695,
      "avg_duration_per_task_ms": 1429250.3825757576,
      "avg_output_tokens_per_task": 44988.86111111111,
      "last_run_timestamp": "2026-10-24T16:04:55.201Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "GLM 5.3",
      "provider_name": "digitalocean",
      "accuracy": 0.823232,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.025912994949494946,
      "avg_duration_per_task_ms": 162312.898989899,
      "avg_output_tokens_per_task": 9036.257575757576,
      "last_run_timestamp": "2026-10-01T23:50:45.024Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "StreamLake",
      "accuracy": 0.82281125,
      "accuracy_std_dev": 0.014500513393928511,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.024772287878787876,
      "avg_duration_per_task_ms": 511120.71338383836,
      "avg_output_tokens_per_task": 23142.588383838385,
      "last_run_timestamp": "2026-08-29T18:09:03.379Z"
    },
    {
      "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": 0.8225588000000001,
      "accuracy_std_dev": 0.009908448060115158,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.02341788282828283,
      "avg_duration_per_task_ms": 349498.4303030303,
      "avg_output_tokens_per_task": 20884.227272727272,
      "last_run_timestamp": "2026-10-04T00:17:30.035Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-2.5-pro",
      "model_slug": "google/gemini-2.5-pro",
      "model_name": "Google: Gemini 2.5 Pro",
      "provider_name": "Google AI Studio",
      "accuracy": 0.8223905000000001,
      "accuracy_std_dev": 0.0244105,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.2415952904040404,
      "avg_duration_per_task_ms": 181953.0101010101,
      "avg_output_tokens_per_task": 24066.49494949495,
      "last_run_timestamp": "2026-10-01T19:12:31.045Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Io Net",
      "accuracy": 0.8223551976047904,
      "accuracy_std_dev": 0.029631373343505716,
      "run_count": 6,
      "total_questions": 1169,
      "avg_cost_per_task_usd": 0.06724194183062446,
      "avg_duration_per_task_ms": 236148.74337040205,
      "avg_output_tokens_per_task": 20314.131736526946,
      "last_run_timestamp": "2026-10-17T08:26:17.960Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "Novita",
      "accuracy": 0.8223140000000001,
      "accuracy_std_dev": 0.020822165248512558,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.012249293847566574,
      "avg_duration_per_task_ms": 720758.7061524334,
      "avg_output_tokens_per_task": 36235.862258953166,
      "last_run_timestamp": "2026-10-17T08:04:21.262Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Chutes",
      "accuracy": 0.8223139999999999,
      "accuracy_std_dev": 0.02011718801965576,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.03311863039485767,
      "avg_duration_per_task_ms": 262977.6432506887,
      "avg_output_tokens_per_task": 14545.10927456382,
      "last_run_timestamp": "2026-10-24T08:29:59.056Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Crusoe",
      "accuracy": 0.8221660379266751,
      "accuracy_std_dev": 0.02808252331410946,
      "run_count": 4,
      "total_questions": 791,
      "avg_cost_per_task_usd": 0.20924120101137803,
      "avg_duration_per_task_ms": 851157.8482932997,
      "avg_output_tokens_per_task": 47928.662452591656,
      "last_run_timestamp": "2026-08-29T18:53:23.077Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "DigitalOcean",
      "accuracy": 0.8217830000000002,
      "accuracy_std_dev": 0.015536946332318115,
      "run_count": 9,
      "total_questions": 1746,
      "avg_cost_per_task_usd": 0.018000982817869417,
      "avg_duration_per_task_ms": 602038.2628865979,
      "avg_output_tokens_per_task": 20185.35166093929,
      "last_run_timestamp": "2026-10-24T16:01:29.842Z"
    },
    {
      "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": 0.8214768773857721,
      "accuracy_std_dev": 0.011242622925586103,
      "run_count": 9,
      "total_questions": 1729,
      "avg_cost_per_task_usd": 0.024795357432041643,
      "avg_duration_per_task_ms": 445258.44707923656,
      "avg_output_tokens_per_task": 24754.15095430885,
      "last_run_timestamp": "2026-10-24T16:11:33.501Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "DeepInfra",
      "accuracy": 0.8214198161925602,
      "accuracy_std_dev": 0.04954945097072217,
      "run_count": 8,
      "total_questions": 1371,
      "avg_cost_per_task_usd": 0.04645164551422319,
      "avg_duration_per_task_ms": 608936.6484318016,
      "avg_output_tokens_per_task": 28638.388767323122,
      "last_run_timestamp": "2026-09-26T09:56:00.145Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Makora",
      "accuracy": 0.8213506405228759,
      "accuracy_std_dev": 0.02248216924587128,
      "run_count": 3,
      "total_questions": 459,
      "avg_cost_per_task_usd": 0.08758356862745097,
      "avg_duration_per_task_ms": 304618.9760348584,
      "avg_output_tokens_per_task": 34720.03050108933,
      "last_run_timestamp": "2026-08-29T10:57:24.797Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "CoreWeave",
      "accuracy": 0.8212682499999999,
      "accuracy_std_dev": 0.0185418798135689,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.04273052441077441,
      "avg_duration_per_task_ms": 239740.20075757575,
      "avg_output_tokens_per_task": 14238.065656565657,
      "last_run_timestamp": "2026-10-24T08:50:39.623Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.8212120999999999,
      "accuracy_std_dev": 0.028508221310527243,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.028071248484848484,
      "avg_duration_per_task_ms": 579831.6848484848,
      "avg_output_tokens_per_task": 50141.56717171717,
      "last_run_timestamp": "2026-10-30T14:53:57.436Z"
    },
    {
      "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": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "Venice",
      "accuracy": 0.8207548040638606,
      "accuracy_std_dev": 0.12376641907798643,
      "run_count": 7,
      "total_questions": 1378,
      "avg_cost_per_task_usd": 0.1265669557329463,
      "avg_duration_per_task_ms": 747147.1386066764,
      "avg_output_tokens_per_task": 38728.26777939042,
      "last_run_timestamp": "2026-10-24T16:03:25.472Z"
    },
    {
      "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": 0.8207070000000001,
      "accuracy_std_dev": 0.024411000000000016,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.04063828030303031,
      "avg_duration_per_task_ms": 83296.42424242424,
      "avg_output_tokens_per_task": 26969.131313131315,
      "last_run_timestamp": "2026-10-01T19:23:01.327Z"
    },
    {
      "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": 0.8202743410372041,
      "accuracy_std_dev": 0.011484132116496031,
      "run_count": 9,
      "total_questions": 1774,
      "avg_cost_per_task_usd": 0.09063781341600902,
      "avg_duration_per_task_ms": 168767.6234498309,
      "avg_output_tokens_per_task": 22236.227170236754,
      "last_run_timestamp": "2026-09-19T10:14:07.966Z"
    },
    {
      "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": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "SambaNova",
      "accuracy": 0.8192149207920791,
      "accuracy_std_dev": 0.018823182524040232,
      "run_count": 6,
      "total_questions": 909,
      "avg_cost_per_task_usd": 0.11224348184818483,
      "avg_duration_per_task_ms": 401079.6094609461,
      "avg_output_tokens_per_task": 46532.80198019802,
      "last_run_timestamp": "2026-10-17T16:06:46.637Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "DigitalOcean",
      "accuracy": 0.8191437142857143,
      "accuracy_std_dev": 0.03986796895724072,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.04288789971139971,
      "avg_duration_per_task_ms": 992008.3816738817,
      "avg_output_tokens_per_task": 26956.146464646463,
      "last_run_timestamp": "2026-10-17T08:20:47.476Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "CoreWeave",
      "accuracy": 0.8190235000000001,
      "accuracy_std_dev": 0.020635683523934946,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.006401880681818182,
      "avg_duration_per_task_ms": 463424.9640151515,
      "avg_output_tokens_per_task": 18789.763257575756,
      "last_run_timestamp": "2026-10-24T16:31:34.552Z"
    },
    {
      "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": 0.8187429166666668,
      "accuracy_std_dev": 0.012566775670117427,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.31407284974747474,
      "avg_duration_per_task_ms": 289933.9659090909,
      "avg_output_tokens_per_task": 20747.42213804714,
      "last_run_timestamp": "2026-10-24T08:12:02.536Z"
    },
    {
      "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": 0.8183921250000001,
      "accuracy_std_dev": 0.03063364984146315,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.08543528156565658,
      "avg_duration_per_task_ms": 987445.6388888889,
      "avg_output_tokens_per_task": 42416.07765151515,
      "last_run_timestamp": "2026-10-24T16:15:14.401Z"
    },
    {
      "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": "auto-routing",
      "accuracy": 0.8183501,
      "accuracy_std_dev": 0.014134440253862193,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.033925831313131316,
      "avg_duration_per_task_ms": 215707.2409090909,
      "avg_output_tokens_per_task": 13837.719191919192,
      "last_run_timestamp": "2026-10-24T08:16:59.511Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "StreamLake",
      "accuracy": 0.81818175,
      "accuracy_std_dev": 0.008828586845441324,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.06985739772727273,
      "avg_duration_per_task_ms": 843652.7916666666,
      "avg_output_tokens_per_task": 40039.878787878784,
      "last_run_timestamp": "2026-08-30T00:37:59.094Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "DigitalOcean",
      "accuracy": 0.8181817499999999,
      "accuracy_std_dev": 0.021788216475597543,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.02338400883838384,
      "avg_duration_per_task_ms": 497693.29671717173,
      "avg_output_tokens_per_task": 22657.377525252527,
      "last_run_timestamp": "2026-08-29T18:16:26.161Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Mancer 2",
      "accuracy": 0.818143387341772,
      "accuracy_std_dev": 0.008051034079545293,
      "run_count": 8,
      "total_questions": 1580,
      "avg_cost_per_task_usd": 0.03837104746835443,
      "avg_duration_per_task_ms": 216584.66329113924,
      "avg_output_tokens_per_task": 14732.209493670885,
      "last_run_timestamp": "2026-10-24T08:26:33.259Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "Parasail",
      "accuracy": 0.8178564226152875,
      "accuracy_std_dev": 0.005702516062011583,
      "run_count": 8,
      "total_questions": 1583,
      "avg_cost_per_task_usd": 0.025747458622867977,
      "avg_duration_per_task_ms": 332030.03348073276,
      "avg_output_tokens_per_task": 25676.125710675933,
      "last_run_timestamp": "2026-10-24T16:05:36.229Z"
    },
    {
      "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": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "Venice",
      "accuracy": 0.8177608749999999,
      "accuracy_std_dev": 0.007516885898387367,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.027347258838383834,
      "avg_duration_per_task_ms": 177583.7866161616,
      "avg_output_tokens_per_task": 23473.940656565657,
      "last_run_timestamp": "2026-10-24T16:20:59.116Z"
    },
    {
      "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": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "DeepInfra",
      "accuracy": 0.8174469182990923,
      "accuracy_std_dev": 0.06189112957986377,
      "run_count": 22,
      "total_questions": 4186,
      "avg_cost_per_task_usd": 0.0075936266125179175,
      "avg_duration_per_task_ms": 503995.6545628285,
      "avg_output_tokens_per_task": 17382.02126134735,
      "last_run_timestamp": "2026-10-24T16:15:47.415Z"
    },
    {
      "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": "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": 0.8169191785714285,
      "accuracy_std_dev": 0.010832356972296234,
      "run_count": 28,
      "total_questions": 5544,
      "avg_cost_per_task_usd": 0.3328294617604617,
      "avg_duration_per_task_ms": 295647.11093073594,
      "avg_output_tokens_per_task": 21018.825216450216,
      "last_run_timestamp": "2026-10-24T08:56:14.444Z"
    },
    {
      "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": 0.8167087499999999,
      "accuracy_std_dev": 0.004328069568237081,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.010634625000000002,
      "avg_duration_per_task_ms": 790260.4886363636,
      "avg_output_tokens_per_task": 18570.635101010103,
      "last_run_timestamp": "2026-10-24T16:12:04.706Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "auto-routing",
      "accuracy": 0.8167087499999999,
      "accuracy_std_dev": 0.01709464425448801,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.007263619318181818,
      "avg_duration_per_task_ms": 512210.79671717173,
      "avg_output_tokens_per_task": 18196.01452020202,
      "last_run_timestamp": "2026-10-24T16:06:30.906Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Darkbloom",
      "accuracy": 0.8164054780192015,
      "accuracy_std_dev": 0.008737950672783622,
      "run_count": 10,
      "total_questions": 1979,
      "avg_cost_per_task_usd": 0.03418838352703386,
      "avg_duration_per_task_ms": 578940.7988883274,
      "avg_output_tokens_per_task": 15946.312784234462,
      "last_run_timestamp": "2026-10-30T01:10:29.825Z"
    },
    {
      "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": 0.8161616,
      "accuracy_std_dev": 0.060936149095918435,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.09838498181818181,
      "avg_duration_per_task_ms": 774603.698989899,
      "avg_output_tokens_per_task": 39333.56565656565,
      "last_run_timestamp": "2026-10-03T17:58:22.995Z"
    },
    {
      "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": 0.81607725,
      "accuracy_std_dev": 0.011564661524986357,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.018877026515151513,
      "avg_duration_per_task_ms": 531080.7954545454,
      "avg_output_tokens_per_task": 18678.97095959596,
      "last_run_timestamp": "2026-08-30T00:16:38.394Z"
    },
    {
      "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": "PrimeIntellect",
      "accuracy": 0.8157141657142857,
      "accuracy_std_dev": 0.010726999999999987,
      "run_count": 2,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.00801732,
      "avg_duration_per_task_ms": 285410.90857142856,
      "avg_output_tokens_per_task": 31549.22857142857,
      "last_run_timestamp": "2026-10-29T23:13:31.837Z"
    },
    {
      "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-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "StreamLake",
      "accuracy": 0.81481475,
      "accuracy_std_dev": 0.005455009183081147,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.006102087121212121,
      "avg_duration_per_task_ms": 532828.8005050505,
      "avg_output_tokens_per_task": 18519.08712121212,
      "last_run_timestamp": "2026-08-29T19:09:30.030Z"
    },
    {
      "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": 0.8146800399999999,
      "accuracy_std_dev": 0.010528738528351803,
      "run_count": 25,
      "total_questions": 4950,
      "avg_cost_per_task_usd": 0.36333395474747476,
      "avg_duration_per_task_ms": 358588.514949495,
      "avg_output_tokens_per_task": 22927.937777777777,
      "last_run_timestamp": "2026-10-24T08:44:41.412Z"
    },
    {
      "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": "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": 0.813973,
      "accuracy_std_dev": 0.01262599999999997,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.04186610353535354,
      "avg_duration_per_task_ms": 76135.87878787878,
      "avg_output_tokens_per_task": 27786.689393939392,
      "last_run_timestamp": "2026-10-01T19:16:58.595Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Tenstorrent",
      "accuracy": 0.8138046,
      "accuracy_std_dev": 0.01492933679169976,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.010332651515151515,
      "avg_duration_per_task_ms": 579885.2626262626,
      "avg_output_tokens_per_task": 12515.30505050505,
      "last_run_timestamp": "2026-10-24T08:57:37.499Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Morph",
      "accuracy": 0.8136272825651302,
      "accuracy_std_dev": 0.008773473238499465,
      "run_count": 3,
      "total_questions": 499,
      "avg_cost_per_task_usd": 0.0068851422845691386,
      "avg_duration_per_task_ms": 1090136,
      "avg_output_tokens_per_task": 17404.21643286573,
      "last_run_timestamp": "2026-08-15T10:05:32.709Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Tenstorrent",
      "accuracy": 0.8134042249778565,
      "accuracy_std_dev": 0.008662954935945464,
      "run_count": 6,
      "total_questions": 1129,
      "avg_cost_per_task_usd": 0.04263479716563331,
      "avg_duration_per_task_ms": 176521.81842338352,
      "avg_output_tokens_per_task": 16879.856510186004,
      "last_run_timestamp": "2026-10-30T18:58:50.063Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "DigitalOcean",
      "accuracy": 0.8130656097285068,
      "accuracy_std_dev": 0.09667022957112152,
      "run_count": 9,
      "total_questions": 1768,
      "avg_cost_per_task_usd": 0.061109888574660626,
      "avg_duration_per_task_ms": 1704990.4128959277,
      "avg_output_tokens_per_task": 19426.679864253394,
      "last_run_timestamp": "2026-10-24T16:10:43.374Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "inclusionai/ling-3.1-flash-20261002",
      "model_slug": "inclusionai/ling-3.1-flash",
      "model_name": "inclusionAI: Ling 3.1 Flash",
      "provider_name": "Novita",
      "accuracy": 0.813051,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 189,
      "avg_cost_per_task_usd": 0.00766168253968254,
      "avg_duration_per_task_ms": 356938.3703703704,
      "avg_output_tokens_per_task": 33904.37566137566,
      "last_run_timestamp": "2026-10-14T18:00:03.925Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Venice",
      "accuracy": 0.812725848879248,
      "accuracy_std_dev": 0.09609347186495672,
      "run_count": 7,
      "total_questions": 1383,
      "avg_cost_per_task_usd": 0.09072004772234273,
      "avg_duration_per_task_ms": 711729.140274765,
      "avg_output_tokens_per_task": 26485.660882140273,
      "last_run_timestamp": "2026-10-24T16:10:43.300Z"
    },
    {
      "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": 0.8127104999999999,
      "accuracy_std_dev": 0.020528377888912718,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.10173348611111112,
      "avg_duration_per_task_ms": 574816.9229797979,
      "avg_output_tokens_per_task": 37826.392676767675,
      "last_run_timestamp": "2026-08-29T18:26:14.942Z"
    },
    {
      "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": 0.8122894999999999,
      "accuracy_std_dev": 0.02853281357051911,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.026932902777777778,
      "avg_duration_per_task_ms": 507868.57196969696,
      "avg_output_tokens_per_task": 24489.684343434343,
      "last_run_timestamp": "2026-08-29T18:05:36.579Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Near AI",
      "accuracy": 0.8122604367816092,
      "accuracy_std_dev": 0.03327589059946405,
      "run_count": 4,
      "total_questions": 783,
      "avg_cost_per_task_usd": 0.024626680715197958,
      "avg_duration_per_task_ms": 153695.50446998724,
      "avg_output_tokens_per_task": 14302.206896551725,
      "last_run_timestamp": "2026-10-24T08:18:45.292Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "DigitalOcean",
      "accuracy": 0.812079,
      "accuracy_std_dev": 0.1279844797709863,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.15244413194444442,
      "avg_duration_per_task_ms": 1980931.956439394,
      "avg_output_tokens_per_task": 55523.080808080806,
      "last_run_timestamp": "2026-10-24T16:14:24.978Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3-ultra-550b-a55b-20260604",
      "model_slug": "nvidia/nemotron-3-ultra-550b-a55b",
      "model_name": "NVIDIA: Nemotron 3 Ultra",
      "provider_name": "Together",
      "accuracy": 0.812016,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 172,
      "avg_cost_per_task_usd": 0.14135672674418603,
      "avg_duration_per_task_ms": 305120.88372093026,
      "avg_output_tokens_per_task": 39843.74418604651,
      "last_run_timestamp": "2026-08-04T03:11:59.879Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.5-sonnet-20250929",
      "model_slug": "anthropic/claude-sonnet-4.5",
      "model_name": "Anthropic: Claude Sonnet 4.5",
      "provider_name": "Anthropic",
      "accuracy": 0.8120089333333333,
      "accuracy_std_dev": 0.016976505295914775,
      "run_count": 15,
      "total_questions": 2970,
      "avg_cost_per_task_usd": 0.3404598653198653,
      "avg_duration_per_task_ms": 353948.01683501684,
      "avg_output_tokens_per_task": 22525.30909090909,
      "last_run_timestamp": "2026-10-24T08:55:49.941Z"
    },
    {
      "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": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Reka",
      "accuracy": 0.8111930073800738,
      "accuracy_std_dev": 0.03224095604623893,
      "run_count": 7,
      "total_questions": 813,
      "avg_cost_per_task_usd": 0.05198734809348094,
      "avg_duration_per_task_ms": 309460.76998769987,
      "avg_output_tokens_per_task": 20725.47355473555,
      "last_run_timestamp": "2026-10-24T16:02:08.817Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Chutes",
      "accuracy": 0.810966857142857,
      "accuracy_std_dev": 0.041886993615231564,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.14839432323232324,
      "avg_duration_per_task_ms": 1654857.906926407,
      "avg_output_tokens_per_task": 48197.77489177489,
      "last_run_timestamp": "2026-10-24T16:00:53.339Z"
    },
    {
      "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": 0.8108163749999999,
      "accuracy_std_dev": 0.048001397940418106,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.22005288573232326,
      "avg_duration_per_task_ms": 1162217.8257575757,
      "avg_output_tokens_per_task": 50375.75252525252,
      "last_run_timestamp": "2026-10-24T16:06:30.902Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "StreamLake",
      "accuracy": 0.8099512222222223,
      "accuracy_std_dev": 0.024479317441363883,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.11572864197530866,
      "avg_duration_per_task_ms": 766947.6907968575,
      "avg_output_tokens_per_task": 60198.64141414141,
      "last_run_timestamp": "2026-10-24T16:03:25.400Z"
    },
    {
      "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": 0.8097645,
      "accuracy_std_dev": 0.021885500000000002,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.044968992424242425,
      "avg_duration_per_task_ms": 208495.61616161617,
      "avg_output_tokens_per_task": 22388.290404040403,
      "last_run_timestamp": "2026-10-01T19:19:28.473Z"
    },
    {
      "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": 0.8094875524100994,
      "accuracy_std_dev": 0.03366403756751318,
      "run_count": 7,
      "total_questions": 1307,
      "avg_cost_per_task_usd": 0.028843539403213467,
      "avg_duration_per_task_ms": 492278.9755164499,
      "avg_output_tokens_per_task": 17132.139250191278,
      "last_run_timestamp": "2026-10-17T16:03:44.233Z"
    },
    {
      "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": 0.8094485,
      "accuracy_std_dev": 0.017925139427630677,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.22312939835858592,
      "avg_duration_per_task_ms": 1023977.4273989899,
      "avg_output_tokens_per_task": 60760.55808080808,
      "last_run_timestamp": "2026-10-24T16:04:12.636Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Novita",
      "accuracy": 0.8093435,
      "accuracy_std_dev": 0.02388992446095214,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.2666720176767677,
      "avg_duration_per_task_ms": 969823.9217171717,
      "avg_output_tokens_per_task": 60642.0404040404,
      "last_run_timestamp": "2026-10-24T16:00:53.339Z"
    },
    {
      "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": 0.8091328750000001,
      "accuracy_std_dev": 0.024870082244724787,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.009033124368686868,
      "avg_duration_per_task_ms": 569112.759469697,
      "avg_output_tokens_per_task": 19095.31818181818,
      "last_run_timestamp": "2026-10-24T16:00:55.675Z"
    },
    {
      "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": "anthropic/claude-5-fable-20260609",
      "model_slug": "anthropic/claude-fable-5",
      "model_name": "Anthropic: Claude Fable 5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.8089225,
      "accuracy_std_dev": 0.0025254999999999583,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.15927496717171716,
      "avg_duration_per_task_ms": 42973.2904040404,
      "avg_output_tokens_per_task": 2992.459595959596,
      "last_run_timestamp": "2026-08-15T09:15:35.213Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "Crusoe",
      "accuracy": 0.8089225,
      "accuracy_std_dev": 0.07491550000000002,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.023557068181818182,
      "avg_duration_per_task_ms": 701295.6338383838,
      "avg_output_tokens_per_task": 58706.843434343435,
      "last_run_timestamp": "2026-10-24T08:01:41.139Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "ModelRun",
      "accuracy": 0.8083212857142857,
      "accuracy_std_dev": 0.009239648782970742,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.045195821067821065,
      "avg_duration_per_task_ms": 59529.89033189033,
      "avg_output_tokens_per_task": 15222.556277056277,
      "last_run_timestamp": "2026-10-28T04:47:40.540Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.5-sonnet-20250929",
      "model_slug": "anthropic/claude-sonnet-4.5",
      "model_name": "Anthropic: Claude Sonnet 4.5",
      "provider_name": "Azure",
      "accuracy": 0.8082103076923077,
      "accuracy_std_dev": 0.02014208867935159,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.3422723123543123,
      "avg_duration_per_task_ms": 356747.24864024867,
      "avg_output_tokens_per_task": 22654.46153846154,
      "last_run_timestamp": "2026-10-24T08:14:36.266Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Alibaba",
      "accuracy": 0.8082103076923076,
      "accuracy_std_dev": 0.09500960411646216,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.09292749378399379,
      "avg_duration_per_task_ms": 514800.35275835276,
      "avg_output_tokens_per_task": 34280.08508158508,
      "last_run_timestamp": "2026-10-28T03:53:08.289Z"
    },
    {
      "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": 0.8080809999999999,
      "accuracy_std_dev": 0.032382769067823704,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.026336486111111113,
      "avg_duration_per_task_ms": 464443.5934343434,
      "avg_output_tokens_per_task": 21920.722222222223,
      "last_run_timestamp": "2026-08-29T19:19:06.625Z"
    },
    {
      "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": 0.8080704302325582,
      "accuracy_std_dev": 0.10379216096569623,
      "run_count": 13,
      "total_questions": 2408,
      "avg_cost_per_task_usd": 0.041565338870431894,
      "avg_duration_per_task_ms": 155784.30357142858,
      "avg_output_tokens_per_task": 19890.849667774088,
      "last_run_timestamp": "2026-10-28T03:27:17.589Z"
    },
    {
      "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": 0.8068179999999999,
      "accuracy_std_dev": 0.027363478050131,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.046258910353535354,
      "avg_duration_per_task_ms": 196125.59974747474,
      "avg_output_tokens_per_task": 21946.10606060606,
      "last_run_timestamp": "2026-10-01T19:34:33.810Z"
    },
    {
      "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": 0.806397,
      "accuracy_std_dev": 0.003367000000000009,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.04417995707070707,
      "avg_duration_per_task_ms": 210130.46464646465,
      "avg_output_tokens_per_task": 21993.772727272728,
      "last_run_timestamp": "2026-10-01T19:26:25.385Z"
    },
    {
      "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": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "AtlasCloud",
      "accuracy": 0.805345125,
      "accuracy_std_dev": 0.036272944774437245,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.19822186489898988,
      "avg_duration_per_task_ms": 832606.8560606061,
      "avg_output_tokens_per_task": 49537.894570707074,
      "last_run_timestamp": "2026-10-24T16:31:00.679Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.5-sonnet-20250929",
      "model_slug": "anthropic/claude-sonnet-4.5",
      "model_name": "Anthropic: Claude Sonnet 4.5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.8048667272727271,
      "accuracy_std_dev": 0.020423284626269588,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.33759371303948577,
      "avg_duration_per_task_ms": 354437.65059687785,
      "avg_output_tokens_per_task": 22371.59963269054,
      "last_run_timestamp": "2026-10-24T08:45:54.077Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": "Particle AI",
      "accuracy": 0.8047135000000001,
      "accuracy_std_dev": 0.0016835000000000044,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.028701856060606063,
      "avg_duration_per_task_ms": 283393.26767676766,
      "avg_output_tokens_per_task": 57180.469696969696,
      "last_run_timestamp": "2026-10-28T23:02:42.227Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "DekaLLM",
      "accuracy": 0.8043401184631804,
      "accuracy_std_dev": 0.024791902944308233,
      "run_count": 5,
      "total_questions": 937,
      "avg_cost_per_task_usd": 0.03799079615795091,
      "avg_duration_per_task_ms": 223986.25720384205,
      "avg_output_tokens_per_task": 37938.2166488794,
      "last_run_timestamp": "2026-10-24T16:04:55.218Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.8042929999999999,
      "accuracy_std_dev": 0.028405059672530177,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.04403506628787878,
      "avg_duration_per_task_ms": 84875.49242424243,
      "avg_output_tokens_per_task": 27183.9375,
      "last_run_timestamp": "2026-10-01T19:25:58.919Z"
    },
    {
      "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": 0.8042647569692637,
      "accuracy_std_dev": 0.11128768555499524,
      "run_count": 8,
      "total_questions": 1399,
      "avg_cost_per_task_usd": 0.04140569907076483,
      "avg_duration_per_task_ms": 141306.68977841316,
      "avg_output_tokens_per_task": 29614.428877769835,
      "last_run_timestamp": "2026-10-24T16:27:33.561Z"
    },
    {
      "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": 0.8040404,
      "accuracy_std_dev": 0.050668612985160726,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.19756500707070707,
      "avg_duration_per_task_ms": 1564421.486868687,
      "avg_output_tokens_per_task": 47084.214141414144,
      "last_run_timestamp": "2026-10-03T17:00:45.414Z"
    },
    {
      "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": 0.8036315714285714,
      "accuracy_std_dev": 0.0246877462424299,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.0294750468975469,
      "avg_duration_per_task_ms": 317654.27489177487,
      "avg_output_tokens_per_task": 19508.803030303032,
      "last_run_timestamp": "2026-10-17T16:09:25.569Z"
    },
    {
      "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": 0.8028197499999999,
      "accuracy_std_dev": 0.04280068601304774,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.026668318181818182,
      "avg_duration_per_task_ms": 335428.4646464646,
      "avg_output_tokens_per_task": 25945.37941919192,
      "last_run_timestamp": "2026-10-24T16:03:25.477Z"
    },
    {
      "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": 0.8026092499999999,
      "accuracy_std_dev": 0.024724444537491615,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.03972177714646465,
      "avg_duration_per_task_ms": 187520.72222222222,
      "avg_output_tokens_per_task": 21966.539772727272,
      "last_run_timestamp": "2026-10-24T16:05:36.261Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Makora",
      "accuracy": 0.8011755539419088,
      "accuracy_std_dev": 0.08103010676902754,
      "run_count": 8,
      "total_questions": 1446,
      "avg_cost_per_task_usd": 0.018679138312586448,
      "avg_duration_per_task_ms": 300257.92392807745,
      "avg_output_tokens_per_task": 39516.01521438451,
      "last_run_timestamp": "2026-10-24T08:32:39.573Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8,
      "accuracy_std_dev": 0.004689918080308032,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.013948154545454547,
      "avg_duration_per_task_ms": 321232.3868686869,
      "avg_output_tokens_per_task": 14312.276767676767,
      "last_run_timestamp": "2026-10-04T00:00:36.942Z"
    },
    {
      "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": 0.7993280088709678,
      "accuracy_std_dev": 0.03778798369706964,
      "run_count": 10,
      "total_questions": 1240,
      "avg_cost_per_task_usd": 0.006506485483870966,
      "avg_duration_per_task_ms": 293916.95967741933,
      "avg_output_tokens_per_task": 25872.287096774195,
      "last_run_timestamp": "2026-10-24T16:40:33.762Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "DigitalOcean",
      "accuracy": 0.798727888888889,
      "accuracy_std_dev": 0.10517670417967454,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.05109221941638608,
      "avg_duration_per_task_ms": 1276720.367003367,
      "avg_output_tokens_per_task": 27966.93265993266,
      "last_run_timestamp": "2026-10-24T16:17:39.036Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "DeepInfra",
      "accuracy": 0.7986353029239767,
      "accuracy_std_dev": 0.024634052579147848,
      "run_count": 6,
      "total_questions": 855,
      "avg_cost_per_task_usd": 0.06986718011695907,
      "avg_duration_per_task_ms": 1584713.8748538012,
      "avg_output_tokens_per_task": 39897.390643274855,
      "last_run_timestamp": "2026-10-24T16:26:52.183Z"
    },
    {
      "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": "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": 0.7971601298174443,
      "accuracy_std_dev": 0.023536053284742144,
      "run_count": 3,
      "total_questions": 493,
      "avg_cost_per_task_usd": 0.04863367951318459,
      "avg_duration_per_task_ms": 714404.5334685599,
      "avg_output_tokens_per_task": 21075.50304259635,
      "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-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "AtlasCloud",
      "accuracy": 0.79629625,
      "accuracy_std_dev": 0.01565761636353056,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.17376329545454547,
      "avg_duration_per_task_ms": 832021.8737373737,
      "avg_output_tokens_per_task": 55481.57828282828,
      "last_run_timestamp": "2026-08-29T19:16:33.097Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-r1-0528",
      "model_slug": "deepseek/deepseek-r1-0528",
      "model_name": "DeepSeek: R1 0528",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7962405172932331,
      "accuracy_std_dev": 0.005792281993642719,
      "run_count": 7,
      "total_questions": 1330,
      "avg_cost_per_task_usd": 0.050177010526315784,
      "avg_duration_per_task_ms": 1098597.4812030075,
      "avg_output_tokens_per_task": 22847.669172932332,
      "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-v3.1-terminus",
      "model_slug": "deepseek/deepseek-v3.1-terminus",
      "model_name": "DeepSeek: DeepSeek V3.1 Terminus",
      "provider_name": "StreamLake",
      "accuracy": 0.7958755,
      "accuracy_std_dev": 0.00853230152127785,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.014628411616161617,
      "avg_duration_per_task_ms": 291373.6123737374,
      "avg_output_tokens_per_task": 13989.849747474747,
      "last_run_timestamp": "2026-08-29T16:10:33.070Z"
    },
    {
      "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": 0.795735,
      "accuracy_std_dev": 0.01152797383180008,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0539772760942761,
      "avg_duration_per_task_ms": 405752.09764309763,
      "avg_output_tokens_per_task": 23428.74074074074,
      "last_run_timestamp": "2026-08-22T17:18:01.115Z"
    },
    {
      "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": 0.7955037897026831,
      "accuracy_std_dev": 0.07014310843946839,
      "run_count": 7,
      "total_questions": 1379,
      "avg_cost_per_task_usd": 0.19247783321247278,
      "avg_duration_per_task_ms": 1684826.8375634518,
      "avg_output_tokens_per_task": 45136.36040609137,
      "last_run_timestamp": "2026-10-24T16:01:08.930Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Mistral",
      "accuracy": 0.7953587113924051,
      "accuracy_std_dev": 0.017304818120613137,
      "run_count": 4,
      "total_questions": 790,
      "avg_cost_per_task_usd": 0.027054707594936712,
      "avg_duration_per_task_ms": 235650.0012658228,
      "avg_output_tokens_per_task": 53875.47088607595,
      "last_run_timestamp": "2026-10-30T16:25:09.907Z"
    },
    {
      "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/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "auto-routing",
      "accuracy": 0.7944022499999999,
      "accuracy_std_dev": 0.03928997983185408,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.04325068055555555,
      "avg_duration_per_task_ms": 495166.12626262626,
      "avg_output_tokens_per_task": 42549.15993265993,
      "last_run_timestamp": "2026-10-24T16:02:47.095Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.7939496780821917,
      "accuracy_std_dev": 0.012044500000000014,
      "run_count": 2,
      "total_questions": 292,
      "avg_cost_per_task_usd": 0.01254563698630137,
      "avg_duration_per_task_ms": 564085.2363013698,
      "avg_output_tokens_per_task": 36747.568493150684,
      "last_run_timestamp": "2026-10-29T23:46:30.580Z"
    },
    {
      "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": 0.7933882727272729,
      "accuracy_std_dev": 0.01128387116915042,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.0045943420569329665,
      "avg_duration_per_task_ms": 250886.0945821855,
      "avg_output_tokens_per_task": 30080.480257116622,
      "last_run_timestamp": "2026-10-17T16:03:34.733Z"
    },
    {
      "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": 0.79335,
      "accuracy_std_dev": 0.008448975529613064,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.013479858585858586,
      "avg_duration_per_task_ms": 277514.2765151515,
      "avg_output_tokens_per_task": 14053.487373737375,
      "last_run_timestamp": "2026-08-29T19:12:57.906Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Venice",
      "accuracy": 0.7929315464581417,
      "accuracy_std_dev": 0.02021084770060026,
      "run_count": 11,
      "total_questions": 2174,
      "avg_cost_per_task_usd": 0.06395397148114075,
      "avg_duration_per_task_ms": 189643.35372585096,
      "avg_output_tokens_per_task": 20031.46780128795,
      "last_run_timestamp": "2026-10-24T08:53:01.970Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "SambaNova",
      "accuracy": 0.7916666329787234,
      "accuracy_std_dev": 0.015009500000000009,
      "run_count": 2,
      "total_questions": 188,
      "avg_cost_per_task_usd": 0.0925983085106383,
      "avg_duration_per_task_ms": 133806.4414893617,
      "avg_output_tokens_per_task": 38480.17021276596,
      "last_run_timestamp": "2026-08-15T14:54:40.470Z"
    },
    {
      "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": 0.7916664999999999,
      "accuracy_std_dev": 0.0223775773990394,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.006407165404040405,
      "avg_duration_per_task_ms": 353276.53661616164,
      "avg_output_tokens_per_task": 15555.522727272728,
      "last_run_timestamp": "2026-08-29T19:20:50.265Z"
    },
    {
      "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": 0.7914561249999998,
      "accuracy_std_dev": 0.012094065181293467,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.017830758207070706,
      "avg_duration_per_task_ms": 61119.01325757576,
      "avg_output_tokens_per_task": 5789.232323232323,
      "last_run_timestamp": "2026-10-24T16:02:28.570Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "DeepInfra",
      "accuracy": 0.7914427791218638,
      "accuracy_std_dev": 0.014961655425666921,
      "run_count": 12,
      "total_questions": 2232,
      "avg_cost_per_task_usd": 0.004660359318996416,
      "avg_duration_per_task_ms": 1335489.0049283155,
      "avg_output_tokens_per_task": 30608.047491039426,
      "last_run_timestamp": "2026-10-24T16:07:10.479Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.7911994699286443,
      "accuracy_std_dev": 0.008413367458990474,
      "run_count": 5,
      "total_questions": 981,
      "avg_cost_per_task_usd": 0.01172353007135576,
      "avg_duration_per_task_ms": 806571.2609582059,
      "avg_output_tokens_per_task": 39025.64220183486,
      "last_run_timestamp": "2026-10-27T15:15:33.419Z"
    },
    {
      "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": 0.7902097692307692,
      "accuracy_std_dev": 0.03221779293198532,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.10744422882672883,
      "avg_duration_per_task_ms": 389672.32750582753,
      "avg_output_tokens_per_task": 29212.660062160063,
      "last_run_timestamp": "2026-10-27T01:36:12.510Z"
    },
    {
      "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": 0.789983125,
      "accuracy_std_dev": 0.044698004036638775,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.17659961553030304,
      "avg_duration_per_task_ms": 551265.0208333334,
      "avg_output_tokens_per_task": 54985.17803030303,
      "last_run_timestamp": "2026-10-24T16:04:55.367Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Phala",
      "accuracy": 0.7893355839160839,
      "accuracy_std_dev": 0.10239971274711347,
      "run_count": 9,
      "total_questions": 1716,
      "avg_cost_per_task_usd": 0.07681546794871794,
      "avg_duration_per_task_ms": 733179.0168997669,
      "avg_output_tokens_per_task": 27895.818764568765,
      "last_run_timestamp": "2026-10-10T16:29:40.552Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "Together",
      "accuracy": 0.7891638401060072,
      "accuracy_std_dev": 0.014821929173009913,
      "run_count": 12,
      "total_questions": 2264,
      "avg_cost_per_task_usd": 0.008643914752650177,
      "avg_duration_per_task_ms": 377329.0070671378,
      "avg_output_tokens_per_task": 34148.09540636042,
      "last_run_timestamp": "2026-10-24T16:13:05.629Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-r1-0528",
      "model_slug": "deepseek/deepseek-r1-0528",
      "model_name": "DeepSeek: R1 0528",
      "provider_name": "DeepInfra",
      "accuracy": 0.7876364502435629,
      "accuracy_std_dev": 0.019910444432115912,
      "run_count": 8,
      "total_questions": 1437,
      "avg_cost_per_task_usd": 0.04065875156576201,
      "avg_duration_per_task_ms": 745848.8225469729,
      "avg_output_tokens_per_task": 18787.501043841337,
      "last_run_timestamp": "2026-10-17T16:06:15.029Z"
    },
    {
      "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": 0.787037125,
      "accuracy_std_dev": 0.02095938087371319,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.3012645997474747,
      "avg_duration_per_task_ms": 971517.2935606061,
      "avg_output_tokens_per_task": 75923.07512626263,
      "last_run_timestamp": "2026-10-24T16:02:08.814Z"
    },
    {
      "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": 0.787037,
      "accuracy_std_dev": 0.014506239226622442,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.028991700757575758,
      "avg_duration_per_task_ms": 504207.9595959596,
      "avg_output_tokens_per_task": 24268.680555555555,
      "last_run_timestamp": "2026-08-29T18:09:57.600Z"
    },
    {
      "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": 0.7861952499999999,
      "accuracy_std_dev": 0.010513514181162274,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.021063606691919187,
      "avg_duration_per_task_ms": 256646.1351010101,
      "avg_output_tokens_per_task": 12520.557449494949,
      "last_run_timestamp": "2026-10-24T16:18:15.010Z"
    },
    {
      "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": 0.786195,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.2643429696969697,
      "avg_duration_per_task_ms": 180778.88888888888,
      "avg_output_tokens_per_task": 17428.555555555555,
      "last_run_timestamp": "2026-09-01T13:19:59.945Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "Venice",
      "accuracy": 0.7854672791234141,
      "accuracy_std_dev": 0.12208478471357521,
      "run_count": 5,
      "total_questions": 867,
      "avg_cost_per_task_usd": 0.023352209919261825,
      "avg_duration_per_task_ms": 590708.8500576701,
      "avg_output_tokens_per_task": 19635.643598615916,
      "last_run_timestamp": "2026-10-17T08:18:07.707Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.1-terminus",
      "model_slug": "deepseek/deepseek-v3.1-terminus",
      "model_name": "DeepSeek: DeepSeek V3.1 Terminus",
      "provider_name": "DeepInfra",
      "accuracy": 0.7853535,
      "accuracy_std_dev": 0.002525500000000014,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.01384800505050505,
      "avg_duration_per_task_ms": 544619.0151515151,
      "avg_output_tokens_per_task": 14426.747474747475,
      "last_run_timestamp": "2026-08-15T14:56:55.269Z"
    },
    {
      "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": 0.7853107987288136,
      "accuracy_std_dev": 0.021945448081549836,
      "run_count": 4,
      "total_questions": 472,
      "avg_cost_per_task_usd": 0.03487327542372881,
      "avg_duration_per_task_ms": 120089.44067796611,
      "avg_output_tokens_per_task": 36221.95338983051,
      "last_run_timestamp": "2026-08-29T19:44:46.543Z"
    },
    {
      "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": 0.7834595000000001,
      "accuracy_std_dev": 0.025125138984690213,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.0537799494949495,
      "avg_duration_per_task_ms": 455539.1212121212,
      "avg_output_tokens_per_task": 23376.137626262625,
      "last_run_timestamp": "2026-10-17T16:03:34.700Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "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": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "Cloudflare",
      "accuracy": 0.7816675629587804,
      "accuracy_std_dev": 0.02991313237433426,
      "run_count": 9,
      "total_questions": 1771,
      "avg_cost_per_task_usd": 0.011191632411067194,
      "avg_duration_per_task_ms": 582007.9429700734,
      "avg_output_tokens_per_task": 37062.36589497459,
      "last_run_timestamp": "2026-10-24T16:18:35.530Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "Venice",
      "accuracy": 0.7814084898477156,
      "accuracy_std_dev": 0.060666856050972155,
      "run_count": 8,
      "total_questions": 1576,
      "avg_cost_per_task_usd": 0.007266644670050761,
      "avg_duration_per_task_ms": 867252.3737309645,
      "avg_output_tokens_per_task": 18458.295050761422,
      "last_run_timestamp": "2026-10-24T16:21:54.634Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Nebius",
      "accuracy": 0.7805836666666667,
      "accuracy_std_dev": 0.009152126540986119,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.03482366666666667,
      "avg_duration_per_task_ms": 531300.7171717172,
      "avg_output_tokens_per_task": 69410.62962962964,
      "last_run_timestamp": "2026-10-24T08:41:52.312Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "auto-routing",
      "accuracy": 0.780302875,
      "accuracy_std_dev": 0.044334001808537145,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.012861900883838383,
      "avg_duration_per_task_ms": 776646.4172979798,
      "avg_output_tokens_per_task": 38634.71654040404,
      "last_run_timestamp": "2026-10-17T08:18:06.210Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Phala",
      "accuracy": 0.7786195454545455,
      "accuracy_std_dev": 0.04555204390657037,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.21966554178145087,
      "avg_duration_per_task_ms": 867448.9150596878,
      "avg_output_tokens_per_task": 73058.02662993572,
      "last_run_timestamp": "2026-10-17T16:06:46.266Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Venice",
      "accuracy": 0.7785086368421053,
      "accuracy_std_dev": 0.010062970594585871,
      "run_count": 4,
      "total_questions": 760,
      "avg_cost_per_task_usd": 0.002249036842105263,
      "avg_duration_per_task_ms": 780019.0210526315,
      "avg_output_tokens_per_task": 14812.598684210527,
      "last_run_timestamp": "2026-10-24T16:06:32.425Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "Google",
      "accuracy": 0.7784476091370558,
      "accuracy_std_dev": 0.018688411382040342,
      "run_count": 8,
      "total_questions": 1576,
      "avg_cost_per_task_usd": 0.022516517131979694,
      "avg_duration_per_task_ms": 871719.5628172589,
      "avg_output_tokens_per_task": 38655.122461928935,
      "last_run_timestamp": "2026-10-24T16:37:21.981Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.776936,
      "accuracy_std_dev": 0.004208999999999963,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.013052893939393939,
      "avg_duration_per_task_ms": 64108.87373737374,
      "avg_output_tokens_per_task": 10319.823232323231,
      "last_run_timestamp": "2026-10-01T19:12:45.924Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "DekaLLM",
      "accuracy": 0.7769359999999998,
      "accuracy_std_dev": 0.005389920933248147,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.00513745202020202,
      "avg_duration_per_task_ms": 836559.9427609427,
      "avg_output_tokens_per_task": 28379.09006734007,
      "last_run_timestamp": "2026-10-24T16:31:50.394Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "DeepInfra",
      "accuracy": 0.7768158778004073,
      "accuracy_std_dev": 0.02614019988234382,
      "run_count": 6,
      "total_questions": 982,
      "avg_cost_per_task_usd": 0.010268270875763747,
      "avg_duration_per_task_ms": 1330121.382892057,
      "avg_output_tokens_per_task": 39680.64867617108,
      "last_run_timestamp": "2026-10-24T08:12:02.532Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "auto-routing",
      "accuracy": 0.7767955833333333,
      "accuracy_std_dev": 0.0196863031719109,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.005944730218855219,
      "avg_duration_per_task_ms": 663975.4595959596,
      "avg_output_tokens_per_task": 36414.121632996634,
      "last_run_timestamp": "2026-10-24T16:20:08.201Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "DeepInfra",
      "accuracy": 0.7765211365054602,
      "accuracy_std_dev": 0.014331339795043158,
      "run_count": 9,
      "total_questions": 1282,
      "avg_cost_per_task_usd": 0.04696385179407177,
      "avg_duration_per_task_ms": 968660.7449297972,
      "avg_output_tokens_per_task": 23871.57488299532,
      "last_run_timestamp": "2026-10-24T16:01:45.820Z"
    },
    {
      "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": 0.7760940000000002,
      "accuracy_std_dev": 0.012598160421267844,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.013424281144781143,
      "avg_duration_per_task_ms": 64995.4595959596,
      "avg_output_tokens_per_task": 10615.195286195287,
      "last_run_timestamp": "2026-10-02T23:20:38.343Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "Darkbloom",
      "accuracy": 0.7753727142857143,
      "accuracy_std_dev": 0.018106406953769444,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.02145574458874459,
      "avg_duration_per_task_ms": 495034.65367965365,
      "avg_output_tokens_per_task": 28526.727272727272,
      "last_run_timestamp": "2026-10-30T01:54:18.125Z"
    },
    {
      "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": 0.7751615438596491,
      "accuracy_std_dev": 0.0257925287798497,
      "run_count": 8,
      "total_questions": 1083,
      "avg_cost_per_task_usd": 0.013254867959372115,
      "avg_duration_per_task_ms": 53661.341643582644,
      "avg_output_tokens_per_task": 17570.362880886427,
      "last_run_timestamp": "2026-10-24T16:14:06.770Z"
    },
    {
      "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": 0.7746153472485767,
      "accuracy_std_dev": 0.05521980939561002,
      "run_count": 8,
      "total_questions": 1581,
      "avg_cost_per_task_usd": 0.15729624731182795,
      "avg_duration_per_task_ms": 963160.6476913346,
      "avg_output_tokens_per_task": 65807.53447185326,
      "last_run_timestamp": "2026-10-24T16:03:59.070Z"
    },
    {
      "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": "openai/gpt-5.4-nano-20260317",
      "model_slug": "openai/gpt-5.4-nano",
      "model_name": "OpenAI: GPT-5.4 Nano",
      "provider_name": "Azure",
      "accuracy": 0.7734006,
      "accuracy_std_dev": 0.005280843894681953,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.014113778787878787,
      "avg_duration_per_task_ms": 123713.42222222222,
      "avg_output_tokens_per_task": 10542.545454545454,
      "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": "BaseTen",
      "accuracy": 0.7713244166666666,
      "accuracy_std_dev": 0.0099082406313998,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.014599125841750842,
      "avg_duration_per_task_ms": 134726.7032828283,
      "avg_output_tokens_per_task": 29087.89814814815,
      "last_run_timestamp": "2026-10-24T16:26:32.305Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "DeepInfra",
      "accuracy": 0.7710211897522523,
      "accuracy_std_dev": 0.007222635018972211,
      "run_count": 9,
      "total_questions": 1776,
      "avg_cost_per_task_usd": 0.011324844594594596,
      "avg_duration_per_task_ms": 964384.0050675676,
      "avg_output_tokens_per_task": 11774.962274774774,
      "last_run_timestamp": "2026-10-24T16:05:13.521Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-thinking-2507",
      "model_slug": "qwen/qwen3-235b-a22b-thinking-2507",
      "model_name": "Qwen: Qwen3 235B A22B Thinking 2507",
      "provider_name": "auto-routing",
      "accuracy": 0.7704826666666666,
      "accuracy_std_dev": 0.023421454841424542,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0568129494949495,
      "avg_duration_per_task_ms": 614174.2895622896,
      "avg_output_tokens_per_task": 22780.829966329966,
      "last_run_timestamp": "2026-08-22T17:09:41.940Z"
    },
    {
      "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": 0.7703704,
      "accuracy_std_dev": 0.025566915297704586,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.011996698989898991,
      "avg_duration_per_task_ms": 914513.3373737374,
      "avg_output_tokens_per_task": 45163.47373737374,
      "last_run_timestamp": "2026-10-27T02:58:53.157Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "CoreWeave",
      "accuracy": 0.769679306122449,
      "accuracy_std_dev": 0.017686462624278497,
      "run_count": 5,
      "total_questions": 686,
      "avg_cost_per_task_usd": 0.009246491253644315,
      "avg_duration_per_task_ms": 700873.1428571428,
      "avg_output_tokens_per_task": 30693.99271137026,
      "last_run_timestamp": "2026-10-24T16:14:24.978Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Sail Research",
      "accuracy": 0.7691307272727272,
      "accuracy_std_dev": 0.12910889092269848,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.016461994031221304,
      "avg_duration_per_task_ms": 677663.191919192,
      "avg_output_tokens_per_task": 35541.38842975206,
      "last_run_timestamp": "2026-10-24T08:57:59.951Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Parasail",
      "accuracy": 0.768994582345191,
      "accuracy_std_dev": 0.016269118990725185,
      "run_count": 8,
      "total_questions": 1518,
      "avg_cost_per_task_usd": 0.021888522397891963,
      "avg_duration_per_task_ms": 236418.71277997366,
      "avg_output_tokens_per_task": 29098.49209486166,
      "last_run_timestamp": "2026-10-24T16:34:51.218Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "Phala",
      "accuracy": 0.7689713726812815,
      "accuracy_std_dev": 0.07621573205272868,
      "run_count": 6,
      "total_questions": 1186,
      "avg_cost_per_task_usd": 0.018058317875210794,
      "avg_duration_per_task_ms": 1377820.0623946036,
      "avg_output_tokens_per_task": 17446.093591905566,
      "last_run_timestamp": "2026-10-24T16:01:29.840Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-r1-0528",
      "model_slug": "deepseek/deepseek-r1-0528",
      "model_name": "DeepSeek: R1 0528",
      "provider_name": "auto-routing",
      "accuracy": 0.7686387142857142,
      "accuracy_std_dev": 0.0359817938793428,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.05187446248196248,
      "avg_duration_per_task_ms": 873111.9227994228,
      "avg_output_tokens_per_task": 23166.116883116883,
      "last_run_timestamp": "2026-10-17T16:03:24.588Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "AkashML",
      "accuracy": 0.7685711884534847,
      "accuracy_std_dev": 0.015738499677663245,
      "run_count": 17,
      "total_questions": 3343,
      "avg_cost_per_task_usd": 0.006987174394256656,
      "avg_duration_per_task_ms": 444537.46365539933,
      "avg_output_tokens_per_task": 28098.178881244392,
      "last_run_timestamp": "2026-10-30T00:43:52.327Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Xiaomi",
      "accuracy": 0.7685185,
      "accuracy_std_dev": 0.018461144134370444,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.01486243308080808,
      "avg_duration_per_task_ms": 1091396.6527777778,
      "avg_output_tokens_per_task": 52782.86363636364,
      "last_run_timestamp": "2026-10-24T08:54:50.797Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Morph",
      "accuracy": 0.7677873143148243,
      "accuracy_std_dev": 0.10954529486385985,
      "run_count": 10,
      "total_questions": 1963,
      "avg_cost_per_task_usd": 0.15105290779419256,
      "avg_duration_per_task_ms": 869942.4396332145,
      "avg_output_tokens_per_task": 39571.91288843607,
      "last_run_timestamp": "2026-10-24T16:19:11.498Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "DekaLLM",
      "accuracy": 0.7671445564924115,
      "accuracy_std_dev": 0.01261567820104104,
      "run_count": 6,
      "total_questions": 1186,
      "avg_cost_per_task_usd": 0.019929450252951095,
      "avg_duration_per_task_ms": 1212404.510961214,
      "avg_output_tokens_per_task": 60253.559021922425,
      "last_run_timestamp": "2026-10-24T16:08:41.765Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "SambaNova",
      "accuracy": 0.767059119032047,
      "accuracy_std_dev": 0.008480289053888434,
      "run_count": 8,
      "total_questions": 1529,
      "avg_cost_per_task_usd": 0.02492288554610857,
      "avg_duration_per_task_ms": 72092.31916285153,
      "avg_output_tokens_per_task": 26100.194244604318,
      "last_run_timestamp": "2026-10-24T16:10:43.382Z"
    },
    {
      "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": 0.7669488305084746,
      "accuracy_std_dev": 0.016365499999999977,
      "run_count": 2,
      "total_questions": 354,
      "avg_cost_per_task_usd": 0.21194037853107345,
      "avg_duration_per_task_ms": 159764.62146892655,
      "avg_output_tokens_per_task": 13982.706214689266,
      "last_run_timestamp": "2026-09-01T13:33:46.786Z"
    },
    {
      "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": 0.7668349999999999,
      "accuracy_std_dev": 0.004208999999999963,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.057140755050505046,
      "avg_duration_per_task_ms": 101024.92676767676,
      "avg_output_tokens_per_task": 22766.989898989897,
      "last_run_timestamp": "2026-10-01T19:11:42.323Z"
    },
    {
      "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Sail Research",
      "accuracy": 0.7657,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 138,
      "avg_cost_per_task_usd": 0.012955934782608695,
      "avg_duration_per_task_ms": 216746.15942028986,
      "avg_output_tokens_per_task": 32299.463768115944,
      "last_run_timestamp": "2026-10-26T18:00:03.751Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Reka",
      "accuracy": 0.7651274895635675,
      "accuracy_std_dev": 0.012168813200753945,
      "run_count": 16,
      "total_questions": 3162,
      "avg_cost_per_task_usd": 0.01713221790006325,
      "avg_duration_per_task_ms": 104832.72896900696,
      "avg_output_tokens_per_task": 5987.342504743833,
      "last_run_timestamp": "2026-10-24T09:05:10.210Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "DeepInfra",
      "accuracy": 0.7647332578268877,
      "accuracy_std_dev": 0.03506906370263671,
      "run_count": 7,
      "total_questions": 1086,
      "avg_cost_per_task_usd": 0.06476073848987109,
      "avg_duration_per_task_ms": 751502.6731123389,
      "avg_output_tokens_per_task": 52466.43278084714,
      "last_run_timestamp": "2026-08-29T19:07:57.500Z"
    },
    {
      "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": 0.7645579698795182,
      "accuracy_std_dev": 0.052301214506824685,
      "run_count": 3,
      "total_questions": 332,
      "avg_cost_per_task_usd": 0.11985862349397591,
      "avg_duration_per_task_ms": 1173733.4337349397,
      "avg_output_tokens_per_task": 34138.882530120485,
      "last_run_timestamp": "2026-08-22T17:30:52.646Z"
    },
    {
      "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": 0.7636309629005059,
      "accuracy_std_dev": 0.012408560333719428,
      "run_count": 3,
      "total_questions": 593,
      "avg_cost_per_task_usd": 0.02061922428330523,
      "avg_duration_per_task_ms": 132670.10623946038,
      "avg_output_tokens_per_task": 11913.527824620573,
      "last_run_timestamp": "2026-08-22T17:12:08.919Z"
    },
    {
      "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": "Parasail",
      "accuracy": 0.7632574999999999,
      "accuracy_std_dev": 0.011447194448859535,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.020826034090909093,
      "avg_duration_per_task_ms": 1150567.0700757576,
      "avg_output_tokens_per_task": 51976.968434343435,
      "last_run_timestamp": "2026-10-24T16:20:24.814Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Nebius",
      "accuracy": 0.7630004444444446,
      "accuracy_std_dev": 0.01096142269873061,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.0185216531986532,
      "avg_duration_per_task_ms": 131653.27160493826,
      "avg_output_tokens_per_task": 30636.1088664422,
      "last_run_timestamp": "2026-10-24T16:05:36.263Z"
    },
    {
      "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": 0.7629069999999999,
      "accuracy_std_dev": 0.00819966182229484,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.2718667558922559,
      "avg_duration_per_task_ms": 264601.48148148146,
      "avg_output_tokens_per_task": 17934.003367003366,
      "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": "Relace",
      "accuracy": 0.7628913204419889,
      "accuracy_std_dev": 0.046742499999999965,
      "run_count": 2,
      "total_questions": 362,
      "avg_cost_per_task_usd": 0.015507096685082874,
      "avg_duration_per_task_ms": 476532.59392265195,
      "avg_output_tokens_per_task": 53949.68508287293,
      "last_run_timestamp": "2026-10-24T08:33:50.635Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.7627229007633587,
      "accuracy_std_dev": 0.011904999999999999,
      "run_count": 2,
      "total_questions": 262,
      "avg_cost_per_task_usd": 0.003081797709923664,
      "avg_duration_per_task_ms": 72385.01526717558,
      "avg_output_tokens_per_task": 3744.9809160305344,
      "last_run_timestamp": "2026-08-15T14:54:48.654Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Crusoe",
      "accuracy": 0.7626265,
      "accuracy_std_dev": 0.007714766182458135,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.0075155000000000005,
      "avg_duration_per_task_ms": 129316.82575757576,
      "avg_output_tokens_per_task": 29867.227272727272,
      "last_run_timestamp": "2026-10-24T16:12:49.068Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-coder-next-2025-02-03",
      "model_slug": "qwen/qwen3-coder-next",
      "model_name": "Qwen: Qwen3 Coder Next",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.7626260000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.022027883838383836,
      "avg_duration_per_task_ms": 90904.42929292929,
      "avg_output_tokens_per_task": 14522.767676767677,
      "last_run_timestamp": "2026-10-04T03:02:47.748Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "InferenceNet",
      "accuracy": 0.7626259999999999,
      "accuracy_std_dev": 0.01346799999999998,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.01717197474747475,
      "avg_duration_per_task_ms": 622307.4974747475,
      "avg_output_tokens_per_task": 61038.34090909091,
      "last_run_timestamp": "2026-10-03T10:05:09.589Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "GMICloud",
      "accuracy": 0.7624391111111111,
      "accuracy_std_dev": 0.12809223425194174,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.0067465028058361414,
      "avg_duration_per_task_ms": 341598.29517396184,
      "avg_output_tokens_per_task": 21125.109427609426,
      "last_run_timestamp": "2026-10-27T02:22:11.895Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Near AI",
      "accuracy": 0.762346,
      "accuracy_std_dev": 0.0014306672103136615,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.004426390572390572,
      "avg_duration_per_task_ms": 502309.91414141416,
      "avg_output_tokens_per_task": 29382.90404040404,
      "last_run_timestamp": "2026-10-24T16:30:53.943Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7619949375000001,
      "accuracy_std_dev": 0.014121240413242524,
      "run_count": 16,
      "total_questions": 3168,
      "avg_cost_per_task_usd": 0.018249934027777776,
      "avg_duration_per_task_ms": 152465.32638888888,
      "avg_output_tokens_per_task": 30190.491161616163,
      "last_run_timestamp": "2026-10-24T16:26:58.593Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "DeepInfra",
      "accuracy": 0.7616160933333332,
      "accuracy_std_dev": 0.0330626227537687,
      "run_count": 9,
      "total_questions": 1650,
      "avg_cost_per_task_usd": 0.0073882921212121205,
      "avg_duration_per_task_ms": 943448.4084848484,
      "avg_output_tokens_per_task": 19044.426666666666,
      "last_run_timestamp": "2026-10-24T16:00:43.926Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "AtlasCloud",
      "accuracy": 0.76094275,
      "accuracy_std_dev": 0.01636554188493312,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.05637325126262625,
      "avg_duration_per_task_ms": 580765.970959596,
      "avg_output_tokens_per_task": 26018.467171717173,
      "last_run_timestamp": "2026-08-29T18:26:43.144Z"
    },
    {
      "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": "deepseek/deepseek-v3.1-terminus",
      "model_slug": "deepseek/deepseek-v3.1-terminus",
      "model_name": "DeepSeek: DeepSeek V3.1 Terminus",
      "provider_name": "auto-routing",
      "accuracy": 0.760101,
      "accuracy_std_dev": 0.012052078264764132,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.016351531565656566,
      "avg_duration_per_task_ms": 481999.97601010103,
      "avg_output_tokens_per_task": 16387.108585858587,
      "last_run_timestamp": "2026-08-29T18:21:38.666Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Groq",
      "accuracy": 0.7588383750000002,
      "accuracy_std_dev": 0.006285020782334392,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.01747037941919192,
      "avg_duration_per_task_ms": 73518.15088383839,
      "avg_output_tokens_per_task": 28877.368055555555,
      "last_run_timestamp": "2026-10-24T16:15:45.667Z"
    },
    {
      "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": 0.7582697022900764,
      "accuracy_std_dev": 0.01511744199591981,
      "run_count": 8,
      "total_questions": 1572,
      "avg_cost_per_task_usd": 0.016854382951653944,
      "avg_duration_per_task_ms": 249305.42239185752,
      "avg_output_tokens_per_task": 28356.20165394402,
      "last_run_timestamp": "2026-10-24T16:36:35.542Z"
    },
    {
      "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": 0.7572390000000001,
      "accuracy_std_dev": 0.007014128456194694,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.022558349494949493,
      "avg_duration_per_task_ms": 96018.14343434344,
      "avg_output_tokens_per_task": 14878.855555555556,
      "last_run_timestamp": "2026-10-03T16:56:57.458Z"
    },
    {
      "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": 0.7570765817490493,
      "accuracy_std_dev": 0.03212040139824491,
      "run_count": 8,
      "total_questions": 1578,
      "avg_cost_per_task_usd": 0.004716984157160963,
      "avg_duration_per_task_ms": 687613.1001267427,
      "avg_output_tokens_per_task": 28235.273764258556,
      "last_run_timestamp": "2026-10-24T16:26:55.134Z"
    },
    {
      "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": 0.7570145555555555,
      "accuracy_std_dev": 0.13531913577753976,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.029267514029180697,
      "avg_duration_per_task_ms": 135665.3411896745,
      "avg_output_tokens_per_task": 22416.14478114478,
      "last_run_timestamp": "2026-10-28T03:24:15.005Z"
    },
    {
      "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": 0.7564585415688118,
      "accuracy_std_dev": 0.10021722464025354,
      "run_count": 11,
      "total_questions": 2129,
      "avg_cost_per_task_usd": 0.006281726632221701,
      "avg_duration_per_task_ms": 548745.0742132457,
      "avg_output_tokens_per_task": 26811.618130577735,
      "last_run_timestamp": "2026-10-27T02:37:18.076Z"
    },
    {
      "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": 0.755892,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.018171858585858586,
      "avg_duration_per_task_ms": 285760.0353535354,
      "avg_output_tokens_per_task": 24027.954545454544,
      "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-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": 0.7544054846967168,
      "accuracy_std_dev": 0.03260817390833722,
      "run_count": 11,
      "total_questions": 1797,
      "avg_cost_per_task_usd": 0.045709284919309964,
      "avg_duration_per_task_ms": 262995.0717863105,
      "avg_output_tokens_per_task": 17224.77796327212,
      "last_run_timestamp": "2026-10-24T08:41:25.770Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "auto-routing",
      "accuracy": 0.752104375,
      "accuracy_std_dev": 0.041320927025350916,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.05484566098484848,
      "avg_duration_per_task_ms": 1035357.8099747475,
      "avg_output_tokens_per_task": 28498.281565656565,
      "last_run_timestamp": "2026-10-24T16:02:06.908Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Phala",
      "accuracy": 0.7516834000000001,
      "accuracy_std_dev": 0.0580270664952141,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.01393037222222222,
      "avg_duration_per_task_ms": 97444.4191919192,
      "avg_output_tokens_per_task": 5974.968181818182,
      "last_run_timestamp": "2026-10-24T08:48:00.943Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "Venice",
      "accuracy": 0.7515440142932769,
      "accuracy_std_dev": 0.14128088782365433,
      "run_count": 10,
      "total_questions": 1889,
      "avg_cost_per_task_usd": 0.1006153043938592,
      "avg_duration_per_task_ms": 662503.1037586024,
      "avg_output_tokens_per_task": 29255.94706193753,
      "last_run_timestamp": "2026-10-24T16:11:49.038Z"
    },
    {
      "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": 0.751402888888889,
      "accuracy_std_dev": 0.03418981154101141,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.31025038888888895,
      "avg_duration_per_task_ms": 925905.71661055,
      "avg_output_tokens_per_task": 79989.34511784512,
      "last_run_timestamp": "2026-10-27T01:41:24.438Z"
    },
    {
      "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": 0.7513468000000002,
      "accuracy_std_dev": 0.014443665156739118,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.01692796111111111,
      "avg_duration_per_task_ms": 1325396.3136363637,
      "avg_output_tokens_per_task": 54802.47171717172,
      "last_run_timestamp": "2026-10-30T15:58:48.139Z"
    },
    {
      "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": "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": 0.7502806666666667,
      "accuracy_std_dev": 0.02200747420511694,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.07882791077441077,
      "avg_duration_per_task_ms": 825090.7154882154,
      "avg_output_tokens_per_task": 22562.79292929293,
      "last_run_timestamp": "2026-08-22T17:38:59.682Z"
    },
    {
      "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": 0.75,
      "accuracy_std_dev": 0.00985231011827516,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.013619438552188553,
      "avg_duration_per_task_ms": 95832.97558922559,
      "avg_output_tokens_per_task": 14970.221380471381,
      "last_run_timestamp": "2026-10-03T16:54:39.591Z"
    },
    {
      "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": 0.7495791249999999,
      "accuracy_std_dev": 0.028467424081911153,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.18216496464646464,
      "avg_duration_per_task_ms": 1070460.8421717172,
      "avg_output_tokens_per_task": 82670.29924242424,
      "last_run_timestamp": "2026-10-24T16:01:48.224Z"
    },
    {
      "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": "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": "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": 0.7479075236336781,
      "accuracy_std_dev": 0.023052021760906493,
      "run_count": 7,
      "total_questions": 1354,
      "avg_cost_per_task_usd": 0.01579615952732644,
      "avg_duration_per_task_ms": 361173.26514032495,
      "avg_output_tokens_per_task": 51714.83751846381,
      "last_run_timestamp": "2026-10-27T15:23:44.931Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "Makora",
      "accuracy": 0.7468014,
      "accuracy_std_dev": 0.009196272649285712,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.015463989898989897,
      "avg_duration_per_task_ms": 342566.1424242424,
      "avg_output_tokens_per_task": 63071.326262626266,
      "last_run_timestamp": "2026-10-24T16:16:25.413Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "Io Net",
      "accuracy": 0.7465941264367815,
      "accuracy_std_dev": 0.10532480903970348,
      "run_count": 8,
      "total_questions": 1566,
      "avg_cost_per_task_usd": 0.04816395977011494,
      "avg_duration_per_task_ms": 468882.4540229885,
      "avg_output_tokens_per_task": 36470.09642401022,
      "last_run_timestamp": "2026-09-19T17:18:00.087Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "DeepInfra",
      "accuracy": 0.7465043071718538,
      "accuracy_std_dev": 0.027768881506803414,
      "run_count": 6,
      "total_questions": 739,
      "avg_cost_per_task_usd": 0.053211952638700946,
      "avg_duration_per_task_ms": 424838.5182679296,
      "avg_output_tokens_per_task": 15554.012178619756,
      "last_run_timestamp": "2026-09-19T08:53:38.031Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-coder-next-2025-02-03",
      "model_slug": "qwen/qwen3-coder-next",
      "model_name": "Qwen: Qwen3 Coder Next",
      "provider_name": "auto-routing",
      "accuracy": 0.7464647999999999,
      "accuracy_std_dev": 0.010410614859843804,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.014634281818181818,
      "avg_duration_per_task_ms": 108043.59494949495,
      "avg_output_tokens_per_task": 14531.531313131312,
      "last_run_timestamp": "2026-10-03T16:43:33.673Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-9b-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": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Novita",
      "accuracy": 0.7459349573170732,
      "accuracy_std_dev": 0.02985061357091343,
      "run_count": 5,
      "total_questions": 984,
      "avg_cost_per_task_usd": 0.014618331300813008,
      "avg_duration_per_task_ms": 1032360.968495935,
      "avg_output_tokens_per_task": 51870.760162601626,
      "last_run_timestamp": "2026-10-24T08:20:32.397Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.7457911666666666,
      "accuracy_std_dev": 0.008528929132012367,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.006685603535353536,
      "avg_duration_per_task_ms": 166915.6186868687,
      "avg_output_tokens_per_task": 7768.992424242424,
      "last_run_timestamp": "2026-10-03T16:58:32.590Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Nebius",
      "accuracy": 0.7443482857142857,
      "accuracy_std_dev": 0.006065167898142083,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.005301684704184704,
      "avg_duration_per_task_ms": 153567.91919191918,
      "avg_output_tokens_per_task": 8565.327561327562,
      "last_run_timestamp": "2026-10-17T16:06:15.944Z"
    },
    {
      "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": 0.7438972500000001,
      "accuracy_std_dev": 0.045129307669046964,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.3025249255050505,
      "avg_duration_per_task_ms": 1004822.8832070708,
      "avg_output_tokens_per_task": 88113.10290404041,
      "last_run_timestamp": "2026-10-24T16:08:29.474Z"
    },
    {
      "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": 0.7430977999999999,
      "accuracy_std_dev": 0.008356521127837845,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.0232193797979798,
      "avg_duration_per_task_ms": 98352.36666666667,
      "avg_output_tokens_per_task": 15371.260606060607,
      "last_run_timestamp": "2026-10-03T16:58:08.462Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "SiliconFlow",
      "accuracy": 0.74242425,
      "accuracy_std_dev": 0.013572888443050747,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.11500917424242424,
      "avg_duration_per_task_ms": 764443.2228535353,
      "avg_output_tokens_per_task": 44885.719696969696,
      "last_run_timestamp": "2026-10-24T16:39:40.802Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.742214,
      "accuracy_std_dev": 0.015132597149861616,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.01671199116161616,
      "avg_duration_per_task_ms": 766495.9797979798,
      "avg_output_tokens_per_task": 59349.530303030304,
      "last_run_timestamp": "2026-10-24T08:26:00.562Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-coder-next-2025-02-03",
      "model_slug": "qwen/qwen3-coder-next",
      "model_name": "Qwen: Qwen3 Coder Next",
      "provider_name": "Parasail",
      "accuracy": 0.7420804924406049,
      "accuracy_std_dev": 0.0038769124828915004,
      "run_count": 5,
      "total_questions": 926,
      "avg_cost_per_task_usd": 0.008837730021598272,
      "avg_duration_per_task_ms": 138969.186825054,
      "avg_output_tokens_per_task": 11241.735421166306,
      "last_run_timestamp": "2026-10-03T17:03:40.665Z"
    },
    {
      "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": 0.7401795555555556,
      "accuracy_std_dev": 0.09837566628221198,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.13554839506172842,
      "avg_duration_per_task_ms": 959035.4539842873,
      "avg_output_tokens_per_task": 71622.32940516275,
      "last_run_timestamp": "2026-10-27T01:16:10.839Z"
    },
    {
      "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": 0.7384959999999999,
      "accuracy_std_dev": 0.008170584087648423,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.06071549915824916,
      "avg_duration_per_task_ms": 125902.70959595959,
      "avg_output_tokens_per_task": 24199.07239057239,
      "last_run_timestamp": "2026-10-01T19:15:31.968Z"
    },
    {
      "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": 0.7373190000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 184,
      "avg_cost_per_task_usd": 0.008365875,
      "avg_duration_per_task_ms": 223563.88043478262,
      "avg_output_tokens_per_task": 10361.597826086956,
      "last_run_timestamp": "2026-08-08T16:46:11.805Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "Novita",
      "accuracy": 0.737284008888889,
      "accuracy_std_dev": 0.05166831288604067,
      "run_count": 7,
      "total_questions": 1350,
      "avg_cost_per_task_usd": 0.017425666666666666,
      "avg_duration_per_task_ms": 990522.262962963,
      "avg_output_tokens_per_task": 44050.16814814815,
      "last_run_timestamp": "2026-10-24T16:13:05.630Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Alibaba",
      "accuracy": 0.736027,
      "accuracy_std_dev": 0.010289969057290673,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.004868150505050505,
      "avg_duration_per_task_ms": 107346.51515151515,
      "avg_output_tokens_per_task": 7938.19797979798,
      "last_run_timestamp": "2026-10-03T16:59:36.480Z"
    },
    {
      "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": 0.7360049885496184,
      "accuracy_std_dev": 0.012867937149534676,
      "run_count": 8,
      "total_questions": 1572,
      "avg_cost_per_task_usd": 0.006530778625954198,
      "avg_duration_per_task_ms": 339780.16603053437,
      "avg_output_tokens_per_task": 25930.9058524173,
      "last_run_timestamp": "2026-10-24T16:16:05.012Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Parasail",
      "accuracy": 0.7353067940797942,
      "accuracy_std_dev": 0.007845382539271094,
      "run_count": 8,
      "total_questions": 1554,
      "avg_cost_per_task_usd": 0.006840141570141572,
      "avg_duration_per_task_ms": 249346.28893178893,
      "avg_output_tokens_per_task": 8460.478764478765,
      "last_run_timestamp": "2026-10-24T16:02:47.278Z"
    },
    {
      "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": "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": 0.7343291640304627,
      "accuracy_std_dev": 0.0944960545028733,
      "run_count": 10,
      "total_questions": 1707,
      "avg_cost_per_task_usd": 0.0041295178676039835,
      "avg_duration_per_task_ms": 550309.4973637961,
      "avg_output_tokens_per_task": 25314.345049794963,
      "last_run_timestamp": "2026-10-27T15:54:59.962Z"
    },
    {
      "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": 0.7325637142857143,
      "accuracy_std_dev": 0.0065623172238881,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.006008777777777779,
      "avg_duration_per_task_ms": 423083.5331890332,
      "avg_output_tokens_per_task": 7933.917027417027,
      "last_run_timestamp": "2026-10-24T16:07:46.992Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Google",
      "accuracy": 0.732043837671728,
      "accuracy_std_dev": 0.018715369477343178,
      "run_count": 14,
      "total_questions": 2766,
      "avg_cost_per_task_usd": 0.006619636659436008,
      "avg_duration_per_task_ms": 127243.25451916124,
      "avg_output_tokens_per_task": 6931.6146059291395,
      "last_run_timestamp": "2026-10-17T16:03:44.090Z"
    },
    {
      "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": 0.7305929722222222,
      "accuracy_std_dev": 0.012805628963003616,
      "run_count": 36,
      "total_questions": 7128,
      "avg_cost_per_task_usd": 0.20431471787317615,
      "avg_duration_per_task_ms": 243944.4814814815,
      "avg_output_tokens_per_task": 38124.83193041526,
      "last_run_timestamp": "2026-10-24T08:29:16.808Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "Z.AI",
      "accuracy": 0.7294042813688214,
      "accuracy_std_dev": 0.0505548117787454,
      "run_count": 8,
      "total_questions": 1578,
      "avg_cost_per_task_usd": 0.3084836787072243,
      "avg_duration_per_task_ms": 1347611.5551330799,
      "avg_output_tokens_per_task": 97858.03675538657,
      "last_run_timestamp": "2026-10-24T16:28:00.946Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "auto-routing",
      "accuracy": 0.72895625,
      "accuracy_std_dev": 0.024279686950978163,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.006164965909090908,
      "avg_duration_per_task_ms": 374409.91414141416,
      "avg_output_tokens_per_task": 19987.00505050505,
      "last_run_timestamp": "2026-10-24T16:02:47.367Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.5-haiku-20251001",
      "model_slug": "anthropic/claude-haiku-4.5",
      "model_name": "Anthropic: Claude Haiku 4.5",
      "provider_name": "auto-routing",
      "accuracy": 0.7286756666666667,
      "accuracy_std_dev": 0.008877283080362397,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.19010534595959594,
      "avg_duration_per_task_ms": 243777.6670875421,
      "avg_output_tokens_per_task": 37829.979797979795,
      "last_run_timestamp": "2026-10-24T08:26:35.486Z"
    },
    {
      "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": 0.7281145,
      "accuracy_std_dev": 0.004208499999999948,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.060785368686868686,
      "avg_duration_per_task_ms": 121296.46212121213,
      "avg_output_tokens_per_task": 24226.65404040404,
      "last_run_timestamp": "2026-10-01T19:12:41.089Z"
    },
    {
      "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": 0.7273304235890933,
      "accuracy_std_dev": 0.048401510033262374,
      "run_count": 8,
      "total_questions": 1577,
      "avg_cost_per_task_usd": 0.3023879137603043,
      "avg_duration_per_task_ms": 1316651.7057704502,
      "avg_output_tokens_per_task": 95464.91819911223,
      "last_run_timestamp": "2026-10-24T16:02:28.413Z"
    },
    {
      "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": 0.72727275,
      "accuracy_std_dev": 0.008934765069892239,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.18649388005050505,
      "avg_duration_per_task_ms": 302584.60984848486,
      "avg_output_tokens_per_task": 37129.88047138047,
      "last_run_timestamp": "2026-10-24T08:28:43.171Z"
    },
    {
      "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": 0.7264923098256736,
      "accuracy_std_dev": 0.012443953888007772,
      "run_count": 13,
      "total_questions": 2524,
      "avg_cost_per_task_usd": 0.18690665134706813,
      "avg_duration_per_task_ms": 345005.80625990493,
      "avg_output_tokens_per_task": 37190.982171156895,
      "last_run_timestamp": "2026-10-24T08:44:41.412Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Tenstorrent",
      "accuracy": 0.7264309999999999,
      "accuracy_std_dev": 0.021044000000000007,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.1303771111111111,
      "avg_duration_per_task_ms": 318087.7601010101,
      "avg_output_tokens_per_task": 43267.416666666664,
      "last_run_timestamp": "2026-10-04T05:57:29.940Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.725799875,
      "accuracy_std_dev": 0.01577998058013301,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.005505772095959596,
      "avg_duration_per_task_ms": 302649.43876262626,
      "avg_output_tokens_per_task": 8026.409090909091,
      "last_run_timestamp": "2026-10-24T16:01:05.272Z"
    },
    {
      "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": 0.7252524,
      "accuracy_std_dev": 0.016713400846027716,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.004543460606060606,
      "avg_duration_per_task_ms": 167068.73636363636,
      "avg_output_tokens_per_task": 7707.889898989899,
      "last_run_timestamp": "2026-10-03T16:57:10.991Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Krea",
      "accuracy": 0.7234044255319149,
      "accuracy_std_dev": 0.1917994784299478,
      "run_count": 3,
      "total_questions": 517,
      "avg_cost_per_task_usd": 0.033565591876208896,
      "avg_duration_per_task_ms": 242772.43907156674,
      "avg_output_tokens_per_task": 37102.30560928433,
      "last_run_timestamp": "2026-09-29T20:57:20.673Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "DeepInfra",
      "accuracy": 0.7226898888888889,
      "accuracy_std_dev": 0.010780440739747587,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.004412828282828282,
      "avg_duration_per_task_ms": 687415.5527497195,
      "avg_output_tokens_per_task": 7898.555555555556,
      "last_run_timestamp": "2026-10-24T16:01:45.799Z"
    },
    {
      "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": 0.7219817142857143,
      "accuracy_std_dev": 0.03351855020183082,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.007928627705627704,
      "avg_duration_per_task_ms": 786446.1406926407,
      "avg_output_tokens_per_task": 27309.676767676767,
      "last_run_timestamp": "2026-10-24T16:14:06.772Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "ModelRun",
      "accuracy": 0.7219408075949366,
      "accuracy_std_dev": 0.025598999999999983,
      "run_count": 2,
      "total_questions": 395,
      "avg_cost_per_task_usd": 0.27939689620253166,
      "avg_duration_per_task_ms": 450249.90379746835,
      "avg_output_tokens_per_task": 45941.00253164557,
      "last_run_timestamp": "2026-08-15T08:16:15.061Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "CoreWeave",
      "accuracy": 0.721866038034188,
      "accuracy_std_dev": 0.03876582276336764,
      "run_count": 12,
      "total_questions": 2340,
      "avg_cost_per_task_usd": 0.010606732051282052,
      "avg_duration_per_task_ms": 258545.00683760684,
      "avg_output_tokens_per_task": 48059.278205128205,
      "last_run_timestamp": "2026-10-24T08:11:11.359Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.5-haiku-20251001",
      "model_slug": "anthropic/claude-haiku-4.5",
      "model_name": "Anthropic: Claude Haiku 4.5",
      "provider_name": "Google",
      "accuracy": 0.721557710526316,
      "accuracy_std_dev": 0.013393277763800398,
      "run_count": 38,
      "total_questions": 7524,
      "avg_cost_per_task_usd": 0.19713929492291335,
      "avg_duration_per_task_ms": 268866.1802232855,
      "avg_output_tokens_per_task": 36797.53282828283,
      "last_run_timestamp": "2026-10-24T09:06:24.922Z"
    },
    {
      "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": 0.7204335,
      "accuracy_std_dev": 0.06469765028190745,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.12717121843434342,
      "avg_duration_per_task_ms": 642265.8712121212,
      "avg_output_tokens_per_task": 38206.32133838384,
      "last_run_timestamp": "2026-10-24T16:25:38.398Z"
    },
    {
      "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": 0.7203351056305258,
      "accuracy_std_dev": 0.11641213306920471,
      "run_count": 12,
      "total_questions": 2149,
      "avg_cost_per_task_usd": 0.033083575151233134,
      "avg_duration_per_task_ms": 1068472.419730107,
      "avg_output_tokens_per_task": 34739.315960912056,
      "last_run_timestamp": "2026-10-24T16:14:06.772Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Cohere",
      "accuracy": 0.719697,
      "accuracy_std_dev": 0.004209000000000018,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.021745989898989897,
      "avg_duration_per_task_ms": 204822.55555555556,
      "avg_output_tokens_per_task": 77279.37878787878,
      "last_run_timestamp": "2026-10-24T08:14:03.073Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "AkashML",
      "accuracy": 0.7194163333333333,
      "accuracy_std_dev": 0.011195005920300179,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.014668272727272728,
      "avg_duration_per_task_ms": 1270908.3501683502,
      "avg_output_tokens_per_task": 60914.632996632994,
      "last_run_timestamp": "2026-10-30T00:57:12.245Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Makora",
      "accuracy": 0.7174523333333331,
      "accuracy_std_dev": 0.013163488932987655,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.018911546296296298,
      "avg_duration_per_task_ms": 531218.9570707071,
      "avg_output_tokens_per_task": 67191.76851851853,
      "last_run_timestamp": "2026-10-24T09:17:16.298Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-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": 0.7164982000000001,
      "accuracy_std_dev": 0.007939319804618048,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.007827072727272727,
      "avg_duration_per_task_ms": 80947.86969696969,
      "avg_output_tokens_per_task": 9940.124242424243,
      "last_run_timestamp": "2026-10-03T16:40:12.993Z"
    },
    {
      "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": 0.7138721298568508,
      "accuracy_std_dev": 0.014796710480373683,
      "run_count": 5,
      "total_questions": 978,
      "avg_cost_per_task_usd": 0.011163598159509203,
      "avg_duration_per_task_ms": 105524.10429447853,
      "avg_output_tokens_per_task": 10243.394683026585,
      "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-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": 0.713468,
      "accuracy_std_dev": 0.009367361122536085,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.010825097979797981,
      "avg_duration_per_task_ms": 93154.30303030302,
      "avg_output_tokens_per_task": 10165.817171717172,
      "last_run_timestamp": "2026-10-03T16:59:33.949Z"
    },
    {
      "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": "Novita",
      "accuracy": 0.7124580000000001,
      "accuracy_std_dev": 0.005150598062361318,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.015565995959595959,
      "avg_duration_per_task_ms": 101217.36060606061,
      "avg_output_tokens_per_task": 10296.08787878788,
      "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-vl-235b-a22b-instruct",
      "model_slug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_name": "Qwen: Qwen3 VL 235B A22B Instruct",
      "provider_name": "Novita",
      "accuracy": 0.7117846000000001,
      "accuracy_std_dev": 0.007633552268767163,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.00877359797979798,
      "avg_duration_per_task_ms": 133676.15555555557,
      "avg_output_tokens_per_task": 5686.586868686869,
      "last_run_timestamp": "2026-10-03T16:45:33.688Z"
    },
    {
      "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "DeepInfra",
      "accuracy": 0.7106959574742269,
      "accuracy_std_dev": 0.04945132031209455,
      "run_count": 16,
      "total_questions": 3104,
      "avg_cost_per_task_usd": 0.007755754832474224,
      "avg_duration_per_task_ms": 541486.2728737113,
      "avg_output_tokens_per_task": 25549.894974226805,
      "last_run_timestamp": "2026-10-24T16:12:04.707Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-thinking-2507",
      "model_slug": "qwen/qwen3-235b-a22b-thinking-2507",
      "model_name": "Qwen: Qwen3 235B A22B Thinking 2507",
      "provider_name": "Novita",
      "accuracy": 0.7098763333333333,
      "accuracy_std_dev": 0.0209969340354464,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0663709898989899,
      "avg_duration_per_task_ms": 1050299.8484848484,
      "avg_output_tokens_per_task": 22129.309764309764,
      "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-vl-235b-a22b-instruct",
      "model_slug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_name": "Qwen: Qwen3 VL 235B A22B Instruct",
      "provider_name": "Alibaba",
      "accuracy": 0.7077440000000002,
      "accuracy_std_dev": 0.006867349678005355,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.0059367939393939395,
      "avg_duration_per_task_ms": 109836.06363636363,
      "avg_output_tokens_per_task": 5507.80101010101,
      "last_run_timestamp": "2026-10-03T17:02:24.058Z"
    },
    {
      "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": "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": 0.70665,
      "accuracy_std_dev": 0.05616899659064598,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.006621275252525252,
      "avg_duration_per_task_ms": 125753.64267676767,
      "avg_output_tokens_per_task": 7346.292929292929,
      "last_run_timestamp": "2026-08-29T18:01:46.940Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "auto-routing",
      "accuracy": 0.7048612500000001,
      "accuracy_std_dev": 0.04887384717246537,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.016058294823232323,
      "avg_duration_per_task_ms": 999836.755050505,
      "avg_output_tokens_per_task": 49417.84027777778,
      "last_run_timestamp": "2026-10-24T16:05:13.521Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5-nano-2025-08-07",
      "model_slug": "openai/gpt-5-nano",
      "model_name": "OpenAI: GPT-5 Nano",
      "provider_name": "auto-routing",
      "accuracy": 0.7045455,
      "accuracy_std_dev": 0.004208500000000004,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.01336774494949495,
      "avg_duration_per_task_ms": 231791.09848484848,
      "avg_output_tokens_per_task": 33323.15909090909,
      "last_run_timestamp": "2026-10-01T19:27:22.032Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_slug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_name": "Qwen: Qwen3 VL 235B A22B Instruct",
      "provider_name": "Parasail",
      "accuracy": 0.7037038000000001,
      "accuracy_std_dev": 0.015684797906253056,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.01112428686868687,
      "avg_duration_per_task_ms": 196538.2898989899,
      "avg_output_tokens_per_task": 5791.187878787879,
      "last_run_timestamp": "2026-10-03T16:59:57.847Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Venice",
      "accuracy": 0.7032012317327765,
      "accuracy_std_dev": 0.20342432055346774,
      "run_count": 5,
      "total_questions": 958,
      "avg_cost_per_task_usd": 0.08802485803757831,
      "avg_duration_per_task_ms": 1349048.8903966597,
      "avg_output_tokens_per_task": 33413.259916492694,
      "last_run_timestamp": "2026-10-24T16:13:05.627Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "auto-routing",
      "accuracy": 0.7030114672897196,
      "accuracy_std_dev": 0.02694021764221664,
      "run_count": 10,
      "total_questions": 1926,
      "avg_cost_per_task_usd": 0.009145764278296989,
      "avg_duration_per_task_ms": 363422.58099688473,
      "avg_output_tokens_per_task": 44741.07476635514,
      "last_run_timestamp": "2026-10-24T08:01:41.230Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "GMICloud",
      "accuracy": 0.7028622499999999,
      "accuracy_std_dev": 0.08200476941122087,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.0070032777777777775,
      "avg_duration_per_task_ms": 172461.12373737374,
      "avg_output_tokens_per_task": 7212.60101010101,
      "last_run_timestamp": "2026-10-03T16:56:36.098Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-next-80b-a3b-instruct-2509",
      "model_slug": "qwen/qwen3-next-80b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Next 80B A3B Instruct",
      "provider_name": "Google",
      "accuracy": 0.7006736,
      "accuracy_std_dev": 0.029399119487494893,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.010841593939393938,
      "avg_duration_per_task_ms": 65671.14141414141,
      "avg_output_tokens_per_task": 8990.147474747475,
      "last_run_timestamp": "2026-10-03T16:36:16.895Z"
    },
    {
      "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": 0.7006460556023589,
      "accuracy_std_dev": 0.008635764504856131,
      "run_count": 6,
      "total_questions": 1187,
      "avg_cost_per_task_usd": 0.010373346251053074,
      "avg_duration_per_task_ms": 142035.05391743893,
      "avg_output_tokens_per_task": 9363.947767481044,
      "last_run_timestamp": "2026-10-08T22:40:37.311Z"
    },
    {
      "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": 0.6936026666666666,
      "accuracy_std_dev": 0.010286382189844758,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.005539732323232322,
      "avg_duration_per_task_ms": 380174.12373737374,
      "avg_output_tokens_per_task": 6110.575757575758,
      "last_run_timestamp": "2026-10-08T23:09:15.796Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.6-sonnet-20260217",
      "model_slug": "anthropic/claude-sonnet-4.6",
      "model_name": "Anthropic: Claude Sonnet 4.6",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.6932966363636364,
      "accuracy_std_dev": 0.13147150091889792,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.609068142332415,
      "avg_duration_per_task_ms": 632474.5367309458,
      "avg_output_tokens_per_task": 40521.938016528926,
      "last_run_timestamp": "2026-10-24T08:08:54.166Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "DeepInfra",
      "accuracy": 0.6932378040665435,
      "accuracy_std_dev": 0.046335214744527585,
      "run_count": 11,
      "total_questions": 2164,
      "avg_cost_per_task_usd": 0.0074276700554528645,
      "avg_duration_per_task_ms": 281567.51894639555,
      "avg_output_tokens_per_task": 40195.87939001848,
      "last_run_timestamp": "2026-10-24T08:42:05.963Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Friendli",
      "accuracy": 0.691171,
      "accuracy_std_dev": 0.011421081968593563,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.006491580808080808,
      "avg_duration_per_task_ms": 106721.03198653199,
      "avg_output_tokens_per_task": 15952.695847362515,
      "last_run_timestamp": "2026-10-28T10:35:05.809Z"
    },
    {
      "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": 0.6905724000000001,
      "accuracy_std_dev": 0.021171864222122725,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.009511277777777778,
      "avg_duration_per_task_ms": 270838.6898989899,
      "avg_output_tokens_per_task": 6542.475757575758,
      "last_run_timestamp": "2026-10-03T17:02:48.284Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "DekaLLM",
      "accuracy": 0.6899941197301855,
      "accuracy_std_dev": 0.005866752821337055,
      "run_count": 3,
      "total_questions": 593,
      "avg_cost_per_task_usd": 0.0068884283305227665,
      "avg_duration_per_task_ms": 1677214.3709949409,
      "avg_output_tokens_per_task": 49011.02360876897,
      "last_run_timestamp": "2026-10-24T16:08:06.921Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5-nano-2025-08-07",
      "model_slug": "openai/gpt-5-nano",
      "model_name": "OpenAI: GPT-5 Nano",
      "provider_name": "Azure",
      "accuracy": 0.689815,
      "accuracy_std_dev": 0.0034448118235979193,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.013852137626262627,
      "avg_duration_per_task_ms": 234308.42045454544,
      "avg_output_tokens_per_task": 32893.11994949495,
      "last_run_timestamp": "2026-10-01T19:27:10.142Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "SambaNova",
      "accuracy": 0.6874404742120345,
      "accuracy_std_dev": 0.03143087716808991,
      "run_count": 6,
      "total_questions": 698,
      "avg_cost_per_task_usd": 0.009229828080229225,
      "avg_duration_per_task_ms": 54932.83810888252,
      "avg_output_tokens_per_task": 5977.71346704871,
      "last_run_timestamp": "2026-10-24T16:45:45.392Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "DeepInfra",
      "accuracy": 0.6855188289251547,
      "accuracy_std_dev": 0.09527891619753961,
      "run_count": 9,
      "total_questions": 1777,
      "avg_cost_per_task_usd": 0.011474473832301631,
      "avg_duration_per_task_ms": 906865.2909397861,
      "avg_output_tokens_per_task": 33586.68429938098,
      "last_run_timestamp": "2026-10-24T16:17:39.034Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Venice",
      "accuracy": 0.6854461436619719,
      "accuracy_std_dev": 0.012155576012164115,
      "run_count": 4,
      "total_questions": 710,
      "avg_cost_per_task_usd": 0.040622619718309855,
      "avg_duration_per_task_ms": 770310.514084507,
      "avg_output_tokens_per_task": 43146.859154929574,
      "last_run_timestamp": "2026-08-29T18:51:07.939Z"
    },
    {
      "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": "anthropic/claude-4.6-sonnet-20260217",
      "model_slug": "anthropic/claude-sonnet-4.6",
      "model_name": "Anthropic: Claude Sonnet 4.6",
      "provider_name": "Google",
      "accuracy": 0.6844233356431357,
      "accuracy_std_dev": 0.1414237811051766,
      "run_count": 35,
      "total_questions": 6927,
      "avg_cost_per_task_usd": 0.6621923740435977,
      "avg_duration_per_task_ms": 602080.901255955,
      "avg_output_tokens_per_task": 41272.465136422696,
      "last_run_timestamp": "2026-10-24T08:43:07.831Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5-nano-2025-08-07",
      "model_slug": "openai/gpt-5-nano",
      "model_name": "OpenAI: GPT-5 Nano",
      "provider_name": "OpenAI",
      "accuracy": 0.6843435000000001,
      "accuracy_std_dev": 0.005892499999999967,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.01356670707070707,
      "avg_duration_per_task_ms": 229761.702020202,
      "avg_output_tokens_per_task": 33820.56313131313,
      "last_run_timestamp": "2026-10-01T19:28:34.261Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.6-sonnet-20260217",
      "model_slug": "anthropic/claude-sonnet-4.6",
      "model_name": "Anthropic: Claude Sonnet 4.6",
      "provider_name": "auto-routing",
      "accuracy": 0.6830809999999999,
      "accuracy_std_dev": 0.14005659745795623,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.6164722958754209,
      "avg_duration_per_task_ms": 635625.3400673401,
      "avg_output_tokens_per_task": 40941.11069023569,
      "last_run_timestamp": "2026-10-24T08:26:35.484Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.6-sonnet-20260217",
      "model_slug": "anthropic/claude-sonnet-4.6",
      "model_name": "Anthropic: Claude Sonnet 4.6",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.6825664722222222,
      "accuracy_std_dev": 0.136152835465126,
      "run_count": 36,
      "total_questions": 7128,
      "avg_cost_per_task_usd": 0.6585870970819304,
      "avg_duration_per_task_ms": 454569.8709315376,
      "avg_output_tokens_per_task": 40985.43560606061,
      "last_run_timestamp": "2026-10-24T08:24:10.735Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.6-sonnet-20260217",
      "model_slug": "anthropic/claude-sonnet-4.6",
      "model_name": "Anthropic: Claude Sonnet 4.6",
      "provider_name": "Azure",
      "accuracy": 0.6805556666666668,
      "accuracy_std_dev": 0.14593486187813,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.6194684494949494,
      "avg_duration_per_task_ms": 636905.944023569,
      "avg_output_tokens_per_task": 41202.27062289562,
      "last_run_timestamp": "2026-10-24T08:14:03.137Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "SiliconFlow",
      "accuracy": 0.680394694980695,
      "accuracy_std_dev": 0.020660473202167934,
      "run_count": 8,
      "total_questions": 1554,
      "avg_cost_per_task_usd": 0.0022303307593307593,
      "avg_duration_per_task_ms": 211402.3449163449,
      "avg_output_tokens_per_task": 4069.488416988417,
      "last_run_timestamp": "2026-10-24T16:39:13.265Z"
    },
    {
      "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": 0.6796669584736251,
      "accuracy_std_dev": 0.014488707679507579,
      "run_count": 12,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.004952217732884399,
      "avg_duration_per_task_ms": 782700.4062850729,
      "avg_output_tokens_per_task": 49359.42255892256,
      "last_run_timestamp": "2026-10-30T00:43:53.430Z"
    },
    {
      "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": 0.6751971669865644,
      "accuracy_std_dev": 0.03877467265315564,
      "run_count": 8,
      "total_questions": 1563,
      "avg_cost_per_task_usd": 0.02805620857325656,
      "avg_duration_per_task_ms": 841992.2539987204,
      "avg_output_tokens_per_task": 18377.529750479847,
      "last_run_timestamp": "2026-10-24T16:22:30.498Z"
    },
    {
      "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": 0.6726822675159235,
      "accuracy_std_dev": 0.0862043071017272,
      "run_count": 3,
      "total_questions": 471,
      "avg_cost_per_task_usd": 0.07077279405520169,
      "avg_duration_per_task_ms": 1745687.9214437366,
      "avg_output_tokens_per_task": 24749.823779193204,
      "last_run_timestamp": "2026-08-22T17:09:42.218Z"
    },
    {
      "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": 0.6691920000000001,
      "accuracy_std_dev": 0.06399689168567932,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.008361025974025976,
      "avg_duration_per_task_ms": 1523891.8455988455,
      "avg_output_tokens_per_task": 65650.8253968254,
      "last_run_timestamp": "2026-10-30T04:19:55.253Z"
    },
    {
      "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": "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": 0.6683503999999999,
      "accuracy_std_dev": 0.046945141186708554,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.015881437373737373,
      "avg_duration_per_task_ms": 296657.8616161616,
      "avg_output_tokens_per_task": 8706.40404040404,
      "last_run_timestamp": "2026-10-03T17:36:48.437Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.6-20251208",
      "model_slug": "z-ai/glm-4.6v",
      "model_name": "Z.ai: GLM 4.6V",
      "provider_name": "auto-routing",
      "accuracy": 0.66835,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.013946772727272728,
      "avg_duration_per_task_ms": 324889.0353535354,
      "avg_output_tokens_per_task": 15315.242424242424,
      "last_run_timestamp": "2026-10-01T08:41:30.639Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Venice",
      "accuracy": 0.66813975,
      "accuracy_std_dev": 0.0169491097609137,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.005768640151515151,
      "avg_duration_per_task_ms": 418288.6957070707,
      "avg_output_tokens_per_task": 15810.410353535353,
      "last_run_timestamp": "2026-10-24T16:11:51.309Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "DeepInfra",
      "accuracy": 0.6671506640851887,
      "accuracy_std_dev": 0.1203125901358104,
      "run_count": 8,
      "total_questions": 1033,
      "avg_cost_per_task_usd": 0.09392116069699902,
      "avg_duration_per_task_ms": 1316783.9874152953,
      "avg_output_tokens_per_task": 26739.650532429816,
      "last_run_timestamp": "2026-10-24T16:06:30.901Z"
    },
    {
      "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": 0.6666666755181347,
      "accuracy_std_dev": 0.016603057166797653,
      "run_count": 9,
      "total_questions": 1544,
      "avg_cost_per_task_usd": 0.006663814119170984,
      "avg_duration_per_task_ms": 477892.86852331605,
      "avg_output_tokens_per_task": 47426.00971502591,
      "last_run_timestamp": "2026-10-24T16:02:28.575Z"
    },
    {
      "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": 0.6662925555555553,
      "accuracy_std_dev": 0.007644274285896932,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.005087682379349046,
      "avg_duration_per_task_ms": 104399.63636363637,
      "avg_output_tokens_per_task": 9580.892817059485,
      "last_run_timestamp": "2026-10-03T09:23:08.256Z"
    },
    {
      "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": "auto-routing",
      "accuracy": 0.6658247500000001,
      "accuracy_std_dev": 0.0064656044332684175,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.0027968030303030305,
      "avg_duration_per_task_ms": 81352.02777777778,
      "avg_output_tokens_per_task": 2993.090909090909,
      "last_run_timestamp": "2026-08-29T19:49:56.943Z"
    },
    {
      "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": 0.6649832,
      "accuracy_std_dev": 0.006981815133616774,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.002071141414141414,
      "avg_duration_per_task_ms": 228896.86666666667,
      "avg_output_tokens_per_task": 3033.7636363636366,
      "last_run_timestamp": "2026-10-05T20:44:48.611Z"
    },
    {
      "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": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Phala",
      "accuracy": 0.6627089841688654,
      "accuracy_std_dev": 0.007909198264520861,
      "run_count": 4,
      "total_questions": 379,
      "avg_cost_per_task_usd": 0.004262828496042217,
      "avg_duration_per_task_ms": 535749.2849604222,
      "avg_output_tokens_per_task": 28502.488126649077,
      "last_run_timestamp": "2026-08-29T17:24:11.557Z"
    },
    {
      "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": 0.6622475,
      "accuracy_std_dev": 0.019541184080039777,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.0077082720959595966,
      "avg_duration_per_task_ms": 563098.2026515151,
      "avg_output_tokens_per_task": 59511.905303030304,
      "last_run_timestamp": "2026-10-24T16:15:47.411Z"
    },
    {
      "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": 0.6616163125,
      "accuracy_std_dev": 0.014371619927476648,
      "run_count": 16,
      "total_questions": 3168,
      "avg_cost_per_task_usd": 0.00900978345959596,
      "avg_duration_per_task_ms": 319538.81597222225,
      "avg_output_tokens_per_task": 59717.22032828283,
      "last_run_timestamp": "2026-10-24T16:22:30.492Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Parasail",
      "accuracy": 0.6601731428571429,
      "accuracy_std_dev": 0.023160494075833857,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.008067838383838384,
      "avg_duration_per_task_ms": 808632.0974025974,
      "avg_output_tokens_per_task": 55153.242424242424,
      "last_run_timestamp": "2026-10-24T16:14:24.977Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.6-20251208",
      "model_slug": "z-ai/glm-4.6v",
      "model_name": "Z.ai: GLM 4.6V",
      "provider_name": "Z.AI",
      "accuracy": 0.659933,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.014984520202020202,
      "avg_duration_per_task_ms": 348987.1767676768,
      "avg_output_tokens_per_task": 16461.863636363636,
      "last_run_timestamp": "2026-10-01T08:28:22.621Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4.1-2025-04-14",
      "model_slug": "openai/gpt-4.1",
      "model_name": "OpenAI: GPT-4.1",
      "provider_name": "OpenAI",
      "accuracy": 0.6599326666666666,
      "accuracy_std_dev": 0.013255807590972711,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.015927558922558923,
      "avg_duration_per_task_ms": 14095.8367003367,
      "avg_output_tokens_per_task": 1794.7289562289561,
      "last_run_timestamp": "2026-10-02T23:17:48.072Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 VL 30B A3B Instruct",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.658249,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0066776414141414145,
      "avg_duration_per_task_ms": 101359.9898989899,
      "avg_output_tokens_per_task": 9307.358585858587,
      "last_run_timestamp": "2026-10-04T03:07:35.168Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "ibm-granite/granite-4.2-8b-20260831",
      "model_slug": "ibm-granite/granite-4.2-8b",
      "model_name": "IBM: Granite 4.2 8B",
      "provider_name": "auto-routing",
      "accuracy": 0.65736,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 197,
      "avg_cost_per_task_usd": 0.014864365482233502,
      "avg_duration_per_task_ms": 1717906.416243655,
      "avg_output_tokens_per_task": 64415.350253807104,
      "last_run_timestamp": "2026-10-01T08:48:38.304Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.6-sonnet-20260217",
      "model_slug": "anthropic/claude-sonnet-4.6",
      "model_name": "Anthropic: Claude Sonnet 4.6",
      "provider_name": "Anthropic",
      "accuracy": 0.6570467857142857,
      "accuracy_std_dev": 0.14261200486794257,
      "run_count": 14,
      "total_questions": 2772,
      "avg_cost_per_task_usd": 0.6177520173160173,
      "avg_duration_per_task_ms": 633109.20995671,
      "avg_output_tokens_per_task": 41054.18795093795,
      "last_run_timestamp": "2026-10-24T08:12:02.408Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "Venice",
      "accuracy": 0.6560171418530352,
      "accuracy_std_dev": 0.056453171596991564,
      "run_count": 8,
      "total_questions": 1565,
      "avg_cost_per_task_usd": 0.05637429073482428,
      "avg_duration_per_task_ms": 1175439.4670926516,
      "avg_output_tokens_per_task": 24772.420447284345,
      "last_run_timestamp": "2026-10-24T16:10:43.378Z"
    },
    {
      "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": 0.65277775,
      "accuracy_std_dev": 0.007840200328276042,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.003156868686868687,
      "avg_duration_per_task_ms": 41742.583333333336,
      "avg_output_tokens_per_task": 2990.8636363636365,
      "last_run_timestamp": "2026-08-29T19:51:51.198Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4.1-mini-2025-04-14",
      "model_slug": "openai/gpt-4.1-mini",
      "model_name": "OpenAI: GPT-4.1 Mini",
      "provider_name": "auto-routing",
      "accuracy": 0.652357,
      "accuracy_std_dev": 0.01262599999999997,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.004161952020202021,
      "avg_duration_per_task_ms": 27160.11111111111,
      "avg_output_tokens_per_task": 2408.8232323232323,
      "last_run_timestamp": "2026-10-01T19:12:31.553Z"
    },
    {
      "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": "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": "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": 0.6492706666666667,
      "accuracy_std_dev": 0.010498645193018418,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0041127255892255896,
      "avg_duration_per_task_ms": 24949.132996632998,
      "avg_output_tokens_per_task": 2375.9343434343436,
      "last_run_timestamp": "2026-10-02T23:15:53.868Z"
    },
    {
      "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": 0.6484773807106599,
      "accuracy_std_dev": 0.0054344006339245585,
      "run_count": 4,
      "total_questions": 788,
      "avg_cost_per_task_usd": 0.00290907614213198,
      "avg_duration_per_task_ms": 58072.35659898477,
      "avg_output_tokens_per_task": 2371.812182741117,
      "last_run_timestamp": "2026-10-27T00:30:02.893Z"
    },
    {
      "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": 0.6477272500000001,
      "accuracy_std_dev": 0.00710503227913143,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.003938823232323232,
      "avg_duration_per_task_ms": 63531.12626262626,
      "avg_output_tokens_per_task": 2389.935606060606,
      "last_run_timestamp": "2026-08-29T16:10:45.295Z"
    },
    {
      "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": "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-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "auto-routing",
      "accuracy": 0.6468855,
      "accuracy_std_dev": 0.133548982014091,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.015117938762626265,
      "avg_duration_per_task_ms": 431100.12184343435,
      "avg_output_tokens_per_task": 12974.737373737375,
      "last_run_timestamp": "2026-10-24T16:02:06.940Z"
    },
    {
      "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": 0.646675125,
      "accuracy_std_dev": 0.02302836491936359,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.006796745580808081,
      "avg_duration_per_task_ms": 647569.8042929292,
      "avg_output_tokens_per_task": 44075.63952020202,
      "last_run_timestamp": "2026-10-24T16:13:05.627Z"
    },
    {
      "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": 0.644781,
      "accuracy_std_dev": 0.0023808285322550727,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.0024933156565656567,
      "avg_duration_per_task_ms": 50490.56060606061,
      "avg_output_tokens_per_task": 2691.5126262626263,
      "last_run_timestamp": "2026-08-29T17:03:33.297Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "Mancer 2",
      "accuracy": 0.6439055298507463,
      "accuracy_std_dev": 0.20314360210271282,
      "run_count": 7,
      "total_questions": 1340,
      "avg_cost_per_task_usd": 0.07648400820895522,
      "avg_duration_per_task_ms": 1782783.4514925373,
      "avg_output_tokens_per_task": 36935.7671641791,
      "last_run_timestamp": "2026-10-24T16:17:39.031Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.64309775,
      "accuracy_std_dev": 0.010030734778045907,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.0023864103535353536,
      "avg_duration_per_task_ms": 55257.29924242424,
      "avg_output_tokens_per_task": 2792.7045454545455,
      "last_run_timestamp": "2026-08-29T16:11:23.388Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 VL 30B A3B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.6419754444444444,
      "accuracy_std_dev": 0.014018014458475362,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.007092590347923681,
      "avg_duration_per_task_ms": 241593.06509539843,
      "avg_output_tokens_per_task": 11377.103254769921,
      "last_run_timestamp": "2026-10-03T10:20:18.247Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "auto-routing",
      "accuracy": 0.641751,
      "accuracy_std_dev": 0.012926960106691747,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.0027229565656565658,
      "avg_duration_per_task_ms": 151188.63838383838,
      "avg_output_tokens_per_task": 9770.50707070707,
      "last_run_timestamp": "2026-10-03T17:16:11.892Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "Alibaba",
      "accuracy": 0.641751,
      "accuracy_std_dev": 0.015110385329302487,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.0044971595959595965,
      "avg_duration_per_task_ms": 76672.9303030303,
      "avg_output_tokens_per_task": 8453.274747474747,
      "last_run_timestamp": "2026-10-03T16:51:25.541Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-30b-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": 0.6401515,
      "accuracy_std_dev": 0.002759842885020819,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.004124544191919192,
      "avg_duration_per_task_ms": 34476.35353535353,
      "avg_output_tokens_per_task": 2259.632575757576,
      "last_run_timestamp": "2026-10-01T19:13:29.560Z"
    },
    {
      "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": 0.6388888630705394,
      "accuracy_std_dev": 0.22030686687149445,
      "run_count": 8,
      "total_questions": 1446,
      "avg_cost_per_task_usd": 0.008508840940525588,
      "avg_duration_per_task_ms": 167853.79668049794,
      "avg_output_tokens_per_task": 14018.163208852005,
      "last_run_timestamp": "2026-10-24T16:26:15.957Z"
    },
    {
      "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": 0.6381674285714285,
      "accuracy_std_dev": 0.012871844288514171,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.001091096681096681,
      "avg_duration_per_task_ms": 351783.54761904763,
      "avg_output_tokens_per_task": 11560.167388167389,
      "last_run_timestamp": "2026-10-30T01:21:27.783Z"
    },
    {
      "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": 0.6378052777777776,
      "accuracy_std_dev": 0.01862890271779676,
      "run_count": 9,
      "total_questions": 1692,
      "avg_cost_per_task_usd": 0.005482690307328605,
      "avg_duration_per_task_ms": 821174.3670212766,
      "avg_output_tokens_per_task": 35064.53132387707,
      "last_run_timestamp": "2026-10-30T02:09:50.419Z"
    },
    {
      "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": 0.6372055000000001,
      "accuracy_std_dev": 0.004208500000000004,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.015519383838383837,
      "avg_duration_per_task_ms": 14506.310606060606,
      "avg_output_tokens_per_task": 1746.2525252525252,
      "last_run_timestamp": "2026-10-01T19:12:45.033Z"
    },
    {
      "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": 0.6367847499999999,
      "accuracy_std_dev": 0.012277609526593496,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.015237281565656567,
      "avg_duration_per_task_ms": 18840.242424242424,
      "avg_output_tokens_per_task": 1618.909090909091,
      "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": "StreamLake",
      "accuracy": 0.6358028333333334,
      "accuracy_std_dev": 0.012808867337599457,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.001631114478114478,
      "avg_duration_per_task_ms": 75943.06734006734,
      "avg_output_tokens_per_task": 8246.542087542088,
      "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": 0.6324910327398047,
      "accuracy_std_dev": 0.011104531477434478,
      "run_count": 9,
      "total_questions": 1741,
      "avg_cost_per_task_usd": 0.01822674784606548,
      "avg_duration_per_task_ms": 490237.7823090178,
      "avg_output_tokens_per_task": 91135.2251579552,
      "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": "qwen/qwen3-32b-04-28",
      "model_slug": "qwen/qwen3-32b",
      "model_name": "Qwen: Qwen3 32B",
      "provider_name": "Groq",
      "accuracy": 0.6315160731707317,
      "accuracy_std_dev": 0.011095205901649571,
      "run_count": 4,
      "total_questions": 697,
      "avg_cost_per_task_usd": 0.008070608321377333,
      "avg_duration_per_task_ms": 35343.74175035868,
      "avg_output_tokens_per_task": 13371.598278335725,
      "last_run_timestamp": "2026-08-29T16:23:45.123Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "SiliconFlow",
      "accuracy": 0.6313132,
      "accuracy_std_dev": 0.007896357575490096,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.0027152242424242426,
      "avg_duration_per_task_ms": 260211.7595959596,
      "avg_output_tokens_per_task": 8830.749494949496,
      "last_run_timestamp": "2026-10-03T17:03:50.406Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "DekaLLM",
      "accuracy": 0.6306148156028369,
      "accuracy_std_dev": 0.05281625698168657,
      "run_count": 3,
      "total_questions": 282,
      "avg_cost_per_task_usd": 0.0037584255319148938,
      "avg_duration_per_task_ms": 208452.59574468085,
      "avg_output_tokens_per_task": 12308.468085106382,
      "last_run_timestamp": "2026-10-23T09:29:14.171Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "SiliconFlow",
      "accuracy": 0.6305678657378088,
      "accuracy_std_dev": 0.007960405916746635,
      "run_count": 8,
      "total_questions": 1579,
      "avg_cost_per_task_usd": 0.00187089613679544,
      "avg_duration_per_task_ms": 238458.68967701076,
      "avg_output_tokens_per_task": 10226.877137428752,
      "last_run_timestamp": "2026-10-24T16:03:25.396Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "Venice",
      "accuracy": 0.6292411856287424,
      "accuracy_std_dev": 0.0117815,
      "run_count": 2,
      "total_questions": 334,
      "avg_cost_per_task_usd": 0.0071094161676646716,
      "avg_duration_per_task_ms": 112258.81736526947,
      "avg_output_tokens_per_task": 28232.077844311378,
      "last_run_timestamp": "2026-08-15T08:02:44.624Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 VL 30B A3B Instruct",
      "provider_name": "Novita",
      "accuracy": 0.6290686666666666,
      "accuracy_std_dev": 0.03535789523021852,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.00678852356902357,
      "avg_duration_per_task_ms": 106935.86756453423,
      "avg_output_tokens_per_task": 9465.799102132436,
      "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-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": "Groq",
      "accuracy": 0.6280566473616473,
      "accuracy_std_dev": 0.014452741297068885,
      "run_count": 8,
      "total_questions": 1554,
      "avg_cost_per_task_usd": 0.01670690218790219,
      "avg_duration_per_task_ms": 81944.96975546975,
      "avg_output_tokens_per_task": 55459.59009009009,
      "last_run_timestamp": "2026-10-24T16:45:19.455Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "CoreWeave",
      "accuracy": 0.6275255000000001,
      "accuracy_std_dev": 0.010334903785231893,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.0036913661616161616,
      "avg_duration_per_task_ms": 150999.66287878787,
      "avg_output_tokens_per_task": 12500.281565656565,
      "last_run_timestamp": "2026-08-29T18:20:19.448Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3-nano-30b-a3b",
      "model_slug": "nvidia/nemotron-3-nano-30b-a3b",
      "model_name": "NVIDIA: Nemotron 3 Nano 30B A3B",
      "provider_name": "Nebius",
      "accuracy": 0.627104375,
      "accuracy_std_dev": 0.008333089267155095,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.024497946969696968,
      "avg_duration_per_task_ms": 1060409.1414141415,
      "avg_output_tokens_per_task": 101874.34027777778,
      "last_run_timestamp": "2026-10-23T09:03:34.804Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.6265433333333333,
      "accuracy_std_dev": 0.01630548430913829,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.004165183501683501,
      "avg_duration_per_task_ms": 241184.85185185185,
      "avg_output_tokens_per_task": 13613.127104377105,
      "last_run_timestamp": "2026-10-23T08:08:11.470Z"
    },
    {
      "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-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "DigitalOcean",
      "accuracy": 0.6199326452702703,
      "accuracy_std_dev": 0.053313699046305146,
      "run_count": 3,
      "total_questions": 592,
      "avg_cost_per_task_usd": 0.037519454391891895,
      "avg_duration_per_task_ms": 2801762.006756757,
      "avg_output_tokens_per_task": 26159.864864864863,
      "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": 0.6184415959279561,
      "accuracy_std_dev": 0.18122512960844483,
      "run_count": 13,
      "total_questions": 2554,
      "avg_cost_per_task_usd": 0.026508938919342206,
      "avg_duration_per_task_ms": 731184.3801879404,
      "avg_output_tokens_per_task": 37523.92443226312,
      "last_run_timestamp": "2026-10-30T02:15:35.781Z"
    },
    {
      "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Google",
      "accuracy": 0.6139909190897598,
      "accuracy_std_dev": 0.11036434192418933,
      "run_count": 8,
      "total_questions": 1582,
      "avg_cost_per_task_usd": 0.009721548672566372,
      "avg_duration_per_task_ms": 349467.04930467764,
      "avg_output_tokens_per_task": 27722.50379266751,
      "last_run_timestamp": "2026-10-24T16:05:40.119Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-14b-04-28",
      "model_slug": "qwen/qwen3-14b",
      "model_name": "Qwen: Qwen3 14B",
      "provider_name": "DeepInfra",
      "accuracy": 0.6123375670628184,
      "accuracy_std_dev": 0.008273645159312946,
      "run_count": 3,
      "total_questions": 589,
      "avg_cost_per_task_usd": 0.003321100169779287,
      "avg_duration_per_task_ms": 426798.44821731746,
      "avg_output_tokens_per_task": 13430.818336162989,
      "last_run_timestamp": "2026-10-08T22:16:42.056Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-32b-04-28",
      "model_slug": "qwen/qwen3-32b",
      "model_name": "Qwen: Qwen3 32B",
      "provider_name": "Nebius",
      "accuracy": 0.611953,
      "accuracy_std_dev": 0.007760307500350757,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.005744295454545455,
      "avg_duration_per_task_ms": 648984.4204545454,
      "avg_output_tokens_per_task": 19126.670454545456,
      "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-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "Novita",
      "accuracy": 0.6104378,
      "accuracy_std_dev": 0.012074440366327535,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.003635177777777778,
      "avg_duration_per_task_ms": 32727.437373737375,
      "avg_output_tokens_per_task": 2146.0969696969696,
      "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": 0.6102123225480283,
      "accuracy_std_dev": 0.005364099163885748,
      "run_count": 5,
      "total_questions": 989,
      "avg_cost_per_task_usd": 0.004066017189079878,
      "avg_duration_per_task_ms": 27933.76643073812,
      "avg_output_tokens_per_task": 2160.4469160768454,
      "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": "DeepInfra",
      "accuracy": 0.609428,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.00778269696969697,
      "avg_duration_per_task_ms": 302947.1666666667,
      "avg_output_tokens_per_task": 15369.68686868687,
      "last_run_timestamp": "2026-10-08T22:57:53.933Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-32b-04-28",
      "model_slug": "qwen/qwen3-32b",
      "model_name": "Qwen: Qwen3 32B",
      "provider_name": "auto-routing",
      "accuracy": 0.6073235,
      "accuracy_std_dev": 0.010738651928896826,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.008191137626262626,
      "avg_duration_per_task_ms": 318721.5845959596,
      "avg_output_tokens_per_task": 17321.883838383837,
      "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-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": 0.6040405999999999,
      "accuracy_std_dev": 0.007924755859961838,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.003472381818181818,
      "avg_duration_per_task_ms": 58103.21111111111,
      "avg_output_tokens_per_task": 2119.49797979798,
      "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": 0.6033671999999999,
      "accuracy_std_dev": 0.005694314722598297,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.010998522222222222,
      "avg_duration_per_task_ms": 31713.89797979798,
      "avg_output_tokens_per_task": 2095.668686868687,
      "last_run_timestamp": "2026-10-03T17:43:44.972Z"
    },
    {
      "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": 0.5997241451304901,
      "accuracy_std_dev": 0.07108217488715086,
      "run_count": 8,
      "total_questions": 1571,
      "avg_cost_per_task_usd": 0.012881001273074474,
      "avg_duration_per_task_ms": 1360475.176957352,
      "avg_output_tokens_per_task": 57168.979630808404,
      "last_run_timestamp": "2026-10-30T01:56:42.191Z"
    },
    {
      "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": 0.5990461666666667,
      "accuracy_std_dev": 0.015102210725769769,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.003087529461279461,
      "avg_duration_per_task_ms": 677793.6026936027,
      "avg_output_tokens_per_task": 13663.569865319865,
      "last_run_timestamp": "2026-10-30T14:52:53.719Z"
    },
    {
      "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": 0.5983168000000001,
      "accuracy_std_dev": 0.003626295707743628,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.003269590909090909,
      "avg_duration_per_task_ms": 65326.75858585859,
      "avg_output_tokens_per_task": 2077.4313131313133,
      "last_run_timestamp": "2026-10-03T16:59:44.383Z"
    },
    {
      "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": 0.5963805,
      "accuracy_std_dev": 0.0012624999999999997,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.01735105808080808,
      "avg_duration_per_task_ms": 112232.29292929293,
      "avg_output_tokens_per_task": 43198.833333333336,
      "last_run_timestamp": "2026-10-01T19:14:48.696Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "DeepInfra",
      "accuracy": 0.5948373333333333,
      "accuracy_std_dev": 0.010316724200162651,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.04306378282828283,
      "avg_duration_per_task_ms": 401108.6296296296,
      "avg_output_tokens_per_task": 21434.68181818182,
      "last_run_timestamp": "2026-08-22T10:29:25.897Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3-nano-30b-a3b",
      "model_slug": "nvidia/nemotron-3-nano-30b-a3b",
      "model_name": "NVIDIA: Nemotron 3 Nano 30B A3B",
      "provider_name": "Novita",
      "accuracy": 0.5942281016850668,
      "accuracy_std_dev": 0.00965016140081236,
      "run_count": 9,
      "total_questions": 1721,
      "avg_cost_per_task_usd": 0.009055742010459035,
      "avg_duration_per_task_ms": 523063.0418361418,
      "avg_output_tokens_per_task": 45085.088901801275,
      "last_run_timestamp": "2026-10-03T17:04:03.802Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Cerebras",
      "accuracy": 0.591751125,
      "accuracy_std_dev": 0.0083327609685731,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.01027929356060606,
      "avg_duration_per_task_ms": 11398.482323232323,
      "avg_output_tokens_per_task": 13251.253787878788,
      "last_run_timestamp": "2026-10-24T16:20:41.333Z"
    },
    {
      "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": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "DeepInfra",
      "accuracy": 0.5878379610492847,
      "accuracy_std_dev": 0.08614130418814413,
      "run_count": 8,
      "total_questions": 1258,
      "avg_cost_per_task_usd": 0.056543784578696354,
      "avg_duration_per_task_ms": 347246.7694753577,
      "avg_output_tokens_per_task": 21441.431637519872,
      "last_run_timestamp": "2026-10-24T16:28:50.076Z"
    },
    {
      "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": 0.5875422,
      "accuracy_std_dev": 0.007299620439447542,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.006416512121212121,
      "avg_duration_per_task_ms": 659187.0757575758,
      "avg_output_tokens_per_task": 42517.9595959596,
      "last_run_timestamp": "2026-10-03T16:12:07.492Z"
    },
    {
      "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": 0.5835020000000001,
      "accuracy_std_dev": 0.037153058162148644,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.02795441818181818,
      "avg_duration_per_task_ms": 1395358.9363636363,
      "avg_output_tokens_per_task": 69775.76262626263,
      "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": 0.5800594410646388,
      "accuracy_std_dev": 0.007005609872630659,
      "run_count": 4,
      "total_questions": 789,
      "avg_cost_per_task_usd": 0.025549254752851712,
      "avg_duration_per_task_ms": 1152259.712294043,
      "avg_output_tokens_per_task": 67901.9062103929,
      "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": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "SambaNova",
      "accuracy": 0.5736152536443148,
      "accuracy_std_dev": 0.03731433483807747,
      "run_count": 7,
      "total_questions": 686,
      "avg_cost_per_task_usd": 0.03136954518950437,
      "avg_duration_per_task_ms": 62083.20991253644,
      "avg_output_tokens_per_task": 6724.625364431487,
      "last_run_timestamp": "2026-10-24T16:01:29.840Z"
    },
    {
      "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": "Fireworks",
      "accuracy": 0.569024,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.019884464646464647,
      "avg_duration_per_task_ms": 348444.5303030303,
      "avg_output_tokens_per_task": 99240.31818181818,
      "last_run_timestamp": "2026-10-30T21:00:04.500Z"
    },
    {
      "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": 0.56060625,
      "accuracy_std_dev": 0.009297342103391676,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.017428513888888886,
      "avg_duration_per_task_ms": 127446.60858585859,
      "avg_output_tokens_per_task": 43389.579545454544,
      "last_run_timestamp": "2026-10-01T19:27:01.395Z"
    },
    {
      "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": 0.5558924000000001,
      "accuracy_std_dev": 0.010399706661247687,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.004659020202020202,
      "avg_duration_per_task_ms": 490788.4282828283,
      "avg_output_tokens_per_task": 16414.2101010101,
      "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": 0.5526449293849659,
      "accuracy_std_dev": 0.02398069638824749,
      "run_count": 7,
      "total_questions": 1317,
      "avg_cost_per_task_usd": 0.005556834472285498,
      "avg_duration_per_task_ms": 809029.5223993926,
      "avg_output_tokens_per_task": 14330.80941533789,
      "last_run_timestamp": "2026-10-30T20:09:39.337Z"
    },
    {
      "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": 0.5525624444444444,
      "accuracy_std_dev": 0.13700937369238736,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.0030681475869809203,
      "avg_duration_per_task_ms": 217856.8911335578,
      "avg_output_tokens_per_task": 7427.39898989899,
      "last_run_timestamp": "2026-10-24T16:03:25.478Z"
    },
    {
      "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": 0.5521885000000001,
      "accuracy_std_dev": 0.005050500000000013,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.01778226515151515,
      "avg_duration_per_task_ms": 116667.00757575757,
      "avg_output_tokens_per_task": 44273.378787878784,
      "last_run_timestamp": "2026-10-01T19:15:36.386Z"
    },
    {
      "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": 0.54798,
      "accuracy_std_dev": 0.06951982075998184,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.005136612373737373,
      "avg_duration_per_task_ms": 120175.0505050505,
      "avg_output_tokens_per_task": 4605.713383838384,
      "last_run_timestamp": "2026-08-29T20:57:58.443Z"
    },
    {
      "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": 0.5420490229357798,
      "accuracy_std_dev": 0.06975384672944968,
      "run_count": 3,
      "total_questions": 218,
      "avg_cost_per_task_usd": 0.03485861009174312,
      "avg_duration_per_task_ms": 397279.0229357798,
      "avg_output_tokens_per_task": 16660.94495412844,
      "last_run_timestamp": "2026-08-29T19:07:49.385Z"
    },
    {
      "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": 0.5415076988564168,
      "accuracy_std_dev": 0.08088323645385365,
      "run_count": 4,
      "total_questions": 787,
      "avg_cost_per_task_usd": 0.0033490050825921223,
      "avg_duration_per_task_ms": 120991.54510800508,
      "avg_output_tokens_per_task": 3568.362134688691,
      "last_run_timestamp": "2026-08-29T20:34:10.117Z"
    },
    {
      "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": 0.5373176666666667,
      "accuracy_std_dev": 0.013403444242266851,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0019087592592592591,
      "avg_duration_per_task_ms": 97924.95622895623,
      "avg_output_tokens_per_task": 1886.641414141414,
      "last_run_timestamp": "2026-10-08T22:10:22.349Z"
    },
    {
      "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": 0.537037,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 126,
      "avg_cost_per_task_usd": 0.007595468253968254,
      "avg_duration_per_task_ms": 1060044.9047619049,
      "avg_output_tokens_per_task": 22408.52380952381,
      "last_run_timestamp": "2026-08-08T10:11:59.748Z"
    },
    {
      "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": "openai/gpt-4o",
      "model_slug": "openai/gpt-4o",
      "model_name": "OpenAI: GPT-4o",
      "provider_name": "OpenAI",
      "accuracy": 0.5266555,
      "accuracy_std_dev": 0.023895888382383546,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.01547567760942761,
      "avg_duration_per_task_ms": 13716.89898989899,
      "avg_output_tokens_per_task": 1354.1801346801346,
      "last_run_timestamp": "2026-10-02T23:17:37.980Z"
    },
    {
      "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": 0.5238700798479087,
      "accuracy_std_dev": 0.008445731536551466,
      "run_count": 4,
      "total_questions": 789,
      "avg_cost_per_task_usd": 0.016655803548795946,
      "avg_duration_per_task_ms": 761511.7452471483,
      "avg_output_tokens_per_task": 42502.96070975919,
      "last_run_timestamp": "2026-08-30T00:11:29.373Z"
    },
    {
      "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": "auto-routing",
      "accuracy": 0.5193605,
      "accuracy_std_dev": 0.0025254999999999583,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.003164820707070707,
      "avg_duration_per_task_ms": 106661.49242424243,
      "avg_output_tokens_per_task": 2729.3585858585857,
      "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",
      "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": "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": "Alibaba",
      "accuracy": 0.5151515,
      "accuracy_std_dev": 0.008417499999999967,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.004014873737373737,
      "avg_duration_per_task_ms": 26643.18181818182,
      "avg_output_tokens_per_task": 2582.7247474747473,
      "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": "auto-routing",
      "accuracy": 0.510943,
      "accuracy_std_dev": 0.0008420000000000094,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.013169762626262625,
      "avg_duration_per_task_ms": 12925.411616161617,
      "avg_output_tokens_per_task": 1121.608585858586,
      "last_run_timestamp": "2026-10-01T16:03:45.472Z"
    },
    {
      "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": 0.5084175,
      "accuracy_std_dev": 0.025252499999999983,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.004674189393939395,
      "avg_duration_per_task_ms": 291399.3181818182,
      "avg_output_tokens_per_task": 2695.719696969697,
      "last_run_timestamp": "2026-08-15T09:11:29.276Z"
    },
    {
      "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": 0.5075759999999999,
      "accuracy_std_dev": 0.0008420000000000094,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.0008372954545454546,
      "avg_duration_per_task_ms": 96066.76515151515,
      "avg_output_tokens_per_task": 2787.2348484848485,
      "last_run_timestamp": "2026-08-15T08:49:03.156Z"
    },
    {
      "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": 0.5050510000000001,
      "accuracy_std_dev": 0.002749143987983661,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0008273518518518518,
      "avg_duration_per_task_ms": 18715.93265993266,
      "avg_output_tokens_per_task": 1873.8602693602693,
      "last_run_timestamp": "2026-10-02T23:16:50.042Z"
    },
    {
      "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-4.1-nano-2025-04-14",
      "model_slug": "openai/gpt-4.1-nano",
      "model_name": "OpenAI: GPT-4.1 Nano",
      "provider_name": "Azure",
      "accuracy": 0.4995792500000001,
      "accuracy_std_dev": 0.02025051336849268,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.0008581388888888889,
      "avg_duration_per_task_ms": 12992.723484848484,
      "avg_output_tokens_per_task": 1853.0088383838383,
      "last_run_timestamp": "2026-10-01T19:12:29.219Z"
    },
    {
      "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": 0.496001625,
      "accuracy_std_dev": 0.012540032196305345,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.0023573857323232325,
      "avg_duration_per_task_ms": 9976.099116161617,
      "avg_output_tokens_per_task": 2337.1717171717173,
      "last_run_timestamp": "2026-10-24T16:02:08.875Z"
    },
    {
      "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": 0.49242400000000003,
      "accuracy_std_dev": 0.0008419999999999817,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.0017634166666666666,
      "avg_duration_per_task_ms": 30891.38888888889,
      "avg_output_tokens_per_task": 1731.7373737373737,
      "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-4.7-flash-20260119",
      "model_slug": "z-ai/glm-4.7-flash",
      "model_name": "Z.ai: GLM 4.7 Flash",
      "provider_name": "Venice",
      "accuracy": 0.49168807800511505,
      "accuracy_std_dev": 0.022345304455019185,
      "run_count": 4,
      "total_questions": 782,
      "avg_cost_per_task_usd": 0.010985424552429669,
      "avg_duration_per_task_ms": 526999.0754475703,
      "avg_output_tokens_per_task": 27562.6368286445,
      "last_run_timestamp": "2026-08-30T00:23:23.003Z"
    },
    {
      "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": 0.49074049999999997,
      "accuracy_std_dev": 0.009259500000000004,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.0008383863636363636,
      "avg_duration_per_task_ms": 22708.954545454544,
      "avg_output_tokens_per_task": 1902.2954545454545,
      "last_run_timestamp": "2026-10-01T19:13:37.127Z"
    },
    {
      "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": 0.48782048076923085,
      "accuracy_std_dev": 0.015008531315555174,
      "run_count": 4,
      "total_questions": 780,
      "avg_cost_per_task_usd": 0.011197282051282051,
      "avg_duration_per_task_ms": 480526.36282051285,
      "avg_output_tokens_per_task": 27926.11923076923,
      "last_run_timestamp": "2026-08-30T01:15:14.669Z"
    },
    {
      "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": 0.4844273749999999,
      "accuracy_std_dev": 0.0055676718190258885,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.0023532323232323235,
      "avg_duration_per_task_ms": 15477.722222222223,
      "avg_output_tokens_per_task": 2211.7329545454545,
      "last_run_timestamp": "2026-10-24T16:21:36.007Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "AkashML",
      "accuracy": 0.4819933646302252,
      "accuracy_std_dev": 0.014921479863223815,
      "run_count": 16,
      "total_questions": 3110,
      "avg_cost_per_task_usd": 0.0012891855305466235,
      "avg_duration_per_task_ms": 59448.33536977492,
      "avg_output_tokens_per_task": 2320.65884244373,
      "last_run_timestamp": "2026-10-30T00:42:34.049Z"
    },
    {
      "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": 0.48148112499999995,
      "accuracy_std_dev": 0.009923980116836928,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.0026001357323232324,
      "avg_duration_per_task_ms": 50611.480429292926,
      "avg_output_tokens_per_task": 2853.9835858585857,
      "last_run_timestamp": "2026-10-24T16:03:59.174Z"
    },
    {
      "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": 0.478956,
      "accuracy_std_dev": 0.004209000000000018,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.048212727272727275,
      "avg_duration_per_task_ms": 51496.83838383838,
      "avg_output_tokens_per_task": 16895.032828282827,
      "last_run_timestamp": "2026-08-15T14:46:10.940Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "Parasail",
      "accuracy": 0.47664124999999996,
      "accuracy_std_dev": 0.014045976548019012,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.0013353724747474747,
      "avg_duration_per_task_ms": 46855.55997474748,
      "avg_output_tokens_per_task": 2393.6155303030305,
      "last_run_timestamp": "2026-10-24T16:03:59.176Z"
    },
    {
      "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": "Nebius",
      "accuracy": 0.47558900000000004,
      "accuracy_std_dev": 0.011198623844026553,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.0010472727272727272,
      "avg_duration_per_task_ms": 222278.38005050505,
      "avg_output_tokens_per_task": 2355.248737373737,
      "last_run_timestamp": "2026-08-29T20:09:29.346Z"
    },
    {
      "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": 0.47474712499999994,
      "accuracy_std_dev": 0.009182339672402404,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.0014147474747474747,
      "avg_duration_per_task_ms": 78596.79671717172,
      "avg_output_tokens_per_task": 2360.6445707070707,
      "last_run_timestamp": "2026-10-24T16:04:55.367Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "Together",
      "accuracy": 0.4728232857142857,
      "accuracy_std_dev": 0.008882080414508524,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.0031378203463203465,
      "avg_duration_per_task_ms": 63855.042568542565,
      "avg_output_tokens_per_task": 2211.1046176046175,
      "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": "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": "Novita",
      "accuracy": 0.46877630000000003,
      "accuracy_std_dev": 0.013737037653161437,
      "run_count": 8,
      "total_questions": 1580,
      "avg_cost_per_task_usd": 0.0013216873417721518,
      "avg_duration_per_task_ms": 88415.32848101266,
      "avg_output_tokens_per_task": 3031.9816455696205,
      "last_run_timestamp": "2026-10-24T16:02:28.569Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "DeepInfra",
      "accuracy": 0.468294,
      "accuracy_std_dev": 0.012796634635715754,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.0007987283950617283,
      "avg_duration_per_task_ms": 163190.25925925927,
      "avg_output_tokens_per_task": 2245.347923681257,
      "last_run_timestamp": "2026-10-24T16:22:30.499Z"
    },
    {
      "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": 0.4637588786346397,
      "accuracy_std_dev": 0.005380526437998433,
      "run_count": 4,
      "total_questions": 791,
      "avg_cost_per_task_usd": 0.0022411302149178258,
      "avg_duration_per_task_ms": 37374.46017699115,
      "avg_output_tokens_per_task": 2724.637168141593,
      "last_run_timestamp": "2026-08-29T16:31:42.504Z"
    },
    {
      "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/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": 0.457071,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0017076717171717172,
      "avg_duration_per_task_ms": 27081.62626262626,
      "avg_output_tokens_per_task": 1337.1363636363637,
      "last_run_timestamp": "2026-09-01T13:03:42.840Z"
    },
    {
      "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": 0.442761,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0013494898989898992,
      "avg_duration_per_task_ms": 69228.69696969698,
      "avg_output_tokens_per_task": 1528.5050505050506,
      "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": 0.43602700000000005,
      "accuracy_std_dev": 0.006734000000000018,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.0010042146464646464,
      "avg_duration_per_task_ms": 21558.090909090908,
      "avg_output_tokens_per_task": 1481.030303030303,
      "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": 0.43139700000000003,
      "accuracy_std_dev": 0.012951914704011906,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.0010746982323232325,
      "avg_duration_per_task_ms": 24227.715909090908,
      "avg_output_tokens_per_task": 1509.2487373737374,
      "last_run_timestamp": "2026-10-01T19:05:18.003Z"
    },
    {
      "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": 0.429293,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0015700353535353536,
      "avg_duration_per_task_ms": 59122.85858585859,
      "avg_output_tokens_per_task": 1664.8737373737374,
      "last_run_timestamp": "2026-09-01T13:21:35.512Z"
    },
    {
      "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": 0.4276093333333334,
      "accuracy_std_dev": 0.00629925838668507,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0010292508417508418,
      "avg_duration_per_task_ms": 16725.542087542086,
      "avg_output_tokens_per_task": 1523.1902356902358,
      "last_run_timestamp": "2026-10-02T23:19:57.879Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "DigitalOcean",
      "accuracy": 0.422559,
      "accuracy_std_dev": 0.09221941654915557,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.019149646464646466,
      "avg_duration_per_task_ms": 3630846.4646464647,
      "avg_output_tokens_per_task": 17674.203703703704,
      "last_run_timestamp": "2026-08-22T17:37:26.215Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "Cloudflare",
      "accuracy": 0.421027614334471,
      "accuracy_std_dev": 0.09192114718900765,
      "run_count": 9,
      "total_questions": 1758,
      "avg_cost_per_task_usd": 0.004391554607508533,
      "avg_duration_per_task_ms": 42970.98634812287,
      "avg_output_tokens_per_task": 1850.198521046644,
      "last_run_timestamp": "2026-10-24T16:02:47.168Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-3-12b-it",
      "model_slug": "google/gemma-3-12b-it",
      "model_name": "Google: Gemma 3 12B",
      "provider_name": "DeepInfra",
      "accuracy": 0.405724,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0003259848484848485,
      "avg_duration_per_task_ms": 29056.348484848484,
      "avg_output_tokens_per_task": 1917.1515151515152,
      "last_run_timestamp": "2026-10-08T22:01:16.852Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "tencent/hy-mt2-30b-a3b-20260521",
      "model_slug": "tencent/hy-mt2-30b-a3b",
      "model_name": "Tencent: Hy-MT2-30B-A3B",
      "provider_name": "Reka",
      "accuracy": 0.388889,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.00023147979797979797,
      "avg_duration_per_task_ms": 8604.353535353535,
      "avg_output_tokens_per_task": 710.540404040404,
      "last_run_timestamp": "2026-09-24T06:00:03.613Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "amazon/nova-micro-v1",
      "model_slug": "amazon/nova-micro-v1",
      "model_name": "Amazon: Nova Micro 1.0",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.3855215,
      "accuracy_std_dev": 0.011293261320362695,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.0001567070707070707,
      "avg_duration_per_task_ms": 7111.70202020202,
      "avg_output_tokens_per_task": 932.0808080808081,
      "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": 0.3804713333333334,
      "accuracy_std_dev": 0.013950341891469493,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.00034128451178451176,
      "avg_duration_per_task_ms": 43460.74410774411,
      "avg_output_tokens_per_task": 1987.3602693602693,
      "last_run_timestamp": "2026-10-30T14:15:27.188Z"
    },
    {
      "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": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "Venice",
      "accuracy": 0.32428450000000003,
      "accuracy_std_dev": 0.09813129033340996,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.008669871843434344,
      "avg_duration_per_task_ms": 575303.3933080808,
      "avg_output_tokens_per_task": 21440.64835858586,
      "last_run_timestamp": "2026-10-24T16:36:16.779Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.28756325,
      "accuracy_std_dev": 0.017090551700793627,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.002848802398989899,
      "avg_duration_per_task_ms": 50657.708333333336,
      "avg_output_tokens_per_task": 35609.12626262626,
      "last_run_timestamp": "2026-10-24T16:02:28.572Z"
    },
    {
      "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": 0.2873176666666667,
      "accuracy_std_dev": 0.0007938452130121062,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.00006128619528619528,
      "avg_duration_per_task_ms": 10231.205387205388,
      "avg_output_tokens_per_task": 580.9023569023569,
      "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-3.2-3b-instruct",
      "model_slug": "meta-llama/llama-3.2-3b-instruct",
      "model_name": "Meta: Llama 3.2 3B Instruct",
      "provider_name": "Cloudflare",
      "accuracy": 0.286195,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0005941969696969697,
      "avg_duration_per_task_ms": 14998.156565656565,
      "avg_output_tokens_per_task": 1642.1363636363637,
      "last_run_timestamp": "2026-10-13T18:56:05.001Z"
    },
    {
      "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": 0.280303,
      "accuracy_std_dev": 0.021444131149337797,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.0013256224747474747,
      "avg_duration_per_task_ms": 144884.44255050505,
      "avg_output_tokens_per_task": 18578.765151515152,
      "last_run_timestamp": "2026-10-24T16:08:06.919Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Coder 30B A3B Instruct",
      "provider_name": "Novita",
      "accuracy": 0.27693599999999996,
      "accuracy_std_dev": 0.010943000000000008,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.0007441464646464646,
      "avg_duration_per_task_ms": 48473.70202020202,
      "avg_output_tokens_per_task": 2545.467171717172,
      "last_run_timestamp": "2026-08-15T09:28:07.208Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "sao10k/l3-lunaris-8b",
      "model_slug": "sao10k/l3-lunaris-8b",
      "model_name": "Sao10K: Llama 3 8B Lunaris",
      "provider_name": "auto-routing",
      "accuracy": 0.27665533333333336,
      "accuracy_std_dev": 0.0007938452130121063,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.00006201683501683503,
      "avg_duration_per_task_ms": 9496.395622895623,
      "avg_output_tokens_per_task": 580.2053872053872,
      "last_run_timestamp": "2026-10-01T20:43:57.173Z"
    },
    {
      "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": 0.2745902896174863,
      "accuracy_std_dev": 0.009941000000000005,
      "run_count": 2,
      "total_questions": 366,
      "avg_cost_per_task_usd": 0.0000641912568306011,
      "avg_duration_per_task_ms": 8433.43169398907,
      "avg_output_tokens_per_task": 532.0136612021857,
      "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": "Parasail",
      "accuracy": 0.27161569453924916,
      "accuracy_std_dev": 0.020450238993441835,
      "run_count": 3,
      "total_questions": 586,
      "avg_cost_per_task_usd": 0.000055156996587030725,
      "avg_duration_per_task_ms": 10158.571672354949,
      "avg_output_tokens_per_task": 550.3515358361775,
      "last_run_timestamp": "2026-10-01T20:47:49.128Z"
    },
    {
      "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": "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": 0.27046308294930876,
      "accuracy_std_dev": 0.013958073966682333,
      "run_count": 8,
      "total_questions": 1519,
      "avg_cost_per_task_usd": 0.00941191836734694,
      "avg_duration_per_task_ms": 387844.9934167215,
      "avg_output_tokens_per_task": 42072.92560895326,
      "last_run_timestamp": "2026-10-24T16:10:43.379Z"
    },
    {
      "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": 0.265432,
      "accuracy_std_dev": 0.010735715625890985,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.00039356734006734006,
      "avg_duration_per_task_ms": 391096.85185185185,
      "avg_output_tokens_per_task": 9436.173961840628,
      "last_run_timestamp": "2026-10-24T16:01:29.844Z"
    },
    {
      "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": 0.254208875,
      "accuracy_std_dev": 0.012051921116957867,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.0004300145202020202,
      "avg_duration_per_task_ms": 75497.70643939394,
      "avg_output_tokens_per_task": 8276.654671717171,
      "last_run_timestamp": "2026-10-24T16:02:28.577Z"
    },
    {
      "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": 0.253367,
      "accuracy_std_dev": 0.23989899999999997,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.016424065656565658,
      "avg_duration_per_task_ms": 206615.3207070707,
      "avg_output_tokens_per_task": 20468.073232323233,
      "last_run_timestamp": "2026-08-15T15:24:35.389Z"
    },
    {
      "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": 0.21801350000000003,
      "accuracy_std_dev": 0.11886565594085619,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.0018304709595959596,
      "avg_duration_per_task_ms": 56840.131313131315,
      "avg_output_tokens_per_task": 1584.419191919192,
      "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-3.1-8b-instruct",
      "model_slug": "meta-llama/llama-3.1-8b-instruct",
      "model_name": "Meta: Llama 3.1 8B Instruct",
      "provider_name": "Cloudflare",
      "accuracy": 0.17239079999999998,
      "accuracy_std_dev": 0.1395319132204529,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.001048532323232323,
      "avg_duration_per_task_ms": 421240.27575757576,
      "avg_output_tokens_per_task": 3399.7606060606063,
      "last_run_timestamp": "2026-10-24T16:03:19.773Z"
    },
    {
      "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": "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": "anthropic/claude-5-fable-20260609",
      "model_slug": "anthropic/claude-fable-5",
      "model_name": "Anthropic: Claude Fable 5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.8233334999999999,
      "accuracy_std_dev": 0.0033335000000000448,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.7708268799999999,
      "avg_duration_per_task_ms": 139666.89,
      "avg_output_tokens_per_task": 5287.85,
      "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-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "DeepSeek",
      "accuracy": 0.813333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.00614692,
      "avg_duration_per_task_ms": 141813.64,
      "avg_output_tokens_per_task": 15680.88,
      "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": "google/gemini-3.7-flash-20260813",
      "model_slug": "google/gemini-3.7-flash",
      "model_name": "Google: Gemini 3.7 Flash",
      "provider_name": "Google AI Studio",
      "accuracy": 0.806667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.20567228,
      "avg_duration_per_task_ms": 69455.26,
      "avg_output_tokens_per_task": 15737.28,
      "last_run_timestamp": "2026-08-15T00:15:11.753Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-5-fable-20260609",
      "model_slug": "anthropic/claude-fable-5",
      "model_name": "Anthropic: Claude Fable 5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.8013334,
      "accuracy_std_dev": 0.021249926537284785,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.8758712919999999,
      "avg_duration_per_task_ms": 139145.43,
      "avg_output_tokens_per_task": 5724.488,
      "last_run_timestamp": "2026-10-24T11:41:42.405Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-5-fable-20260609",
      "model_slug": "anthropic/claude-fable-5",
      "model_name": "Anthropic: Claude Fable 5",
      "provider_name": "Azure",
      "accuracy": 0.8,
      "accuracy_std_dev": 0.02326845525989678,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.8815157218181818,
      "avg_duration_per_task_ms": 126457.99454545455,
      "avg_output_tokens_per_task": 5745.552727272728,
      "last_run_timestamp": "2026-10-24T11:46:02.314Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.1-20251113",
      "model_slug": "openai/gpt-5.1",
      "model_name": "OpenAI: GPT-5.1",
      "provider_name": "auto-routing",
      "accuracy": 0.8,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.39313749000000003,
      "avg_duration_per_task_ms": 477750.16,
      "avg_output_tokens_per_task": 34082.08,
      "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": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Reka",
      "accuracy": 0.7996450638297874,
      "accuracy_std_dev": 0.011700000000000044,
      "run_count": 2,
      "total_questions": 94,
      "avg_cost_per_task_usd": 0.06506034042553191,
      "avg_duration_per_task_ms": 304476.60638297873,
      "avg_output_tokens_per_task": 17465.68085106383,
      "last_run_timestamp": "2026-10-24T19:25:11.376Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "Phala",
      "accuracy": 0.7981576231155779,
      "accuracy_std_dev": 0.02702867818632647,
      "run_count": 8,
      "total_questions": 398,
      "avg_cost_per_task_usd": 0.20028302512562812,
      "avg_duration_per_task_ms": 842159.8542713568,
      "avg_output_tokens_per_task": 34248.3391959799,
      "last_run_timestamp": "2026-10-24T19:22:19.879Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "auto-routing",
      "accuracy": 0.7975757272727273,
      "accuracy_std_dev": 0.026707871060082322,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.8921026254545454,
      "avg_duration_per_task_ms": 120870.10727272727,
      "avg_output_tokens_per_task": 5752.245454545455,
      "last_run_timestamp": "2026-10-24T11:40:21.012Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-5-fable-20260609",
      "model_slug": "anthropic/claude-fable-5",
      "model_name": "Anthropic: Claude Fable 5",
      "provider_name": "Anthropic",
      "accuracy": 0.7939394545454546,
      "accuracy_std_dev": 0.01716339453797282,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.8242855072727272,
      "avg_duration_per_task_ms": 128426.18181818182,
      "avg_output_tokens_per_task": 5739.361818181818,
      "last_run_timestamp": "2026-10-24T12:06:18.636Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "anthropic/claude-5-fable-20260609",
      "model_slug": "anthropic/claude-fable-5",
      "model_name": "Anthropic: Claude Fable 5",
      "provider_name": "Google",
      "accuracy": 0.7936365,
      "accuracy_std_dev": 0.0214125331263343,
      "run_count": 22,
      "total_questions": 1100,
      "avg_cost_per_task_usd": 0.9285239381818181,
      "avg_duration_per_task_ms": 94655,
      "avg_output_tokens_per_task": 5727.099090909091,
      "last_run_timestamp": "2026-10-24T11:52:13.306Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "inclusionai/ling-3.1-flash-20261002",
      "model_slug": "inclusionai/ling-3.1-flash",
      "model_name": "inclusionAI: Ling 3.1 Flash",
      "provider_name": "Novita",
      "accuracy": 0.793333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.00828808,
      "avg_duration_per_task_ms": 196026.56,
      "avg_output_tokens_per_task": 16303,
      "last_run_timestamp": "2026-10-14T19:21:35.384Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "inclusionai/ling-3.1-flash-20261002",
      "model_slug": "inclusionai/ling-3.1-flash",
      "model_name": "inclusionAI: Ling 3.1 Flash",
      "provider_name": "DeepInfra",
      "accuracy": 0.793333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.03146312,
      "avg_duration_per_task_ms": 636210.72,
      "avg_output_tokens_per_task": 16745.44,
      "last_run_timestamp": "2026-10-14T19:17:59.742Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "Xiaomi",
      "accuracy": 0.7920001999999999,
      "accuracy_std_dev": 0.014847359662916514,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.05966906,
      "avg_duration_per_task_ms": 918022.212,
      "avg_output_tokens_per_task": 41767.624,
      "last_run_timestamp": "2026-10-24T12:47:38.794Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "anthropic/claude-opus-5-20260723",
      "model_slug": "anthropic/claude-opus-5",
      "model_name": "Anthropic: Claude Opus 5",
      "provider_name": "auto-routing",
      "accuracy": 0.7890909090909091,
      "accuracy_std_dev": 0.021324416182798293,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.47806725999999994,
      "avg_duration_per_task_ms": 120590.92909090909,
      "avg_output_tokens_per_task": 7057.872727272727,
      "last_run_timestamp": "2026-10-24T11:55:24.959Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Crusoe",
      "accuracy": 0.7889448542713567,
      "accuracy_std_dev": 0.022421675465662684,
      "run_count": 4,
      "total_questions": 199,
      "avg_cost_per_task_usd": 0.02856220603015075,
      "avg_duration_per_task_ms": 373105.72361809044,
      "avg_output_tokens_per_task": 11144.979899497488,
      "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": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "Novita",
      "accuracy": 0.7883330000000001,
      "accuracy_std_dev": 0.02598076211353313,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.06002512,
      "avg_duration_per_task_ms": 888300.44,
      "avg_output_tokens_per_task": 41523.41,
      "last_run_timestamp": "2026-10-24T12:33:32.948Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Tenstorrent",
      "accuracy": 0.7880001999999999,
      "accuracy_std_dev": 0.03709450273773731,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.043486136,
      "avg_duration_per_task_ms": 113472.604,
      "avg_output_tokens_per_task": 4980.336,
      "last_run_timestamp": "2026-10-30T19:04:49.114Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Chutes",
      "accuracy": 0.7876542314814815,
      "accuracy_std_dev": 0.022797342804780508,
      "run_count": 11,
      "total_questions": 540,
      "avg_cost_per_task_usd": 0.07938424444444445,
      "avg_duration_per_task_ms": 1140818.088888889,
      "avg_output_tokens_per_task": 26072.82222222222,
      "last_run_timestamp": "2026-10-24T19:36:21.865Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Wafer",
      "accuracy": 0.7874881680993314,
      "accuracy_std_dev": 0.015935514274356157,
      "run_count": 21,
      "total_questions": 1047,
      "avg_cost_per_task_usd": 0.08563230085959884,
      "avg_duration_per_task_ms": 141393.38681948424,
      "avg_output_tokens_per_task": 12084.589302769818,
      "last_run_timestamp": "2026-10-29T22:52:40.244Z"
    },
    {
      "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": 0.786667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.01468154,
      "avg_duration_per_task_ms": 178223.4,
      "avg_output_tokens_per_task": 11352.52,
      "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": "google/gemini-4-argon-20261013",
      "model_slug": "google/gemini-4-argon",
      "model_name": "Google: Gemini 4 Argon",
      "provider_name": "Google",
      "accuracy": 0.786667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.49828548,
      "avg_duration_per_task_ms": 348417.68,
      "avg_output_tokens_per_task": 21971.04,
      "last_run_timestamp": "2026-10-14T15:16:05.578Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-sonnet-5.5-20260928",
      "model_slug": "anthropic/claude-sonnet-5.5",
      "model_name": "Anthropic: Claude Sonnet 5.5",
      "provider_name": "auto-routing",
      "accuracy": 0.78666675,
      "accuracy_std_dev": 0.024944159861328273,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.15998305999999998,
      "avg_duration_per_task_ms": 73714.275,
      "avg_output_tokens_per_task": 6299.305,
      "last_run_timestamp": "2026-10-24T11:26:51.227Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "Darkbloom",
      "accuracy": 0.7862320000000002,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 46,
      "avg_cost_per_task_usd": 0.029582717391304348,
      "avg_duration_per_task_ms": 2435445.304347826,
      "avg_output_tokens_per_task": 108721.58695652174,
      "last_run_timestamp": "2026-10-03T16:12:07.492Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-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": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "Alibaba",
      "accuracy": 0.78583325,
      "accuracy_std_dev": 0.02190571019021068,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.18706350749999998,
      "avg_duration_per_task_ms": 433314.44,
      "avg_output_tokens_per_task": 30794.75,
      "last_run_timestamp": "2026-10-24T18:46:07.800Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Together",
      "accuracy": 0.78583325,
      "accuracy_std_dev": 0.016811747713652493,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.07747623,
      "avg_duration_per_task_ms": 47946.0625,
      "avg_output_tokens_per_task": 4735.27,
      "last_run_timestamp": "2026-10-24T11:58:13.073Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-sonnet-5.5-20260928",
      "model_slug": "anthropic/claude-sonnet-5.5",
      "model_name": "Anthropic: Claude Sonnet 5.5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7857144285714285,
      "accuracy_std_dev": 0.013058229959434387,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.17514343714285713,
      "avg_duration_per_task_ms": 72438.17428571428,
      "avg_output_tokens_per_task": 6286.754285714285,
      "last_run_timestamp": "2026-10-29T00:13:53.374Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-sonnet-5.5-20260928",
      "model_slug": "anthropic/claude-sonnet-5.5",
      "model_name": "Anthropic: Claude Sonnet 5.5",
      "provider_name": "Azure",
      "accuracy": 0.7853333,
      "accuracy_std_dev": 0.011469724399914758,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.17582804400000004,
      "avg_duration_per_task_ms": 67798.288,
      "avg_output_tokens_per_task": 6240.81,
      "last_run_timestamp": "2026-10-24T12:30:32.152Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Fireworks",
      "accuracy": 0.7842432787356322,
      "accuracy_std_dev": 0.019450509073790118,
      "run_count": 14,
      "total_questions": 696,
      "avg_cost_per_task_usd": 0.09254919540229886,
      "avg_duration_per_task_ms": 69468.46551724138,
      "avg_output_tokens_per_task": 4697.962643678161,
      "last_run_timestamp": "2026-10-24T13:01:18.254Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "auto-routing",
      "accuracy": 0.783333375,
      "accuracy_std_dev": 0.023804813940343558,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.19070967249999998,
      "avg_duration_per_task_ms": 778974.4025,
      "avg_output_tokens_per_task": 30803.295,
      "last_run_timestamp": "2026-10-24T19:20:47.328Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.7833330000000001,
      "accuracy_std_dev": 0.009999999999999953,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.07443635,
      "avg_duration_per_task_ms": 98397.15,
      "avg_output_tokens_per_task": 21518.67,
      "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": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "DeepInfra",
      "accuracy": 0.7833330000000001,
      "accuracy_std_dev": 0.009999999999999953,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.06795213,
      "avg_duration_per_task_ms": 191712.46,
      "avg_output_tokens_per_task": 8639.07,
      "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-sonnet-5.5-20260928",
      "model_slug": "anthropic/claude-sonnet-5.5",
      "model_name": "Anthropic: Claude Sonnet 5.5",
      "provider_name": "Google",
      "accuracy": 0.7831111333333334,
      "accuracy_std_dev": 0.01851801018780244,
      "run_count": 15,
      "total_questions": 750,
      "avg_cost_per_task_usd": 0.181601536,
      "avg_duration_per_task_ms": 65069.36666666667,
      "avg_output_tokens_per_task": 6347.210666666667,
      "last_run_timestamp": "2026-10-24T12:24:09.591Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.7-flash-20260813",
      "model_slug": "google/gemini-3.7-flash",
      "model_name": "Google: Gemini 3.7 Flash",
      "provider_name": "Google",
      "accuracy": 0.7831052054794521,
      "accuracy_std_dev": 0.017155537965592597,
      "run_count": 3,
      "total_questions": 146,
      "avg_cost_per_task_usd": 0.14992966438356164,
      "avg_duration_per_task_ms": 183129.3904109589,
      "avg_output_tokens_per_task": 16313.164383561643,
      "last_run_timestamp": "2026-10-15T21:26:28.195Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Novita",
      "accuracy": 0.7826668,
      "accuracy_std_dev": 0.01717857063204037,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.01173768,
      "avg_duration_per_task_ms": 599702.892,
      "avg_output_tokens_per_task": 22568,
      "last_run_timestamp": "2026-10-24T14:27:48.875Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "DeepInfra",
      "accuracy": 0.7817936907630522,
      "accuracy_std_dev": 0.018188779316930522,
      "run_count": 5,
      "total_questions": 249,
      "avg_cost_per_task_usd": 0.08406518072289156,
      "avg_duration_per_task_ms": 255202.53012048194,
      "avg_output_tokens_per_task": 12331.429718875503,
      "last_run_timestamp": "2026-10-24T19:28:54.737Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.78166675,
      "accuracy_std_dev": 0.015183827290492351,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.014457230000000001,
      "avg_duration_per_task_ms": 401220.695,
      "avg_output_tokens_per_task": 21060.795,
      "last_run_timestamp": "2026-10-24T13:36:02.112Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Novita",
      "accuracy": 0.7809524285714285,
      "accuracy_std_dev": 0.025053077443455044,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.08778324,
      "avg_duration_per_task_ms": 269653.90285714285,
      "avg_output_tokens_per_task": 20106.945714285714,
      "last_run_timestamp": "2026-10-24T12:40:07.007Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Crusoe",
      "accuracy": 0.7809524285714285,
      "accuracy_std_dev": 0.011507741653303821,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.08243106857142857,
      "avg_duration_per_task_ms": 42328.47714285714,
      "avg_output_tokens_per_task": 4763.78,
      "last_run_timestamp": "2026-10-24T12:26:24.296Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "Alibaba",
      "accuracy": 0.7807406666666666,
      "accuracy_std_dev": 0.036709605237140126,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.0812698177777778,
      "avg_duration_per_task_ms": 207249.8177777778,
      "avg_output_tokens_per_task": 25401.606666666667,
      "last_run_timestamp": "2026-10-28T03:33:36.433Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Z.AI",
      "accuracy": 0.7807406666666666,
      "accuracy_std_dev": 0.01923075098204262,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.08175265555555557,
      "avg_duration_per_task_ms": 137863.5,
      "avg_output_tokens_per_task": 4670.331111111111,
      "last_run_timestamp": "2026-10-24T13:21:36.785Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "BaseTen",
      "accuracy": 0.7804090721205597,
      "accuracy_std_dev": 0.030493720143653286,
      "run_count": 19,
      "total_questions": 929,
      "avg_cost_per_task_usd": 0.07802858557588804,
      "avg_duration_per_task_ms": 76989.98815931109,
      "avg_output_tokens_per_task": 4254.709364908504,
      "last_run_timestamp": "2026-10-24T12:32:51.108Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "BaseTen",
      "accuracy": 0.7801419645390072,
      "accuracy_std_dev": 0.026804102476217243,
      "run_count": 3,
      "total_questions": 141,
      "avg_cost_per_task_usd": 0.09157175886524822,
      "avg_duration_per_task_ms": 67531.68794326241,
      "avg_output_tokens_per_task": 7567.7943262411345,
      "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": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "auto-routing",
      "accuracy": 0.7800001111111111,
      "accuracy_std_dev": 0.0235177999322709,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.06914864222222221,
      "avg_duration_per_task_ms": 267980.70666666667,
      "avg_output_tokens_per_task": 14178.462222222222,
      "last_run_timestamp": "2026-10-24T12:19:13.995Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "DeepSeek",
      "accuracy": 0.78,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.00686524,
      "avg_duration_per_task_ms": 150471.32,
      "avg_output_tokens_per_task": 16577.68,
      "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": "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": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "auto-routing",
      "accuracy": 0.7793333,
      "accuracy_std_dev": 0.02393506864853326,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.09867859,
      "avg_duration_per_task_ms": 301600.814,
      "avg_output_tokens_per_task": 20737.49,
      "last_run_timestamp": "2026-10-24T11:44:08.040Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Nebius",
      "accuracy": 0.7792640401337794,
      "accuracy_std_dev": 0.028913483122066055,
      "run_count": 6,
      "total_questions": 299,
      "avg_cost_per_task_usd": 0.1886598294314381,
      "avg_duration_per_task_ms": 56557.29765886287,
      "avg_output_tokens_per_task": 4653.71237458194,
      "last_run_timestamp": "2026-10-24T12:16:13.900Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "Parasail",
      "accuracy": 0.779214709770115,
      "accuracy_std_dev": 0.033260406159852735,
      "run_count": 7,
      "total_questions": 348,
      "avg_cost_per_task_usd": 0.19238710632183906,
      "avg_duration_per_task_ms": 621470.8362068966,
      "avg_output_tokens_per_task": 29821.287356321838,
      "last_run_timestamp": "2026-10-24T19:18:27.046Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-opus-5-20260723",
      "model_slug": "anthropic/claude-opus-5",
      "model_name": "Anthropic: Claude Opus 5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7783333125,
      "accuracy_std_dev": 0.02006927412650602,
      "run_count": 32,
      "total_questions": 1600,
      "avg_cost_per_task_usd": 0.49599691625,
      "avg_duration_per_task_ms": 120067.109375,
      "avg_output_tokens_per_task": 7093.541875,
      "last_run_timestamp": "2026-10-24T11:54:38.630Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Near AI",
      "accuracy": 0.77833325,
      "accuracy_std_dev": 0.011902275693643617,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.0336993,
      "avg_duration_per_task_ms": 220470.06,
      "avg_output_tokens_per_task": 12832.385,
      "last_run_timestamp": "2026-10-24T11:48:43.119Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "ModelRun",
      "accuracy": 0.77833325,
      "accuracy_std_dev": 0.005527984369324845,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.12165100000000001,
      "avg_duration_per_task_ms": 123453.655,
      "avg_output_tokens_per_task": 4597.39,
      "last_run_timestamp": "2026-10-28T05:13:02.136Z"
    },
    {
      "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": 0.7781817878787878,
      "accuracy_std_dev": 0.022993088488233407,
      "run_count": 33,
      "total_questions": 1650,
      "avg_cost_per_task_usd": 0.5197950163636365,
      "avg_duration_per_task_ms": 109495.11454545455,
      "avg_output_tokens_per_task": 7111.084848484848,
      "last_run_timestamp": "2026-10-24T11:58:58.417Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7781817272727273,
      "accuracy_std_dev": 0.03929598310743943,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.1061598909090909,
      "avg_duration_per_task_ms": 180685.46181818182,
      "avg_output_tokens_per_task": 6783.716363636364,
      "last_run_timestamp": "2026-10-24T11:43:03.112Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Ambient",
      "accuracy": 0.777778,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 48,
      "avg_cost_per_task_usd": 0.012714583333333333,
      "avg_duration_per_task_ms": 516568.7291666667,
      "avg_output_tokens_per_task": 14230.208333333334,
      "last_run_timestamp": "2026-08-03T17:20:47.207Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7771429999999999,
      "accuracy_std_dev": 0.017036713725699246,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.15415246857142859,
      "avg_duration_per_task_ms": 315516.36,
      "avg_output_tokens_per_task": 19966.505714285715,
      "last_run_timestamp": "2026-10-24T14:03:17.639Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7771428571428571,
      "accuracy_std_dev": 0.018120213341763423,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.09363316857142856,
      "avg_duration_per_task_ms": 130500.70857142858,
      "avg_output_tokens_per_task": 5167.157142857143,
      "last_run_timestamp": "2026-10-24T13:07:35.631Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-opus-5-20260723",
      "model_slug": "anthropic/claude-opus-5",
      "model_name": "Anthropic: Claude Opus 5",
      "provider_name": "Azure",
      "accuracy": 0.7769697272727273,
      "accuracy_std_dev": 0.02235858980790933,
      "run_count": 22,
      "total_questions": 1100,
      "avg_cost_per_task_usd": 0.47973416363636356,
      "avg_duration_per_task_ms": 102499.14090909091,
      "avg_output_tokens_per_task": 7159.897272727273,
      "last_run_timestamp": "2026-10-24T11:37:51.375Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Baidu",
      "accuracy": 0.77666675,
      "accuracy_std_dev": 0.015986122456290018,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.040007865,
      "avg_duration_per_task_ms": 202967.295,
      "avg_output_tokens_per_task": 13307.9975,
      "last_run_timestamp": "2026-10-24T19:48:57.247Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "Cerebras",
      "accuracy": 0.7766665,
      "accuracy_std_dev": 0.016666499999999973,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.41419341000000004,
      "avg_duration_per_task_ms": 38158.51,
      "avg_output_tokens_per_task": 9152,
      "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": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Alibaba",
      "accuracy": 0.7764583125,
      "accuracy_std_dev": 0.027168853280730192,
      "run_count": 16,
      "total_questions": 800,
      "avg_cost_per_task_usd": 0.13175464625,
      "avg_duration_per_task_ms": 318529.455,
      "avg_output_tokens_per_task": 19311.1725,
      "last_run_timestamp": "2026-10-28T04:30:12.019Z"
    },
    {
      "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": "Anthropic",
      "accuracy": 0.7764101538461539,
      "accuracy_std_dev": 0.017853329682318696,
      "run_count": 13,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.44019213846153843,
      "avg_duration_per_task_ms": 91000.5123076923,
      "avg_output_tokens_per_task": 5712.275384615385,
      "last_run_timestamp": "2026-10-24T12:19:07.891Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "StreamLake",
      "accuracy": 0.7759999000000001,
      "accuracy_std_dev": 0.02174607193242035,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.08178571999999999,
      "avg_duration_per_task_ms": 336667.87,
      "avg_output_tokens_per_task": 20194.232,
      "last_run_timestamp": "2026-10-24T12:55:53.240Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.8-opus-20260528",
      "model_slug": "anthropic/claude-opus-4.8",
      "model_name": "Anthropic: Claude Opus 4.8",
      "provider_name": "Anthropic",
      "accuracy": 0.7758974615384615,
      "accuracy_std_dev": 0.022726804434125412,
      "run_count": 13,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.508684303076923,
      "avg_duration_per_task_ms": 124701.7123076923,
      "avg_output_tokens_per_task": 8041.3538461538465,
      "last_run_timestamp": "2026-10-24T12:26:01.097Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "GMICloud",
      "accuracy": 0.7755556666666666,
      "accuracy_std_dev": 0.029313124352981113,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.07678593777777779,
      "avg_duration_per_task_ms": 340037.8177777778,
      "avg_output_tokens_per_task": 17962.597777777777,
      "last_run_timestamp": "2026-10-27T03:02:28.656Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7755444170854271,
      "accuracy_std_dev": 0.02087028870637634,
      "run_count": 8,
      "total_questions": 398,
      "avg_cost_per_task_usd": 0.15095997738693467,
      "avg_duration_per_task_ms": 341068.3417085427,
      "avg_output_tokens_per_task": 20195.442211055277,
      "last_run_timestamp": "2026-10-24T13:15:20.470Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "DeepInfra",
      "accuracy": 0.7753625384615385,
      "accuracy_std_dev": 0.017323958686038136,
      "run_count": 6,
      "total_questions": 299,
      "avg_cost_per_task_usd": 0.06710363879598663,
      "avg_duration_per_task_ms": 1125092.9698996656,
      "avg_output_tokens_per_task": 36573.67558528428,
      "last_run_timestamp": "2026-10-24T13:26:35.887Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Friendli",
      "accuracy": 0.7753334,
      "accuracy_std_dev": 0.019561963092696,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.06818830799999999,
      "avg_duration_per_task_ms": 45366.578,
      "avg_output_tokens_per_task": 4737.112,
      "last_run_timestamp": "2026-10-28T11:10:06.273Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.8-opus-20260528",
      "model_slug": "anthropic/claude-opus-4.8",
      "model_name": "Anthropic: Claude Opus 4.8",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.7753333,
      "accuracy_std_dev": 0.025307041621058755,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.48832906800000003,
      "avg_duration_per_task_ms": 125426.254,
      "avg_output_tokens_per_task": 7994.034,
      "last_run_timestamp": "2026-10-24T12:05:00.275Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-fable-5.1-20260831",
      "model_slug": "anthropic/claude-fable-5.1",
      "model_name": "Anthropic: Claude Fable 5.1",
      "provider_name": "Google",
      "accuracy": 0.7746668,
      "accuracy_std_dev": 0.012927338819726197,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.758212416,
      "avg_duration_per_task_ms": 112131.876,
      "avg_output_tokens_per_task": 6866.944,
      "last_run_timestamp": "2026-10-01T08:05:27.141Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-sonnet-5.5-20260928",
      "model_slug": "anthropic/claude-sonnet-5.5",
      "model_name": "Anthropic: Claude Sonnet 5.5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.7746666,
      "accuracy_std_dev": 0.02508632562652409,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.16812989199999998,
      "avg_duration_per_task_ms": 82939.64,
      "avg_output_tokens_per_task": 6304.564,
      "last_run_timestamp": "2026-10-24T11:50:50.507Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Cloudflare",
      "accuracy": 0.7745533125,
      "accuracy_std_dev": 0.016606913050345267,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.08764765178571429,
      "avg_duration_per_task_ms": 134355.6294642857,
      "avg_output_tokens_per_task": 4417.004464285715,
      "last_run_timestamp": "2026-10-24T12:55:13.106Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-opus-5-20260723",
      "model_slug": "anthropic/claude-opus-5",
      "model_name": "Anthropic: Claude Opus 5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.7745455454545455,
      "accuracy_std_dev": 0.01924194798002831,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.4526700363636364,
      "avg_duration_per_task_ms": 121754.2,
      "avg_output_tokens_per_task": 7111.852727272727,
      "last_run_timestamp": "2026-10-24T11:43:26.286Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.8-opus-20260528",
      "model_slug": "anthropic/claude-opus-4.8",
      "model_name": "Anthropic: Claude Opus 4.8",
      "provider_name": "Azure",
      "accuracy": 0.7745454545454545,
      "accuracy_std_dev": 0.0297895320002411,
      "run_count": 22,
      "total_questions": 1100,
      "avg_cost_per_task_usd": 0.5158095636363637,
      "avg_duration_per_task_ms": 126197.32727272727,
      "avg_output_tokens_per_task": 7983.286363636364,
      "last_run_timestamp": "2026-10-24T11:35:03.985Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "DeepInfra",
      "accuracy": 0.7742857142857142,
      "accuracy_std_dev": 0.026945723560596418,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.21623159142857146,
      "avg_duration_per_task_ms": 562306.5571428571,
      "avg_output_tokens_per_task": 12602.945714285714,
      "last_run_timestamp": "2026-10-10T11:06:43.832Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "DekaLLM",
      "accuracy": 0.7740741111111111,
      "accuracy_std_dev": 0.022098307604602994,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.05655496222222222,
      "avg_duration_per_task_ms": 155063.69555555555,
      "avg_output_tokens_per_task": 16071.293333333333,
      "last_run_timestamp": "2026-10-24T11:33:51.404Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.7-opus-20260416",
      "model_slug": "anthropic/claude-opus-4.7",
      "model_name": "Anthropic: Claude Opus 4.7",
      "provider_name": "Azure",
      "accuracy": 0.7739393636363636,
      "accuracy_std_dev": 0.019582366837506575,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.43740748545454544,
      "avg_duration_per_task_ms": 94129.65818181819,
      "avg_output_tokens_per_task": 5651.554545454545,
      "last_run_timestamp": "2026-10-24T11:36:40.498Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "DeepInfra",
      "accuracy": 0.7738693316582915,
      "accuracy_std_dev": 0.030268145815516595,
      "run_count": 16,
      "total_questions": 796,
      "avg_cost_per_task_usd": 0.03491607663316582,
      "avg_duration_per_task_ms": 506838.68467336684,
      "avg_output_tokens_per_task": 12387.751256281406,
      "last_run_timestamp": "2026-10-24T19:21:21.689Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.7736362727272728,
      "accuracy_std_dev": 0.032503472161631956,
      "run_count": 22,
      "total_questions": 1100,
      "avg_cost_per_task_usd": 0.44677050818181824,
      "avg_duration_per_task_ms": 79384.7109090909,
      "avg_output_tokens_per_task": 5652.894545454546,
      "last_run_timestamp": "2026-10-24T12:17:57.222Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Relace",
      "accuracy": 0.7733335,
      "accuracy_std_dev": 0.02708037419885479,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.13855224,
      "avg_duration_per_task_ms": 310185.72,
      "avg_output_tokens_per_task": 21233.035,
      "last_run_timestamp": "2026-10-24T12:39:23.138Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Reka",
      "accuracy": 0.7733335,
      "accuracy_std_dev": 0.00666650000000002,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.11407636,
      "avg_duration_per_task_ms": 360681.75,
      "avg_output_tokens_per_task": 19997.86,
      "last_run_timestamp": "2026-10-01T22:40:52.402Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7733335,
      "accuracy_std_dev": 0.013333499999999998,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01881542,
      "avg_duration_per_task_ms": 96553.18,
      "avg_output_tokens_per_task": 11186.19,
      "last_run_timestamp": "2026-10-24T11:34:09.882Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Io Net",
      "accuracy": 0.7733335,
      "accuracy_std_dev": 0.013333499999999998,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.10867763,
      "avg_duration_per_task_ms": 521628.03,
      "avg_output_tokens_per_task": 12092.47,
      "last_run_timestamp": "2026-09-19T08:11:59.798Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "DeepInfra",
      "accuracy": 0.7733335,
      "accuracy_std_dev": 0.03126950946609175,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.068859715,
      "avg_duration_per_task_ms": 704699.005,
      "avg_output_tokens_per_task": 13178.825,
      "last_run_timestamp": "2026-10-24T12:26:26.393Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-sonnet-5.5-20260928",
      "model_slug": "anthropic/claude-sonnet-5.5",
      "model_name": "Anthropic: Claude Sonnet 5.5",
      "provider_name": "Anthropic",
      "accuracy": 0.7733334,
      "accuracy_std_dev": 0.013984054413509715,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.17623406799999997,
      "avg_duration_per_task_ms": 74817.192,
      "avg_output_tokens_per_task": 6393.728,
      "last_run_timestamp": "2026-10-24T11:36:43.533Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Baidu",
      "accuracy": 0.7733332857142857,
      "accuracy_std_dev": 0.022537305934283946,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.08668051142857143,
      "avg_duration_per_task_ms": 364495.4428571428,
      "avg_output_tokens_per_task": 20483.94,
      "last_run_timestamp": "2026-10-24T12:45:56.375Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Reka",
      "accuracy": 0.773333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.012345539999999999,
      "avg_duration_per_task_ms": 657478.84,
      "avg_output_tokens_per_task": 14196.68,
      "last_run_timestamp": "2026-10-03T00:00:06.435Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "CoreWeave",
      "accuracy": 0.773333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.16341722,
      "avg_duration_per_task_ms": 174236.4,
      "avg_output_tokens_per_task": 10664.8,
      "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-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "DeepInfra",
      "accuracy": 0.773333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.0051899,
      "avg_duration_per_task_ms": 123477.22,
      "avg_output_tokens_per_task": 11624.46,
      "last_run_timestamp": "2026-10-20T20:17:11.291Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "GMICloud",
      "accuracy": 0.773333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.26629643999999997,
      "avg_duration_per_task_ms": 438741.58,
      "avg_output_tokens_per_task": 27876.64,
      "last_run_timestamp": "2026-08-29T16:58:30.455Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "Makora",
      "accuracy": 0.773333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.11485508,
      "avg_duration_per_task_ms": 381375.48,
      "avg_output_tokens_per_task": 14199.1,
      "last_run_timestamp": "2026-10-13T03:50:19.636Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "StreamLake",
      "accuracy": 0.773333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.057549260000000005,
      "avg_duration_per_task_ms": 140710.34,
      "avg_output_tokens_per_task": 4743.94,
      "last_run_timestamp": "2026-09-11T04:09:34.110Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.7727272727272727,
      "accuracy_std_dev": 0.028632777073235846,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.1722844727272727,
      "avg_duration_per_task_ms": 588293.6218181818,
      "avg_output_tokens_per_task": 36404.07818181818,
      "last_run_timestamp": "2026-10-24T18:49:47.645Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Parasail",
      "accuracy": 0.7726668,
      "accuracy_std_dev": 0.0325508757356849,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.065897038,
      "avg_duration_per_task_ms": 245761.102,
      "avg_output_tokens_per_task": 11799.874,
      "last_run_timestamp": "2026-10-24T12:58:13.946Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "ModelRun",
      "accuracy": 0.7723808571428572,
      "accuracy_std_dev": 0.03175867478536295,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.0666134542857143,
      "avg_duration_per_task_ms": 68063.51142857142,
      "avg_output_tokens_per_task": 12368.354285714286,
      "last_run_timestamp": "2026-10-28T04:52:11.699Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "DeepInfra",
      "accuracy": 0.7723808571428572,
      "accuracy_std_dev": 0.03553283429420726,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.0410159,
      "avg_duration_per_task_ms": 732886.1685714286,
      "avg_output_tokens_per_task": 12629.145714285714,
      "last_run_timestamp": "2026-10-24T12:41:08.943Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Cerebras",
      "accuracy": 0.7722223333333333,
      "accuracy_std_dev": 0.051949695516164705,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.2672586,
      "avg_duration_per_task_ms": 26529.913333333334,
      "avg_output_tokens_per_task": 12210.046666666667,
      "last_run_timestamp": "2026-10-24T12:34:49.317Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta/muse-glimmer-30b-20260810",
      "model_slug": "meta/muse-glimmer-30b",
      "model_name": "Meta: Muse Glimmer 30B",
      "provider_name": "DeepInfra",
      "accuracy": 0.7722222222222223,
      "accuracy_std_dev": 0.03095202181433487,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.03593436666666667,
      "avg_duration_per_task_ms": 446735.1666666667,
      "avg_output_tokens_per_task": 13815.786666666667,
      "last_run_timestamp": "2026-10-08T22:42:47.098Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-fable-5.1-20260831",
      "model_slug": "anthropic/claude-fable-5.1",
      "model_name": "Anthropic: Claude Fable 5.1",
      "provider_name": "Azure",
      "accuracy": 0.772,
      "accuracy_std_dev": 0.02679938639596063,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.808827636,
      "avg_duration_per_task_ms": 183544.5,
      "avg_output_tokens_per_task": 6940.7,
      "last_run_timestamp": "2026-10-01T08:03:00.737Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "DeepInfra",
      "accuracy": 0.771929874493927,
      "accuracy_std_dev": 0.030837182396580943,
      "run_count": 5,
      "total_questions": 247,
      "avg_cost_per_task_usd": 0.011004716599190284,
      "avg_duration_per_task_ms": 1072772.7165991904,
      "avg_output_tokens_per_task": 18798.943319838058,
      "last_run_timestamp": "2026-10-17T12:03:02.755Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Novita",
      "accuracy": 0.7718518888888889,
      "accuracy_std_dev": 0.02102986445480936,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.05397537555555555,
      "avg_duration_per_task_ms": 110633.31777777777,
      "avg_output_tokens_per_task": 4708.497777777778,
      "last_run_timestamp": "2026-10-24T13:04:30.805Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Venice",
      "accuracy": 0.7714285714285715,
      "accuracy_std_dev": 0.02435571983191243,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.21782835142857143,
      "avg_duration_per_task_ms": 377656.6857142857,
      "avg_output_tokens_per_task": 17802.317142857144,
      "last_run_timestamp": "2026-10-24T12:37:29.521Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "DeepInfra",
      "accuracy": 0.7714285714285715,
      "accuracy_std_dev": 0.023289574128825135,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.044844714285714286,
      "avg_duration_per_task_ms": 100224.04857142858,
      "avg_output_tokens_per_task": 4956.551428571429,
      "last_run_timestamp": "2026-10-24T12:38:46.757Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7714284285714286,
      "accuracy_std_dev": 0.043456125344526336,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.12002842571428574,
      "avg_duration_per_task_ms": 270068.88285714283,
      "avg_output_tokens_per_task": 11820.128571428571,
      "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": "Chutes",
      "accuracy": 0.7713333,
      "accuracy_std_dev": 0.0351567708046402,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.060528674000000005,
      "avg_duration_per_task_ms": 302911.6,
      "avg_output_tokens_per_task": 12614.202,
      "last_run_timestamp": "2026-10-24T12:18:13.146Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.5-opus-20251124",
      "model_slug": "anthropic/claude-opus-4.5",
      "model_name": "Anthropic: Claude Opus 4.5",
      "provider_name": "Anthropic",
      "accuracy": 0.7712821538461538,
      "accuracy_std_dev": 0.029045614221682316,
      "run_count": 13,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.5353010338461538,
      "avg_duration_per_task_ms": 182572.49076923076,
      "avg_output_tokens_per_task": 10548.853846153846,
      "last_run_timestamp": "2026-10-24T12:18:32.884Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-122b-a10b-20260224",
      "model_slug": "qwen/qwen3.5-122b-a10b",
      "model_name": "Qwen: Qwen3.5-122B-A10B",
      "provider_name": "Novita",
      "accuracy": 0.771253766972477,
      "accuracy_std_dev": 0.0197324823544269,
      "run_count": 11,
      "total_questions": 545,
      "avg_cost_per_task_usd": 0.20062177064220185,
      "avg_duration_per_task_ms": 536809.0183486239,
      "avg_output_tokens_per_task": 28063.546788990825,
      "last_run_timestamp": "2026-10-24T19:14:04.604Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Alibaba",
      "accuracy": 0.7711113333333333,
      "accuracy_std_dev": 0.021999034771755084,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.029255286666666668,
      "avg_duration_per_task_ms": 133490.13333333333,
      "avg_output_tokens_per_task": 12074.193333333333,
      "last_run_timestamp": "2026-10-28T03:37:31.074Z"
    },
    {
      "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": 0.7711112499999999,
      "accuracy_std_dev": 0.014989728995354355,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.13635980666666667,
      "avg_duration_per_task_ms": 232657.155,
      "avg_output_tokens_per_task": 24827.668333333335,
      "last_run_timestamp": "2026-10-29T23:10:20.098Z"
    },
    {
      "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": 0.7709090909090909,
      "accuracy_std_dev": 0.02562719156909333,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.47683379272727267,
      "avg_duration_per_task_ms": 109957.77818181818,
      "avg_output_tokens_per_task": 7177.963636363636,
      "last_run_timestamp": "2026-10-24T11:37:27.338Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "anthropic/claude-4.8-opus-20260528",
      "model_slug": "anthropic/claude-opus-4.8",
      "model_name": "Anthropic: Claude Opus 4.8",
      "provider_name": "Google",
      "accuracy": 0.7707070606060606,
      "accuracy_std_dev": 0.02775262413138882,
      "run_count": 33,
      "total_questions": 1650,
      "avg_cost_per_task_usd": 0.5647490236363636,
      "avg_duration_per_task_ms": 115196.03454545455,
      "avg_output_tokens_per_task": 8110.365454545455,
      "last_run_timestamp": "2026-10-24T12:27:09.155Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta/muse-glimmer-30b-20260810",
      "model_slug": "meta/muse-glimmer-30b",
      "model_name": "Meta: Muse Glimmer 30B",
      "provider_name": "Parasail",
      "accuracy": 0.7706935637583894,
      "accuracy_std_dev": 0.013644816256244217,
      "run_count": 3,
      "total_questions": 149,
      "avg_cost_per_task_usd": 0.09587820805369127,
      "avg_duration_per_task_ms": 288021.6644295302,
      "avg_output_tokens_per_task": 13114.979865771813,
      "last_run_timestamp": "2026-08-29T08:48:12.957Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "NextBit",
      "accuracy": 0.7706666,
      "accuracy_std_dev": 0.0204828947622156,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.10916058,
      "avg_duration_per_task_ms": 258128.42,
      "avg_output_tokens_per_task": 11359.708,
      "last_run_timestamp": "2026-10-03T16:07:05.245Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Parasail",
      "accuracy": 0.770416625,
      "accuracy_std_dev": 0.0211762125800242,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.079187595,
      "avg_duration_per_task_ms": 95610.7625,
      "avg_output_tokens_per_task": 4413.6475,
      "last_run_timestamp": "2026-10-24T13:02:59.051Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.7703030909090909,
      "accuracy_std_dev": 0.021151618478604965,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.5374423745454546,
      "avg_duration_per_task_ms": 175150.91636363635,
      "avg_output_tokens_per_task": 10708.305454545454,
      "last_run_timestamp": "2026-10-24T11:55:03.241Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Parasail",
      "accuracy": 0.77,
      "accuracy_std_dev": 0.027688565631863254,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.021299527500000002,
      "avg_duration_per_task_ms": 231325.375,
      "avg_output_tokens_per_task": 12153.2125,
      "last_run_timestamp": "2026-10-24T19:11:05.620Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Relace",
      "accuracy": 0.77,
      "accuracy_std_dev": 0.018855956915521412,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.065556992,
      "avg_duration_per_task_ms": 170454.924,
      "avg_output_tokens_per_task": 10274.532,
      "last_run_timestamp": "2026-10-24T19:27:30.834Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.77,
      "accuracy_std_dev": 0.0033330000000000304,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.35129818,
      "avg_duration_per_task_ms": 177012.35,
      "avg_output_tokens_per_task": 12679.64,
      "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": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "StreamLake",
      "accuracy": 0.77,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.05546486,
      "avg_duration_per_task_ms": 677432.76,
      "avg_output_tokens_per_task": 33432.56,
      "last_run_timestamp": "2026-10-24T13:08:41.848Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Inceptron",
      "accuracy": 0.77,
      "accuracy_std_dev": 0.022852121225391757,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.047772247500000004,
      "avg_duration_per_task_ms": 117312.3025,
      "avg_output_tokens_per_task": 4701.9325,
      "last_run_timestamp": "2026-10-24T12:40:31.665Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7699997500000001,
      "accuracy_std_dev": 0.029999861111136825,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.08088121,
      "avg_duration_per_task_ms": 203815.155,
      "avg_output_tokens_per_task": 9386.985,
      "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": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Alibaba",
      "accuracy": 0.7697777333333333,
      "accuracy_std_dev": 0.019147228922106605,
      "run_count": 15,
      "total_questions": 750,
      "avg_cost_per_task_usd": 0.12604373866666665,
      "avg_duration_per_task_ms": 93167.28,
      "avg_output_tokens_per_task": 4490.572,
      "last_run_timestamp": "2026-10-28T04:16:35.871Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Phala",
      "accuracy": 0.7696297777777777,
      "accuracy_std_dev": 0.028477632067516422,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.1409867,
      "avg_duration_per_task_ms": 323727.4288888889,
      "avg_output_tokens_per_task": 20180.546666666665,
      "last_run_timestamp": "2026-10-24T12:56:30.925Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "StreamLake",
      "accuracy": 0.7696296666666667,
      "accuracy_std_dev": 0.03639449000421171,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.01613732888888889,
      "avg_duration_per_task_ms": 298156.97555555555,
      "avg_output_tokens_per_task": 11483.206666666667,
      "last_run_timestamp": "2026-10-24T19:22:55.840Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Mistral",
      "accuracy": 0.7696296296296297,
      "accuracy_std_dev": 0.0190587770392855,
      "run_count": 27,
      "total_questions": 1350,
      "avg_cost_per_task_usd": 0.06694535481481483,
      "avg_duration_per_task_ms": 71687.36444444445,
      "avg_output_tokens_per_task": 4656.992592592593,
      "last_run_timestamp": "2026-10-30T16:33:36.217Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Venice",
      "accuracy": 0.7695238571428571,
      "accuracy_std_dev": 0.029354336946199627,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.12993965428571427,
      "avg_duration_per_task_ms": 178407.48,
      "avg_output_tokens_per_task": 8114.634285714285,
      "last_run_timestamp": "2026-10-17T12:47:59.459Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Reka",
      "accuracy": 0.7695237142857143,
      "accuracy_std_dev": 0.02386662296246901,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.11461242285714283,
      "avg_duration_per_task_ms": 472929.0742857143,
      "avg_output_tokens_per_task": 11699.722857142857,
      "last_run_timestamp": "2026-10-23T07:26:58.509Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "AkashML",
      "accuracy": 0.7691498636363635,
      "accuracy_std_dev": 0.024155174859563606,
      "run_count": 16,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.05303868308080808,
      "avg_duration_per_task_ms": 267537.58207070705,
      "avg_output_tokens_per_task": 11036.6148989899,
      "last_run_timestamp": "2026-10-30T00:40:23.011Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "DekaLLM",
      "accuracy": 0.769144,
      "accuracy_std_dev": 0.05227625153611027,
      "run_count": 3,
      "total_questions": 148,
      "avg_cost_per_task_usd": 0.03692847297297297,
      "avg_duration_per_task_ms": 119447.77702702703,
      "avg_output_tokens_per_task": 4304.4121621621625,
      "last_run_timestamp": "2026-10-24T12:53:05.664Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Modal",
      "accuracy": 0.7688888888888888,
      "accuracy_std_dev": 0.02914421358022683,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.08110119333333334,
      "avg_duration_per_task_ms": 48308.12,
      "avg_output_tokens_per_task": 4858.82,
      "last_run_timestamp": "2026-10-24T12:49:25.219Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Sail Research",
      "accuracy": 0.7688235294117647,
      "accuracy_std_dev": 0.026394292226378947,
      "run_count": 17,
      "total_questions": 850,
      "avg_cost_per_task_usd": 0.04698485764705882,
      "avg_duration_per_task_ms": 91736.21058823529,
      "avg_output_tokens_per_task": 4343.287058823529,
      "last_run_timestamp": "2026-10-24T12:48:21.011Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "auto-routing",
      "accuracy": 0.768333375,
      "accuracy_std_dev": 0.032787198976649036,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.05900729999999999,
      "avg_duration_per_task_ms": 114247.045,
      "avg_output_tokens_per_task": 6145.8525,
      "last_run_timestamp": "2026-10-24T12:32:52.472Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "auto-routing",
      "accuracy": 0.76833325,
      "accuracy_std_dev": 0.023746338900712666,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.05045579,
      "avg_duration_per_task_ms": 229267.9725,
      "avg_output_tokens_per_task": 12600.1625,
      "last_run_timestamp": "2026-10-24T19:41:33.487Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.768080787878788,
      "accuracy_std_dev": 0.02061397303207526,
      "run_count": 33,
      "total_questions": 1650,
      "avg_cost_per_task_usd": 0.5330831266666668,
      "avg_duration_per_task_ms": 119035.52545454545,
      "avg_output_tokens_per_task": 8065.8436363636365,
      "last_run_timestamp": "2026-10-24T12:05:52.446Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "CoreWeave",
      "accuracy": 0.7680176486486487,
      "accuracy_std_dev": 0.018065309426264085,
      "run_count": 3,
      "total_questions": 148,
      "avg_cost_per_task_usd": 0.06668277702702702,
      "avg_duration_per_task_ms": 92879.05405405405,
      "avg_output_tokens_per_task": 4281.108108108108,
      "last_run_timestamp": "2026-10-24T13:05:47.020Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Cloudflare",
      "accuracy": 0.768,
      "accuracy_std_dev": 0.028875333573830795,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.15511289399999997,
      "avg_duration_per_task_ms": 603929.03,
      "avg_output_tokens_per_task": 12401.186,
      "last_run_timestamp": "2026-10-24T12:59:12.979Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Relace",
      "accuracy": 0.768,
      "accuracy_std_dev": 0.011469922022402766,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.030947572,
      "avg_duration_per_task_ms": 58222.324,
      "avg_output_tokens_per_task": 4803.824,
      "last_run_timestamp": "2026-10-24T12:19:36.645Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "auto-routing",
      "accuracy": 0.7679998,
      "accuracy_std_dev": 0.04721585044622197,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.07315105200000001,
      "avg_duration_per_task_ms": 1015313.192,
      "avg_output_tokens_per_task": 46954.436,
      "last_run_timestamp": "2026-10-24T13:12:06.662Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.767879,
      "accuracy_std_dev": 0.0204628104530234,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.49366084363636353,
      "avg_duration_per_task_ms": 125790.01454545454,
      "avg_output_tokens_per_task": 8065.552727272728,
      "last_run_timestamp": "2026-10-24T11:39:26.193Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Phala",
      "accuracy": 0.767500125,
      "accuracy_std_dev": 0.01839017891727471,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.05273846000000001,
      "avg_duration_per_task_ms": 114111.54,
      "avg_output_tokens_per_task": 4459.3875,
      "last_run_timestamp": "2026-10-24T13:02:56.928Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Novita",
      "accuracy": 0.7675,
      "accuracy_std_dev": 0.018085537502380192,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.1069855275,
      "avg_duration_per_task_ms": 249511.1875,
      "avg_output_tokens_per_task": 18702.1275,
      "last_run_timestamp": "2026-10-24T19:31:34.515Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Decart",
      "accuracy": 0.7675,
      "accuracy_std_dev": 0.019561171859068173,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0578507225,
      "avg_duration_per_task_ms": 50974.42,
      "avg_output_tokens_per_task": 4677.1725,
      "last_run_timestamp": "2026-10-24T12:36:03.010Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.7672728181818181,
      "accuracy_std_dev": 0.023690757149795346,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.41836188,
      "avg_duration_per_task_ms": 93044.15818181819,
      "avg_output_tokens_per_task": 5611.463636363636,
      "last_run_timestamp": "2026-10-24T12:23:35.338Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "amazon/nova-micro-v1",
      "model_slug": "amazon/nova-micro-v1",
      "model_name": "Amazon: Nova Micro 1.0",
      "provider_name": "auto-routing",
      "accuracy": 0.7666669999999999,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 1.2750355,
      "avg_duration_per_task_ms": 98712.99,
      "avg_output_tokens_per_task": 5708.24,
      "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": "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-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "ModelRun",
      "accuracy": 0.7666669999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.01300872,
      "avg_duration_per_task_ms": 90405.02,
      "avg_output_tokens_per_task": 13808.24,
      "last_run_timestamp": "2026-10-29T03:35:28.909Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta/muse-spark-open-20261015",
      "model_slug": "meta/muse-spark-open",
      "model_name": "Meta: Muse Spark Open",
      "provider_name": "Modal",
      "accuracy": 0.7666669999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.30783086,
      "avg_duration_per_task_ms": 74384.36,
      "avg_output_tokens_per_task": 8511.9,
      "last_run_timestamp": "2026-10-24T11:55:43.315Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7666669999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.06503968,
      "avg_duration_per_task_ms": 307621.28,
      "avg_output_tokens_per_task": 13420.2,
      "last_run_timestamp": "2026-10-27T04:07:26.250Z"
    },
    {
      "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": "google/gemini-3.6-flash-20260721",
      "model_slug": "google/gemini-3.6-flash",
      "model_name": "Google: Gemini 3.6 Flash",
      "provider_name": "Google",
      "accuracy": 0.7666668,
      "accuracy_std_dev": 0.007302967435227963,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.220517392,
      "avg_duration_per_task_ms": 158612.068,
      "avg_output_tokens_per_task": 23228.836,
      "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": "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": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "GMICloud",
      "accuracy": 0.76666675,
      "accuracy_std_dev": 0.0141420177728463,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.01053462,
      "avg_duration_per_task_ms": 598882.16,
      "avg_output_tokens_per_task": 21582.78,
      "last_run_timestamp": "2026-10-27T03:23:59.523Z"
    },
    {
      "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": 0.7666666666666667,
      "accuracy_std_dev": 0.009428090415820642,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.08312547333333332,
      "avg_duration_per_task_ms": 156615.94,
      "avg_output_tokens_per_task": 8699.633333333333,
      "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": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Friendli",
      "accuracy": 0.766666625,
      "accuracy_std_dev": 0.017320700526663898,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.029551312500000003,
      "avg_duration_per_task_ms": 128791.4,
      "avg_output_tokens_per_task": 12071.38,
      "last_run_timestamp": "2026-10-28T10:41:11.460Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.7-opus-20260416",
      "model_slug": "anthropic/claude-opus-4.7",
      "model_name": "Anthropic: Claude Opus 4.7",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.7666666,
      "accuracy_std_dev": 0.03169302061400902,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.435956584,
      "avg_duration_per_task_ms": 91965.526,
      "avg_output_tokens_per_task": 5586.83,
      "last_run_timestamp": "2026-10-24T11:44:03.067Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.8-flash-20260902",
      "model_slug": "google/gemini-3.8-flash",
      "model_name": "Google: Gemini 3.8 Flash",
      "provider_name": "Google",
      "accuracy": 0.7666665,
      "accuracy_std_dev": 0.013333499999999998,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.27949224,
      "avg_duration_per_task_ms": 363191.33,
      "avg_output_tokens_per_task": 35460.86,
      "last_run_timestamp": "2026-10-15T21:31:19.497Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "ParetoAlpha",
      "accuracy": 0.7666665,
      "accuracy_std_dev": 0.00666650000000002,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.29799199,
      "avg_duration_per_task_ms": 1675038.6,
      "avg_output_tokens_per_task": 11020.82,
      "last_run_timestamp": "2026-10-30T07:12:44.253Z"
    },
    {
      "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": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Venice",
      "accuracy": 0.7658335,
      "accuracy_std_dev": 0.028369961618761493,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.1399948425,
      "avg_duration_per_task_ms": 291221.6025,
      "avg_output_tokens_per_task": 11931.6425,
      "last_run_timestamp": "2026-10-24T19:21:41.094Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "DekaLLM",
      "accuracy": 0.7658335,
      "accuracy_std_dev": 0.029095183832552104,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.11472753,
      "avg_duration_per_task_ms": 363943.795,
      "avg_output_tokens_per_task": 19034.395,
      "last_run_timestamp": "2026-10-24T12:25:43.493Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "AtlasCloud",
      "accuracy": 0.76583325,
      "accuracy_std_dev": 0.029142835503559023,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.10458012499999998,
      "avg_duration_per_task_ms": 230385.535,
      "avg_output_tokens_per_task": 11203.13,
      "last_run_timestamp": "2026-10-24T19:49:41.079Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7657141428571429,
      "accuracy_std_dev": 0.02265784142050462,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.01868964,
      "avg_duration_per_task_ms": 296461.9485714286,
      "avg_output_tokens_per_task": 14208.894285714287,
      "last_run_timestamp": "2026-10-24T13:09:04.430Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "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": "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": 0.7655204498567336,
      "accuracy_std_dev": 0.018269240228608635,
      "run_count": 7,
      "total_questions": 349,
      "avg_cost_per_task_usd": 0.012520114613180514,
      "avg_duration_per_task_ms": 376237.81088825216,
      "avg_output_tokens_per_task": 17341.69340974212,
      "last_run_timestamp": "2026-10-27T15:46:29.407Z"
    },
    {
      "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": 0.7654902352941176,
      "accuracy_std_dev": 0.022548186516434778,
      "run_count": 34,
      "total_questions": 1700,
      "avg_cost_per_task_usd": 0.5963552552941177,
      "avg_duration_per_task_ms": 71890.29588235295,
      "avg_output_tokens_per_task": 5710.195294117647,
      "last_run_timestamp": "2026-10-24T12:22:02.797Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Morph",
      "accuracy": 0.7654544545454546,
      "accuracy_std_dev": 0.0392678247038871,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.09729278181818182,
      "avg_duration_per_task_ms": 320080.3927272727,
      "avg_output_tokens_per_task": 16582.44,
      "last_run_timestamp": "2026-10-24T19:50:25.118Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.010908814536340851,
      "avg_duration_per_task_ms": 294625.86215538846,
      "avg_output_tokens_per_task": 12715.822055137845,
      "last_run_timestamp": "2026-10-24T19:59:46.522Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Z.AI",
      "accuracy": 0.7651851111111111,
      "accuracy_std_dev": 0.030107298743897986,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.009121904444444445,
      "avg_duration_per_task_ms": 193621.67333333334,
      "avg_output_tokens_per_task": 4487.197777777777,
      "last_run_timestamp": "2026-10-24T13:10:16.181Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Makora",
      "accuracy": 0.7651515252525253,
      "accuracy_std_dev": 0.02642279772753065,
      "run_count": 8,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.07224725252525253,
      "avg_duration_per_task_ms": 56760.18939393939,
      "avg_output_tokens_per_task": 4516.429292929293,
      "last_run_timestamp": "2026-10-24T12:38:39.898Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Ionstream",
      "accuracy": 0.765,
      "accuracy_std_dev": 0.03840572873934304,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.08522504,
      "avg_duration_per_task_ms": 558480.23,
      "avg_output_tokens_per_task": 11022.365,
      "last_run_timestamp": "2026-08-29T17:19:42.622Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "Tencent",
      "accuracy": 0.765,
      "accuracy_std_dev": 0.02682235800596212,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.093536965,
      "avg_duration_per_task_ms": 395836.625,
      "avg_output_tokens_per_task": 14656.925,
      "last_run_timestamp": "2026-10-24T12:32:30.905Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7648483636363637,
      "accuracy_std_dev": 0.02528071162430767,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.021531063636363635,
      "avg_duration_per_task_ms": 268995.7272727273,
      "avg_output_tokens_per_task": 15699.76909090909,
      "last_run_timestamp": "2026-10-24T12:30:46.104Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Ionstream",
      "accuracy": 0.7645742242152466,
      "accuracy_std_dev": 0.03220926464716497,
      "run_count": 9,
      "total_questions": 446,
      "avg_cost_per_task_usd": 0.05124673542600897,
      "avg_duration_per_task_ms": 396898.4529147982,
      "avg_output_tokens_per_task": 12964.300448430493,
      "last_run_timestamp": "2026-10-24T12:42:05.792Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.7645455,
      "accuracy_std_dev": 0.032559987304205146,
      "run_count": 22,
      "total_questions": 1100,
      "avg_cost_per_task_usd": 0.5562211936363636,
      "avg_duration_per_task_ms": 182577.9390909091,
      "avg_output_tokens_per_task": 10629.811818181819,
      "last_run_timestamp": "2026-10-24T12:21:43.024Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "GMICloud",
      "accuracy": 0.7644446666666667,
      "accuracy_std_dev": 0.003142853940113798,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.01656822,
      "avg_duration_per_task_ms": 126283.51333333334,
      "avg_output_tokens_per_task": 12553.306666666667,
      "last_run_timestamp": "2026-10-27T00:46:05.392Z"
    },
    {
      "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": 0.7644445555555556,
      "accuracy_std_dev": 0.027216570843314122,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.04539358888888889,
      "avg_duration_per_task_ms": 189556.9111111111,
      "avg_output_tokens_per_task": 9729.92888888889,
      "last_run_timestamp": "2026-10-24T19:32:27.920Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-flash-20260826",
      "model_slug": "qwen/qwen3.8-flash",
      "model_name": "Qwen: Qwen3.8 Flash",
      "provider_name": "Alibaba",
      "accuracy": 0.7644443333333334,
      "accuracy_std_dev": 0.04565023765059233,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.01549936,
      "avg_duration_per_task_ms": 297786.8,
      "avg_output_tokens_per_task": 16565.086666666666,
      "last_run_timestamp": "2026-10-28T04:10:41.154Z"
    },
    {
      "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": 0.7641668749999999,
      "accuracy_std_dev": 0.02623356496188377,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.334670415,
      "avg_duration_per_task_ms": 656088.585,
      "avg_output_tokens_per_task": 28108.1825,
      "last_run_timestamp": "2026-10-24T18:52:57.046Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Cohere",
      "accuracy": 0.7636566923076923,
      "accuracy_std_dev": 0.016849090853317095,
      "run_count": 6,
      "total_questions": 299,
      "avg_cost_per_task_usd": 0.028887086956521736,
      "avg_duration_per_task_ms": 106220.35785953177,
      "avg_output_tokens_per_task": 18019.585284280936,
      "last_run_timestamp": "2026-10-24T11:41:59.518Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Alibaba",
      "accuracy": 0.7636092192393736,
      "accuracy_std_dev": 0.031035672419704556,
      "run_count": 9,
      "total_questions": 447,
      "avg_cost_per_task_usd": 0.11008267114093959,
      "avg_duration_per_task_ms": 162725.5794183445,
      "avg_output_tokens_per_task": 9299.664429530201,
      "last_run_timestamp": "2026-10-28T04:07:24.080Z"
    },
    {
      "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": 0.7633820364963503,
      "accuracy_std_dev": 0.022249781810054704,
      "run_count": 11,
      "total_questions": 548,
      "avg_cost_per_task_usd": 0.09270584489051097,
      "avg_duration_per_task_ms": 50291.538321167885,
      "avg_output_tokens_per_task": 4353.293795620438,
      "last_run_timestamp": "2026-10-29T23:26:26.346Z"
    },
    {
      "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": 0.7633335,
      "accuracy_std_dev": 0.0033334999999999892,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.021769780000000002,
      "avg_duration_per_task_ms": 1044368.78,
      "avg_output_tokens_per_task": 12563.69,
      "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": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "GMICloud",
      "accuracy": 0.7633334166666667,
      "accuracy_std_dev": 0.02501853017084981,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.021799716666666667,
      "avg_duration_per_task_ms": 153361.12,
      "avg_output_tokens_per_task": 16488.393333333333,
      "last_run_timestamp": "2026-10-27T00:49:28.181Z"
    },
    {
      "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": 0.7633333333333333,
      "accuracy_std_dev": 0.017105316480231798,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.15788212333333335,
      "avg_duration_per_task_ms": 383915.8333333333,
      "avg_output_tokens_per_task": 18570.623333333333,
      "last_run_timestamp": "2026-10-24T20:11:33.520Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Sail Research",
      "accuracy": 0.7633332777777778,
      "accuracy_std_dev": 0.01808766376046748,
      "run_count": 18,
      "total_questions": 900,
      "avg_cost_per_task_usd": 0.010060852222222218,
      "avg_duration_per_task_ms": 327666.52,
      "avg_output_tokens_per_task": 14799.732222222223,
      "last_run_timestamp": "2026-10-24T13:02:06.972Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "CoreWeave",
      "accuracy": 0.7633332500000001,
      "accuracy_std_dev": 0.03605539719220273,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.157135485,
      "avg_duration_per_task_ms": 411704.8,
      "avg_output_tokens_per_task": 26150.13,
      "last_run_timestamp": "2026-08-29T08:26:04.467Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Xiaomi",
      "accuracy": 0.7633332500000001,
      "accuracy_std_dev": 0.013743584364622665,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.01189864,
      "avg_duration_per_task_ms": 642869.1,
      "avg_output_tokens_per_task": 23140.98,
      "last_run_timestamp": "2026-10-24T13:28:08.470Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Novita",
      "accuracy": 0.762963,
      "accuracy_std_dev": 0.025745351040959277,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.08904649555555556,
      "avg_duration_per_task_ms": 300656.5133333333,
      "avg_output_tokens_per_task": 12058.504444444445,
      "last_run_timestamp": "2026-10-24T12:17:31.834Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7629628888888889,
      "accuracy_std_dev": 0.02354127566374736,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.01074868,
      "avg_duration_per_task_ms": 1059485.7177777777,
      "avg_output_tokens_per_task": 13128.448888888888,
      "last_run_timestamp": "2026-10-24T19:49:45.575Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Novita",
      "accuracy": 0.7629628888888889,
      "accuracy_std_dev": 0.031128845034949126,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.008403426666666667,
      "avg_duration_per_task_ms": 205005.57333333333,
      "avg_output_tokens_per_task": 4237.573333333334,
      "last_run_timestamp": "2026-10-24T13:15:23.369Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Together",
      "accuracy": 0.7629022732240437,
      "accuracy_std_dev": 0.028478290118661355,
      "run_count": 11,
      "total_questions": 549,
      "avg_cost_per_task_usd": 0.01388206375227687,
      "avg_duration_per_task_ms": 226746.868852459,
      "avg_output_tokens_per_task": 15915.701275045538,
      "last_run_timestamp": "2026-10-17T11:00:48.267Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "NextBit",
      "accuracy": 0.7628571428571429,
      "accuracy_std_dev": 0.023328310442945685,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.11986269428571429,
      "avg_duration_per_task_ms": 415507.14,
      "avg_output_tokens_per_task": 18962.734285714287,
      "last_run_timestamp": "2026-10-03T08:33:31.578Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "GMICloud",
      "accuracy": 0.7627778333333333,
      "accuracy_std_dev": 0.03313374040368652,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.07174823666666666,
      "avg_duration_per_task_ms": 893696.3066666666,
      "avg_output_tokens_per_task": 51834.78666666667,
      "last_run_timestamp": "2026-10-27T03:58:39.326Z"
    },
    {
      "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": 0.7625640769230769,
      "accuracy_std_dev": 0.027985121905594868,
      "run_count": 13,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.5249607092307692,
      "avg_duration_per_task_ms": 215709.9046153846,
      "avg_output_tokens_per_task": 9987.407692307692,
      "last_run_timestamp": "2026-10-24T12:29:32.505Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Novita",
      "accuracy": 0.7625627587939698,
      "accuracy_std_dev": 0.017640643984489356,
      "run_count": 8,
      "total_questions": 398,
      "avg_cost_per_task_usd": 0.01927761809045226,
      "avg_duration_per_task_ms": 121120.84924623116,
      "avg_output_tokens_per_task": 14540.939698492462,
      "last_run_timestamp": "2026-10-24T11:49:30.492Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "AtlasCloud",
      "accuracy": 0.762500125,
      "accuracy_std_dev": 0.06959710991384181,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.25472334,
      "avg_duration_per_task_ms": 482869.145,
      "avg_output_tokens_per_task": 29497.4675,
      "last_run_timestamp": "2026-10-24T19:02:15.249Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7624327096774194,
      "accuracy_std_dev": 0.039878336715063745,
      "run_count": 10,
      "total_questions": 496,
      "avg_cost_per_task_usd": 0.09450495362903226,
      "avg_duration_per_task_ms": 748971.7096774194,
      "avg_output_tokens_per_task": 10301.362903225807,
      "last_run_timestamp": "2026-10-24T19:52:35.108Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-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": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Morph",
      "accuracy": 0.762143986599665,
      "accuracy_std_dev": 0.04074106053036531,
      "run_count": 12,
      "total_questions": 597,
      "avg_cost_per_task_usd": 0.06808563986599665,
      "avg_duration_per_task_ms": 207102.986599665,
      "avg_output_tokens_per_task": 11782.765494137353,
      "last_run_timestamp": "2026-10-24T12:36:04.401Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.5-opus-20251124",
      "model_slug": "anthropic/claude-opus-4.5",
      "model_name": "Anthropic: Claude Opus 4.5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.762,
      "accuracy_std_dev": 0.02386079071196091,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.5457165,
      "avg_duration_per_task_ms": 180407.204,
      "avg_output_tokens_per_task": 10745.142,
      "last_run_timestamp": "2026-10-24T11:51:51.137Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "auto-routing",
      "accuracy": 0.7618181818181818,
      "accuracy_std_dev": 0.03607974434957347,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.14144602545454546,
      "avg_duration_per_task_ms": 856323.6963636364,
      "avg_output_tokens_per_task": 31885.77090909091,
      "last_run_timestamp": "2026-10-24T19:43:40.189Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Baidu",
      "accuracy": 0.7617847878787878,
      "accuracy_std_dev": 0.03876959831491683,
      "run_count": 4,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.02413183333333333,
      "avg_duration_per_task_ms": 113963.81313131313,
      "avg_output_tokens_per_task": 15997.81313131313,
      "last_run_timestamp": "2026-10-24T12:54:33.303Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.7617776666666667,
      "accuracy_std_dev": 0.02671855322596808,
      "run_count": 15,
      "total_questions": 750,
      "avg_cost_per_task_usd": 0.011852372,
      "avg_duration_per_task_ms": 180670.42666666667,
      "avg_output_tokens_per_task": 5590.144,
      "last_run_timestamp": "2026-10-29T22:50:44.363Z"
    },
    {
      "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": 0.007206587248322147,
      "avg_duration_per_task_ms": 288157.1812080537,
      "avg_output_tokens_per_task": 12110.157718120805,
      "last_run_timestamp": "2026-10-24T20:14:29.971Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "SiliconFlow",
      "accuracy": 0.76166675,
      "accuracy_std_dev": 0.02229839929653024,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.065100265,
      "avg_duration_per_task_ms": 116150.4225,
      "avg_output_tokens_per_task": 4604.0675,
      "last_run_timestamp": "2026-10-24T12:22:07.097Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Alibaba",
      "accuracy": 0.7616666428571429,
      "accuracy_std_dev": 0.025346699169734525,
      "run_count": 14,
      "total_questions": 700,
      "avg_cost_per_task_usd": 0.09051784142857143,
      "avg_duration_per_task_ms": 191721.94857142857,
      "avg_output_tokens_per_task": 11065.187142857143,
      "last_run_timestamp": "2026-10-28T04:31:19.438Z"
    },
    {
      "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": 0.761666625,
      "accuracy_std_dev": 0.015545671961493828,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.08149474,
      "avg_duration_per_task_ms": 103926,
      "avg_output_tokens_per_task": 22421.02,
      "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/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "DeepInfra",
      "accuracy": 0.7615039124579126,
      "accuracy_std_dev": 0.01612860952882591,
      "run_count": 6,
      "total_questions": 297,
      "avg_cost_per_task_usd": 0.1581539494949495,
      "avg_duration_per_task_ms": 807649.4747474748,
      "avg_output_tokens_per_task": 30684.649831649833,
      "last_run_timestamp": "2026-10-24T18:54:54.259Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.5-opus-20251124",
      "model_slug": "anthropic/claude-opus-4.5",
      "model_name": "Anthropic: Claude Opus 4.5",
      "provider_name": "Google",
      "accuracy": 0.7612120909090909,
      "accuracy_std_dev": 0.030986186680386097,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.5405228218181818,
      "avg_duration_per_task_ms": 191201.12545454546,
      "avg_output_tokens_per_task": 10721.865454545454,
      "last_run_timestamp": "2026-10-24T12:29:33.066Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Baidu",
      "accuracy": 0.761111,
      "accuracy_std_dev": 0.016063062451890448,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.05401205,
      "avg_duration_per_task_ms": 73585.80666666667,
      "avg_output_tokens_per_task": 4477.26,
      "last_run_timestamp": "2026-10-24T12:38:41.606Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "BaseTen",
      "accuracy": 0.7610546224489795,
      "accuracy_std_dev": 0.016475935184003962,
      "run_count": 4,
      "total_questions": 196,
      "avg_cost_per_task_usd": 0.2764416326530612,
      "avg_duration_per_task_ms": 159672.02551020408,
      "avg_output_tokens_per_task": 7623.877551020408,
      "last_run_timestamp": "2026-10-24T12:38:26.300Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "Darkbloom",
      "accuracy": 0.7609524285714285,
      "accuracy_std_dev": 0.025053305529593985,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.06623187714285714,
      "avg_duration_per_task_ms": 1264087.4542857143,
      "avg_output_tokens_per_task": 64691.61714285714,
      "last_run_timestamp": "2026-10-30T02:58:57.544Z"
    },
    {
      "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": 0.7609091818181818,
      "accuracy_std_dev": 0.03684790544884393,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.19680807272727271,
      "avg_duration_per_task_ms": 478180.9890909091,
      "avg_output_tokens_per_task": 27737.445454545454,
      "last_run_timestamp": "2026-10-28T04:04:29.913Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Azure",
      "accuracy": 0.760833375,
      "accuracy_std_dev": 0.017460477198071506,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.17206684999999997,
      "avg_duration_per_task_ms": 195271.6725,
      "avg_output_tokens_per_task": 11186.875,
      "last_run_timestamp": "2026-10-24T19:43:42.996Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "OpenInference",
      "accuracy": 0.76083325,
      "accuracy_std_dev": 0.022897649686976593,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.018742625,
      "avg_duration_per_task_ms": 681240.3,
      "avg_output_tokens_per_task": 12523.52,
      "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": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Relace",
      "accuracy": 0.7607406666666667,
      "accuracy_std_dev": 0.034633054297881376,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.006436806666666667,
      "avg_duration_per_task_ms": 125547.04444444444,
      "avg_output_tokens_per_task": 4262.8533333333335,
      "last_run_timestamp": "2026-10-24T14:25:23.734Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Together",
      "accuracy": 0.7606666,
      "accuracy_std_dev": 0.031048465341140474,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.027007422000000003,
      "avg_duration_per_task_ms": 150781.166,
      "avg_output_tokens_per_task": 8984.718,
      "last_run_timestamp": "2026-10-24T19:16:17.608Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "InferenceNet",
      "accuracy": 0.7606059090909091,
      "accuracy_std_dev": 0.02402931176124294,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.03610860181818182,
      "avg_duration_per_task_ms": 72225.87454545454,
      "avg_output_tokens_per_task": 4745.927272727273,
      "last_run_timestamp": "2026-10-24T12:22:54.608Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-122b-a10b-20260224",
      "model_slug": "qwen/qwen3.5-122b-a10b",
      "model_name": "Qwen: Qwen3.5-122B-A10B",
      "provider_name": "DeepInfra",
      "accuracy": 0.760476,
      "accuracy_std_dev": 0.023999188385086207,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.21848949428571432,
      "avg_duration_per_task_ms": 711094.3942857143,
      "avg_output_tokens_per_task": 57307.56,
      "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": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Crusoe",
      "accuracy": 0.7603748473895582,
      "accuracy_std_dev": 0.02026019621425223,
      "run_count": 5,
      "total_questions": 249,
      "avg_cost_per_task_usd": 0.03634712449799197,
      "avg_duration_per_task_ms": 152002.0281124498,
      "avg_output_tokens_per_task": 23794.738955823294,
      "last_run_timestamp": "2026-10-24T12:53:48.922Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "DekaLLM",
      "accuracy": 0.7603120367892977,
      "accuracy_std_dev": 0.027860534151775026,
      "run_count": 6,
      "total_questions": 299,
      "avg_cost_per_task_usd": 0.06456417725752508,
      "avg_duration_per_task_ms": 376665.0802675585,
      "avg_output_tokens_per_task": 52393.28428093645,
      "last_run_timestamp": "2026-10-24T19:21:23.089Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Novita",
      "accuracy": 0.7600000909090909,
      "accuracy_std_dev": 0.023441265523758994,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.03510028181818181,
      "avg_duration_per_task_ms": 167423.50181818183,
      "avg_output_tokens_per_task": 16452.421818181818,
      "last_run_timestamp": "2026-10-24T12:19:39.208Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Cloudflare",
      "accuracy": 0.76,
      "accuracy_std_dev": 0.02160235326879922,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.027435835,
      "avg_duration_per_task_ms": 200904.6775,
      "avg_output_tokens_per_task": 13494.075,
      "last_run_timestamp": "2026-10-24T20:44:23.337Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "SiliconFlow",
      "accuracy": 0.76,
      "accuracy_std_dev": 0.01333300000000004,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.054461335,
      "avg_duration_per_task_ms": 91417,
      "avg_output_tokens_per_task": 15101.33,
      "last_run_timestamp": "2026-10-03T08:13:12.597Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "DekaLLM",
      "accuracy": 0.76,
      "accuracy_std_dev": 0.014401594564491825,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.019761673333333334,
      "avg_duration_per_task_ms": 291193.25333333336,
      "avg_output_tokens_per_task": 17811.393333333333,
      "last_run_timestamp": "2026-10-10T11:23:23.352Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "GLM 5.3",
      "provider_name": "digitalocean",
      "accuracy": 0.76,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.01597332,
      "avg_duration_per_task_ms": 42256.58,
      "avg_output_tokens_per_task": 1764.66,
      "last_run_timestamp": "2026-10-01T23:50:45.025Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.76,
      "accuracy_std_dev": 0.01333300000000004,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.3443486,
      "avg_duration_per_task_ms": 184596.7,
      "avg_output_tokens_per_task": 12502.76,
      "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": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Venice",
      "accuracy": 0.76,
      "accuracy_std_dev": 0.02494442712310708,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.022151744999999997,
      "avg_duration_per_task_ms": 485324.6075,
      "avg_output_tokens_per_task": 14080.3675,
      "last_run_timestamp": "2026-10-24T18:47:38.219Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "openai/gpt-5.1-20251113",
      "model_slug": "openai/gpt-5.1",
      "model_name": "OpenAI: GPT-5.1",
      "provider_name": "OpenAI",
      "accuracy": 0.76,
      "accuracy_std_dev": 0.005443582705045132,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.40005456666666667,
      "avg_duration_per_task_ms": 394539.76666666666,
      "avg_output_tokens_per_task": 34690.113333333335,
      "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": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "Novita",
      "accuracy": 0.76,
      "accuracy_std_dev": 0.02108195647941624,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.08341662,
      "avg_duration_per_task_ms": 533566.115,
      "avg_output_tokens_per_task": 14923.395,
      "last_run_timestamp": "2026-10-24T12:36:09.902Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "Cohere",
      "accuracy": 0.76,
      "accuracy_std_dev": 0.0066669999999999785,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.10151575999999998,
      "avg_duration_per_task_ms": 434936.03,
      "avg_output_tokens_per_task": 13962.46,
      "last_run_timestamp": "2026-10-24T12:32:07.608Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Makora",
      "accuracy": 0.76,
      "accuracy_std_dev": 0.0066669999999999785,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.07038507,
      "avg_duration_per_task_ms": 256302.49,
      "avg_output_tokens_per_task": 10045.27,
      "last_run_timestamp": "2026-08-26T00:00:05.638Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "GMICloud",
      "accuracy": 0.7599998888888889,
      "accuracy_std_dev": 0.03899989237148808,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.009081111111111111,
      "avg_duration_per_task_ms": 187614.08,
      "avg_output_tokens_per_task": 15582.557777777778,
      "last_run_timestamp": "2026-10-27T01:09:17.824Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Reka",
      "accuracy": 0.7595238571428571,
      "accuracy_std_dev": 0.02174373003832582,
      "run_count": 14,
      "total_questions": 700,
      "avg_cost_per_task_usd": 0.04527854285714286,
      "avg_duration_per_task_ms": 136160.14428571428,
      "avg_output_tokens_per_task": 4500.364285714286,
      "last_run_timestamp": "2026-10-24T12:17:41.928Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Venice",
      "accuracy": 0.7594445,
      "accuracy_std_dev": 0.03225570769776,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.01983395166666667,
      "avg_duration_per_task_ms": 192524.21833333332,
      "avg_output_tokens_per_task": 14719.435,
      "last_run_timestamp": "2026-10-24T12:16:56.429Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "DeepInfra",
      "accuracy": 0.759408620967742,
      "accuracy_std_dev": 0.023253272624729635,
      "run_count": 5,
      "total_questions": 248,
      "avg_cost_per_task_usd": 0.012160826612903226,
      "avg_duration_per_task_ms": 309190.9919354839,
      "avg_output_tokens_per_task": 4125.725806451613,
      "last_run_timestamp": "2026-10-17T11:47:18.976Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Cloudflare",
      "accuracy": 0.7593939090909091,
      "accuracy_std_dev": 0.017397280855741624,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.031511507272727265,
      "avg_duration_per_task_ms": 229078.9090909091,
      "avg_output_tokens_per_task": 15370.054545454546,
      "last_run_timestamp": "2026-10-24T13:28:03.080Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "CoreWeave",
      "accuracy": 0.7592594444444444,
      "accuracy_std_dev": 0.017340275597650393,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.15033249555555556,
      "avg_duration_per_task_ms": 165755.92888888888,
      "avg_output_tokens_per_task": 20125.08,
      "last_run_timestamp": "2026-10-24T11:39:34.603Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Reka",
      "accuracy": 0.7592592436412315,
      "accuracy_std_dev": 0.02502658689065603,
      "run_count": 15,
      "total_questions": 747,
      "avg_cost_per_task_usd": 0.05607525167336011,
      "avg_duration_per_task_ms": 275516.45113788487,
      "avg_output_tokens_per_task": 12015.718875502009,
      "last_run_timestamp": "2026-10-24T11:39:58.119Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Azure",
      "accuracy": 0.759166625,
      "accuracy_std_dev": 0.02777639760631993,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0257473675,
      "avg_duration_per_task_ms": 253688.2675,
      "avg_output_tokens_per_task": 13099.1275,
      "last_run_timestamp": "2026-10-24T19:09:33.707Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Alibaba",
      "accuracy": 0.7591665000000001,
      "accuracy_std_dev": 0.030264945803685155,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.03163724750000001,
      "avg_duration_per_task_ms": 184467.965,
      "avg_output_tokens_per_task": 14426.225,
      "last_run_timestamp": "2026-10-28T03:36:02.600Z"
    },
    {
      "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": 0.7591665000000001,
      "accuracy_std_dev": 0.003632454025861855,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.09515249499999999,
      "avg_duration_per_task_ms": 195938.94,
      "avg_output_tokens_per_task": 9656.23,
      "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-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Wafer",
      "accuracy": 0.7588889999999999,
      "accuracy_std_dev": 0.019019067756683928,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.009268686666666666,
      "avg_duration_per_task_ms": 94260.51333333334,
      "avg_output_tokens_per_task": 13026.133333333333,
      "last_run_timestamp": "2026-10-29T22:30:49.139Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Makora",
      "accuracy": 0.7588886666666668,
      "accuracy_std_dev": 0.011811189506377245,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.010121153333333334,
      "avg_duration_per_task_ms": 244060.09,
      "avg_output_tokens_per_task": 16283.283333333333,
      "last_run_timestamp": "2026-10-24T13:01:16.778Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Darkbloom",
      "accuracy": 0.7588075528455285,
      "accuracy_std_dev": 0.013043211721044773,
      "run_count": 5,
      "total_questions": 246,
      "avg_cost_per_task_usd": 0.0340294837398374,
      "avg_duration_per_task_ms": 524064.9918699187,
      "avg_output_tokens_per_task": 12340.776422764227,
      "last_run_timestamp": "2026-10-30T01:17:31.545Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "CoreWeave",
      "accuracy": 0.7584615384615384,
      "accuracy_std_dev": 0.02110672856848985,
      "run_count": 13,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.021854490769230767,
      "avg_duration_per_task_ms": 124647.87076923077,
      "avg_output_tokens_per_task": 17096.62,
      "last_run_timestamp": "2026-10-24T12:20:00.386Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "auto-routing",
      "accuracy": 0.7583335,
      "accuracy_std_dev": 0.034760921546328406,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.12354291499999999,
      "avg_duration_per_task_ms": 1021866.15,
      "avg_output_tokens_per_task": 26151.615,
      "last_run_timestamp": "2026-10-24T12:47:02.906Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta/muse-glimmer-30b-20260810",
      "model_slug": "meta/muse-glimmer-30b",
      "model_name": "Meta: Muse Glimmer 30B",
      "provider_name": "Phala",
      "accuracy": 0.758333375,
      "accuracy_std_dev": 0.012360224066916245,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.034803532500000005,
      "avg_duration_per_task_ms": 275650.13,
      "avg_output_tokens_per_task": 13717.48,
      "last_run_timestamp": "2026-10-03T08:20:09.018Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "auto-routing",
      "accuracy": 0.758333375,
      "accuracy_std_dev": 0.03329163433558611,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.044925605,
      "avg_duration_per_task_ms": 183156.9125,
      "avg_output_tokens_per_task": 9429.4925,
      "last_run_timestamp": "2026-10-24T19:48:57.855Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.1-20251113",
      "model_slug": "openai/gpt-5.1",
      "model_name": "OpenAI: GPT-5.1",
      "provider_name": "Azure",
      "accuracy": 0.7583332500000001,
      "accuracy_std_dev": 0.014433785597946933,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.42515429,
      "avg_duration_per_task_ms": 460450.54,
      "avg_output_tokens_per_task": 34342.89,
      "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": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Mancer 2",
      "accuracy": 0.7583332500000001,
      "accuracy_std_dev": 0.03460084697211181,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0886916525,
      "avg_duration_per_task_ms": 581833.97,
      "avg_output_tokens_per_task": 12683.3375,
      "last_run_timestamp": "2026-10-24T12:19:32.894Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7583332500000001,
      "accuracy_std_dev": 0.029767707149485012,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.1867523675,
      "avg_duration_per_task_ms": 181894.3925,
      "avg_output_tokens_per_task": 9712.265,
      "last_run_timestamp": "2026-10-24T19:21:02.236Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "CoreWeave",
      "accuracy": 0.7581818181818182,
      "accuracy_std_dev": 0.034942042529720004,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.0916818309090909,
      "avg_duration_per_task_ms": 244976.56,
      "avg_output_tokens_per_task": 12409.656363636364,
      "last_run_timestamp": "2026-10-24T12:33:35.006Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "SambaNova",
      "accuracy": 0.7579907534246576,
      "accuracy_std_dev": 0.01133630408123485,
      "run_count": 3,
      "total_questions": 146,
      "avg_cost_per_task_usd": 0.06852380136986301,
      "avg_duration_per_task_ms": 156936.7876712329,
      "avg_output_tokens_per_task": 10408.410958904109,
      "last_run_timestamp": "2026-10-24T18:33:39.875Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "StreamLake",
      "accuracy": 0.7579486923076924,
      "accuracy_std_dev": 0.021623878857143158,
      "run_count": 13,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.0068238523076923074,
      "avg_duration_per_task_ms": 261267.2046153846,
      "avg_output_tokens_per_task": 15563.432307692308,
      "last_run_timestamp": "2026-10-24T13:09:56.475Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.757778,
      "accuracy_std_dev": 0.025724014817546128,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.03362461666666666,
      "avg_duration_per_task_ms": 226709.77666666667,
      "avg_output_tokens_per_task": 35231.33666666667,
      "last_run_timestamp": "2026-10-24T11:55:02.754Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "AkashML",
      "accuracy": 0.7575758811188811,
      "accuracy_std_dev": 0.010732768960317529,
      "run_count": 3,
      "total_questions": 143,
      "avg_cost_per_task_usd": 0.27724311888111886,
      "avg_duration_per_task_ms": 338699.5594405594,
      "avg_output_tokens_per_task": 6540.4965034965035,
      "last_run_timestamp": "2026-10-24T11:53:36.261Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Cloudflare",
      "accuracy": 0.7575504530201342,
      "accuracy_std_dev": 0.028065656837905205,
      "run_count": 12,
      "total_questions": 596,
      "avg_cost_per_task_usd": 0.19407245637583895,
      "avg_duration_per_task_ms": 482897.4211409396,
      "avg_output_tokens_per_task": 20396.08389261745,
      "last_run_timestamp": "2026-10-24T14:01:08.267Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "SiliconFlow",
      "accuracy": 0.7573333999999999,
      "accuracy_std_dev": 0.032000033334982636,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.028732296,
      "avg_duration_per_task_ms": 171140.192,
      "avg_output_tokens_per_task": 15777.572,
      "last_run_timestamp": "2026-10-24T12:23:36.194Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "NextBit",
      "accuracy": 0.7573333999999999,
      "accuracy_std_dev": 0.03991088294989224,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.023026752,
      "avg_duration_per_task_ms": 161149.652,
      "avg_output_tokens_per_task": 18473.992,
      "last_run_timestamp": "2026-10-03T08:55:31.322Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7571428571428571,
      "accuracy_std_dev": 0.024651550802730272,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.012830634285714286,
      "avg_duration_per_task_ms": 221282.58285714284,
      "avg_output_tokens_per_task": 4997.4742857142855,
      "last_run_timestamp": "2026-10-24T12:58:11.932Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Sail Research",
      "accuracy": 0.7568475720930232,
      "accuracy_std_dev": 0.02459803490467476,
      "run_count": 13,
      "total_questions": 645,
      "avg_cost_per_task_usd": 0.09061050697674419,
      "avg_duration_per_task_ms": 702536.334883721,
      "avg_output_tokens_per_task": 21508.96434108527,
      "last_run_timestamp": "2026-10-24T12:57:12.879Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.7567265784753363,
      "accuracy_std_dev": 0.024115065606584538,
      "run_count": 9,
      "total_questions": 446,
      "avg_cost_per_task_usd": 0.015529553811659193,
      "avg_duration_per_task_ms": 194340.35874439462,
      "avg_output_tokens_per_task": 5359.600896860987,
      "last_run_timestamp": "2026-10-30T14:59:37.060Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7566669999999999,
      "accuracy_std_dev": 0.010000000000000009,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.023368849999999997,
      "avg_duration_per_task_ms": 302412.09,
      "avg_output_tokens_per_task": 11606.21,
      "last_run_timestamp": "2026-10-24T12:08:01.600Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7566666666666667,
      "accuracy_std_dev": 0.04895203245581898,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.02191566,
      "avg_duration_per_task_ms": 332956.19333333336,
      "avg_output_tokens_per_task": 12348.723333333333,
      "last_run_timestamp": "2026-10-24T12:55:32.102Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.7566666,
      "accuracy_std_dev": 0.015846499791436604,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.175855144,
      "avg_duration_per_task_ms": 317393.27,
      "avg_output_tokens_per_task": 10799.458,
      "last_run_timestamp": "2026-10-28T03:37:05.929Z"
    },
    {
      "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": 0.7566665,
      "accuracy_std_dev": 0.028087175801422264,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0129076525,
      "avg_duration_per_task_ms": 225184.265,
      "avg_output_tokens_per_task": 14188.9,
      "last_run_timestamp": "2026-10-24T19:54:36.988Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Parasail",
      "accuracy": 0.7566665,
      "accuracy_std_dev": 0.0033334999999999892,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.018845109999999998,
      "avg_duration_per_task_ms": 92258.91,
      "avg_output_tokens_per_task": 11470.03,
      "last_run_timestamp": "2026-10-24T12:23:52.812Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Alibaba",
      "accuracy": 0.7564972372881356,
      "accuracy_std_dev": 0.0180813407986352,
      "run_count": 12,
      "total_questions": 590,
      "avg_cost_per_task_usd": 0.032619,
      "avg_duration_per_task_ms": 164901.18983050846,
      "avg_output_tokens_per_task": 16019.486440677965,
      "last_run_timestamp": "2026-10-28T03:43:00.533Z"
    },
    {
      "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": 0.7563636363636363,
      "accuracy_std_dev": 0.028620044734329858,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.1009524109090909,
      "avg_duration_per_task_ms": 283080.54363636364,
      "avg_output_tokens_per_task": 12195.158181818182,
      "last_run_timestamp": "2026-10-28T03:44:06.462Z"
    },
    {
      "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": "Wafer",
      "accuracy": 0.756,
      "accuracy_std_dev": 0.04249977777824259,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.03474274,
      "avg_duration_per_task_ms": 260320.88,
      "avg_output_tokens_per_task": 13874.396,
      "last_run_timestamp": "2026-10-29T23:09:03.074Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "SiliconFlow",
      "accuracy": 0.756,
      "accuracy_std_dev": 0.024073867807230344,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.07011482399999999,
      "avg_duration_per_task_ms": 334056.692,
      "avg_output_tokens_per_task": 14633.816,
      "last_run_timestamp": "2026-10-24T12:59:11.337Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "GMICloud",
      "accuracy": 0.756,
      "accuracy_std_dev": 0.022151057162131126,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.064562922,
      "avg_duration_per_task_ms": 168412.394,
      "avg_output_tokens_per_task": 4657.626,
      "last_run_timestamp": "2026-10-27T01:58:28.298Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "InferenceNet",
      "accuracy": 0.7559648947368421,
      "accuracy_std_dev": 0.020191449941144954,
      "run_count": 19,
      "total_questions": 950,
      "avg_cost_per_task_usd": 0.29131978947368414,
      "avg_duration_per_task_ms": 264712.04,
      "avg_output_tokens_per_task": 11238.59157894737,
      "last_run_timestamp": "2026-10-24T12:24:12.062Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Modal",
      "accuracy": 0.755892303030303,
      "accuracy_std_dev": 0.022789500000000018,
      "run_count": 2,
      "total_questions": 99,
      "avg_cost_per_task_usd": 0.027549767676767676,
      "avg_duration_per_task_ms": 104945.30303030302,
      "avg_output_tokens_per_task": 10081.373737373737,
      "last_run_timestamp": "2026-10-24T12:19:36.298Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.75583325,
      "accuracy_std_dev": 0.019982613972839007,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.008674565,
      "avg_duration_per_task_ms": 119249.875,
      "avg_output_tokens_per_task": 4714.72,
      "last_run_timestamp": "2026-10-24T12:46:25.479Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "ModelRun",
      "accuracy": 0.7557780862865947,
      "accuracy_std_dev": 0.028031620161237676,
      "run_count": 13,
      "total_questions": 649,
      "avg_cost_per_task_usd": 0.09079595223420647,
      "avg_duration_per_task_ms": 135126.0123266564,
      "avg_output_tokens_per_task": 12093.19722650231,
      "last_run_timestamp": "2026-10-28T04:53:14.750Z"
    },
    {
      "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": 0.5314156181818182,
      "avg_duration_per_task_ms": 187359.03636363638,
      "avg_output_tokens_per_task": 10435.296363636364,
      "last_run_timestamp": "2026-10-24T11:54:04.367Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Venice",
      "accuracy": 0.7556052242152467,
      "accuracy_std_dev": 0.025010430481878735,
      "run_count": 9,
      "total_questions": 446,
      "avg_cost_per_task_usd": 0.020433513452914796,
      "avg_duration_per_task_ms": 233457.43946188342,
      "avg_output_tokens_per_task": 4244.6569506726455,
      "last_run_timestamp": "2026-10-24T12:41:47.992Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "StreamLake",
      "accuracy": 0.7555555555555555,
      "accuracy_std_dev": 0.02330683151215116,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.007330602222222223,
      "avg_duration_per_task_ms": 193346.8088888889,
      "avg_output_tokens_per_task": 4415.6,
      "last_run_timestamp": "2026-10-24T14:44:52.720Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-sonnet-5-20260630",
      "model_slug": "anthropic/claude-sonnet-5",
      "model_name": "Anthropic: Claude Sonnet 5",
      "provider_name": "Google",
      "accuracy": 0.7554901764705882,
      "accuracy_std_dev": 0.03402120717371878,
      "run_count": 34,
      "total_questions": 1700,
      "avg_cost_per_task_usd": 0.2160354817647059,
      "avg_duration_per_task_ms": 94519.24058823529,
      "avg_output_tokens_per_task": 7506.205882352941,
      "last_run_timestamp": "2026-10-24T12:46:28.470Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-sonnet-5-20260630",
      "model_slug": "anthropic/claude-sonnet-5",
      "model_name": "Anthropic: Claude Sonnet 5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7554166562500002,
      "accuracy_std_dev": 0.029647766247654916,
      "run_count": 32,
      "total_questions": 1600,
      "avg_cost_per_task_usd": 0.20572255374999998,
      "avg_duration_per_task_ms": 117215.80375,
      "avg_output_tokens_per_task": 7761.9825,
      "last_run_timestamp": "2026-10-24T12:18:37.751Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-sonnet-5-20260630",
      "model_slug": "anthropic/claude-sonnet-5",
      "model_name": "Anthropic: Claude Sonnet 5",
      "provider_name": "auto-routing",
      "accuracy": 0.7551515454545454,
      "accuracy_std_dev": 0.030363520838375783,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.2013825509090909,
      "avg_duration_per_task_ms": 122033.56545454546,
      "avg_output_tokens_per_task": 7806.483636363636,
      "last_run_timestamp": "2026-10-24T11:58:02.887Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "CoreWeave",
      "accuracy": 0.755000125,
      "accuracy_std_dev": 0.02576585003176445,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.025328775,
      "avg_duration_per_task_ms": 341291.3375,
      "avg_output_tokens_per_task": 10920.815,
      "last_run_timestamp": "2026-10-24T19:18:15.431Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Together",
      "accuracy": 0.755,
      "accuracy_std_dev": 0.04014562657874455,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.161091306,
      "avg_duration_per_task_ms": 311751.64,
      "avg_output_tokens_per_task": 23196.802,
      "last_run_timestamp": "2026-10-24T12:52:46.316Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "AtlasCloud",
      "accuracy": 0.755,
      "accuracy_std_dev": 0.008333000000000035,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.07072132,
      "avg_duration_per_task_ms": 164484.88,
      "avg_output_tokens_per_task": 7418.93,
      "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": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Parasail",
      "accuracy": 0.7548484545454546,
      "accuracy_std_dev": 0.021805424301655014,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.01682924181818182,
      "avg_duration_per_task_ms": 119778.08181818182,
      "avg_output_tokens_per_task": 4470.658181818182,
      "last_run_timestamp": "2026-10-24T13:29:07.045Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "InferenceNet",
      "accuracy": 0.7545454545454545,
      "accuracy_std_dev": 0.03163137195128692,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.008343416363636365,
      "avg_duration_per_task_ms": 139678.11272727273,
      "avg_output_tokens_per_task": 4440.807272727273,
      "last_run_timestamp": "2026-10-24T13:04:30.804Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "Parasail",
      "accuracy": 0.7545453636363637,
      "accuracy_std_dev": 0.01578076033582857,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.03706595454545454,
      "avg_duration_per_task_ms": 272204.82727272727,
      "avg_output_tokens_per_task": 21398.014545454545,
      "last_run_timestamp": "2026-10-24T19:02:49.020Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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-sonnet-5-20260630",
      "model_slug": "anthropic/claude-sonnet-5",
      "model_name": "Anthropic: Claude Sonnet 5",
      "provider_name": "Azure",
      "accuracy": 0.7542424545454546,
      "accuracy_std_dev": 0.03230505110621568,
      "run_count": 22,
      "total_questions": 1100,
      "avg_cost_per_task_usd": 0.2011599981818182,
      "avg_duration_per_task_ms": 97937.50545454545,
      "avg_output_tokens_per_task": 7615.459090909091,
      "last_run_timestamp": "2026-10-24T12:48:21.016Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-122b-a10b-20260224",
      "model_slug": "qwen/qwen3.5-122b-a10b",
      "model_name": "Qwen: Qwen3.5-122B-A10B",
      "provider_name": "Alibaba",
      "accuracy": 0.7540288473282444,
      "accuracy_std_dev": 0.03768553089030821,
      "run_count": 8,
      "total_questions": 393,
      "avg_cost_per_task_usd": 0.12041368702290077,
      "avg_duration_per_task_ms": 478479.56488549616,
      "avg_output_tokens_per_task": 22514.213740458013,
      "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": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Parasail",
      "accuracy": 0.7539394545454545,
      "accuracy_std_dev": 0.02627826791498197,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.01945476181818182,
      "avg_duration_per_task_ms": 219416.4818181818,
      "avg_output_tokens_per_task": 15698.581818181818,
      "last_run_timestamp": "2026-10-24T12:17:05.499Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-122b-a10b-20260224",
      "model_slug": "qwen/qwen3.5-122b-a10b",
      "model_name": "Qwen: Qwen3.5-122B-A10B",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7539394545454545,
      "accuracy_std_dev": 0.10561804966556154,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.15414004727272726,
      "avg_duration_per_task_ms": 605208.28,
      "avg_output_tokens_per_task": 28608.598181818183,
      "last_run_timestamp": "2026-10-24T19:26:44.780Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "Darkbloom",
      "accuracy": 0.753478061769616,
      "accuracy_std_dev": 0.01967899930322652,
      "run_count": 12,
      "total_questions": 599,
      "avg_cost_per_task_usd": 0.027070736227045072,
      "avg_duration_per_task_ms": 546959.5626043406,
      "avg_output_tokens_per_task": 22325.539232053423,
      "last_run_timestamp": "2026-10-30T02:35:47.337Z"
    },
    {
      "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": 0.7533335,
      "accuracy_std_dev": 0.026666500000000037,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.16966541999999996,
      "avg_duration_per_task_ms": 147439,
      "avg_output_tokens_per_task": 22903.86,
      "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": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Morph",
      "accuracy": 0.7533335,
      "accuracy_std_dev": 0.013333499999999998,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.13189366,
      "avg_duration_per_task_ms": 651730.99,
      "avg_output_tokens_per_task": 24955.07,
      "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": "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": "anthropic/claude-4.5-sonnet-20250929",
      "model_slug": "anthropic/claude-sonnet-4.5",
      "model_name": "Anthropic: Claude Sonnet 4.5",
      "provider_name": "Anthropic",
      "accuracy": 0.7533334615384615,
      "accuracy_std_dev": 0.02958406824787076,
      "run_count": 13,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.34445012307692313,
      "avg_duration_per_task_ms": 228394.87538461538,
      "avg_output_tokens_per_task": 10886.436923076923,
      "last_run_timestamp": "2026-10-24T11:48:43.120Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "auto-routing",
      "accuracy": 0.753333375,
      "accuracy_std_dev": 0.01452975873799614,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.1921822125,
      "avg_duration_per_task_ms": 294611.6675,
      "avg_output_tokens_per_task": 11751.955,
      "last_run_timestamp": "2026-10-24T11:55:23.806Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-sonnet-5-20260630",
      "model_slug": "anthropic/claude-sonnet-5",
      "model_name": "Anthropic: Claude Sonnet 5",
      "provider_name": "Anthropic",
      "accuracy": 0.7533333636363636,
      "accuracy_std_dev": 0.03856000964087188,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.19848266727272726,
      "avg_duration_per_task_ms": 100895.72727272728,
      "avg_output_tokens_per_task": 7619.894545454546,
      "last_run_timestamp": "2026-10-24T12:18:01.499Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta/muse-spark-open-20261015",
      "model_slug": "meta/muse-spark-open",
      "model_name": "Meta: Muse Spark Open",
      "provider_name": "Fireworks",
      "accuracy": 0.7533333333333333,
      "accuracy_std_dev": 0.024944382578492966,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.09710899333333334,
      "avg_duration_per_task_ms": 165189.49333333335,
      "avg_output_tokens_per_task": 9718.706666666667,
      "last_run_timestamp": "2026-10-30T01:11:01.826Z"
    },
    {
      "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": 0.7533332500000001,
      "accuracy_std_dev": 0.030731887168501403,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.029856812500000003,
      "avg_duration_per_task_ms": 481555.77,
      "avg_output_tokens_per_task": 11765.71,
      "last_run_timestamp": "2026-10-24T19:23:19.868Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Phala",
      "accuracy": 0.7533331428571429,
      "accuracy_std_dev": 0.0230940520074423,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.011956362857142858,
      "avg_duration_per_task_ms": 437970.4885714286,
      "avg_output_tokens_per_task": 6008.582857142857,
      "last_run_timestamp": "2026-10-24T12:34:15.321Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Morph",
      "accuracy": 0.753333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.02472732,
      "avg_duration_per_task_ms": 1279934.76,
      "avg_output_tokens_per_task": 22994.42,
      "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-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "BaseTen",
      "accuracy": 0.753333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.024162740000000002,
      "avg_duration_per_task_ms": 65962.82,
      "avg_output_tokens_per_task": 8135.38,
      "last_run_timestamp": "2026-10-24T12:37:26.070Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Decart",
      "accuracy": 0.753333,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01197713,
      "avg_duration_per_task_ms": 212560.45,
      "avg_output_tokens_per_task": 13957.42,
      "last_run_timestamp": "2026-10-24T11:49:49.163Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Baidu",
      "accuracy": 0.753333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.0185477,
      "avg_duration_per_task_ms": 97588.58,
      "avg_output_tokens_per_task": 10717.4,
      "last_run_timestamp": "2026-10-24T11:41:40.972Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Morph",
      "accuracy": 0.7533329999999999,
      "accuracy_std_dev": 0.03999999999999998,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.0194427,
      "avg_duration_per_task_ms": 127110.83,
      "avg_output_tokens_per_task": 12815.84,
      "last_run_timestamp": "2026-10-24T12:03:02.865Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Alibaba",
      "accuracy": 0.7531401420289855,
      "accuracy_std_dev": 0.02334674149443907,
      "run_count": 7,
      "total_questions": 345,
      "avg_cost_per_task_usd": 0.0109654,
      "avg_duration_per_task_ms": 132639.57101449274,
      "avg_output_tokens_per_task": 12361.4,
      "last_run_timestamp": "2026-10-24T19:21:18.651Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "Phala",
      "accuracy": 0.7527273636363636,
      "accuracy_std_dev": 0.022822115642797527,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.07911564,
      "avg_duration_per_task_ms": 552538.9854545455,
      "avg_output_tokens_per_task": 35028.91454545454,
      "last_run_timestamp": "2026-10-24T19:16:16.979Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Baidu",
      "accuracy": 0.7527272727272727,
      "accuracy_std_dev": 0.029875735292863623,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.013207607272727274,
      "avg_duration_per_task_ms": 136477.7381818182,
      "avg_output_tokens_per_task": 16262.556363636364,
      "last_run_timestamp": "2026-10-24T11:51:53.770Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Fireworks",
      "accuracy": 0.7525337229729728,
      "accuracy_std_dev": 0.02409323915008227,
      "run_count": 12,
      "total_questions": 592,
      "avg_cost_per_task_usd": 0.014006842905405404,
      "avg_duration_per_task_ms": 57600.26858108108,
      "avg_output_tokens_per_task": 4072.2314189189187,
      "last_run_timestamp": "2026-10-24T12:20:04.669Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "SiliconFlow",
      "accuracy": 0.752500125,
      "accuracy_std_dev": 0.02039545667199868,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0145751425,
      "avg_duration_per_task_ms": 232212.94,
      "avg_output_tokens_per_task": 14651.365,
      "last_run_timestamp": "2026-10-24T19:19:42.648Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "DeepInfra",
      "accuracy": 0.7522221666666666,
      "accuracy_std_dev": 0.023227349249083266,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.0077021400000000005,
      "avg_duration_per_task_ms": 223752.14166666666,
      "avg_output_tokens_per_task": 14385.541666666666,
      "last_run_timestamp": "2026-10-24T12:31:25.936Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-opus-5.5-20260921",
      "model_slug": "anthropic/claude-opus-5.5",
      "model_name": "Anthropic: Claude Opus 5.5",
      "provider_name": "auto-routing",
      "accuracy": 0.752,
      "accuracy_std_dev": 0.01857135189478677,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.33952648,
      "avg_duration_per_task_ms": 120001.212,
      "avg_output_tokens_per_task": 7107.392,
      "last_run_timestamp": "2026-10-24T12:22:52.909Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-opus-5.5-20260921",
      "model_slug": "anthropic/claude-opus-5.5",
      "model_name": "Anthropic: Claude Opus 5.5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7518518333333334,
      "accuracy_std_dev": 0.017715982107345284,
      "run_count": 18,
      "total_questions": 900,
      "avg_cost_per_task_usd": 0.36871217222222225,
      "avg_duration_per_task_ms": 149259.27444444445,
      "avg_output_tokens_per_task": 7150.745555555555,
      "last_run_timestamp": "2026-10-24T12:24:28.843Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Wafer",
      "accuracy": 0.7518183181818181,
      "accuracy_std_dev": 0.02334910466725432,
      "run_count": 22,
      "total_questions": 1100,
      "avg_cost_per_task_usd": 0.2542307654545455,
      "avg_duration_per_task_ms": 230926.2109090909,
      "avg_output_tokens_per_task": 10132.794545454546,
      "last_run_timestamp": "2026-10-30T00:59:12.977Z"
    },
    {
      "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": 0.7517172727272728,
      "accuracy_std_dev": 0.023898191710814292,
      "run_count": 33,
      "total_questions": 1650,
      "avg_cost_per_task_usd": 0.37267475515151516,
      "avg_duration_per_task_ms": 157429.93757575756,
      "avg_output_tokens_per_task": 11808.122424242425,
      "last_run_timestamp": "2026-10-24T12:39:24.516Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "CoreWeave",
      "accuracy": 0.751666875,
      "accuracy_std_dev": 0.03278711001841082,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.08480622499999999,
      "avg_duration_per_task_ms": 70521.9675,
      "avg_output_tokens_per_task": 11906.8925,
      "last_run_timestamp": "2026-10-24T19:37:32.789Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Ambient",
      "accuracy": 0.7516667499999999,
      "accuracy_std_dev": 0.01723988431479461,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.066217935,
      "avg_duration_per_task_ms": 295605.795,
      "avg_output_tokens_per_task": 8579.39,
      "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": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Moonshot AI",
      "accuracy": 0.751666625,
      "accuracy_std_dev": 0.046874494071769725,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.11106282249999999,
      "avg_duration_per_task_ms": 257380.8975,
      "avg_output_tokens_per_task": 11828.0925,
      "last_run_timestamp": "2026-10-24T19:31:21.574Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.7515151818181818,
      "accuracy_std_dev": 0.027131013635248374,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.5026920418181818,
      "avg_duration_per_task_ms": 194031.15818181817,
      "avg_output_tokens_per_task": 10111.854545454546,
      "last_run_timestamp": "2026-10-24T12:42:05.323Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Sail Research",
      "accuracy": 0.7514285714285714,
      "accuracy_std_dev": 0.021296004758486267,
      "run_count": 14,
      "total_questions": 700,
      "avg_cost_per_task_usd": 0.009337331428571431,
      "avg_duration_per_task_ms": 230829.7457142857,
      "avg_output_tokens_per_task": 4265.0371428571425,
      "last_run_timestamp": "2026-10-24T13:34:36.032Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "CoreWeave",
      "accuracy": 0.7512689898477157,
      "accuracy_std_dev": 0.021764084474197413,
      "run_count": 4,
      "total_questions": 197,
      "avg_cost_per_task_usd": 0.12904325888324872,
      "avg_duration_per_task_ms": 273181.16751269036,
      "avg_output_tokens_per_task": 46030.85279187817,
      "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": "anthropic/claude-opus-5.5-20260921",
      "model_slug": "anthropic/claude-opus-5.5",
      "model_name": "Anthropic: Claude Opus 5.5",
      "provider_name": "Anthropic",
      "accuracy": 0.7511111666666667,
      "accuracy_std_dev": 0.02454511593248011,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.3494615833333333,
      "avg_duration_per_task_ms": 100555.84,
      "avg_output_tokens_per_task": 7111.15,
      "last_run_timestamp": "2026-10-24T11:44:08.041Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-opus-5.5-20260921",
      "model_slug": "anthropic/claude-opus-5.5",
      "model_name": "Anthropic: Claude Opus 5.5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.751111,
      "accuracy_std_dev": 0.01523486904002351,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.3475124233333333,
      "avg_duration_per_task_ms": 124013.22666666667,
      "avg_output_tokens_per_task": 7125.08,
      "last_run_timestamp": "2026-10-24T11:31:04.549Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Fireworks",
      "accuracy": 0.7510038032128515,
      "accuracy_std_dev": 0.01043774738916402,
      "run_count": 5,
      "total_questions": 249,
      "avg_cost_per_task_usd": 0.023900196787148595,
      "avg_duration_per_task_ms": 72851.72690763052,
      "avg_output_tokens_per_task": 12280.232931726907,
      "last_run_timestamp": "2026-10-28T00:15:49.021Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Decart",
      "accuracy": 0.7506668,
      "accuracy_std_dev": 0.025854254755455656,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.264866352,
      "avg_duration_per_task_ms": 264537.464,
      "avg_output_tokens_per_task": 10440.212,
      "last_run_timestamp": "2026-10-24T11:49:28.605Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Relace",
      "accuracy": 0.7506666,
      "accuracy_std_dev": 0.024073823498563778,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.022084748,
      "avg_duration_per_task_ms": 137222.84,
      "avg_output_tokens_per_task": 13584.26,
      "last_run_timestamp": "2026-10-24T19:40:37.108Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Phala",
      "accuracy": 0.7506666,
      "accuracy_std_dev": 0.012364666507431593,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.17681883199999998,
      "avg_duration_per_task_ms": 536508.956,
      "avg_output_tokens_per_task": 9711.396,
      "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-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "GMICloud",
      "accuracy": 0.7505555833333333,
      "accuracy_std_dev": 0.034043327967112576,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.102866665,
      "avg_duration_per_task_ms": 304845.985,
      "avg_output_tokens_per_task": 18629.698333333334,
      "last_run_timestamp": "2026-10-27T01:19:50.371Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Phala",
      "accuracy": 0.7505555,
      "accuracy_std_dev": 0.03574174773170259,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.03120006,
      "avg_duration_per_task_ms": 258878.755,
      "avg_output_tokens_per_task": 17037.54,
      "last_run_timestamp": "2026-10-24T12:34:23.300Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "auto-routing",
      "accuracy": 0.7503030909090909,
      "accuracy_std_dev": 0.028193269749597228,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.02549516181818182,
      "avg_duration_per_task_ms": 155201.13454545455,
      "avg_output_tokens_per_task": 9057.756363636363,
      "last_run_timestamp": "2026-10-24T19:12:17.342Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Inceptron",
      "accuracy": 0.7502788193979932,
      "accuracy_std_dev": 0.03287689055865162,
      "run_count": 6,
      "total_questions": 299,
      "avg_cost_per_task_usd": 0.01826405016722408,
      "avg_duration_per_task_ms": 135784.91304347827,
      "avg_output_tokens_per_task": 4013.7859531772574,
      "last_run_timestamp": "2026-10-24T12:16:33.907Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-opus-5.5-20260921",
      "model_slug": "anthropic/claude-opus-5.5",
      "model_name": "Anthropic: Claude Opus 5.5",
      "provider_name": "Azure",
      "accuracy": 0.7500001666666666,
      "accuracy_std_dev": 0.023333261905247805,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.3534387216666667,
      "avg_duration_per_task_ms": 87888.735,
      "avg_output_tokens_per_task": 7005.235,
      "last_run_timestamp": "2026-10-24T12:23:35.337Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-sonnet-5-20260630",
      "model_slug": "anthropic/claude-sonnet-5",
      "model_name": "Anthropic: Claude Sonnet 5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.7500001,
      "accuracy_std_dev": 0.02333341428702625,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.196243942,
      "avg_duration_per_task_ms": 117146.822,
      "avg_output_tokens_per_task": 7664.984,
      "last_run_timestamp": "2026-10-24T12:32:11.296Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "BaseTen",
      "accuracy": 0.75,
      "accuracy_std_dev": 0.0033330000000000304,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.28014248,
      "avg_duration_per_task_ms": 223020.56,
      "avg_output_tokens_per_task": 18715.38,
      "last_run_timestamp": "2026-09-12T08:12:13.731Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.75,
      "accuracy_std_dev": 0.010000000000000009,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.300415095,
      "avg_duration_per_task_ms": 278099.495,
      "avg_output_tokens_per_task": 10648.38,
      "last_run_timestamp": "2026-10-24T12:46:10.016Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.4-20260305",
      "model_slug": "openai/gpt-5.4",
      "model_name": "OpenAI: GPT-5.4",
      "provider_name": "auto-routing",
      "accuracy": 0.75,
      "accuracy_std_dev": 0.0033330000000000304,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.29130333999999997,
      "avg_duration_per_task_ms": 176962.37,
      "avg_output_tokens_per_task": 10386.26,
      "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": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Venice",
      "accuracy": 0.75,
      "accuracy_std_dev": 0.016666999999999987,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01952235,
      "avg_duration_per_task_ms": 2267441.92,
      "avg_output_tokens_per_task": 19667.19,
      "last_run_timestamp": "2026-10-03T08:13:52.062Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "AkashML",
      "accuracy": 0.75,
      "accuracy_std_dev": 0.01333300000000004,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.08838628,
      "avg_duration_per_task_ms": 204069.19,
      "avg_output_tokens_per_task": 7577,
      "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-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.75,
      "accuracy_std_dev": 0.010000000000000009,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.09412079000000001,
      "avg_duration_per_task_ms": 469968.05,
      "avg_output_tokens_per_task": 4703.42,
      "last_run_timestamp": "2026-10-24T12:22:38.011Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Novita",
      "accuracy": 0.7493334,
      "accuracy_std_dev": 0.035427579965896655,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.014355716000000001,
      "avg_duration_per_task_ms": 155681.656,
      "avg_output_tokens_per_task": 14723.468,
      "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-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "StreamLake",
      "accuracy": 0.7492560625,
      "accuracy_std_dev": 0.02911246180199882,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.0047560535714285725,
      "avg_duration_per_task_ms": 260988.63169642858,
      "avg_output_tokens_per_task": 12223.955357142857,
      "last_run_timestamp": "2026-10-24T20:11:39.941Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "auto-routing",
      "accuracy": 0.74916675,
      "accuracy_std_dev": 0.022095025207894678,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.03993733,
      "avg_duration_per_task_ms": 621867.6175,
      "avg_output_tokens_per_task": 16182.08,
      "last_run_timestamp": "2026-10-24T19:37:04.226Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Novita",
      "accuracy": 0.749166625,
      "accuracy_std_dev": 0.022837048057583416,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0985293125,
      "avg_duration_per_task_ms": 261632.96,
      "avg_output_tokens_per_task": 12654.5375,
      "last_run_timestamp": "2026-10-24T19:44:46.340Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.6-opus-20260205",
      "model_slug": "anthropic/claude-opus-4.6",
      "model_name": "Anthropic: Claude Opus 4.6",
      "provider_name": "Google",
      "accuracy": 0.7490908636363637,
      "accuracy_std_dev": 0.027527419506071103,
      "run_count": 22,
      "total_questions": 1100,
      "avg_cost_per_task_usd": 0.5507198090909092,
      "avg_duration_per_task_ms": 199727.5418181818,
      "avg_output_tokens_per_task": 9759.008181818182,
      "last_run_timestamp": "2026-10-24T12:58:30.427Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "DeepInfra",
      "accuracy": 0.7488888888888889,
      "accuracy_std_dev": 0.016329946740366617,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.01262068,
      "avg_duration_per_task_ms": 336899.4511111111,
      "avg_output_tokens_per_task": 16364.997777777779,
      "last_run_timestamp": "2026-10-24T12:24:10.218Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Wafer",
      "accuracy": 0.7486274705882353,
      "accuracy_std_dev": 0.025128604245121908,
      "run_count": 17,
      "total_questions": 850,
      "avg_cost_per_task_usd": 0.021811628235294114,
      "avg_duration_per_task_ms": 107903.25411764705,
      "avg_output_tokens_per_task": 14472.312941176471,
      "last_run_timestamp": "2026-10-29T22:41:53.110Z"
    },
    {
      "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": 0.748333375,
      "accuracy_std_dev": 0.024664456546503812,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0074231424999999995,
      "avg_duration_per_task_ms": 153071.365,
      "avg_output_tokens_per_task": 12771.8775,
      "last_run_timestamp": "2026-10-24T19:02:53.466Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "DeepInfra",
      "accuracy": 0.7483333,
      "accuracy_std_dev": 0.02761851834566802,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.045017131999999994,
      "avg_duration_per_task_ms": 459054.072,
      "avg_output_tokens_per_task": 23025.238,
      "last_run_timestamp": "2026-10-24T19:38:23.863Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Nebius",
      "accuracy": 0.748275551724138,
      "accuracy_std_dev": 0.011091913731282908,
      "run_count": 3,
      "total_questions": 145,
      "avg_cost_per_task_usd": 0.00999303448275862,
      "avg_duration_per_task_ms": 128162.52413793103,
      "avg_output_tokens_per_task": 3554.88275862069,
      "last_run_timestamp": "2026-10-24T12:44:22.556Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "auto-routing",
      "accuracy": 0.7480002,
      "accuracy_std_dev": 0.020396130352593862,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.030171100000000003,
      "avg_duration_per_task_ms": 197857.312,
      "avg_output_tokens_per_task": 18663.832,
      "last_run_timestamp": "2026-10-03T08:18:33.163Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.6-sonnet-20260217",
      "model_slug": "anthropic/claude-sonnet-4.6",
      "model_name": "Anthropic: Claude Sonnet 4.6",
      "provider_name": "Azure",
      "accuracy": 0.7478788181818182,
      "accuracy_std_dev": 0.018817276123132574,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.3490534545454545,
      "avg_duration_per_task_ms": 215062.8381818182,
      "avg_output_tokens_per_task": 11920.356363636363,
      "last_run_timestamp": "2026-10-24T12:39:07.663Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "CoreWeave",
      "accuracy": 0.7477778333333334,
      "accuracy_std_dev": 0.022580213671978926,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.018034246666666667,
      "avg_duration_per_task_ms": 75225.08333333333,
      "avg_output_tokens_per_task": 4268.986666666667,
      "last_run_timestamp": "2026-10-24T12:38:26.301Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Ionstream",
      "accuracy": 0.747691088161209,
      "accuracy_std_dev": 0.03008101142513954,
      "run_count": 8,
      "total_questions": 397,
      "avg_cost_per_task_usd": 0.09758945340050378,
      "avg_duration_per_task_ms": 397184.00755667506,
      "avg_output_tokens_per_task": 25302.57934508816,
      "last_run_timestamp": "2026-10-24T13:02:31.690Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "auto-routing",
      "accuracy": 0.747619,
      "accuracy_std_dev": 0.020292645810736476,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.010855531428571428,
      "avg_duration_per_task_ms": 226454.37142857144,
      "avg_output_tokens_per_task": 13871.608571428571,
      "last_run_timestamp": "2026-10-24T19:37:39.139Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "x-ai/grok-4.6-20260810",
      "model_slug": "x-ai/grok-4.6",
      "model_name": "SpaceXAI: Grok 4.6",
      "provider_name": "xAI",
      "accuracy": 0.747619,
      "accuracy_std_dev": 0.025305279849978234,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.33171734571428574,
      "avg_duration_per_task_ms": 235739.5942857143,
      "avg_output_tokens_per_task": 12164.34,
      "last_run_timestamp": "2026-10-01T16:01:04.865Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Reka",
      "accuracy": 0.7475554666666667,
      "accuracy_std_dev": 0.02281875035978576,
      "run_count": 15,
      "total_questions": 750,
      "avg_cost_per_task_usd": 0.013525546666666666,
      "avg_duration_per_task_ms": 139030.82666666666,
      "avg_output_tokens_per_task": 15549.14,
      "last_run_timestamp": "2026-10-24T12:22:53.795Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "Baidu",
      "accuracy": 0.7475,
      "accuracy_std_dev": 0.02271496080780242,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.06216827,
      "avg_duration_per_task_ms": 205983.465,
      "avg_output_tokens_per_task": 10804.545,
      "last_run_timestamp": "2026-10-24T20:08:17.474Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Mara",
      "accuracy": 0.7473276647230321,
      "accuracy_std_dev": 0.024479710940170173,
      "run_count": 7,
      "total_questions": 343,
      "avg_cost_per_task_usd": 0.13171293294460643,
      "avg_duration_per_task_ms": 209506.55976676385,
      "avg_output_tokens_per_task": 11363.093294460641,
      "last_run_timestamp": "2026-10-24T19:49:56.630Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7472726363636364,
      "accuracy_std_dev": 0.03984348764688863,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.04194175272727273,
      "avg_duration_per_task_ms": 190838.81636363638,
      "avg_output_tokens_per_task": 19126.147272727274,
      "last_run_timestamp": "2026-10-24T12:19:21.144Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "Modal",
      "accuracy": 0.7472221666666666,
      "accuracy_std_dev": 0.011613686586906374,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.17439429333333334,
      "avg_duration_per_task_ms": 161420.59666666668,
      "avg_output_tokens_per_task": 12499.9,
      "last_run_timestamp": "2026-10-10T11:19:48.441Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "BaseTen",
      "accuracy": 0.747143,
      "accuracy_std_dev": 0.033305956785965055,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.014612914285714287,
      "avg_duration_per_task_ms": 50786.11142857143,
      "avg_output_tokens_per_task": 4096.908571428571,
      "last_run_timestamp": "2026-10-24T13:27:44.702Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "DeepInfra",
      "accuracy": 0.7468778674351586,
      "accuracy_std_dev": 0.03317731677283287,
      "run_count": 7,
      "total_questions": 347,
      "avg_cost_per_task_usd": 0.06897791354466859,
      "avg_duration_per_task_ms": 486659.6887608069,
      "avg_output_tokens_per_task": 10547.132564841499,
      "last_run_timestamp": "2026-10-24T19:42:30.512Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "StreamLake",
      "accuracy": 0.746667,
      "accuracy_std_dev": 0.014142135623730963,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.01997175,
      "avg_duration_per_task_ms": 504783.88,
      "avg_output_tokens_per_task": 14879.43,
      "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-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Cloudflare",
      "accuracy": 0.746667,
      "accuracy_std_dev": 0.020000000000000018,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.00955731,
      "avg_duration_per_task_ms": 207347.51,
      "avg_output_tokens_per_task": 13127.37,
      "last_run_timestamp": "2026-10-24T12:22:57.425Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta/muse-spark-open-20261015",
      "model_slug": "meta/muse-spark-open",
      "model_name": "Meta: Muse Spark Open",
      "provider_name": "auto-routing",
      "accuracy": 0.746667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.08652918,
      "avg_duration_per_task_ms": 60257.98,
      "avg_output_tokens_per_task": 8644.22,
      "last_run_timestamp": "2026-10-24T12:40:48.538Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Relace",
      "accuracy": 0.746667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.00930136,
      "avg_duration_per_task_ms": 89811.58,
      "avg_output_tokens_per_task": 12020.36,
      "last_run_timestamp": "2026-10-24T15:18:33.444Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "GMICloud",
      "accuracy": 0.746667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.07358494,
      "avg_duration_per_task_ms": 796913.16,
      "avg_output_tokens_per_task": 13587.26,
      "last_run_timestamp": "2026-10-27T10:59:10.499Z"
    },
    {
      "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-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7466667142857143,
      "accuracy_std_dev": 0.025448285565123643,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.04929559142857143,
      "avg_duration_per_task_ms": 346170.3714285714,
      "avg_output_tokens_per_task": 11334.602857142858,
      "last_run_timestamp": "2026-10-24T21:05:35.420Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Wafer",
      "accuracy": 0.7466666666666667,
      "accuracy_std_dev": 0.019626097517789383,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.05652010888888889,
      "avg_duration_per_task_ms": 356610.63333333336,
      "avg_output_tokens_per_task": 14359.217777777778,
      "last_run_timestamp": "2026-10-29T22:50:33.321Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Crusoe",
      "accuracy": 0.7466666000000001,
      "accuracy_std_dev": 0.030983694638309366,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.022173396,
      "avg_duration_per_task_ms": 122667.268,
      "avg_output_tokens_per_task": 7404.18,
      "last_run_timestamp": "2026-10-24T12:20:04.753Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7466665714285714,
      "accuracy_std_dev": 0.02760272588431981,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.056462091428571434,
      "avg_duration_per_task_ms": 739730.7085714285,
      "avg_output_tokens_per_task": 16026.454285714286,
      "last_run_timestamp": "2026-10-24T19:50:11.361Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta/muse-glimmer-30b-20260810",
      "model_slug": "meta/muse-glimmer-30b",
      "model_name": "Meta: Muse Glimmer 30B",
      "provider_name": "auto-routing",
      "accuracy": 0.7466665714285714,
      "accuracy_std_dev": 0.03380611384628165,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.03743281428571429,
      "avg_duration_per_task_ms": 197093.65714285715,
      "avg_output_tokens_per_task": 13601.3,
      "last_run_timestamp": "2026-10-03T08:15:57.213Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "CoreWeave",
      "accuracy": 0.7466665714285714,
      "accuracy_std_dev": 0.029814069333871523,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.02163002857142857,
      "avg_duration_per_task_ms": 164450.89714285714,
      "avg_output_tokens_per_task": 8529.557142857142,
      "last_run_timestamp": "2026-10-24T18:59:48.400Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Together",
      "accuracy": 0.7466665,
      "accuracy_std_dev": 0.026666500000000037,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.019887960000000003,
      "avg_duration_per_task_ms": 51072.42,
      "avg_output_tokens_per_task": 13340.71,
      "last_run_timestamp": "2026-10-24T12:26:21.165Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Relace",
      "accuracy": 0.7466665,
      "accuracy_std_dev": 0.0466665,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.011416649999999999,
      "avg_duration_per_task_ms": 95996.46,
      "avg_output_tokens_per_task": 13669.03,
      "last_run_timestamp": "2026-10-24T12:07:39.217Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta/muse-glimmer-30b-20260810",
      "model_slug": "meta/muse-glimmer-30b",
      "model_name": "Meta: Muse Glimmer 30B",
      "provider_name": "Fireworks",
      "accuracy": 0.7466665,
      "accuracy_std_dev": 0.011547101609639251,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.04422623666666667,
      "avg_duration_per_task_ms": 266653.04,
      "avg_output_tokens_per_task": 13360.296666666667,
      "last_run_timestamp": "2026-09-19T08:01:36.427Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.5-20260423",
      "model_slug": "openai/gpt-5.5",
      "model_name": "OpenAI: GPT-5.5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7466665,
      "accuracy_std_dev": 0.00666650000000002,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.58250779,
      "avg_duration_per_task_ms": 102747.13,
      "avg_output_tokens_per_task": 7546.85,
      "last_run_timestamp": "2026-10-01T19:04:42.459Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "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-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "Google",
      "accuracy": 0.7461904285714286,
      "accuracy_std_dev": 0.021560263063973885,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.12111663714285714,
      "avg_duration_per_task_ms": 120926.64285714286,
      "avg_output_tokens_per_task": 10598.828571428572,
      "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": "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": 0.7460607272727272,
      "accuracy_std_dev": 0.03259221120867796,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.5475128981818181,
      "avg_duration_per_task_ms": 181106.1872727273,
      "avg_output_tokens_per_task": 10316.370909090909,
      "last_run_timestamp": "2026-10-24T12:39:07.404Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.6-opus-20260205",
      "model_slug": "anthropic/claude-opus-4.6",
      "model_name": "Anthropic: Claude Opus 4.6",
      "provider_name": "Azure",
      "accuracy": 0.7460605454545455,
      "accuracy_std_dev": 0.027035941116708276,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.52474048,
      "avg_duration_per_task_ms": 221889.97272727272,
      "avg_output_tokens_per_task": 10006.24909090909,
      "last_run_timestamp": "2026-10-24T12:47:43.608Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "Xiaomi",
      "accuracy": 0.7460000999999999,
      "accuracy_std_dev": 0.04038166415701563,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.024082161999999997,
      "avg_duration_per_task_ms": 403869.142,
      "avg_output_tokens_per_task": 15439.562,
      "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": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Alibaba",
      "accuracy": 0.7459999,
      "accuracy_std_dev": 0.035080781064423305,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.36040775999999997,
      "avg_duration_per_task_ms": 382799.108,
      "avg_output_tokens_per_task": 9698.124,
      "last_run_timestamp": "2026-10-28T03:50:02.255Z"
    },
    {
      "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": 0.745975738515901,
      "accuracy_std_dev": 0.015812418783302782,
      "run_count": 17,
      "total_questions": 849,
      "avg_cost_per_task_usd": 0.044388724381625444,
      "avg_duration_per_task_ms": 325041.3757361602,
      "avg_output_tokens_per_task": 25777.905771495876,
      "last_run_timestamp": "2026-10-30T01:06:37.160Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "Novita",
      "accuracy": 0.7459258888888889,
      "accuracy_std_dev": 0.025809336792221565,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.17245649777777777,
      "avg_duration_per_task_ms": 324708.56444444443,
      "avg_output_tokens_per_task": 11055.957777777778,
      "last_run_timestamp": "2026-10-24T12:33:16.196Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "AtlasCloud",
      "accuracy": 0.745833375,
      "accuracy_std_dev": 0.0427442661854707,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.11668912499999999,
      "avg_duration_per_task_ms": 240026.82,
      "avg_output_tokens_per_task": 9661.8575,
      "last_run_timestamp": "2026-10-24T19:48:41.460Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Decart",
      "accuracy": 0.74583325,
      "accuracy_std_dev": 0.03402396010442498,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.02499106,
      "avg_duration_per_task_ms": 929627.78,
      "avg_output_tokens_per_task": 52531.49,
      "last_run_timestamp": "2026-10-24T12:31:49.408Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.5-20260423",
      "model_slug": "openai/gpt-5.5",
      "model_name": "OpenAI: GPT-5.5",
      "provider_name": "Azure",
      "accuracy": 0.7457142857142857,
      "accuracy_std_dev": 0.029692331634924993,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.5093009542857143,
      "avg_duration_per_task_ms": 175938.91142857142,
      "avg_output_tokens_per_task": 7450.982857142857,
      "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": "anthropic/claude-4.6-sonnet-20260217",
      "model_slug": "anthropic/claude-sonnet-4.6",
      "model_name": "Anthropic: Claude Sonnet 4.6",
      "provider_name": "Anthropic",
      "accuracy": 0.745641,
      "accuracy_std_dev": 0.025997527320290117,
      "run_count": 13,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.34545777384615384,
      "avg_duration_per_task_ms": 208026.4569230769,
      "avg_output_tokens_per_task": 11868.513846153846,
      "last_run_timestamp": "2026-10-24T13:09:08.696Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "InferenceNet",
      "accuracy": 0.7455556666666666,
      "accuracy_std_dev": 0.020846357411627908,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.018734011111111108,
      "avg_duration_per_task_ms": 376517.3044444444,
      "avg_output_tokens_per_task": 24556.922222222223,
      "last_run_timestamp": "2026-10-24T11:55:52.478Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Parasail",
      "accuracy": 0.7454488757062147,
      "accuracy_std_dev": 0.03153896451944291,
      "run_count": 11,
      "total_questions": 531,
      "avg_cost_per_task_usd": 0.022084410546139358,
      "avg_duration_per_task_ms": 94955.74764595104,
      "avg_output_tokens_per_task": 7543.256120527307,
      "last_run_timestamp": "2026-10-24T18:57:18.438Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.6-sonnet-20260217",
      "model_slug": "anthropic/claude-sonnet-4.6",
      "model_name": "Anthropic: Claude Sonnet 4.6",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.7453334,
      "accuracy_std_dev": 0.02543827737170895,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.346131274,
      "avg_duration_per_task_ms": 200655.086,
      "avg_output_tokens_per_task": 11839.45,
      "last_run_timestamp": "2026-10-24T13:00:12.600Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "Novita",
      "accuracy": 0.7453333,
      "accuracy_std_dev": 0.027936434196403816,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.02067147,
      "avg_duration_per_task_ms": 399151.052,
      "avg_output_tokens_per_task": 11925.912,
      "last_run_timestamp": "2026-10-17T11:29:07.053Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "StreamLake",
      "accuracy": 0.7451852222222222,
      "accuracy_std_dev": 0.02994276662337214,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.06914028222222222,
      "avg_duration_per_task_ms": 256444.49333333335,
      "avg_output_tokens_per_task": 9708.04,
      "last_run_timestamp": "2026-10-24T19:33:48.037Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "DeepInfra",
      "accuracy": 0.7451851111111111,
      "accuracy_std_dev": 0.030270966502370805,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.4099032177777778,
      "avg_duration_per_task_ms": 619397.1977777778,
      "avg_output_tokens_per_task": 9848.733333333334,
      "last_run_timestamp": "2026-10-24T12:05:22.704Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "StreamLake",
      "accuracy": 0.745000125,
      "accuracy_std_dev": 0.03312433437609538,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.015775572499999998,
      "avg_duration_per_task_ms": 148938.23,
      "avg_output_tokens_per_task": 15126.975,
      "last_run_timestamp": "2026-10-24T12:05:54.269Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "SambaNova",
      "accuracy": 0.745,
      "accuracy_std_dev": 0.026822730826670128,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.15284815,
      "avg_duration_per_task_ms": 494527.49,
      "avg_output_tokens_per_task": 28246.935,
      "last_run_timestamp": "2026-10-17T19:47:21.934Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Phala",
      "accuracy": 0.745,
      "accuracy_std_dev": 0.03362364412879724,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.1356366175,
      "avg_duration_per_task_ms": 342258.37,
      "avg_output_tokens_per_task": 11856.1675,
      "last_run_timestamp": "2026-10-24T19:38:37.255Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Relace",
      "accuracy": 0.745,
      "accuracy_std_dev": 0.028332999999999997,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01250761,
      "avg_duration_per_task_ms": 355978.98,
      "avg_output_tokens_per_task": 23604.2,
      "last_run_timestamp": "2026-10-24T12:29:47.596Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Near AI",
      "accuracy": 0.745,
      "accuracy_std_dev": 0.040380248197602775,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.007515549999999999,
      "avg_duration_per_task_ms": 328112.605,
      "avg_output_tokens_per_task": 3998.705,
      "last_run_timestamp": "2026-10-24T13:13:17.945Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Morph",
      "accuracy": 0.7448485454545455,
      "accuracy_std_dev": 0.021944100032763564,
      "run_count": 22,
      "total_questions": 1100,
      "avg_cost_per_task_usd": 0.4515615263636363,
      "avg_duration_per_task_ms": 587567.8727272727,
      "avg_output_tokens_per_task": 10197.047272727274,
      "last_run_timestamp": "2026-10-24T11:44:25.226Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Venice",
      "accuracy": 0.7448484545454546,
      "accuracy_std_dev": 0.024468606225356824,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.14367746363636363,
      "avg_duration_per_task_ms": 628527.9981818182,
      "avg_output_tokens_per_task": 26968.087272727273,
      "last_run_timestamp": "2026-10-24T20:02:49.919Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "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": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "DeepInfra",
      "accuracy": 0.7445560603015074,
      "accuracy_std_dev": 0.018800290223823686,
      "run_count": 8,
      "total_questions": 398,
      "avg_cost_per_task_usd": 0.0472956608040201,
      "avg_duration_per_task_ms": 744919.2989949749,
      "avg_output_tokens_per_task": 14283.391959798995,
      "last_run_timestamp": "2026-10-24T19:19:10.662Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Inceptron",
      "accuracy": 0.7442960934891485,
      "accuracy_std_dev": 0.02614118042622386,
      "run_count": 12,
      "total_questions": 599,
      "avg_cost_per_task_usd": 0.011909569282136896,
      "avg_duration_per_task_ms": 437548.9916527546,
      "avg_output_tokens_per_task": 14218.791318864774,
      "last_run_timestamp": "2026-10-24T12:42:25.209Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Relace",
      "accuracy": 0.7442423636363636,
      "accuracy_std_dev": 0.04102458399827359,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.011285496363636365,
      "avg_duration_per_task_ms": 301862.60363636364,
      "avg_output_tokens_per_task": 16012.072727272727,
      "last_run_timestamp": "2026-10-24T12:46:15.681Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7441665000000001,
      "accuracy_std_dev": 0.036045969878753456,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0744832625,
      "avg_duration_per_task_ms": 216241.835,
      "avg_output_tokens_per_task": 9512.8075,
      "last_run_timestamp": "2026-10-24T19:32:20.110Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "Together",
      "accuracy": 0.744131569416499,
      "accuracy_std_dev": 0.0294471423722235,
      "run_count": 10,
      "total_questions": 497,
      "avg_cost_per_task_usd": 0.06635609456740442,
      "avg_duration_per_task_ms": 1016068.6076458753,
      "avg_output_tokens_per_task": 74326.28571428571,
      "last_run_timestamp": "2026-10-24T19:38:23.433Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "ModelRun",
      "accuracy": 0.7440001333333333,
      "accuracy_std_dev": 0.026195840213964414,
      "run_count": 15,
      "total_questions": 750,
      "avg_cost_per_task_usd": 0.059455711999999994,
      "avg_duration_per_task_ms": 139058.504,
      "avg_output_tokens_per_task": 9866.470666666666,
      "last_run_timestamp": "2026-10-28T07:08:54.400Z"
    },
    {
      "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": "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": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "GMICloud",
      "accuracy": 0.7437037777777777,
      "accuracy_std_dev": 0.029832554190115407,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.020828822222222224,
      "avg_duration_per_task_ms": 153282.48666666666,
      "avg_output_tokens_per_task": 16327.455555555556,
      "last_run_timestamp": "2026-10-27T00:50:04.492Z"
    },
    {
      "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": 0.7436363636363637,
      "accuracy_std_dev": 0.02775984111584323,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.32704656909090907,
      "avg_duration_per_task_ms": 196939.35090909092,
      "avg_output_tokens_per_task": 11566.430909090908,
      "last_run_timestamp": "2026-10-24T12:37:04.719Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Mancer 2",
      "accuracy": 0.7433335,
      "accuracy_std_dev": 0.0033334999999999892,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.06141422999999999,
      "avg_duration_per_task_ms": 1060280.1,
      "avg_output_tokens_per_task": 12094.68,
      "last_run_timestamp": "2026-10-24T12:42:07.189Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "InferenceNet",
      "accuracy": 0.7433335,
      "accuracy_std_dev": 0.023333285714832386,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.006871515,
      "avg_duration_per_task_ms": 97438.585,
      "avg_output_tokens_per_task": 9380.645,
      "last_run_timestamp": "2026-10-24T12:26:07.751Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "StreamLake",
      "accuracy": 0.7433335,
      "accuracy_std_dev": 0.03666649999999999,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.03822515,
      "avg_duration_per_task_ms": 176603.12,
      "avg_output_tokens_per_task": 16434.51,
      "last_run_timestamp": "2026-10-24T18:56:11.637Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "DeepInfra",
      "accuracy": 0.7433335,
      "accuracy_std_dev": 0.01666650000000003,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.03870923,
      "avg_duration_per_task_ms": 112805.69,
      "avg_output_tokens_per_task": 14592.13,
      "last_run_timestamp": "2026-10-24T12:30:47.912Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "x-ai/grok-4.6-20260810",
      "model_slug": "x-ai/grok-4.6",
      "model_name": "SpaceXAI: Grok 4.6",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7433335,
      "accuracy_std_dev": 0.043333500000000025,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.64823893,
      "avg_duration_per_task_ms": 264133.83,
      "avg_output_tokens_per_task": 10511,
      "last_run_timestamp": "2026-10-01T16:00:37.003Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Cloudflare",
      "accuracy": 0.7433333333333333,
      "accuracy_std_dev": 0.017532868606008445,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.012350623333333333,
      "avg_duration_per_task_ms": 125234.99333333333,
      "avg_output_tokens_per_task": 4198.8,
      "last_run_timestamp": "2026-10-03T08:23:09.816Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.7433333,
      "accuracy_std_dev": 0.03441571669470216,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.028436006,
      "avg_duration_per_task_ms": 438756.644,
      "avg_output_tokens_per_task": 14726.93,
      "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": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "Google",
      "accuracy": 0.7433332500000001,
      "accuracy_std_dev": 0.01774311029726281,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.013128406666666667,
      "avg_duration_per_task_ms": 72447.74833333334,
      "avg_output_tokens_per_task": 10872.788333333334,
      "last_run_timestamp": "2026-10-24T12:05:47.786Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.7433332500000001,
      "accuracy_std_dev": 0.023804656407675817,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.055247185,
      "avg_duration_per_task_ms": 380831.14,
      "avg_output_tokens_per_task": 8536.295,
      "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": "Cloudflare",
      "accuracy": 0.742963111111111,
      "accuracy_std_dev": 0.025936488854483866,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.07987847333333332,
      "avg_duration_per_task_ms": 298551.6066666667,
      "avg_output_tokens_per_task": 11298.702222222222,
      "last_run_timestamp": "2026-10-24T23:01:57.882Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-opus-5.5-20260921",
      "model_slug": "anthropic/claude-opus-5.5",
      "model_name": "Anthropic: Claude Opus 5.5",
      "provider_name": "Google",
      "accuracy": 0.7429629444444444,
      "accuracy_std_dev": 0.018622075637599298,
      "run_count": 18,
      "total_questions": 900,
      "avg_cost_per_task_usd": 0.35968263222222224,
      "avg_duration_per_task_ms": 87204.23444444445,
      "avg_output_tokens_per_task": 7123.5233333333335,
      "last_run_timestamp": "2026-10-24T12:18:54.175Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "GMICloud",
      "accuracy": 0.7428572857142857,
      "accuracy_std_dev": 0.03884962929300717,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.025685354285714286,
      "avg_duration_per_task_ms": 140747.64,
      "avg_output_tokens_per_task": 11913.908571428572,
      "last_run_timestamp": "2026-10-27T01:02:57.660Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "AkashML",
      "accuracy": 0.7426668,
      "accuracy_std_dev": 0.03690245640821219,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.011645404,
      "avg_duration_per_task_ms": 585840.916,
      "avg_output_tokens_per_task": 18794.172,
      "last_run_timestamp": "2026-08-29T08:43:07.210Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.6-opus-20260205",
      "model_slug": "anthropic/claude-opus-4.6",
      "model_name": "Anthropic: Claude Opus 4.6",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.7426666000000001,
      "accuracy_std_dev": 0.02908997383704564,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.504919826,
      "avg_duration_per_task_ms": 191444.46,
      "avg_output_tokens_per_task": 9912.906,
      "last_run_timestamp": "2026-10-24T11:43:06.361Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7426265425709516,
      "accuracy_std_dev": 0.027093567829324792,
      "run_count": 12,
      "total_questions": 599,
      "avg_cost_per_task_usd": 0.04797976794657762,
      "avg_duration_per_task_ms": 180887.38898163606,
      "avg_output_tokens_per_task": 9161.572621035059,
      "last_run_timestamp": "2026-10-24T19:33:32.803Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Sail Research",
      "accuracy": 0.7425926666666667,
      "accuracy_std_dev": 0.03165311687654157,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.013586744444444443,
      "avg_duration_per_task_ms": 236886.98,
      "avg_output_tokens_per_task": 22693.746666666666,
      "last_run_timestamp": "2026-10-24T12:29:34.092Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "DeepInfra",
      "accuracy": 0.7425595803571429,
      "accuracy_std_dev": 0.02504151285854947,
      "run_count": 7,
      "total_questions": 336,
      "avg_cost_per_task_usd": 0.15002946726190475,
      "avg_duration_per_task_ms": 711525.6666666666,
      "avg_output_tokens_per_task": 29928.91369047619,
      "last_run_timestamp": "2026-10-24T19:30:34.911Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "Together",
      "accuracy": 0.7425,
      "accuracy_std_dev": 0.01697637247028941,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.22007109249999998,
      "avg_duration_per_task_ms": 261814.77,
      "avg_output_tokens_per_task": 11695.5675,
      "last_run_timestamp": "2026-10-24T11:58:56.378Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Venice",
      "accuracy": 0.7425,
      "accuracy_std_dev": 0.03281884621296123,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.101450235,
      "avg_duration_per_task_ms": 312758.3475,
      "avg_output_tokens_per_task": 7418.4,
      "last_run_timestamp": "2026-10-24T19:37:38.705Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.7422716733870968,
      "accuracy_std_dev": 0.017259512677071726,
      "run_count": 10,
      "total_questions": 496,
      "avg_cost_per_task_usd": 0.01751647782258065,
      "avg_duration_per_task_ms": 79788.1754032258,
      "avg_output_tokens_per_task": 4101.441532258064,
      "last_run_timestamp": "2026-10-28T10:48:35.415Z"
    },
    {
      "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": "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": 0.7422223333333333,
      "accuracy_std_dev": 0.016629618082071397,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.27693500666666665,
      "avg_duration_per_task_ms": 100083.4,
      "avg_output_tokens_per_task": 5197.013333333333,
      "last_run_timestamp": "2026-10-30T00:15:31.633Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Reka",
      "accuracy": 0.7417778,
      "accuracy_std_dev": 0.0192398742345162,
      "run_count": 15,
      "total_questions": 750,
      "avg_cost_per_task_usd": 0.016580628,
      "avg_duration_per_task_ms": 168506.08666666667,
      "avg_output_tokens_per_task": 4247.114666666666,
      "last_run_timestamp": "2026-10-24T13:06:52.407Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "CoreWeave",
      "accuracy": 0.7416667499999999,
      "accuracy_std_dev": 0.02074966532712037,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.019690735,
      "avg_duration_per_task_ms": 215383.61,
      "avg_output_tokens_per_task": 12170.26,
      "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": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "Anthropic",
      "accuracy": 0.7416665,
      "accuracy_std_dev": 0.03140250343921645,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.012469425000000001,
      "avg_duration_per_task_ms": 90319.205,
      "avg_output_tokens_per_task": 10893.135,
      "last_run_timestamp": "2026-10-24T12:16:35.471Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.7413793735632184,
      "accuracy_std_dev": 0.05076740924399175,
      "run_count": 7,
      "total_questions": 348,
      "avg_cost_per_task_usd": 0.08319576436781609,
      "avg_duration_per_task_ms": 840263.8620689656,
      "avg_output_tokens_per_task": 51575.90517241379,
      "last_run_timestamp": "2026-10-24T18:56:22.719Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Relace",
      "accuracy": 0.7410191879699248,
      "accuracy_std_dev": 0.023857300276526983,
      "run_count": 8,
      "total_questions": 399,
      "avg_cost_per_task_usd": 0.22161520300751877,
      "avg_duration_per_task_ms": 375984.74686716794,
      "avg_output_tokens_per_task": 9942.047619047618,
      "last_run_timestamp": "2026-10-24T12:46:10.015Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-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-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "Novita",
      "accuracy": 0.7406668999999999,
      "accuracy_std_dev": 0.04294179161015525,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.009957568,
      "avg_duration_per_task_ms": 419095.964,
      "avg_output_tokens_per_task": 14787.174,
      "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": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "auto-routing",
      "accuracy": 0.7406667,
      "accuracy_std_dev": 0.05001767643153769,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.009102672,
      "avg_duration_per_task_ms": 299089.77,
      "avg_output_tokens_per_task": 15164.772,
      "last_run_timestamp": "2026-10-24T11:53:54.891Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7405555,
      "accuracy_std_dev": 0.01855080944101005,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.0133585,
      "avg_duration_per_task_ms": 101370.275,
      "avg_output_tokens_per_task": 10830.071666666667,
      "last_run_timestamp": "2026-10-24T11:56:06.063Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Fireworks",
      "accuracy": 0.7403685728643216,
      "accuracy_std_dev": 0.02451522166373828,
      "run_count": 12,
      "total_questions": 597,
      "avg_cost_per_task_usd": 0.05593858793969849,
      "avg_duration_per_task_ms": 169768.54773869348,
      "avg_output_tokens_per_task": 30975.941373534337,
      "last_run_timestamp": "2026-10-24T12:06:17.420Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7403274419642857,
      "accuracy_std_dev": 0.036586619279744714,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.07693905803571428,
      "avg_duration_per_task_ms": 873819.4196428572,
      "avg_output_tokens_per_task": 14604.90625,
      "last_run_timestamp": "2026-10-24T20:35:30.856Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.6-sonnet-20260217",
      "model_slug": "anthropic/claude-sonnet-4.6",
      "model_name": "Anthropic: Claude Sonnet 4.6",
      "provider_name": "Google",
      "accuracy": 0.7402083125000002,
      "accuracy_std_dev": 0.030068559537369668,
      "run_count": 32,
      "total_questions": 1600,
      "avg_cost_per_task_usd": 0.37145770687499996,
      "avg_duration_per_task_ms": 200089.441875,
      "avg_output_tokens_per_task": 11938.27625,
      "last_run_timestamp": "2026-10-24T13:14:09.404Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "AkashML",
      "accuracy": 0.7401633919821825,
      "accuracy_std_dev": 0.02335439453822158,
      "run_count": 9,
      "total_questions": 449,
      "avg_cost_per_task_usd": 0.061262200445434296,
      "avg_duration_per_task_ms": 141198.91759465478,
      "avg_output_tokens_per_task": 4110.719376391982,
      "last_run_timestamp": "2026-10-30T02:02:14.900Z"
    },
    {
      "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": 0.7400001666666666,
      "accuracy_std_dev": 0.01154690915954955,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.034900296666666664,
      "avg_duration_per_task_ms": 129527.63333333333,
      "avg_output_tokens_per_task": 16132.923333333334,
      "last_run_timestamp": "2026-10-27T00:42:49.766Z"
    },
    {
      "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": 0.740000125,
      "accuracy_std_dev": 0.02924983380908983,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.072280915,
      "avg_duration_per_task_ms": 322638.715,
      "avg_output_tokens_per_task": 11472.485,
      "last_run_timestamp": "2026-10-24T22:27:21.158Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Modal",
      "accuracy": 0.7400001111111111,
      "accuracy_std_dev": 0.021314690929781274,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.3873409377777778,
      "avg_duration_per_task_ms": 192992.06,
      "avg_output_tokens_per_task": 11071.72,
      "last_run_timestamp": "2026-10-24T12:36:29.628Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "amazon/nova-micro-v1",
      "model_slug": "amazon/nova-micro-v1",
      "model_name": "Amazon: Nova Micro 1.0",
      "provider_name": "Anthropic",
      "accuracy": 0.74,
      "accuracy_std_dev": 0.0066669999999999785,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 1.23335095,
      "avg_duration_per_task_ms": 96091.26,
      "avg_output_tokens_per_task": 5555.9,
      "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": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Mistral",
      "accuracy": 0.74,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.047416770000000004,
      "avg_duration_per_task_ms": 82837.68,
      "avg_output_tokens_per_task": 13780.79,
      "last_run_timestamp": "2026-10-30T16:15:51.889Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "OpenInference",
      "accuracy": 0.74,
      "accuracy_std_dev": 0.020000000000000018,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.02650346,
      "avg_duration_per_task_ms": 245657.46,
      "avg_output_tokens_per_task": 33745.71,
      "last_run_timestamp": "2026-10-24T12:08:32.744Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.74,
      "accuracy_std_dev": 0.0066669999999999785,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.37915447,
      "avg_duration_per_task_ms": 148084.91,
      "avg_output_tokens_per_task": 12525.4,
      "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": "meta/muse-spark-open-20261015",
      "model_slug": "meta/muse-spark-open",
      "model_name": "Meta: Muse Spark Open",
      "provider_name": "Together",
      "accuracy": 0.74,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.08715152,
      "avg_duration_per_task_ms": 69751.7,
      "avg_output_tokens_per_task": 8640,
      "last_run_timestamp": "2026-10-24T11:33:58.313Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Cohere",
      "accuracy": 0.74,
      "accuracy_std_dev": 0.01333300000000004,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01884005,
      "avg_duration_per_task_ms": 67349.35,
      "avg_output_tokens_per_task": 22014.7,
      "last_run_timestamp": "2026-10-24T11:48:47.797Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.74,
      "accuracy_std_dev": 0.009427854713560266,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.022918865,
      "avg_duration_per_task_ms": 72639.31,
      "avg_output_tokens_per_task": 4366.75,
      "last_run_timestamp": "2026-10-29T23:37:44.528Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.5-sonnet-20250929",
      "model_slug": "anthropic/claude-sonnet-4.5",
      "model_name": "Anthropic: Claude Sonnet 4.5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.7399999,
      "accuracy_std_dev": 0.028906131607152153,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.34949705000000003,
      "avg_duration_per_task_ms": 220413.082,
      "avg_output_tokens_per_task": 10856.716,
      "last_run_timestamp": "2026-10-24T12:39:24.524Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Morph",
      "accuracy": 0.7399998571428572,
      "accuracy_std_dev": 0.023094052007442304,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.01886730285714286,
      "avg_duration_per_task_ms": 206247.68,
      "avg_output_tokens_per_task": 4550.928571428572,
      "last_run_timestamp": "2026-10-10T11:50:34.352Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Makora",
      "accuracy": 0.7395835,
      "accuracy_std_dev": 0.022012461748973016,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.042535722500000005,
      "avg_duration_per_task_ms": 134957.37,
      "avg_output_tokens_per_task": 25971.215,
      "last_run_timestamp": "2026-10-23T03:54:04.118Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Mancer 2",
      "accuracy": 0.7393333,
      "accuracy_std_dev": 0.04912106993144592,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.04842688399999999,
      "avg_duration_per_task_ms": 284804.346,
      "avg_output_tokens_per_task": 14992.81,
      "last_run_timestamp": "2026-10-24T12:17:04.178Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7392592222222222,
      "accuracy_std_dev": 0.04548160272700695,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.16905808444444445,
      "avg_duration_per_task_ms": 289473.89555555553,
      "avg_output_tokens_per_task": 9765.786666666667,
      "last_run_timestamp": "2026-10-24T11:42:23.888Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Phala",
      "accuracy": 0.7392592222222222,
      "accuracy_std_dev": 0.03955854406875342,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.12468993111111111,
      "avg_duration_per_task_ms": 675089.5488888889,
      "avg_output_tokens_per_task": 35068.42222222222,
      "last_run_timestamp": "2026-10-24T11:39:42.358Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Z.AI",
      "accuracy": 0.739166625,
      "accuracy_std_dev": 0.013919448327228176,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.11955471250000001,
      "avg_duration_per_task_ms": 375155.945,
      "avg_output_tokens_per_task": 10031.9425,
      "last_run_timestamp": "2026-10-24T20:00:28.580Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "auto-routing",
      "accuracy": 0.7391665000000001,
      "accuracy_std_dev": 0.021196743723978007,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.06600346500000001,
      "avg_duration_per_task_ms": 259839.1925,
      "avg_output_tokens_per_task": 12680.6925,
      "last_run_timestamp": "2026-10-24T20:26:13.288Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Novita",
      "accuracy": 0.7387879090909091,
      "accuracy_std_dev": 0.03704407867651368,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.02604207454545454,
      "avg_duration_per_task_ms": 112310.20909090909,
      "avg_output_tokens_per_task": 9031.76,
      "last_run_timestamp": "2026-10-24T19:07:29.704Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "Groq",
      "accuracy": 0.7385729799331104,
      "accuracy_std_dev": 0.028249376510327148,
      "run_count": 6,
      "total_questions": 299,
      "avg_cost_per_task_usd": 0.1314999698996656,
      "avg_duration_per_task_ms": 63487.39464882943,
      "avg_output_tokens_per_task": 9641.89297658863,
      "last_run_timestamp": "2026-10-24T18:49:16.415Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Makora",
      "accuracy": 0.7384824959349594,
      "accuracy_std_dev": 0.030529621622286764,
      "run_count": 5,
      "total_questions": 246,
      "avg_cost_per_task_usd": 0.012032963414634147,
      "avg_duration_per_task_ms": 61127.26829268293,
      "avg_output_tokens_per_task": 4200.292682926829,
      "last_run_timestamp": "2026-10-24T12:38:23.545Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "DeepInfra",
      "accuracy": 0.7383335,
      "accuracy_std_dev": 0.0016665000000000152,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.08170183999999998,
      "avg_duration_per_task_ms": 327224.56,
      "avg_output_tokens_per_task": 8775.96,
      "last_run_timestamp": "2026-08-22T16:12:20.461Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "Novita",
      "accuracy": 0.73833325,
      "accuracy_std_dev": 0.017240319347027768,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.063434195,
      "avg_duration_per_task_ms": 520281.665,
      "avg_output_tokens_per_task": 14632.895,
      "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-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "DeepInfra",
      "accuracy": 0.73833325,
      "accuracy_std_dev": 0.03210231473254696,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.03571221,
      "avg_duration_per_task_ms": 388203.89,
      "avg_output_tokens_per_task": 8383.965,
      "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": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Parasail",
      "accuracy": 0.73833325,
      "accuracy_std_dev": 0.02153795745161318,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.34754801999999996,
      "avg_duration_per_task_ms": 197375.055,
      "avg_output_tokens_per_task": 8811.4275,
      "last_run_timestamp": "2026-10-24T12:17:34.198Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "StreamLake",
      "accuracy": 0.7382021797752809,
      "accuracy_std_dev": 0.021880590311850887,
      "run_count": 9,
      "total_questions": 445,
      "avg_cost_per_task_usd": 0.020498339325842697,
      "avg_duration_per_task_ms": 307841.11235955055,
      "avg_output_tokens_per_task": 10994.101123595505,
      "last_run_timestamp": "2026-10-17T11:09:59.070Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Venice",
      "accuracy": 0.7380952857142857,
      "accuracy_std_dev": 0.031414080354035444,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.08033971999999999,
      "avg_duration_per_task_ms": 526120.5771428571,
      "avg_output_tokens_per_task": 11993.431428571428,
      "last_run_timestamp": "2026-10-24T23:08:37.435Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "Venice",
      "accuracy": 0.7380950401002507,
      "accuracy_std_dev": 0.022649397541656617,
      "run_count": 8,
      "total_questions": 399,
      "avg_cost_per_task_usd": 0.053091303258145366,
      "avg_duration_per_task_ms": 777194.3383458647,
      "avg_output_tokens_per_task": 14178.676691729323,
      "last_run_timestamp": "2026-10-24T19:40:41.765Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Together",
      "accuracy": 0.7377776666666667,
      "accuracy_std_dev": 0.017777750001367207,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.014543953333333333,
      "avg_duration_per_task_ms": 73473.39777777778,
      "avg_output_tokens_per_task": 4452.611111111111,
      "last_run_timestamp": "2026-10-24T12:18:30.719Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.7375757272727272,
      "accuracy_std_dev": 0.02655621201797806,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.07909413272727274,
      "avg_duration_per_task_ms": 154073.8890909091,
      "avg_output_tokens_per_task": 8791.48909090909,
      "last_run_timestamp": "2026-10-24T19:32:01.303Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Together",
      "accuracy": 0.7375756363636364,
      "accuracy_std_dev": 0.03206980449035491,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.3100524036363636,
      "avg_duration_per_task_ms": 339490.60909090907,
      "avg_output_tokens_per_task": 9723.709090909091,
      "last_run_timestamp": "2026-10-24T13:17:19.762Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "Venice",
      "accuracy": 0.737500125,
      "accuracy_std_dev": 0.03799676247207615,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.042760110000000004,
      "avg_duration_per_task_ms": 594181.5325,
      "avg_output_tokens_per_task": 9082.065,
      "last_run_timestamp": "2026-10-24T19:13:45.181Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Moonshot AI",
      "accuracy": 0.7374243541247485,
      "accuracy_std_dev": 0.021992186721879213,
      "run_count": 10,
      "total_questions": 497,
      "avg_cost_per_task_usd": 0.34455653319919516,
      "avg_duration_per_task_ms": 409342.0885311871,
      "avg_output_tokens_per_task": 9871.545271629779,
      "last_run_timestamp": "2026-10-24T12:29:09.762Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Ionstream",
      "accuracy": 0.7373334,
      "accuracy_std_dev": 0.020482933819157834,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.039714615999999994,
      "avg_duration_per_task_ms": 174773.632,
      "avg_output_tokens_per_task": 27822.08,
      "last_run_timestamp": "2026-10-21T12:36:51.217Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7371794615384615,
      "accuracy_std_dev": 0.04268194927847135,
      "run_count": 10,
      "total_questions": 494,
      "avg_cost_per_task_usd": 0.025897730769230763,
      "avg_duration_per_task_ms": 112126.24089068826,
      "avg_output_tokens_per_task": 9272.680161943319,
      "last_run_timestamp": "2026-10-24T19:18:47.147Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Phala",
      "accuracy": 0.7371428571428571,
      "accuracy_std_dev": 0.023866691370602513,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.022481400000000002,
      "avg_duration_per_task_ms": 144173.6,
      "avg_output_tokens_per_task": 16498.17142857143,
      "last_run_timestamp": "2026-10-24T12:33:37.198Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Wafer",
      "accuracy": 0.7366890939597315,
      "accuracy_std_dev": 0.025312184492760705,
      "run_count": 15,
      "total_questions": 745,
      "avg_cost_per_task_usd": 0.0245328711409396,
      "avg_duration_per_task_ms": 223382.59731543623,
      "avg_output_tokens_per_task": 20623.554362416107,
      "last_run_timestamp": "2026-10-29T23:36:12.571Z"
    },
    {
      "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": 0.736667,
      "accuracy_std_dev": 0.010000000000000009,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.47698523000000004,
      "avg_duration_per_task_ms": 128774.34,
      "avg_output_tokens_per_task": 7206.5,
      "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": "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": "SiliconFlow",
      "accuracy": 0.736666625,
      "accuracy_std_dev": 0.03018462173001967,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.16366423000000002,
      "avg_duration_per_task_ms": 225879.0775,
      "avg_output_tokens_per_task": 9618.825,
      "last_run_timestamp": "2026-10-24T19:37:14.525Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Fireworks",
      "accuracy": 0.7366665,
      "accuracy_std_dev": 0.01666650000000003,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.17216302,
      "avg_duration_per_task_ms": 234612.69,
      "avg_output_tokens_per_task": 17080.8,
      "last_run_timestamp": "2026-09-19T08:12:02.180Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Phala",
      "accuracy": 0.7366665,
      "accuracy_std_dev": 0.0033334999999999892,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.020005850000000002,
      "avg_duration_per_task_ms": 329585.74,
      "avg_output_tokens_per_task": 11689.47,
      "last_run_timestamp": "2026-08-29T16:43:42.382Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "reflection/beam-1.0-501b-a23b-20261016",
      "model_slug": "reflection/beam-1.0-501b-a23b",
      "model_name": "Reflection: Beam 1.0 501B A23B",
      "provider_name": "Io Net",
      "accuracy": 0.7366665,
      "accuracy_std_dev": 0.0033334999999999892,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.02539905,
      "avg_duration_per_task_ms": 216631.23,
      "avg_output_tokens_per_task": 14336.27,
      "last_run_timestamp": "2026-10-24T12:02:36.303Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7366665,
      "accuracy_std_dev": 0.014529701381996819,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.057118495000000005,
      "avg_duration_per_task_ms": 215486.77,
      "avg_output_tokens_per_task": 9145.625,
      "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": "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": 0.7365744652049571,
      "accuracy_std_dev": 0.02859181594676679,
      "run_count": 21,
      "total_questions": 1049,
      "avg_cost_per_task_usd": 0.13010050238322207,
      "avg_duration_per_task_ms": 159321.12011439467,
      "avg_output_tokens_per_task": 9364.21639656816,
      "last_run_timestamp": "2026-10-24T19:48:18.775Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Makora",
      "accuracy": 0.7365591209677419,
      "accuracy_std_dev": 0.016931426040354632,
      "run_count": 5,
      "total_questions": 248,
      "avg_cost_per_task_usd": 0.23607216935483874,
      "avg_duration_per_task_ms": 378114.625,
      "avg_output_tokens_per_task": 11505.427419354839,
      "last_run_timestamp": "2026-10-24T11:40:39.669Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "StreamLake",
      "accuracy": 0.7362962222222222,
      "accuracy_std_dev": 0.03939194698236847,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.06971035555555556,
      "avg_duration_per_task_ms": 294980.4511111111,
      "avg_output_tokens_per_task": 12708.4,
      "last_run_timestamp": "2026-10-24T19:51:25.882Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Inceptron",
      "accuracy": 0.7362762003395587,
      "accuracy_std_dev": 0.03269045093786308,
      "run_count": 12,
      "total_questions": 589,
      "avg_cost_per_task_usd": 0.06473759422750425,
      "avg_duration_per_task_ms": 132369.44312393887,
      "avg_output_tokens_per_task": 8373.801358234296,
      "last_run_timestamp": "2026-10-24T19:13:30.742Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7361904285714286,
      "accuracy_std_dev": 0.035964376429450086,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.024463299999999997,
      "avg_duration_per_task_ms": 153502.6,
      "avg_output_tokens_per_task": 16394.091428571428,
      "last_run_timestamp": "2026-10-24T11:47:20.359Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "Xiaomi",
      "accuracy": 0.7359999,
      "accuracy_std_dev": 0.025157337786220563,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.008132438,
      "avg_duration_per_task_ms": 403849.82,
      "avg_output_tokens_per_task": 14485.44,
      "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": "google/gemini-3.5-flash-20260519",
      "model_slug": "google/gemini-3.5-flash",
      "model_name": "Google: Gemini 3.5 Flash",
      "provider_name": "Google",
      "accuracy": 0.7359998000000001,
      "accuracy_std_dev": 0.019596013200648757,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.3276280200000001,
      "avg_duration_per_task_ms": 156068.02,
      "avg_output_tokens_per_task": 21546.7,
      "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": "openai/gpt-5.5-20260423",
      "model_slug": "openai/gpt-5.5",
      "model_name": "OpenAI: GPT-5.5",
      "provider_name": "OpenAI",
      "accuracy": 0.7355556666666666,
      "accuracy_std_dev": 0.04157400066013479,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.48569376,
      "avg_duration_per_task_ms": 125885.67333333334,
      "avg_output_tokens_per_task": 7296.606666666667,
      "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-v2.5-20260422",
      "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": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "DeepInfra",
      "accuracy": 0.7353332,
      "accuracy_std_dev": 0.03370472568290686,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.033665999999999995,
      "avg_duration_per_task_ms": 426833.488,
      "avg_output_tokens_per_task": 9005.292,
      "last_run_timestamp": "2026-10-24T20:40:57.027Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.4-20260305",
      "model_slug": "openai/gpt-5.4",
      "model_name": "OpenAI: GPT-5.4",
      "provider_name": "Azure",
      "accuracy": 0.7352380000000001,
      "accuracy_std_dev": 0.01698343330593503,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.3052330542857143,
      "avg_duration_per_task_ms": 191492.84285714285,
      "avg_output_tokens_per_task": 10672.025714285714,
      "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-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Phala",
      "accuracy": 0.7351191205357143,
      "accuracy_std_dev": 0.049992352433912764,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.30342583258928574,
      "avg_duration_per_task_ms": 387074.1026785714,
      "avg_output_tokens_per_task": 11534.020089285714,
      "last_run_timestamp": "2026-10-24T13:33:47.625Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.735,
      "accuracy_std_dev": 0.026822730826670128,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.012519140000000002,
      "avg_duration_per_task_ms": 95304.945,
      "avg_output_tokens_per_task": 10824.09,
      "last_run_timestamp": "2026-10-24T11:50:10.636Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Crusoe",
      "accuracy": 0.735,
      "accuracy_std_dev": 0.03227493867926013,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0975708025,
      "avg_duration_per_task_ms": 334496.3,
      "avg_output_tokens_per_task": 11110.8,
      "last_run_timestamp": "2026-10-24T20:22:47.990Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.8-flash-20260902",
      "model_slug": "google/gemini-3.8-flash",
      "model_name": "Google: Gemini 3.8 Flash",
      "provider_name": "Google AI Studio",
      "accuracy": 0.734694,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 49,
      "avg_cost_per_task_usd": 0.25339910204081634,
      "avg_duration_per_task_ms": 146531.16326530612,
      "avg_output_tokens_per_task": 29831.102040816328,
      "last_run_timestamp": "2026-10-01T08:03:06.861Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Relace",
      "accuracy": 0.73458325,
      "accuracy_std_dev": 0.027785801480927293,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0151816525,
      "avg_duration_per_task_ms": 268037.2675,
      "avg_output_tokens_per_task": 21770.385,
      "last_run_timestamp": "2026-10-24T12:56:13.097Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Baidu",
      "accuracy": 0.73416675,
      "accuracy_std_dev": 0.015789390779491794,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.07111907,
      "avg_duration_per_task_ms": 230769.835,
      "avg_output_tokens_per_task": 13083.0325,
      "last_run_timestamp": "2026-10-24T19:41:01.989Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Mistral",
      "accuracy": 0.73416675,
      "accuracy_std_dev": 0.033530506426051786,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.02004302,
      "avg_duration_per_task_ms": 88766.04,
      "avg_output_tokens_per_task": 5752.91,
      "last_run_timestamp": "2026-10-30T16:30:54.187Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "Azure",
      "accuracy": 0.7341665,
      "accuracy_std_dev": 0.02066601477668108,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.012598340000000001,
      "avg_duration_per_task_ms": 74617.86,
      "avg_output_tokens_per_task": 10550.06,
      "last_run_timestamp": "2026-10-24T11:45:42.778Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Morph",
      "accuracy": 0.7340742222222222,
      "accuracy_std_dev": 0.027296881245453555,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.017404215555555554,
      "avg_duration_per_task_ms": 575071.7644444444,
      "avg_output_tokens_per_task": 18088.03111111111,
      "last_run_timestamp": "2026-10-24T11:35:52.428Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "GMICloud",
      "accuracy": 0.7340741111111111,
      "accuracy_std_dev": 0.03257589639747104,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.05069578000000001,
      "avg_duration_per_task_ms": 248940.97333333333,
      "avg_output_tokens_per_task": 10685.184444444445,
      "last_run_timestamp": "2026-10-27T01:22:44.294Z"
    },
    {
      "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": 0.7339394545454545,
      "accuracy_std_dev": 0.03041204391613666,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.026801145454545455,
      "avg_duration_per_task_ms": 98249,
      "avg_output_tokens_per_task": 9318.925454545455,
      "last_run_timestamp": "2026-10-24T19:08:17.063Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Inceptron",
      "accuracy": 0.7339181879699248,
      "accuracy_std_dev": 0.03240656404548892,
      "run_count": 8,
      "total_questions": 399,
      "avg_cost_per_task_usd": 0.06522364912280701,
      "avg_duration_per_task_ms": 152721.04010025063,
      "avg_output_tokens_per_task": 11337.393483709273,
      "last_run_timestamp": "2026-10-24T19:41:50.588Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.7338889166666667,
      "accuracy_std_dev": 0.03382511166391603,
      "run_count": 24,
      "total_questions": 1200,
      "avg_cost_per_task_usd": 0.3533906716666666,
      "avg_duration_per_task_ms": 207450.37666666668,
      "avg_output_tokens_per_task": 10346.453333333333,
      "last_run_timestamp": "2026-10-24T12:18:44.336Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "Novita",
      "accuracy": 0.733796,
      "accuracy_std_dev": 0.0059017228557995535,
      "run_count": 3,
      "total_questions": 144,
      "avg_cost_per_task_usd": 0.20467558333333333,
      "avg_duration_per_task_ms": 370954.7361111111,
      "avg_output_tokens_per_task": 23212.9375,
      "last_run_timestamp": "2026-10-24T19:18:46.938Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Ionstream",
      "accuracy": 0.7333335,
      "accuracy_std_dev": 0.033333500000000016,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.018280730000000002,
      "avg_duration_per_task_ms": 88642.31,
      "avg_output_tokens_per_task": 11899.17,
      "last_run_timestamp": "2026-10-24T11:54:38.633Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3-flash-preview-20251217",
      "model_slug": "google/gemini-3-flash-preview",
      "model_name": "Google: Gemini 3 Flash Preview",
      "provider_name": "Google",
      "accuracy": 0.7333335,
      "accuracy_std_dev": 0.00666650000000002,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.12301775,
      "avg_duration_per_task_ms": 155307.55,
      "avg_output_tokens_per_task": 18812.75,
      "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": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Phala",
      "accuracy": 0.7333335,
      "accuracy_std_dev": 0.013333499999999998,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.08394800999999999,
      "avg_duration_per_task_ms": 134148.76,
      "avg_output_tokens_per_task": 8723.94,
      "last_run_timestamp": "2026-10-17T19:55:39.408Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "Together",
      "accuracy": 0.7333335,
      "accuracy_std_dev": 0.013333499999999998,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.047618709999999995,
      "avg_duration_per_task_ms": 115029.78,
      "avg_output_tokens_per_task": 15522.32,
      "last_run_timestamp": "2026-10-24T12:31:48.669Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.7333335,
      "accuracy_std_dev": 0.026666499999999982,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.12687428,
      "avg_duration_per_task_ms": 110480.49,
      "avg_output_tokens_per_task": 5152.54,
      "last_run_timestamp": "2026-10-01T19:04:55.696Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "Novita",
      "accuracy": 0.733333375,
      "accuracy_std_dev": 0.027688655921412557,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.05093239000000001,
      "avg_duration_per_task_ms": 359484.675,
      "avg_output_tokens_per_task": 10187.6175,
      "last_run_timestamp": "2026-10-24T20:19:38.537Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Decart",
      "accuracy": 0.7333333333333333,
      "accuracy_std_dev": 0.005443174461857933,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.008480786666666665,
      "avg_duration_per_task_ms": 397846.0266666667,
      "avg_output_tokens_per_task": 19127.326666666668,
      "last_run_timestamp": "2026-08-22T08:38:00.859Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "InferenceNet",
      "accuracy": 0.7333333333333333,
      "accuracy_std_dev": 0.05760637824832321,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.04898614,
      "avg_duration_per_task_ms": 106829.63333333333,
      "avg_output_tokens_per_task": 9897.326666666666,
      "last_run_timestamp": "2026-10-10T11:51:48.927Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.1-terminus",
      "model_slug": "deepseek/deepseek-v3.1-terminus",
      "model_name": "DeepSeek: DeepSeek V3.1 Terminus",
      "provider_name": "Novita",
      "accuracy": 0.73333325,
      "accuracy_std_dev": 0.02357013176007932,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.06276583000000001,
      "avg_duration_per_task_ms": 510113.315,
      "avg_output_tokens_per_task": 14674.185,
      "last_run_timestamp": "2026-08-29T16:33:37.924Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "CoreWeave",
      "accuracy": 0.73333325,
      "accuracy_std_dev": 0.02357013176007932,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.04045966999999999,
      "avg_duration_per_task_ms": 165902.775,
      "avg_output_tokens_per_task": 35276.36,
      "last_run_timestamp": "2026-10-24T12:26:22.907Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "auto-routing",
      "accuracy": 0.7333332,
      "accuracy_std_dev": 0.026499341711823724,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.29219796600000003,
      "avg_duration_per_task_ms": 358228.216,
      "avg_output_tokens_per_task": 10539.826,
      "last_run_timestamp": "2026-10-24T12:45:08.706Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "NextBit",
      "accuracy": 0.7333330000000001,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01186947,
      "avg_duration_per_task_ms": 176514.74,
      "avg_output_tokens_per_task": 14201.61,
      "last_run_timestamp": "2026-08-29T16:50:07.898Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7333330000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.058005,
      "avg_duration_per_task_ms": 191243.62,
      "avg_output_tokens_per_task": 30812.32,
      "last_run_timestamp": "2026-09-12T08:09:12.993Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Particle AI",
      "accuracy": 0.7333330000000001,
      "accuracy_std_dev": 0.020000000000000018,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.027065799999999997,
      "avg_duration_per_task_ms": 71689.04,
      "avg_output_tokens_per_task": 9149.63,
      "last_run_timestamp": "2026-10-28T22:54:17.860Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "auto-routing",
      "accuracy": 0.7333330000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.01490002,
      "avg_duration_per_task_ms": 87918.52,
      "avg_output_tokens_per_task": 12302.3,
      "last_run_timestamp": "2026-10-24T11:33:56.850Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.7333330000000001,
      "accuracy_std_dev": 0.020000000000000018,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.12365041999999998,
      "avg_duration_per_task_ms": 156560.04,
      "avg_output_tokens_per_task": 19704.13,
      "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": "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": 0.7333330000000001,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.10928178,
      "avg_duration_per_task_ms": 75479.44,
      "avg_output_tokens_per_task": 18824.05,
      "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": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "auto-routing",
      "accuracy": 0.7333330000000001,
      "accuracy_std_dev": 0.020000000000000018,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.038872460000000005,
      "avg_duration_per_task_ms": 158714.4,
      "avg_output_tokens_per_task": 14542.76,
      "last_run_timestamp": "2026-10-24T13:34:38.188Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "CoreWeave",
      "accuracy": 0.7333330000000001,
      "accuracy_std_dev": 0.020000000000000018,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.02595887,
      "avg_duration_per_task_ms": 77664,
      "avg_output_tokens_per_task": 16596.59,
      "last_run_timestamp": "2026-10-24T12:17:54.586Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "BaseTen",
      "accuracy": 0.7332197098976109,
      "accuracy_std_dev": 0.022906695901829432,
      "run_count": 6,
      "total_questions": 293,
      "avg_cost_per_task_usd": 0.04657865870307167,
      "avg_duration_per_task_ms": 71132.01023890785,
      "avg_output_tokens_per_task": 19231.02389078498,
      "last_run_timestamp": "2026-10-24T12:19:39.574Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Baidu",
      "accuracy": 0.73291675,
      "accuracy_std_dev": 0.033600323453614245,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.08912411,
      "avg_duration_per_task_ms": 173126.64,
      "avg_output_tokens_per_task": 10513.8925,
      "last_run_timestamp": "2026-10-24T19:53:44.150Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.5-sonnet-20250929",
      "model_slug": "anthropic/claude-sonnet-4.5",
      "model_name": "Anthropic: Claude Sonnet 4.5",
      "provider_name": "Google",
      "accuracy": 0.7327272727272728,
      "accuracy_std_dev": 0.025098504195164927,
      "run_count": 22,
      "total_questions": 1100,
      "avg_cost_per_task_usd": 0.36473022363636365,
      "avg_duration_per_task_ms": 228969.27363636365,
      "avg_output_tokens_per_task": 10924.667272727273,
      "last_run_timestamp": "2026-10-24T12:16:43.052Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7327272727272728,
      "accuracy_std_dev": 0.06263266295137199,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.15801903636363634,
      "avg_duration_per_task_ms": 1429178.5054545454,
      "avg_output_tokens_per_task": 28482.28181818182,
      "last_run_timestamp": "2026-10-24T20:45:53.033Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.7326668,
      "accuracy_std_dev": 0.024801517876130098,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.045864716,
      "avg_duration_per_task_ms": 120831.192,
      "avg_output_tokens_per_task": 26031.788,
      "last_run_timestamp": "2026-10-29T23:56:26.476Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "OpenInference",
      "accuracy": 0.7325,
      "accuracy_std_dev": 0.04596937990662913,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.012367795,
      "avg_duration_per_task_ms": 953069.875,
      "avg_output_tokens_per_task": 22737,
      "last_run_timestamp": "2026-10-17T10:32:07.729Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Nebius",
      "accuracy": 0.7325,
      "accuracy_std_dev": 0.029474565306379014,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.08144697000000001,
      "avg_duration_per_task_ms": 228561.56,
      "avg_output_tokens_per_task": 29757.14,
      "last_run_timestamp": "2026-10-24T12:23:33.838Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Sail Research",
      "accuracy": 0.7315151818181818,
      "accuracy_std_dev": 0.012091042761393707,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.23347585818181818,
      "avg_duration_per_task_ms": 257016.06181818183,
      "avg_output_tokens_per_task": 10120.247272727273,
      "last_run_timestamp": "2026-10-24T12:33:16.806Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7315151818181818,
      "accuracy_std_dev": 0.04321351479007923,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.08041129999999999,
      "avg_duration_per_task_ms": 162692.34181818183,
      "avg_output_tokens_per_task": 8235.127272727274,
      "last_run_timestamp": "2026-10-24T19:35:51.798Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Reka",
      "accuracy": 0.7311113333333332,
      "accuracy_std_dev": 0.0332591470759883,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.03472362,
      "avg_duration_per_task_ms": 267399.06,
      "avg_output_tokens_per_task": 33925.37333333334,
      "last_run_timestamp": "2026-10-23T07:22:37.672Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "Z.AI",
      "accuracy": 0.7311111666666666,
      "accuracy_std_dev": 0.03022542058387644,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.05052855666666667,
      "avg_duration_per_task_ms": 436577.48333333334,
      "avg_output_tokens_per_task": 9693.223333333333,
      "last_run_timestamp": "2026-10-24T20:04:11.026Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-sol-20260709",
      "model_slug": "openai/gpt-5.6-sol",
      "model_name": "OpenAI: GPT-5.6 Sol",
      "provider_name": "OpenAI",
      "accuracy": 0.7311110000000001,
      "accuracy_std_dev": 0.013698950130089036,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.12928292,
      "avg_duration_per_task_ms": 109953.54,
      "avg_output_tokens_per_task": 5131.753333333333,
      "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-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Modal",
      "accuracy": 0.7309522857142857,
      "accuracy_std_dev": 0.01941313112092864,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.04753026285714287,
      "avg_duration_per_task_ms": 161306.4342857143,
      "avg_output_tokens_per_task": 28780.68857142857,
      "last_run_timestamp": "2026-10-24T12:41:08.949Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta/muse-spark-open-20261015",
      "model_slug": "meta/muse-spark-open",
      "model_name": "Meta: Muse Spark Open",
      "provider_name": "Nebius",
      "accuracy": 0.7306399292929292,
      "accuracy_std_dev": 0.010748499999999994,
      "run_count": 2,
      "total_questions": 99,
      "avg_cost_per_task_usd": 0.08618277777777777,
      "avg_duration_per_task_ms": 204214.55555555556,
      "avg_output_tokens_per_task": 8369.424242424242,
      "last_run_timestamp": "2026-10-24T12:28:29.312Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "ModelRun",
      "accuracy": 0.7305555,
      "accuracy_std_dev": 0.026346395026960852,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.06789053999999999,
      "avg_duration_per_task_ms": 79373.93333333333,
      "avg_output_tokens_per_task": 8526.28,
      "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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Crusoe",
      "accuracy": 0.7300001666666667,
      "accuracy_std_dev": 0.03605552816336059,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.11662036,
      "avg_duration_per_task_ms": 82058.93666666666,
      "avg_output_tokens_per_task": 9160.97,
      "last_run_timestamp": "2026-09-12T08:12:28.966Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Crusoe",
      "accuracy": 0.73,
      "accuracy_std_dev": 0.04333300000000001,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.023320379999999998,
      "avg_duration_per_task_ms": 139023.54,
      "avg_output_tokens_per_task": 12073.4,
      "last_run_timestamp": "2026-10-24T12:04:17.803Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Novita",
      "accuracy": 0.73,
      "accuracy_std_dev": 0.023332999999999993,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.0186985,
      "avg_duration_per_task_ms": 114277.39,
      "avg_output_tokens_per_task": 11475.52,
      "last_run_timestamp": "2026-10-24T11:36:27.069Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.73,
      "accuracy_std_dev": 0.030000000000000027,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.12856867000000002,
      "avg_duration_per_task_ms": 121477.86,
      "avg_output_tokens_per_task": 19799.68,
      "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-3.6-flash-20260721",
      "model_slug": "google/gemini-3.6-flash",
      "model_name": "Google: Gemini 3.6 Flash",
      "provider_name": "Google AI Studio",
      "accuracy": 0.73,
      "accuracy_std_dev": 0.04333300000000001,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.24468226999999998,
      "avg_duration_per_task_ms": 158524.06,
      "avg_output_tokens_per_task": 22552.35,
      "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": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Cohere",
      "accuracy": 0.73,
      "accuracy_std_dev": 0.010000000000000009,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.02344918,
      "avg_duration_per_task_ms": 35583.1,
      "avg_output_tokens_per_task": 4177.41,
      "last_run_timestamp": "2026-10-24T11:37:33.416Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "DeepInfra",
      "accuracy": 0.729931893877551,
      "accuracy_std_dev": 0.03636312637494197,
      "run_count": 5,
      "total_questions": 245,
      "avg_cost_per_task_usd": 0.018753575510204082,
      "avg_duration_per_task_ms": 310156.1551020408,
      "avg_output_tokens_per_task": 6681.187755102041,
      "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": "anthropic/claude-4.5-sonnet-20250929",
      "model_slug": "anthropic/claude-sonnet-4.5",
      "model_name": "Anthropic: Claude Sonnet 4.5",
      "provider_name": "Azure",
      "accuracy": 0.7296969090909091,
      "accuracy_std_dev": 0.03079598974617119,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.3506678872727273,
      "avg_duration_per_task_ms": 233835.19090909092,
      "avg_output_tokens_per_task": 10922.210909090909,
      "last_run_timestamp": "2026-10-24T11:45:41.975Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Parasail",
      "accuracy": 0.7295238571428572,
      "accuracy_std_dev": 0.035920148907695515,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.07603518571428572,
      "avg_duration_per_task_ms": 216305.50285714286,
      "avg_output_tokens_per_task": 11169.142857142857,
      "last_run_timestamp": "2026-10-24T19:35:24.250Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "DeepInfra",
      "accuracy": 0.7292263868194842,
      "accuracy_std_dev": 0.037395839263812465,
      "run_count": 7,
      "total_questions": 349,
      "avg_cost_per_task_usd": 0.07898853868194843,
      "avg_duration_per_task_ms": 295613.8395415473,
      "avg_output_tokens_per_task": 13730.816618911174,
      "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": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "Novita",
      "accuracy": 0.729166625,
      "accuracy_std_dev": 0.029237916546573137,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.09863991500000001,
      "avg_duration_per_task_ms": 220840.6075,
      "avg_output_tokens_per_task": 12590.2025,
      "last_run_timestamp": "2026-10-24T19:51:30.269Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "ModelRun",
      "accuracy": 0.7288890000000001,
      "accuracy_std_dev": 0.03142665378305492,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.02386675333333333,
      "avg_duration_per_task_ms": 151228.52,
      "avg_output_tokens_per_task": 22554.02666666667,
      "last_run_timestamp": "2026-10-28T06:40:49.113Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Venice",
      "accuracy": 0.7285714285714285,
      "accuracy_std_dev": 0.026598516660344282,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.04723099714285714,
      "avg_duration_per_task_ms": 142250.88571428572,
      "avg_output_tokens_per_task": 24794.32857142857,
      "last_run_timestamp": "2026-10-24T13:07:38.963Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Together",
      "accuracy": 0.7285712857142858,
      "accuracy_std_dev": 0.024871537523352083,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.03253488857142857,
      "avg_duration_per_task_ms": 92457.76285714285,
      "avg_output_tokens_per_task": 23099.80285714286,
      "last_run_timestamp": "2026-10-24T11:48:04.395Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.7284848181818182,
      "accuracy_std_dev": 0.03400415574554003,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.3294088909090909,
      "avg_duration_per_task_ms": 208197.42363636364,
      "avg_output_tokens_per_task": 10397.776363636363,
      "last_run_timestamp": "2026-10-24T12:22:38.802Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7284848181818182,
      "accuracy_std_dev": 0.02941747886057535,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.08185737454545454,
      "avg_duration_per_task_ms": 229135.09818181818,
      "avg_output_tokens_per_task": 8804.125454545454,
      "last_run_timestamp": "2026-10-24T19:45:17.304Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.72833325,
      "accuracy_std_dev": 0.02640480280531367,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.02797866,
      "avg_duration_per_task_ms": 99430.655,
      "avg_output_tokens_per_task": 13207.28,
      "last_run_timestamp": "2026-10-29T23:09:25.660Z"
    },
    {
      "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": 0.72833325,
      "accuracy_std_dev": 0.03411588502717615,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.055064995000000005,
      "avg_duration_per_task_ms": 291890.35,
      "avg_output_tokens_per_task": 8526.425,
      "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": "Fireworks",
      "accuracy": 0.7281586250000001,
      "accuracy_std_dev": 0.032945022045826604,
      "run_count": 10,
      "total_questions": 496,
      "avg_cost_per_task_usd": 0.10040358467741935,
      "avg_duration_per_task_ms": 94375.23588709677,
      "avg_output_tokens_per_task": 7576.909274193548,
      "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": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Venice",
      "accuracy": 0.7278787272727273,
      "accuracy_std_dev": 0.02618002940164649,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.08620709454545455,
      "avg_duration_per_task_ms": 694656.5654545454,
      "avg_output_tokens_per_task": 9307.56,
      "last_run_timestamp": "2026-10-24T19:18:30.102Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Alibaba",
      "accuracy": 0.7278287449541284,
      "accuracy_std_dev": 0.03216567357904057,
      "run_count": 11,
      "total_questions": 545,
      "avg_cost_per_task_usd": 0.07681454862385322,
      "avg_duration_per_task_ms": 195964.3486238532,
      "avg_output_tokens_per_task": 8232.398165137614,
      "last_run_timestamp": "2026-10-28T03:37:10.887Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta/muse-spark-open-20261015",
      "model_slug": "meta/muse-spark-open",
      "model_name": "Meta: Muse Spark Open",
      "provider_name": "Makora",
      "accuracy": 0.7276631288659794,
      "accuracy_std_dev": 0.019224116824642454,
      "run_count": 4,
      "total_questions": 194,
      "avg_cost_per_task_usd": 0.07521926804123712,
      "avg_duration_per_task_ms": 73859.24742268042,
      "avg_output_tokens_per_task": 8316.994845360825,
      "last_run_timestamp": "2026-10-24T11:42:23.362Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "Z.AI",
      "accuracy": 0.7275001249999999,
      "accuracy_std_dev": 0.04047472891335252,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.10270093000000001,
      "avg_duration_per_task_ms": 213296.5425,
      "avg_output_tokens_per_task": 11976.1525,
      "last_run_timestamp": "2026-10-24T20:41:03.023Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "StreamLake",
      "accuracy": 0.7275,
      "accuracy_std_dev": 0.022220636056377873,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.06640188,
      "avg_duration_per_task_ms": 199646.51,
      "avg_output_tokens_per_task": 9529.36,
      "last_run_timestamp": "2026-10-24T19:34:44.643Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Fireworks",
      "accuracy": 0.7274752368672397,
      "accuracy_std_dev": 0.02732433470092314,
      "run_count": 21,
      "total_questions": 1047,
      "avg_cost_per_task_usd": 0.4330738662846227,
      "avg_duration_per_task_ms": 236263.65329512893,
      "avg_output_tokens_per_task": 10771.178605539637,
      "last_run_timestamp": "2026-10-24T12:16:44.774Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7273502461538462,
      "accuracy_std_dev": 0.043866981805083166,
      "run_count": 4,
      "total_questions": 195,
      "avg_cost_per_task_usd": 0.07634109230769232,
      "avg_duration_per_task_ms": 620581.2615384615,
      "avg_output_tokens_per_task": 7602.702564102564,
      "last_run_timestamp": "2026-08-29T17:13:58.312Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7272047264573992,
      "accuracy_std_dev": 0.08169417811313268,
      "run_count": 9,
      "total_questions": 446,
      "avg_cost_per_task_usd": 0.009946286995515695,
      "avg_duration_per_task_ms": 180161.19058295965,
      "avg_output_tokens_per_task": 4234.017937219731,
      "last_run_timestamp": "2026-10-24T12:25:45.296Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-sol-20260709",
      "model_slug": "openai/gpt-5.6-sol",
      "model_name": "OpenAI: GPT-5.6 Sol",
      "provider_name": "Azure",
      "accuracy": 0.7271428571428571,
      "accuracy_std_dev": 0.01855313018202104,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.31224118,
      "avg_duration_per_task_ms": 126481.89142857143,
      "avg_output_tokens_per_task": 5238.86,
      "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": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Fireworks",
      "accuracy": 0.726667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.21341358,
      "avg_duration_per_task_ms": 101630.84,
      "avg_output_tokens_per_task": 8827.48,
      "last_run_timestamp": "2026-08-26T21:00:03.913Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-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": "qwen/qwen3.7-plus-20260602",
      "model_slug": "qwen/qwen3.7-plus",
      "model_name": "Qwen: Qwen3.7 Plus",
      "provider_name": "Alibaba",
      "accuracy": 0.726667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.05448868,
      "avg_duration_per_task_ms": 350679.6,
      "avg_output_tokens_per_task": 17629,
      "last_run_timestamp": "2026-10-01T08:04:38.753Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "auto-routing",
      "accuracy": 0.7266666666666667,
      "accuracy_std_dev": 0.047453712340436424,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.011438460000000001,
      "avg_duration_per_task_ms": 86937.99333333333,
      "avg_output_tokens_per_task": 10682.4,
      "last_run_timestamp": "2026-10-24T11:51:31.389Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Nebius",
      "accuracy": 0.7266666666666667,
      "accuracy_std_dev": 0.010886757164351354,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.02419578,
      "avg_duration_per_task_ms": 133541.80666666667,
      "avg_output_tokens_per_task": 11460.326666666666,
      "last_run_timestamp": "2026-10-24T12:24:12.065Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.4-20260305",
      "model_slug": "openai/gpt-5.4",
      "model_name": "OpenAI: GPT-5.4",
      "provider_name": "OpenAI",
      "accuracy": 0.7266666666666667,
      "accuracy_std_dev": 0.014401903168987377,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.30249026,
      "avg_duration_per_task_ms": 183822.13333333333,
      "avg_output_tokens_per_task": 10747.233333333334,
      "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": "openai/gpt-5.6-luna-20260709",
      "model_slug": "openai/gpt-5.6-luna",
      "model_name": "OpenAI: GPT-5.6 Luna",
      "provider_name": "OpenAI",
      "accuracy": 0.7266666666666667,
      "accuracy_std_dev": 0.014401440259764123,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.013071513333333333,
      "avg_duration_per_task_ms": 97597.57333333333,
      "avg_output_tokens_per_task": 4935.306666666666,
      "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": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Phala",
      "accuracy": 0.7266665,
      "accuracy_std_dev": 0.00666650000000002,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01900955,
      "avg_duration_per_task_ms": 214869.47,
      "avg_output_tokens_per_task": 11727.86,
      "last_run_timestamp": "2026-10-24T12:08:01.610Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Modal",
      "accuracy": 0.7266665,
      "accuracy_std_dev": 0.026666500000000037,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.018476000000000003,
      "avg_duration_per_task_ms": 104934.72,
      "avg_output_tokens_per_task": 10917.83,
      "last_run_timestamp": "2026-10-24T11:48:43.775Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.7266665,
      "accuracy_std_dev": 0.026666500000000037,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.43167194999999997,
      "avg_duration_per_task_ms": 195075.98,
      "avg_output_tokens_per_task": 31329.14,
      "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": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "Io Net",
      "accuracy": 0.7266665,
      "accuracy_std_dev": 0.013333499999999998,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.06725601,
      "avg_duration_per_task_ms": 640810.38,
      "avg_output_tokens_per_task": 15599.02,
      "last_run_timestamp": "2026-10-27T14:13:59.381Z"
    },
    {
      "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": 0.7266665,
      "accuracy_std_dev": 0.036055327855256024,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.09256564750000001,
      "avg_duration_per_task_ms": 228748.9725,
      "avg_output_tokens_per_task": 10105.375,
      "last_run_timestamp": "2026-10-24T20:31:27.967Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Nebius",
      "accuracy": 0.7263158421052631,
      "accuracy_std_dev": 0.007406999999999997,
      "run_count": 2,
      "total_questions": 95,
      "avg_cost_per_task_usd": 0.013109768421052631,
      "avg_duration_per_task_ms": 155953.90526315788,
      "avg_output_tokens_per_task": 8164.357894736842,
      "last_run_timestamp": "2026-10-24T13:07:13.980Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Nebius",
      "accuracy": 0.7262383654618474,
      "accuracy_std_dev": 0.017933902883644697,
      "run_count": 5,
      "total_questions": 249,
      "avg_cost_per_task_usd": 0.17520732931726907,
      "avg_duration_per_task_ms": 546387.9477911646,
      "avg_output_tokens_per_task": 8780.429718875503,
      "last_run_timestamp": "2026-10-24T21:19:54.882Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "auto-routing",
      "accuracy": 0.7253334,
      "accuracy_std_dev": 0.039305065549366525,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.00982467,
      "avg_duration_per_task_ms": 372894.758,
      "avg_output_tokens_per_task": 14244.68,
      "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": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "Venice",
      "accuracy": 0.7253334,
      "accuracy_std_dev": 0.037214139488103176,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.08614405199999998,
      "avg_duration_per_task_ms": 329579.82,
      "avg_output_tokens_per_task": 9375.922,
      "last_run_timestamp": "2026-10-24T19:37:38.817Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Venice",
      "accuracy": 0.7253333,
      "accuracy_std_dev": 0.02809501972610092,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.028140498,
      "avg_duration_per_task_ms": 228402.532,
      "avg_output_tokens_per_task": 9725.512,
      "last_run_timestamp": "2026-10-24T19:14:23.110Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5-mini-2025-08-07",
      "model_slug": "openai/gpt-5-mini",
      "model_name": "OpenAI: GPT-5 Mini",
      "provider_name": "OpenAI",
      "accuracy": 0.7244443333333334,
      "accuracy_std_dev": 0.02199873172606905,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.12152117999999999,
      "avg_duration_per_task_ms": 471173.63333333336,
      "avg_output_tokens_per_task": 54771.27333333333,
      "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": "Parasail",
      "accuracy": 0.7242855714285714,
      "accuracy_std_dev": 0.024734242792748124,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.04387056285714286,
      "avg_duration_per_task_ms": 165104.78285714285,
      "avg_output_tokens_per_task": 29616.194285714286,
      "last_run_timestamp": "2026-10-24T12:35:19.896Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Phala",
      "accuracy": 0.7233335,
      "accuracy_std_dev": 0.023333500000000007,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.02063872,
      "avg_duration_per_task_ms": 638594.61,
      "avg_output_tokens_per_task": 13867.61,
      "last_run_timestamp": "2026-08-29T16:33:36.025Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-terra-20260709",
      "model_slug": "openai/gpt-5.6-terra",
      "model_name": "OpenAI: GPT-5.6 Terra",
      "provider_name": "auto-routing",
      "accuracy": 0.7233335,
      "accuracy_std_dev": 0.0033334999999999892,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.11131322,
      "avg_duration_per_task_ms": 83286.18,
      "avg_output_tokens_per_task": 3736.91,
      "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-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Particle AI",
      "accuracy": 0.7233335,
      "accuracy_std_dev": 0.023333500000000007,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.0109795,
      "avg_duration_per_task_ms": 64854.76,
      "avg_output_tokens_per_task": 4155.27,
      "last_run_timestamp": "2026-10-28T23:08:11.828Z"
    },
    {
      "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": 0.72333325,
      "accuracy_std_dev": 0.024267113092705673,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.07263319833333333,
      "avg_duration_per_task_ms": 224600.79833333334,
      "avg_output_tokens_per_task": 8546.035,
      "last_run_timestamp": "2026-10-24T19:52:00.596Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.4-20260305",
      "model_slug": "openai/gpt-5.4",
      "model_name": "OpenAI: GPT-5.4",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7233330000000001,
      "accuracy_std_dev": 0.010000000000000009,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.33860155,
      "avg_duration_per_task_ms": 113561.71,
      "avg_output_tokens_per_task": 10711.4,
      "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": "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": 0.7232361605839416,
      "accuracy_std_dev": 0.08188802109119138,
      "run_count": 11,
      "total_questions": 548,
      "avg_cost_per_task_usd": 0.04157626824817519,
      "avg_duration_per_task_ms": 270001.6879562044,
      "avg_output_tokens_per_task": 23794.242700729927,
      "last_run_timestamp": "2026-10-24T19:21:10.232Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "Venice",
      "accuracy": 0.722963,
      "accuracy_std_dev": 0.024567656927124495,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.2753419711111111,
      "avg_duration_per_task_ms": 183403.79555555555,
      "avg_output_tokens_per_task": 11487.52,
      "last_run_timestamp": "2026-10-24T12:33:54.727Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "GMICloud",
      "accuracy": 0.7229628888888889,
      "accuracy_std_dev": 0.0711188714906161,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.009601128888888888,
      "avg_duration_per_task_ms": 242853.24666666667,
      "avg_output_tokens_per_task": 7420.495555555555,
      "last_run_timestamp": "2026-10-27T05:56:39.952Z"
    },
    {
      "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": 0.7228571428571429,
      "accuracy_std_dev": 0.019471637611287507,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.10722213142857143,
      "avg_duration_per_task_ms": 326175.72,
      "avg_output_tokens_per_task": 14461.362857142858,
      "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": "OpenInference",
      "accuracy": 0.7223148931552588,
      "accuracy_std_dev": 0.03960298431691171,
      "run_count": 12,
      "total_questions": 599,
      "avg_cost_per_task_usd": 0.025822385642737894,
      "avg_duration_per_task_ms": 283083.48080133554,
      "avg_output_tokens_per_task": 7842.495826377295,
      "last_run_timestamp": "2026-10-24T12:07:12.594Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Io Net",
      "accuracy": 0.7222223333333333,
      "accuracy_std_dev": 0.011330936570096713,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.018792733333333332,
      "avg_duration_per_task_ms": 301192.17333333334,
      "avg_output_tokens_per_task": 19692.3,
      "last_run_timestamp": "2026-10-01T16:00:06.404Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Reka",
      "accuracy": 0.72166675,
      "accuracy_std_dev": 0.012801898733683979,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.01389351,
      "avg_duration_per_task_ms": 148770.74,
      "avg_output_tokens_per_task": 13265.9,
      "last_run_timestamp": "2026-10-24T12:22:53.798Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "Venice",
      "accuracy": 0.721666625,
      "accuracy_std_dev": 0.038405766711971455,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.074211325,
      "avg_duration_per_task_ms": 267437.405,
      "avg_output_tokens_per_task": 27915.47,
      "last_run_timestamp": "2026-10-24T18:56:04.577Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Novita",
      "accuracy": 0.721666625,
      "accuracy_std_dev": 0.02466430450538541,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.12928377000000002,
      "avg_duration_per_task_ms": 250878.7025,
      "avg_output_tokens_per_task": 10320.3,
      "last_run_timestamp": "2026-10-24T22:56:39.524Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Novita",
      "accuracy": 0.720606,
      "accuracy_std_dev": 0.021168874530996952,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.08199455272727274,
      "avg_duration_per_task_ms": 233558.22181818183,
      "avg_output_tokens_per_task": 9247.22,
      "last_run_timestamp": "2026-10-24T19:37:32.748Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "SiliconFlow",
      "accuracy": 0.720606,
      "accuracy_std_dev": 0.033928552923084485,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.08656796545454544,
      "avg_duration_per_task_ms": 136294.9909090909,
      "avg_output_tokens_per_task": 9196.996363636363,
      "last_run_timestamp": "2026-10-24T18:58:26.138Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Sail Research",
      "accuracy": 0.7200002499999999,
      "accuracy_std_dev": 0.016329931619192377,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.005028675,
      "avg_duration_per_task_ms": 140463.585,
      "avg_output_tokens_per_task": 11883.17,
      "last_run_timestamp": "2026-10-24T12:43:45.098Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Fireworks",
      "accuracy": 0.72,
      "accuracy_std_dev": 0.030000000000000027,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.026063740000000002,
      "avg_duration_per_task_ms": 127597.92,
      "avg_output_tokens_per_task": 12968.38,
      "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": "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": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Venice",
      "accuracy": 0.72,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.027683759999999998,
      "avg_duration_per_task_ms": 68973.72,
      "avg_output_tokens_per_task": 10080.22,
      "last_run_timestamp": "2026-10-29T00:22:34.385Z"
    },
    {
      "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": "StreamLake",
      "accuracy": 0.72,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.0210898,
      "avg_duration_per_task_ms": 122128,
      "avg_output_tokens_per_task": 11448.46,
      "last_run_timestamp": "2026-10-26T03:28:50.420Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "inclusionai/ling-3.0-flash-vl-20260910",
      "model_slug": "inclusionai/ling-3.0-flash-vl",
      "model_name": "inclusionAI: Ling 3.0 Flash VL",
      "provider_name": "Novita",
      "accuracy": 0.72,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.00317077,
      "avg_duration_per_task_ms": 128719.23,
      "avg_output_tokens_per_task": 15118.03,
      "last_run_timestamp": "2026-10-01T08:00:27.045Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "GMICloud",
      "accuracy": 0.72,
      "accuracy_std_dev": 0.016329931618554536,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.01825590666666667,
      "avg_duration_per_task_ms": 233637.58,
      "avg_output_tokens_per_task": 15649.073333333334,
      "last_run_timestamp": "2026-10-27T01:30:02.265Z"
    },
    {
      "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-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "auto-routing",
      "accuracy": 0.72,
      "accuracy_std_dev": 0.014142017775055995,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.05195973499999999,
      "avg_duration_per_task_ms": 385642.415,
      "avg_output_tokens_per_task": 9384.3275,
      "last_run_timestamp": "2026-10-24T20:09:28.564Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "Novita",
      "accuracy": 0.72,
      "accuracy_std_dev": 0.020816740060585853,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.058833675,
      "avg_duration_per_task_ms": 335844.4,
      "avg_output_tokens_per_task": 10101.9275,
      "last_run_timestamp": "2026-10-24T20:09:59.536Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7199998461538462,
      "accuracy_std_dev": 0.060595320121208315,
      "run_count": 13,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.28468824153846156,
      "avg_duration_per_task_ms": 424519.81384615385,
      "avg_output_tokens_per_task": 10050.18923076923,
      "last_run_timestamp": "2026-10-24T12:30:07.991Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta/muse-spark-open-20261015",
      "model_slug": "meta/muse-spark-open",
      "model_name": "Meta: Muse Spark Open",
      "provider_name": "BaseTen",
      "accuracy": 0.719858,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 47,
      "avg_cost_per_task_usd": 0.0657194255319149,
      "avg_duration_per_task_ms": 49852.255319148935,
      "avg_output_tokens_per_task": 5715.191489361702,
      "last_run_timestamp": "2026-10-24T12:16:29.775Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Alibaba",
      "accuracy": 0.719770172413793,
      "accuracy_std_dev": 0.03281128869666338,
      "run_count": 29,
      "total_questions": 1450,
      "avg_cost_per_task_usd": 0.11837045241379313,
      "avg_duration_per_task_ms": 130253.44965517242,
      "avg_output_tokens_per_task": 8577.781379310345,
      "last_run_timestamp": "2026-10-28T04:07:30.740Z"
    },
    {
      "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": "Baidu",
      "accuracy": 0.7192592777777778,
      "accuracy_std_dev": 0.027923028635887925,
      "run_count": 18,
      "total_questions": 900,
      "avg_cost_per_task_usd": 0.06975042666666668,
      "avg_duration_per_task_ms": 117431.64666666667,
      "avg_output_tokens_per_task": 8825.745555555555,
      "last_run_timestamp": "2026-10-24T19:51:31.088Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Chutes",
      "accuracy": 0.7191665,
      "accuracy_std_dev": 0.032393007648256436,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.05382420250000001,
      "avg_duration_per_task_ms": 373003.49,
      "avg_output_tokens_per_task": 11827.6175,
      "last_run_timestamp": "2026-10-24T20:11:31.655Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Novita",
      "accuracy": 0.7187878181818182,
      "accuracy_std_dev": 0.02571315171028306,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.04521527272727272,
      "avg_duration_per_task_ms": 179484.99272727274,
      "avg_output_tokens_per_task": 9729.592727272728,
      "last_run_timestamp": "2026-10-24T19:18:28.440Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "Darkbloom",
      "accuracy": 0.7187496874999999,
      "accuracy_std_dev": 0.035822999999999994,
      "run_count": 2,
      "total_questions": 96,
      "avg_cost_per_task_usd": 0.015383489583333333,
      "avg_duration_per_task_ms": 816064.3020833334,
      "avg_output_tokens_per_task": 9819.75,
      "last_run_timestamp": "2026-10-30T05:52:34.884Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Venice",
      "accuracy": 0.7185183333333334,
      "accuracy_std_dev": 0.033815486586933986,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.24051653333333337,
      "avg_duration_per_task_ms": 300854.7,
      "avg_output_tokens_per_task": 21246.093333333334,
      "last_run_timestamp": "2026-10-24T12:24:27.575Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Chutes",
      "accuracy": 0.7180951428571429,
      "accuracy_std_dev": 0.022737622035664484,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.09258931428571428,
      "avg_duration_per_task_ms": 898101.5942857143,
      "avg_output_tokens_per_task": 9610.791428571429,
      "last_run_timestamp": "2026-10-24T20:25:57.765Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "Minimax",
      "accuracy": 0.7179166875,
      "accuracy_std_dev": 0.01925252734291899,
      "run_count": 16,
      "total_questions": 800,
      "avg_cost_per_task_usd": 0.038559541249999996,
      "avg_duration_per_task_ms": 195166.50625,
      "avg_output_tokens_per_task": 9795.555,
      "last_run_timestamp": "2026-10-24T19:52:24.184Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.1-terminus",
      "model_slug": "deepseek/deepseek-v3.1-terminus",
      "model_name": "DeepSeek: DeepSeek V3.1 Terminus",
      "provider_name": "auto-routing",
      "accuracy": 0.7177776666666666,
      "accuracy_std_dev": 0.017497908167041637,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.057615893333333335,
      "avg_duration_per_task_ms": 509994.74666666664,
      "avg_output_tokens_per_task": 13843.966666666667,
      "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": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "Venice",
      "accuracy": 0.7177776666666666,
      "accuracy_std_dev": 0.013698571612479267,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.10180220666666666,
      "avg_duration_per_task_ms": 128107.32666666666,
      "avg_output_tokens_per_task": 16602.8,
      "last_run_timestamp": "2026-10-24T12:39:45.200Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.717499875,
      "accuracy_std_dev": 0.03196146008256467,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.09209893,
      "avg_duration_per_task_ms": 157130.905,
      "avg_output_tokens_per_task": 48565.055,
      "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": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "auto-routing",
      "accuracy": 0.7167414529147983,
      "accuracy_std_dev": 0.04866792401901878,
      "run_count": 9,
      "total_questions": 446,
      "avg_cost_per_task_usd": 0.034667679372197314,
      "avg_duration_per_task_ms": 1113044.8408071748,
      "avg_output_tokens_per_task": 50703.41479820628,
      "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": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "SambaNova",
      "accuracy": 0.7166669999999999,
      "accuracy_std_dev": 0.009999999999999953,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.04331858,
      "avg_duration_per_task_ms": 119875.46,
      "avg_output_tokens_per_task": 9185.81,
      "last_run_timestamp": "2026-10-24T12:22:21.794Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5-mini-2025-08-07",
      "model_slug": "openai/gpt-5-mini",
      "model_name": "OpenAI: GPT-5 Mini",
      "provider_name": "auto-routing",
      "accuracy": 0.7166669999999999,
      "accuracy_std_dev": 0.009999999999999953,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.1280142,
      "avg_duration_per_task_ms": 531647.04,
      "avg_output_tokens_per_task": 56970.52,
      "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": "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": 0.7166665000000001,
      "accuracy_std_dev": 0.03666649999999999,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.71669186,
      "avg_duration_per_task_ms": 277432.68,
      "avg_output_tokens_per_task": 22919.22,
      "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": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "Phala",
      "accuracy": 0.716250125,
      "accuracy_std_dev": 0.028550118578727003,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.2090713325,
      "avg_duration_per_task_ms": 1406671.1525,
      "avg_output_tokens_per_task": 13817.695,
      "last_run_timestamp": "2026-10-24T19:32:17.023Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Nebius",
      "accuracy": 0.7155556666666666,
      "accuracy_std_dev": 0.03651479659596153,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.3018651377777778,
      "avg_duration_per_task_ms": 498139.0266666667,
      "avg_output_tokens_per_task": 9517.055555555555,
      "last_run_timestamp": "2026-10-24T22:09:08.519Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "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": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "CoreWeave",
      "accuracy": 0.7144445,
      "accuracy_std_dev": 0.03942493727431069,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.08227690833333334,
      "avg_duration_per_task_ms": 100916.12166666667,
      "avg_output_tokens_per_task": 8631.778333333334,
      "last_run_timestamp": "2026-10-24T23:37:48.880Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.7139393636363637,
      "accuracy_std_dev": 0.03259241406516308,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.04342848545454545,
      "avg_duration_per_task_ms": 118501.67636363636,
      "avg_output_tokens_per_task": 8446.745454545455,
      "last_run_timestamp": "2026-10-24T19:43:58.683Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "InferenceNet",
      "accuracy": 0.71333325,
      "accuracy_std_dev": 0.016996797086142445,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.01124519,
      "avg_duration_per_task_ms": 291598.365,
      "avg_output_tokens_per_task": 13318.715,
      "last_run_timestamp": "2026-10-24T13:02:30.014Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "auto-routing",
      "accuracy": 0.71333325,
      "accuracy_std_dev": 0.023570237826282113,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0612510275,
      "avg_duration_per_task_ms": 315642.135,
      "avg_output_tokens_per_task": 9841.58,
      "last_run_timestamp": "2026-10-24T20:27:01.220Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Io Net",
      "accuracy": 0.7133330000000001,
      "accuracy_std_dev": 0.020000000000000018,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.12082516,
      "avg_duration_per_task_ms": 865226.27,
      "avg_output_tokens_per_task": 27649.46,
      "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": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Crusoe",
      "accuracy": 0.713333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.01217508,
      "avg_duration_per_task_ms": 145478.36,
      "avg_output_tokens_per_task": 13792.14,
      "last_run_timestamp": "2026-10-26T04:07:05.258Z"
    },
    {
      "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": "Crusoe",
      "accuracy": 0.713333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.14050459999999998,
      "avg_duration_per_task_ms": 414099.52,
      "avg_output_tokens_per_task": 22497.98,
      "last_run_timestamp": "2026-10-26T05:10:12.373Z"
    },
    {
      "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": "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": 0.713333,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.08007529,
      "avg_duration_per_task_ms": 140906.22,
      "avg_output_tokens_per_task": 42659.23,
      "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": "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-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "Moonshot AI",
      "accuracy": 0.713333,
      "accuracy_std_dev": 0.024494897427831803,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.046480935,
      "avg_duration_per_task_ms": 164672.7,
      "avg_output_tokens_per_task": 8116.895,
      "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": "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": "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": "Nebius",
      "accuracy": 0.713333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.07153616,
      "avg_duration_per_task_ms": 149810.8,
      "avg_output_tokens_per_task": 15322.1,
      "last_run_timestamp": "2026-10-27T00:56:39.531Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-flash-20260826",
      "model_slug": "qwen/qwen3.8-flash",
      "model_name": "Qwen: Qwen3.8 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.713333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.01665106,
      "avg_duration_per_task_ms": 327560.6,
      "avg_output_tokens_per_task": 17227.24,
      "last_run_timestamp": "2026-10-01T08:00:41.459Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.8-flash-20260902",
      "model_slug": "google/gemini-3.8-flash",
      "model_name": "Google: Gemini 3.8 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.712766,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 47,
      "avg_cost_per_task_usd": 0.21022357446808512,
      "avg_duration_per_task_ms": 261220.25531914894,
      "avg_output_tokens_per_task": 27609.17021276596,
      "last_run_timestamp": "2026-10-01T08:02:55.088Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "GMICloud",
      "accuracy": 0.7125925555555556,
      "accuracy_std_dev": 0.02047446802516685,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.023023279999999997,
      "avg_duration_per_task_ms": 256611.78222222222,
      "avg_output_tokens_per_task": 9809.96,
      "last_run_timestamp": "2026-10-27T04:17:23.321Z"
    },
    {
      "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": 0.712381,
      "accuracy_std_dev": 0.016106167044245584,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.11192505714285715,
      "avg_duration_per_task_ms": 96202.06,
      "avg_output_tokens_per_task": 3694.017142857143,
      "last_run_timestamp": "2026-10-01T19:04:52.651Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "Novita",
      "accuracy": 0.71166675,
      "accuracy_std_dev": 0.03411566518752785,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.025674655,
      "avg_duration_per_task_ms": 206861.205,
      "avg_output_tokens_per_task": 9942.77,
      "last_run_timestamp": "2026-10-24T19:49:21.066Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "Venice",
      "accuracy": 0.71166675,
      "accuracy_std_dev": 0.03460084697211182,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.072092705,
      "avg_duration_per_task_ms": 320824.8375,
      "avg_output_tokens_per_task": 11065.2975,
      "last_run_timestamp": "2026-10-24T22:01:31.925Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Wafer",
      "accuracy": 0.7114285238095238,
      "accuracy_std_dev": 0.02603554463314856,
      "run_count": 21,
      "total_questions": 1050,
      "avg_cost_per_task_usd": 0.1029916761904762,
      "avg_duration_per_task_ms": 108696.98,
      "avg_output_tokens_per_task": 9711.373333333333,
      "last_run_timestamp": "2026-10-29T23:00:08.664Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Nebius",
      "accuracy": 0.7113332,
      "accuracy_std_dev": 0.017587783639788142,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.28712958000000005,
      "avg_duration_per_task_ms": 155529.044,
      "avg_output_tokens_per_task": 10850.46,
      "last_run_timestamp": "2026-10-24T19:05:28.397Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "DeepInfra",
      "accuracy": 0.7108432289156627,
      "accuracy_std_dev": 0.016820726607373437,
      "run_count": 5,
      "total_questions": 249,
      "avg_cost_per_task_usd": 0.04982926104417671,
      "avg_duration_per_task_ms": 412025.6064257028,
      "avg_output_tokens_per_task": 8776.425702811244,
      "last_run_timestamp": "2026-10-24T21:22:48.118Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Phala",
      "accuracy": 0.7104167499999999,
      "accuracy_std_dev": 0.11449707753448339,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.22292825749999998,
      "avg_duration_per_task_ms": 1269690.4575,
      "avg_output_tokens_per_task": 60617.9,
      "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": "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": 0.7103030909090909,
      "accuracy_std_dev": 0.03170096156687463,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.12382668181818182,
      "avg_duration_per_task_ms": 164491.53272727272,
      "avg_output_tokens_per_task": 11816.025454545455,
      "last_run_timestamp": "2026-10-24T19:52:24.144Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Venice",
      "accuracy": 0.710303,
      "accuracy_std_dev": 0.045426193987410164,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.1190600290909091,
      "avg_duration_per_task_ms": 130010.17272727273,
      "avg_output_tokens_per_task": 9680.261818181818,
      "last_run_timestamp": "2026-10-24T19:35:31.077Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Novita",
      "accuracy": 0.71,
      "accuracy_std_dev": 0.003332999999999975,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01784487,
      "avg_duration_per_task_ms": 239614.14,
      "avg_output_tokens_per_task": 16662.97,
      "last_run_timestamp": "2026-10-24T11:51:42.268Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-terra-20260709",
      "model_slug": "openai/gpt-5.6-terra",
      "model_name": "OpenAI: GPT-5.6 Terra",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.71,
      "accuracy_std_dev": 0.010000000000000009,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.12724396999999998,
      "avg_duration_per_task_ms": 73218.95,
      "avg_output_tokens_per_task": 3826.45,
      "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": "Ambient",
      "accuracy": 0.71,
      "accuracy_std_dev": 0.046187895240528104,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0660468475,
      "avg_duration_per_task_ms": 246966.18,
      "avg_output_tokens_per_task": 8220.67,
      "last_run_timestamp": "2026-10-24T19:39:33.038Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "StreamLake",
      "accuracy": 0.7099999166666667,
      "accuracy_std_dev": 0.03061105767381654,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.03565966833333333,
      "avg_duration_per_task_ms": 184659.72166666668,
      "avg_output_tokens_per_task": 9078.276666666667,
      "last_run_timestamp": "2026-10-24T19:49:36.340Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "OpenInference",
      "accuracy": 0.70990758186398,
      "accuracy_std_dev": 0.05927908384168378,
      "run_count": 8,
      "total_questions": 397,
      "avg_cost_per_task_usd": 0.04869609571788413,
      "avg_duration_per_task_ms": 986154.4534005037,
      "avg_output_tokens_per_task": 53874.76574307305,
      "last_run_timestamp": "2026-10-24T12:28:11.805Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "auto-routing",
      "accuracy": 0.709166625,
      "accuracy_std_dev": 0.01198972650790563,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.02624361,
      "avg_duration_per_task_ms": 203778.2175,
      "avg_output_tokens_per_task": 10089.3,
      "last_run_timestamp": "2026-10-24T19:34:31.773Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Phala",
      "accuracy": 0.7090847895791583,
      "accuracy_std_dev": 0.018714036978695964,
      "run_count": 10,
      "total_questions": 499,
      "avg_cost_per_task_usd": 0.0960923246492986,
      "avg_duration_per_task_ms": 181290.61923847697,
      "avg_output_tokens_per_task": 10798.256513026052,
      "last_run_timestamp": "2026-10-17T19:53:39.136Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Together",
      "accuracy": 0.7084848181818182,
      "accuracy_std_dev": 0.037721935609784936,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.10331338545454545,
      "avg_duration_per_task_ms": 98153.38727272727,
      "avg_output_tokens_per_task": 9101.983636363637,
      "last_run_timestamp": "2026-10-24T19:12:36.271Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "StreamLake",
      "accuracy": 0.7083335,
      "accuracy_std_dev": 0.041466338423473054,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.156516845,
      "avg_duration_per_task_ms": 345707.575,
      "avg_output_tokens_per_task": 10875.9,
      "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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "DeepInfra",
      "accuracy": 0.7077778333333333,
      "accuracy_std_dev": 0.033091888529853064,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.04299519499999999,
      "avg_duration_per_task_ms": 138463.85833333334,
      "avg_output_tokens_per_task": 8481.851666666667,
      "last_run_timestamp": "2026-10-24T19:47:13.308Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7076191428571428,
      "accuracy_std_dev": 0.061867939965308524,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.10378842857142857,
      "avg_duration_per_task_ms": 217807.51714285713,
      "avg_output_tokens_per_task": 11023.12,
      "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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "CoreWeave",
      "accuracy": 0.7075,
      "accuracy_std_dev": 0.038935452283234105,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.06791916749999999,
      "avg_duration_per_task_ms": 69428.1925,
      "avg_output_tokens_per_task": 9662.645,
      "last_run_timestamp": "2026-10-24T19:30:31.496Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "Makora",
      "accuracy": 0.7073334,
      "accuracy_std_dev": 0.025681872553223187,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.024487520000000002,
      "avg_duration_per_task_ms": 117939.624,
      "avg_output_tokens_per_task": 30855.352,
      "last_run_timestamp": "2026-10-24T19:30:58.294Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Nebius",
      "accuracy": 0.706667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.3541182,
      "avg_duration_per_task_ms": 1646927.84,
      "avg_output_tokens_per_task": 24534.74,
      "last_run_timestamp": "2026-09-10T00:00:02.948Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Cohere",
      "accuracy": 0.706667,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.07705153,
      "avg_duration_per_task_ms": 159916.36,
      "avg_output_tokens_per_task": 34851.38,
      "last_run_timestamp": "2026-10-24T12:16:31.422Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.706667,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.44632813,
      "avg_duration_per_task_ms": 132811.5,
      "avg_output_tokens_per_task": 20668.24,
      "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": "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": 0.706667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.71176192,
      "avg_duration_per_task_ms": 285827.28,
      "avg_output_tokens_per_task": 22895.76,
      "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": "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-6.1-sol-20260929",
      "model_slug": "openai/gpt-6.1-sol",
      "model_name": "OpenAI: GPT-6.1 Sol",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7066669999999999,
      "accuracy_std_dev": 0.019999999999999962,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.1042244,
      "avg_duration_per_task_ms": 226024.67,
      "avg_output_tokens_per_task": 3543.23,
      "last_run_timestamp": "2026-10-30T00:17:12.390Z"
    },
    {
      "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": 0.70666675,
      "accuracy_std_dev": 0.027080004924067123,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.04331178,
      "avg_duration_per_task_ms": 178565.92,
      "avg_output_tokens_per_task": 8366.805,
      "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": "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": "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": 0.7066666666666667,
      "accuracy_std_dev": 0.01962587106064057,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.08575882,
      "avg_duration_per_task_ms": 316120.8933333333,
      "avg_output_tokens_per_task": 32977.22666666667,
      "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": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Makora",
      "accuracy": 0.7066665000000001,
      "accuracy_std_dev": 0.027888567949071897,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.021449605000000004,
      "avg_duration_per_task_ms": 108412.625,
      "avg_output_tokens_per_task": 11247.685,
      "last_run_timestamp": "2026-10-24T11:28:47.983Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7060606363636364,
      "accuracy_std_dev": 0.15377933291705106,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.06115002181818183,
      "avg_duration_per_task_ms": 147923.08545454545,
      "avg_output_tokens_per_task": 17075.432727272728,
      "last_run_timestamp": "2026-10-24T18:32:59.224Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-luna-20260709",
      "model_slug": "openai/gpt-5.6-luna",
      "model_name": "OpenAI: GPT-5.6 Luna",
      "provider_name": "Azure",
      "accuracy": 0.7057145714285713,
      "accuracy_std_dev": 0.030327119046712443,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.012995502857142857,
      "avg_duration_per_task_ms": 102610.46857142857,
      "avg_output_tokens_per_task": 4898.9228571428575,
      "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": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7050002499999999,
      "accuracy_std_dev": 0.029580300314356156,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.050410445,
      "avg_duration_per_task_ms": 854714.71,
      "avg_output_tokens_per_task": 8739.19,
      "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": "AtlasCloud",
      "accuracy": 0.705000125,
      "accuracy_std_dev": 0.054746931675294576,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.031620825,
      "avg_duration_per_task_ms": 323083.205,
      "avg_output_tokens_per_task": 10027.1975,
      "last_run_timestamp": "2026-10-24T19:49:45.253Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5-mini-2025-08-07",
      "model_slug": "openai/gpt-5-mini",
      "model_name": "OpenAI: GPT-5 Mini",
      "provider_name": "Azure",
      "accuracy": 0.70499975,
      "accuracy_std_dev": 0.011902485735656231,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.13260128,
      "avg_duration_per_task_ms": 477326.39,
      "avg_output_tokens_per_task": 55745.89,
      "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": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "Chutes",
      "accuracy": 0.7040814081632653,
      "accuracy_std_dev": 0.029861000000000026,
      "run_count": 2,
      "total_questions": 98,
      "avg_cost_per_task_usd": 0.12006706122448979,
      "avg_duration_per_task_ms": 975720.112244898,
      "avg_output_tokens_per_task": 30678.173469387755,
      "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": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7033335,
      "accuracy_std_dev": 0.016666499999999973,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01719149,
      "avg_duration_per_task_ms": 73722.59,
      "avg_output_tokens_per_task": 12133.38,
      "last_run_timestamp": "2026-10-24T12:37:04.711Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.7033335,
      "accuracy_std_dev": 0.0033335000000000448,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.08534018,
      "avg_duration_per_task_ms": 347658.09,
      "avg_output_tokens_per_task": 32908.85,
      "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-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Parasail",
      "accuracy": 0.703333,
      "accuracy_std_dev": 0.010000000000000009,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.012750089999999999,
      "avg_duration_per_task_ms": 89612.54,
      "avg_output_tokens_per_task": 11420.52,
      "last_run_timestamp": "2026-10-24T12:33:10.739Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.7-plus-20260602",
      "model_slug": "qwen/qwen3.7-plus",
      "model_name": "Qwen: Qwen3.7 Plus",
      "provider_name": "auto-routing",
      "accuracy": 0.703333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.05425128,
      "avg_duration_per_task_ms": 396980.46,
      "avg_output_tokens_per_task": 20673.9,
      "last_run_timestamp": "2026-10-01T08:05:20.277Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "BaseTen",
      "accuracy": 0.7030284301606922,
      "accuracy_std_dev": 0.030250110911983902,
      "run_count": 33,
      "total_questions": 1618,
      "avg_cost_per_task_usd": 0.10101714894932014,
      "avg_duration_per_task_ms": 71347.47466007416,
      "avg_output_tokens_per_task": 8063.588998763906,
      "last_run_timestamp": "2026-10-24T19:37:14.677Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Parasail",
      "accuracy": 0.7027272727272728,
      "accuracy_std_dev": 0.026469593069959645,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.11361526181818182,
      "avg_duration_per_task_ms": 93548.7109090909,
      "avg_output_tokens_per_task": 9867.885454545454,
      "last_run_timestamp": "2026-10-24T19:14:21.866Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Inceptron",
      "accuracy": 0.7022816510851418,
      "accuracy_std_dev": 0.032548814933197716,
      "run_count": 12,
      "total_questions": 599,
      "avg_cost_per_task_usd": 0.07496608848080134,
      "avg_duration_per_task_ms": 173732.56427378964,
      "avg_output_tokens_per_task": 9603.57429048414,
      "last_run_timestamp": "2026-10-24T19:09:54.424Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "ModelRun",
      "accuracy": 0.7022223333333333,
      "accuracy_std_dev": 0.013698571612479225,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.013673380000000002,
      "avg_duration_per_task_ms": 132826.08666666667,
      "avg_output_tokens_per_task": 12739.9,
      "last_run_timestamp": "2026-10-28T05:04:05.678Z"
    },
    {
      "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": 0.7022223333333333,
      "accuracy_std_dev": 0.021999169459070246,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.10818771333333332,
      "avg_duration_per_task_ms": 79804.25333333333,
      "avg_output_tokens_per_task": 3542.24,
      "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": "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-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Decart",
      "accuracy": 0.7018970975609756,
      "accuracy_std_dev": 0.11960704720475296,
      "run_count": 5,
      "total_questions": 246,
      "avg_cost_per_task_usd": 0.0732234268292683,
      "avg_duration_per_task_ms": 155630.15853658537,
      "avg_output_tokens_per_task": 15582.174796747968,
      "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": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "Venice",
      "accuracy": 0.70166675,
      "accuracy_std_dev": 0.02939198209014661,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.046523760000000004,
      "avg_duration_per_task_ms": 539638.46,
      "avg_output_tokens_per_task": 9099.4325,
      "last_run_timestamp": "2026-10-24T20:03:06.027Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Novita",
      "accuracy": 0.7015283553008596,
      "accuracy_std_dev": 0.030155364434940605,
      "run_count": 7,
      "total_questions": 349,
      "avg_cost_per_task_usd": 0.02074343553008596,
      "avg_duration_per_task_ms": 613410.9627507164,
      "avg_output_tokens_per_task": 9252.200573065902,
      "last_run_timestamp": "2026-10-24T19:39:11.501Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Parasail",
      "accuracy": 0.701111,
      "accuracy_std_dev": 0.005665795148667743,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.05215248,
      "avg_duration_per_task_ms": 222421.00666666665,
      "avg_output_tokens_per_task": 6776.133333333333,
      "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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Decart",
      "accuracy": 0.701111,
      "accuracy_std_dev": 0.02276020351988678,
      "run_count": 15,
      "total_questions": 750,
      "avg_cost_per_task_usd": 0.14173139733333334,
      "avg_duration_per_task_ms": 80397.728,
      "avg_output_tokens_per_task": 10481.090666666667,
      "last_run_timestamp": "2026-10-24T19:06:47.526Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Venice",
      "accuracy": 0.7010051055276382,
      "accuracy_std_dev": 0.03189651765083454,
      "run_count": 4,
      "total_questions": 199,
      "avg_cost_per_task_usd": 0.016903984924623115,
      "avg_duration_per_task_ms": 263739.4221105528,
      "avg_output_tokens_per_task": 8905.97487437186,
      "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": "anthropic/claude-4-sonnet-20250522",
      "model_slug": "anthropic/claude-sonnet-4",
      "model_name": "Anthropic: Claude Sonnet 4",
      "provider_name": "auto-routing",
      "accuracy": 0.7,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.29769207999999997,
      "avg_duration_per_task_ms": 128375.6,
      "avg_output_tokens_per_task": 8500.86,
      "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": "z-ai/glm-5.3-flash",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "GLM 5.3 Flash",
      "provider_name": "digitalocean",
      "accuracy": 0.7,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.00357748,
      "avg_duration_per_task_ms": 48793.08,
      "avg_output_tokens_per_task": 1567.76,
      "last_run_timestamp": "2026-10-01T23:50:45.022Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "inclusionai/ling-3.0-flash-vl-20260910",
      "model_slug": "inclusionai/ling-3.0-flash-vl",
      "model_name": "inclusionAI: Ling 3.0 Flash VL",
      "provider_name": "auto-routing",
      "accuracy": 0.7,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.0030046400000000003,
      "avg_duration_per_task_ms": 98848.12,
      "avg_output_tokens_per_task": 17113.74,
      "last_run_timestamp": "2026-10-01T08:00:19.901Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.7,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.12726106,
      "avg_duration_per_task_ms": 181226.44,
      "avg_output_tokens_per_task": 9838.56,
      "last_run_timestamp": "2026-10-04T03:00:16.405Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "Phala",
      "accuracy": 0.7,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.07376152,
      "avg_duration_per_task_ms": 750952.84,
      "avg_output_tokens_per_task": 11282.28,
      "last_run_timestamp": "2026-08-29T16:44:16.336Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5-2025-08-07",
      "model_slug": "openai/gpt-5",
      "model_name": "OpenAI: GPT-5",
      "provider_name": "Azure",
      "accuracy": 0.7,
      "accuracy_std_dev": 0.004714280910170736,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.634261665,
      "avg_duration_per_task_ms": 673734.145,
      "avg_output_tokens_per_task": 52923.67,
      "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": "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": "openai/gpt-6.1-terra-20261026",
      "model_slug": "openai/gpt-6.1-terra",
      "model_name": "OpenAI: GPT-6.1 Terra",
      "provider_name": "OpenAI",
      "accuracy": 0.7,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.05778106,
      "avg_duration_per_task_ms": 103456.94,
      "avg_output_tokens_per_task": 4693.4,
      "last_run_timestamp": "2026-10-28T00:06:02.481Z"
    },
    {
      "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": 0.7,
      "accuracy_std_dev": 0.04203161508614677,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.06961394250000001,
      "avg_duration_per_task_ms": 353777.6075,
      "avg_output_tokens_per_task": 9996.6825,
      "last_run_timestamp": "2026-10-24T20:13:56.072Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "AtlasCloud",
      "accuracy": 0.69916675,
      "accuracy_std_dev": 0.10463910531530503,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.09495143499999999,
      "avg_duration_per_task_ms": 230681.74,
      "avg_output_tokens_per_task": 21996.935,
      "last_run_timestamp": "2026-10-24T19:19:49.639Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "DeepInfra",
      "accuracy": 0.6987950281124499,
      "accuracy_std_dev": 0.0269759350636822,
      "run_count": 5,
      "total_questions": 249,
      "avg_cost_per_task_usd": 0.02643562248995984,
      "avg_duration_per_task_ms": 223792.88353413655,
      "avg_output_tokens_per_task": 9633.240963855422,
      "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": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "auto-routing",
      "accuracy": 0.698333375,
      "accuracy_std_dev": 0.030867217869195378,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.051756159999999996,
      "avg_duration_per_task_ms": 379051.165,
      "avg_output_tokens_per_task": 9248.245,
      "last_run_timestamp": "2026-10-24T19:50:05.039Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Cloudflare",
      "accuracy": 0.6977777333333334,
      "accuracy_std_dev": 0.035311727846834236,
      "run_count": 15,
      "total_questions": 750,
      "avg_cost_per_task_usd": 0.12898300933333331,
      "avg_duration_per_task_ms": 132102.74533333333,
      "avg_output_tokens_per_task": 9039.205333333333,
      "last_run_timestamp": "2026-10-24T19:19:29.660Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-sol-20260922",
      "model_slug": "openai/gpt-6-sol",
      "model_name": "OpenAI: GPT-6 Sol",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.6977776666666667,
      "accuracy_std_dev": 0.01911642045177101,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.14221055333333335,
      "avg_duration_per_task_ms": 130884.03333333334,
      "avg_output_tokens_per_task": 4996.426666666666,
      "last_run_timestamp": "2026-10-30T00:14:00.742Z"
    },
    {
      "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": 0.6970368888888889,
      "accuracy_std_dev": 0.01221661375563299,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.021826557777777776,
      "avg_duration_per_task_ms": 527673.02,
      "avg_output_tokens_per_task": 18763.36,
      "last_run_timestamp": "2026-10-24T19:03:53.319Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.696667,
      "accuracy_std_dev": 0.010000000000000009,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01321477,
      "avg_duration_per_task_ms": 116452.62,
      "avg_output_tokens_per_task": 5139.02,
      "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-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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Reka",
      "accuracy": 0.6958335,
      "accuracy_std_dev": 0.03781341986980283,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.132227115,
      "avg_duration_per_task_ms": 205681.995,
      "avg_output_tokens_per_task": 9407.695,
      "last_run_timestamp": "2026-10-01T22:50:26.018Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.7-flash-20260119",
      "model_slug": "z-ai/glm-4.7-flash",
      "model_name": "Z.ai: GLM 4.7 Flash",
      "provider_name": "Venice",
      "accuracy": 0.6948196351351352,
      "accuracy_std_dev": 0.0038228551575432913,
      "run_count": 3,
      "total_questions": 148,
      "avg_cost_per_task_usd": 0.005936054054054054,
      "avg_duration_per_task_ms": 331167.7297297297,
      "avg_output_tokens_per_task": 9197.358108108108,
      "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": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Mancer 2",
      "accuracy": 0.6947621428571428,
      "accuracy_std_dev": 0.08007667414141911,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.051026225714285715,
      "avg_duration_per_task_ms": 483301.40285714285,
      "avg_output_tokens_per_task": 16016.288571428571,
      "last_run_timestamp": "2026-10-24T19:51:41.428Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Together",
      "accuracy": 0.6933334999999999,
      "accuracy_std_dev": 0.006666499999999964,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01276862,
      "avg_duration_per_task_ms": 66785.21,
      "avg_output_tokens_per_task": 8630.75,
      "last_run_timestamp": "2026-10-24T12:58:47.706Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.6933334999999999,
      "accuracy_std_dev": 0.018257236010689022,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.08853248000000001,
      "avg_duration_per_task_ms": 338024.125,
      "avg_output_tokens_per_task": 32557.16,
      "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": "openai/gpt-6-luna-pro-20260922",
      "model_slug": "openai/gpt-6-luna-pro",
      "model_name": "OpenAI: GPT-6 Luna Pro",
      "provider_name": "OpenAI",
      "accuracy": 0.6933333333333334,
      "accuracy_std_dev": 0.014401903168987411,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.03953364,
      "avg_duration_per_task_ms": 309829.13333333336,
      "avg_output_tokens_per_task": 32261.04,
      "last_run_timestamp": "2026-10-02T23:10:33.891Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Cohere",
      "accuracy": 0.693333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.01518678,
      "avg_duration_per_task_ms": 46885.36,
      "avg_output_tokens_per_task": 12120.66,
      "last_run_timestamp": "2026-10-28T21:15:52.006Z"
    },
    {
      "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": 0.693333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.05167764,
      "avg_duration_per_task_ms": 69227.88,
      "avg_output_tokens_per_task": 9318.98,
      "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": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "BaseTen",
      "accuracy": 0.693333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.03237594,
      "avg_duration_per_task_ms": 69982.76,
      "avg_output_tokens_per_task": 15576.66,
      "last_run_timestamp": "2026-10-20T06:20:53.664Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-sol-20260922",
      "model_slug": "openai/gpt-6-sol",
      "model_name": "OpenAI: GPT-6 Sol",
      "provider_name": "auto-routing",
      "accuracy": 0.693333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.12999802000000002,
      "avg_duration_per_task_ms": 96250.54,
      "avg_output_tokens_per_task": 4982.06,
      "last_run_timestamp": "2026-10-01T08:05:35.867Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "reflection/beam-1.0-501b-a23b-20261016",
      "model_slug": "reflection/beam-1.0-501b-a23b",
      "model_name": "Reflection: Beam 1.0 501B A23B",
      "provider_name": "auto-routing",
      "accuracy": 0.693333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.02754302,
      "avg_duration_per_task_ms": 209408.8,
      "avg_output_tokens_per_task": 14914.54,
      "last_run_timestamp": "2026-10-24T12:03:24.771Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Mistral",
      "accuracy": 0.6920511153846154,
      "accuracy_std_dev": 0.02287235899731531,
      "run_count": 26,
      "total_questions": 1300,
      "avg_cost_per_task_usd": 0.10611875923076924,
      "avg_duration_per_task_ms": 75613.00384615385,
      "avg_output_tokens_per_task": 9658.539230769231,
      "last_run_timestamp": "2026-10-30T16:20:14.521Z"
    },
    {
      "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": 0.69,
      "accuracy_std_dev": 0.003332999999999975,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.10882898999999999,
      "avg_duration_per_task_ms": 135482.83,
      "avg_output_tokens_per_task": 51335.48,
      "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": "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": "thinkingmachines/inkling-20260715",
      "model_slug": "thinkingmachines/inkling",
      "model_name": "Thinking Machines: Inkling",
      "provider_name": "BaseTen",
      "accuracy": 0.69,
      "accuracy_std_dev": 0.04333300000000001,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.07366,
      "avg_duration_per_task_ms": 63569.01,
      "avg_output_tokens_per_task": 3649.5,
      "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": "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": 0.6897523648648648,
      "accuracy_std_dev": 0.024522787726702033,
      "run_count": 6,
      "total_questions": 296,
      "avg_cost_per_task_usd": 0.02191278716216216,
      "avg_duration_per_task_ms": 348901.27364864864,
      "avg_output_tokens_per_task": 16679.054054054053,
      "last_run_timestamp": "2026-10-27T15:35:36.855Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "Parasail",
      "accuracy": 0.6886632385786803,
      "accuracy_std_dev": 0.0900995905873051,
      "run_count": 8,
      "total_questions": 394,
      "avg_cost_per_task_usd": 0.1329306802030457,
      "avg_duration_per_task_ms": 1780497.6497461929,
      "avg_output_tokens_per_task": 98848.87817258883,
      "last_run_timestamp": "2026-10-24T19:18:22.830Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "GMICloud",
      "accuracy": 0.6877776666666667,
      "accuracy_std_dev": 0.0470485263023426,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.09563632666666667,
      "avg_duration_per_task_ms": 192574.105,
      "avg_output_tokens_per_task": 7480.165,
      "last_run_timestamp": "2026-10-27T01:50:56.696Z"
    },
    {
      "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": 0.6874997500000001,
      "accuracy_std_dev": 0.045023109806936926,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0177425775,
      "avg_duration_per_task_ms": 384968.0575,
      "avg_output_tokens_per_task": 20103.015,
      "last_run_timestamp": "2026-10-24T21:02:58.473Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "openai/gpt-6.1-terra-20261026",
      "model_slug": "openai/gpt-6.1-terra",
      "model_name": "OpenAI: GPT-6.1 Terra",
      "provider_name": "Azure",
      "accuracy": 0.686667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.06838028,
      "avg_duration_per_task_ms": 78879.18,
      "avg_output_tokens_per_task": 4783.1,
      "last_run_timestamp": "2026-10-28T00:05:23.989Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Tenstorrent",
      "accuracy": 0.6866665000000001,
      "accuracy_std_dev": 0.006666499999999964,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01639777,
      "avg_duration_per_task_ms": 134478.47,
      "avg_output_tokens_per_task": 16372.87,
      "last_run_timestamp": "2026-10-30T18:45:46.265Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Wafer",
      "accuracy": 0.6866665000000001,
      "accuracy_std_dev": 0.026666499999999982,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01561806,
      "avg_duration_per_task_ms": 117250.14,
      "avg_output_tokens_per_task": 14519.81,
      "last_run_timestamp": "2026-10-24T12:06:50.408Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Friendli",
      "accuracy": 0.6861111666666666,
      "accuracy_std_dev": 0.0669138081126675,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.11072275,
      "avg_duration_per_task_ms": 76500.85166666667,
      "avg_output_tokens_per_task": 9058.545,
      "last_run_timestamp": "2026-10-28T11:20:49.843Z"
    },
    {
      "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": 0.6854528208092485,
      "accuracy_std_dev": 0.016921495753349107,
      "run_count": 7,
      "total_questions": 346,
      "avg_cost_per_task_usd": 0.03335441329479769,
      "avg_duration_per_task_ms": 533340.3294797688,
      "avg_output_tokens_per_task": 18704.07225433526,
      "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": "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": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Minimax",
      "accuracy": 0.685000125,
      "accuracy_std_dev": 0.03693083437805562,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0358575875,
      "avg_duration_per_task_ms": 157897.08,
      "avg_output_tokens_per_task": 9891.505,
      "last_run_timestamp": "2026-08-29T17:02:26.936Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "DigitalOcean",
      "accuracy": 0.6837050185873607,
      "accuracy_std_dev": 0.11487260701179351,
      "run_count": 11,
      "total_questions": 538,
      "avg_cost_per_task_usd": 0.32785563568773235,
      "avg_duration_per_task_ms": 1690067.0799256505,
      "avg_output_tokens_per_task": 54783.62267657993,
      "last_run_timestamp": "2026-10-24T12:39:42.207Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.6834170150753769,
      "accuracy_std_dev": 0.019760471659034357,
      "run_count": 4,
      "total_questions": 199,
      "avg_cost_per_task_usd": 0.006376396984924624,
      "avg_duration_per_task_ms": 294466.5879396985,
      "avg_output_tokens_per_task": 9614.21608040201,
      "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": "DeepInfra",
      "accuracy": 0.6833333333333333,
      "accuracy_std_dev": 0.07779388246014092,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.00909372,
      "avg_duration_per_task_ms": 253755.2,
      "avg_output_tokens_per_task": 15941.146666666667,
      "last_run_timestamp": "2026-10-08T22:38:42.670Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6.1-sol-20260929",
      "model_slug": "openai/gpt-6.1-sol",
      "model_name": "OpenAI: GPT-6.1 Sol",
      "provider_name": "Azure",
      "accuracy": 0.6833333333333333,
      "accuracy_std_dev": 0.012619766977070334,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.10710260999999997,
      "avg_duration_per_task_ms": 135398.91,
      "avg_output_tokens_per_task": 4027.3166666666666,
      "last_run_timestamp": "2026-10-01T08:06:09.665Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "Darkbloom",
      "accuracy": 0.683333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.00634196,
      "avg_duration_per_task_ms": 404260.22,
      "avg_output_tokens_per_task": 17374,
      "last_run_timestamp": "2026-10-17T11:31:13.423Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.6826667,
      "accuracy_std_dev": 0.03427992350647823,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.01766094,
      "avg_duration_per_task_ms": 313229.622,
      "avg_output_tokens_per_task": 20074.712,
      "last_run_timestamp": "2026-10-30T16:21:36.179Z"
    },
    {
      "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": 0.6822223333333333,
      "accuracy_std_dev": 0.021998563367234276,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.014046139999999999,
      "avg_duration_per_task_ms": 71689.16,
      "avg_output_tokens_per_task": 5022.28,
      "last_run_timestamp": "2026-10-30T00:15:40.872Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6.1-sol-20260929",
      "model_slug": "openai/gpt-6.1-sol",
      "model_name": "OpenAI: GPT-6.1 Sol",
      "provider_name": "OpenAI",
      "accuracy": 0.6822223333333333,
      "accuracy_std_dev": 0.011330936570096727,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.09746202,
      "avg_duration_per_task_ms": 132162.47333333333,
      "avg_output_tokens_per_task": 3507.6066666666666,
      "last_run_timestamp": "2026-10-02T23:10:34.290Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "DekaLLM",
      "accuracy": 0.6822220000000001,
      "accuracy_std_dev": 0.020607975203789406,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.015206260000000001,
      "avg_duration_per_task_ms": 252877.44,
      "avg_output_tokens_per_task": 17489.156666666666,
      "last_run_timestamp": "2026-10-24T20:33:46.929Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Minimax",
      "accuracy": 0.68,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.01807364,
      "avg_duration_per_task_ms": 208774.72,
      "avg_output_tokens_per_task": 16251.38,
      "last_run_timestamp": "2026-10-24T12:17:54.200Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-astra-20260903",
      "model_slug": "openai/gpt-6-astra",
      "model_name": "OpenAI: GPT-6 Astra",
      "provider_name": "auto-routing",
      "accuracy": 0.68,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.58024224,
      "avg_duration_per_task_ms": 141487.98,
      "avg_output_tokens_per_task": 3731.58,
      "last_run_timestamp": "2026-10-01T08:05:19.702Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-luna-20260922",
      "model_slug": "openai/gpt-6-luna",
      "model_name": "OpenAI: GPT-6 Luna",
      "provider_name": "OpenAI",
      "accuracy": 0.68,
      "accuracy_std_dev": 0.03810292105162893,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.00883898,
      "avg_duration_per_task_ms": 114396.12,
      "avg_output_tokens_per_task": 8841.28,
      "last_run_timestamp": "2026-10-02T23:08:44.329Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.68,
      "accuracy_std_dev": 0.06,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.049401919999999995,
      "avg_duration_per_task_ms": 283175.04,
      "avg_output_tokens_per_task": 24143.94,
      "last_run_timestamp": "2026-10-29T23:29:22.134Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "CoreWeave",
      "accuracy": 0.6785714285714286,
      "accuracy_std_dev": 0.020303791411297286,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.017469171428571427,
      "avg_duration_per_task_ms": 119261.06857142856,
      "avg_output_tokens_per_task": 21365.42,
      "last_run_timestamp": "2026-10-24T11:42:22.586Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "Fireworks",
      "accuracy": 0.6777776666666667,
      "accuracy_std_dev": 0.013698571612479227,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.0340762,
      "avg_duration_per_task_ms": 244093.44,
      "avg_output_tokens_per_task": 27086.806666666667,
      "last_run_timestamp": "2026-09-19T08:01:53.393Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "DigitalOcean",
      "accuracy": 0.676667,
      "accuracy_std_dev": 0.010000000000000009,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01545023,
      "avg_duration_per_task_ms": 128578.81,
      "avg_output_tokens_per_task": 15653.52,
      "last_run_timestamp": "2026-10-24T12:17:01.920Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.67666675,
      "accuracy_std_dev": 0.028867484592314224,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.008505605000000001,
      "avg_duration_per_task_ms": 317641.53,
      "avg_output_tokens_per_task": 13432.655,
      "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": "openai/gpt-6-astra-20260903",
      "model_slug": "openai/gpt-6-astra",
      "model_name": "OpenAI: GPT-6 Astra",
      "provider_name": "OpenAI",
      "accuracy": 0.6766665000000001,
      "accuracy_std_dev": 0.0033335000000000448,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.5967202500000001,
      "avg_duration_per_task_ms": 137724.15,
      "avg_output_tokens_per_task": 4015.19,
      "last_run_timestamp": "2026-10-02T23:06:01.512Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Tenstorrent",
      "accuracy": 0.676,
      "accuracy_std_dev": 0.19009238689542513,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.011841847999999999,
      "avg_duration_per_task_ms": 391925.292,
      "avg_output_tokens_per_task": 10879.944,
      "last_run_timestamp": "2026-10-24T13:02:10.008Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "Google",
      "accuracy": 0.6757320878661087,
      "accuracy_std_dev": 0.02326584314053542,
      "run_count": 5,
      "total_questions": 239,
      "avg_cost_per_task_usd": 0.14251672803347282,
      "avg_duration_per_task_ms": 407673.5857740586,
      "avg_output_tokens_per_task": 10868.456066945606,
      "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": "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": "openai/gpt-6-astra-20260903",
      "model_slug": "openai/gpt-6-astra",
      "model_name": "OpenAI: GPT-6 Astra",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.6755556666666667,
      "accuracy_std_dev": 0.006285236475706844,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.63842362,
      "avg_duration_per_task_ms": 88201.35333333333,
      "avg_output_tokens_per_task": 3898.48,
      "last_run_timestamp": "2026-10-30T00:17:00.667Z"
    },
    {
      "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": "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": "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": 0.6741405246636771,
      "accuracy_std_dev": 0.030345316137851863,
      "run_count": 9,
      "total_questions": 446,
      "avg_cost_per_task_usd": 0.022004533632286993,
      "avg_duration_per_task_ms": 284802.9730941704,
      "avg_output_tokens_per_task": 16879.719730941702,
      "last_run_timestamp": "2026-10-24T19:52:46.621Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Morph",
      "accuracy": 0.6733334999999999,
      "accuracy_std_dev": 0.013333499999999998,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.09937637,
      "avg_duration_per_task_ms": 637553.26,
      "avg_output_tokens_per_task": 9917.08,
      "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-5-2025-08-07",
      "model_slug": "openai/gpt-5",
      "model_name": "OpenAI: GPT-5",
      "provider_name": "auto-routing",
      "accuracy": 0.6733334999999999,
      "accuracy_std_dev": 0.026666499999999982,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.6292479799999999,
      "avg_duration_per_task_ms": 729264.79,
      "avg_output_tokens_per_task": 54931.75,
      "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": "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": 0.6733330000000001,
      "accuracy_std_dev": 0.019999999999999962,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 1.5819717300000002,
      "avg_duration_per_task_ms": 314318.37,
      "avg_output_tokens_per_task": 23725.55,
      "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": "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": "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": "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": 0.6719998,
      "accuracy_std_dev": 0.012927091279943813,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.015563263999999999,
      "avg_duration_per_task_ms": 230227.66,
      "avg_output_tokens_per_task": 10531.424,
      "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-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": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "GMICloud",
      "accuracy": 0.669999875,
      "accuracy_std_dev": 0.10765177786669097,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.017295960000000003,
      "avg_duration_per_task_ms": 256892.225,
      "avg_output_tokens_per_task": 10151.75,
      "last_run_timestamp": "2026-10-27T02:33:21.663Z"
    },
    {
      "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": 0.6686666,
      "accuracy_std_dev": 0.08499931398452577,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.05116424000000001,
      "avg_duration_per_task_ms": 341233.54,
      "avg_output_tokens_per_task": 7834.636,
      "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": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "DigitalOcean",
      "accuracy": 0.66833325,
      "accuracy_std_dev": 0.005527532106419614,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.027697885000000002,
      "avg_duration_per_task_ms": 410531.75,
      "avg_output_tokens_per_task": 10158.81,
      "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": "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": 0.6670706666666666,
      "accuracy_std_dev": 0.03017076900885787,
      "run_count": 33,
      "total_questions": 1650,
      "avg_cost_per_task_usd": 0.12485057212121214,
      "avg_duration_per_task_ms": 121821.06121212122,
      "avg_output_tokens_per_task": 12714.262424242424,
      "last_run_timestamp": "2026-10-24T13:04:43.619Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "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": 0.66666675,
      "accuracy_std_dev": 0.016996797086142445,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.008543925,
      "avg_duration_per_task_ms": 288167.26,
      "avg_output_tokens_per_task": 9621.62,
      "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": "GMICloud",
      "accuracy": 0.6666667,
      "accuracy_std_dev": 0.1539986724332713,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.12421460999999999,
      "avg_duration_per_task_ms": 361200.082,
      "avg_output_tokens_per_task": 8728.562,
      "last_run_timestamp": "2026-10-27T01:27:10.502Z"
    },
    {
      "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": 0.6666665,
      "accuracy_std_dev": 0.013333499999999998,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.08889954,
      "avg_duration_per_task_ms": 108433.06,
      "avg_output_tokens_per_task": 12861.73,
      "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": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Friendli",
      "accuracy": 0.6651852222222222,
      "accuracy_std_dev": 0.07567163233159697,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.09013567777777777,
      "avg_duration_per_task_ms": 120445.85555555555,
      "avg_output_tokens_per_task": 8441.84,
      "last_run_timestamp": "2026-10-28T11:09:01.971Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "DekaLLM",
      "accuracy": 0.6649998333333333,
      "accuracy_std_dev": 0.024626931209664655,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.008708743333333335,
      "avg_duration_per_task_ms": 613076.7966666666,
      "avg_output_tokens_per_task": 17440.79,
      "last_run_timestamp": "2026-10-24T19:39:37.861Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "Io Net",
      "accuracy": 0.66493075,
      "accuracy_std_dev": 0.03659399999999996,
      "run_count": 2,
      "total_questions": 96,
      "avg_cost_per_task_usd": 0.0157241875,
      "avg_duration_per_task_ms": 279786.6875,
      "avg_output_tokens_per_task": 12241.416666666666,
      "last_run_timestamp": "2026-10-27T13:23:22.200Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "SambaNova",
      "accuracy": 0.6632998585858586,
      "accuracy_std_dev": 0.0034015000000000017,
      "run_count": 2,
      "total_questions": 99,
      "avg_cost_per_task_usd": 0.07516567676767677,
      "avg_duration_per_task_ms": 78894.82828282828,
      "avg_output_tokens_per_task": 13736.282828282829,
      "last_run_timestamp": "2026-08-15T14:12:46.929Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "DigitalOcean",
      "accuracy": 0.662963,
      "accuracy_std_dev": 0.06555468465673864,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.05622624222222223,
      "avg_duration_per_task_ms": 819076.5222222222,
      "avg_output_tokens_per_task": 18202.83777777778,
      "last_run_timestamp": "2026-10-05T20:43:11.175Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.5-haiku-20251001",
      "model_slug": "anthropic/claude-haiku-4.5",
      "model_name": "Anthropic: Claude Haiku 4.5",
      "provider_name": "auto-routing",
      "accuracy": 0.6618180909090909,
      "accuracy_std_dev": 0.03062851381683575,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.11597669454545456,
      "avg_duration_per_task_ms": 125484.61818181818,
      "avg_output_tokens_per_task": 12643.88,
      "last_run_timestamp": "2026-10-24T12:09:26.665Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.4-mini-20260317",
      "model_slug": "openai/gpt-5.4-mini",
      "model_name": "OpenAI: GPT-5.4 Mini",
      "provider_name": "Azure",
      "accuracy": 0.6613334000000001,
      "accuracy_std_dev": 0.01758814752724117,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.10186603200000001,
      "avg_duration_per_task_ms": 156255.32,
      "avg_output_tokens_per_task": 13208.728,
      "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": "anthropic/claude-4.5-haiku-20251001",
      "model_slug": "anthropic/claude-haiku-4.5",
      "model_name": "Anthropic: Claude Haiku 4.5",
      "provider_name": "Anthropic",
      "accuracy": 0.6612121818181818,
      "accuracy_std_dev": 0.027386941316879746,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.11405556727272728,
      "avg_duration_per_task_ms": 127222.07090909091,
      "avg_output_tokens_per_task": 12352.41090909091,
      "last_run_timestamp": "2026-10-24T12:22:57.423Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.66,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.03137252,
      "avg_duration_per_task_ms": 226181.32,
      "avg_output_tokens_per_task": 15432.64,
      "last_run_timestamp": "2026-10-29T23:53:36.758Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.66,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.013329200000000001,
      "avg_duration_per_task_ms": 278453.68,
      "avg_output_tokens_per_task": 15070.16,
      "last_run_timestamp": "2026-10-24T12:29:49.995Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.5-haiku-20251001",
      "model_slug": "anthropic/claude-haiku-4.5",
      "model_name": "Anthropic: Claude Haiku 4.5",
      "provider_name": "Google",
      "accuracy": 0.6590195882352939,
      "accuracy_std_dev": 0.028566314925427196,
      "run_count": 34,
      "total_questions": 1700,
      "avg_cost_per_task_usd": 0.12125589529411765,
      "avg_duration_per_task_ms": 121695.54823529412,
      "avg_output_tokens_per_task": 12641.248823529411,
      "last_run_timestamp": "2026-10-24T12:56:50.982Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Near AI",
      "accuracy": 0.6588886666666667,
      "accuracy_std_dev": 0.03965918385723817,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.006265106666666667,
      "avg_duration_per_task_ms": 362004.0733333333,
      "avg_output_tokens_per_task": 19036.246666666666,
      "last_run_timestamp": "2026-10-24T19:34:04.376Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Parasail",
      "accuracy": 0.6585712857142858,
      "accuracy_std_dev": 0.014125931834894325,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.023240128571428573,
      "avg_duration_per_task_ms": 158768.29142857142,
      "avg_output_tokens_per_task": 17172.022857142856,
      "last_run_timestamp": "2026-10-24T19:17:27.524Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "auto-routing",
      "accuracy": 0.6583334999999999,
      "accuracy_std_dev": 0.028037782049406136,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.02298765,
      "avg_duration_per_task_ms": 355829.865,
      "avg_output_tokens_per_task": 10214.775,
      "last_run_timestamp": "2026-08-29T16:11:34.887Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Friendli",
      "accuracy": 0.65666675,
      "accuracy_std_dev": 0.0074535971978300555,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.02864462,
      "avg_duration_per_task_ms": 223453.205,
      "avg_output_tokens_per_task": 10372.695,
      "last_run_timestamp": "2026-08-29T16:14:58.314Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "auto-routing",
      "accuracy": 0.6566381031518624,
      "accuracy_std_dev": 0.020222279406836508,
      "run_count": 7,
      "total_questions": 349,
      "avg_cost_per_task_usd": 0.01698823782234957,
      "avg_duration_per_task_ms": 325791.0286532951,
      "avg_output_tokens_per_task": 22916.045845272205,
      "last_run_timestamp": "2026-10-24T12:47:29.776Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "AkashML",
      "accuracy": 0.6558186153846154,
      "accuracy_std_dev": 0.022915354823914075,
      "run_count": 17,
      "total_questions": 845,
      "avg_cost_per_task_usd": 0.010244520710059171,
      "avg_duration_per_task_ms": 311624.14082840236,
      "avg_output_tokens_per_task": 17072.48402366864,
      "last_run_timestamp": "2026-10-30T00:56:40.041Z"
    },
    {
      "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-5.4-nano-20260317",
      "model_slug": "openai/gpt-5.4-nano",
      "model_name": "OpenAI: GPT-5.4 Nano",
      "provider_name": "OpenAI",
      "accuracy": 0.6555556666666666,
      "accuracy_std_dev": 0.012570472951413636,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.030086579999999998,
      "avg_duration_per_task_ms": 125366.48,
      "avg_output_tokens_per_task": 16133.046666666667,
      "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": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "DigitalOcean",
      "accuracy": 0.654618485943775,
      "accuracy_std_dev": 0.11865847802618236,
      "run_count": 10,
      "total_questions": 498,
      "avg_cost_per_task_usd": 0.04366364257028112,
      "avg_duration_per_task_ms": 1933703.5220883535,
      "avg_output_tokens_per_task": 71251.46586345382,
      "last_run_timestamp": "2026-10-24T12:24:29.901Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "SiliconFlow",
      "accuracy": 0.654279400900901,
      "accuracy_std_dev": 0.11858723742179567,
      "run_count": 9,
      "total_questions": 444,
      "avg_cost_per_task_usd": 0.025987085585585587,
      "avg_duration_per_task_ms": 1029787.9324324324,
      "avg_output_tokens_per_task": 34152.32432432433,
      "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "auto-routing",
      "accuracy": 0.65416675,
      "accuracy_std_dev": 0.03373725891262509,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.013133230000000001,
      "avg_duration_per_task_ms": 415006.955,
      "avg_output_tokens_per_task": 18075.7325,
      "last_run_timestamp": "2026-10-24T19:51:37.271Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "inclusionai/ling-3.0-flash-fin-20260827",
      "model_slug": "inclusionai/ling-3.0-flash-fin",
      "model_name": "inclusionAI: Ling 3.0 Flash Fin",
      "provider_name": "Novita",
      "accuracy": 0.6533335,
      "accuracy_std_dev": 0.00666650000000002,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.00488981,
      "avg_duration_per_task_ms": 111736.2,
      "avg_output_tokens_per_task": 17377.29,
      "last_run_timestamp": "2026-10-01T08:04:38.584Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4-sonnet-20250522",
      "model_slug": "anthropic/claude-sonnet-4",
      "model_name": "Anthropic: Claude Sonnet 4",
      "provider_name": "Google",
      "accuracy": 0.6533333333333333,
      "accuracy_std_dev": 0.014401440259764123,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.2856746933333333,
      "avg_duration_per_task_ms": 183725.1,
      "avg_output_tokens_per_task": 8628.813333333334,
      "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": "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": "DeepInfra",
      "accuracy": 0.65166675,
      "accuracy_std_dev": 0.038657908491504044,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.021260025000000002,
      "avg_duration_per_task_ms": 249515.925,
      "avg_output_tokens_per_task": 26907.185,
      "last_run_timestamp": "2026-10-24T12:38:08.218Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.6515152727272727,
      "accuracy_std_dev": 0.024959225520075408,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.11565049454545455,
      "avg_duration_per_task_ms": 139062.33454545453,
      "avg_output_tokens_per_task": 12670.038181818181,
      "last_run_timestamp": "2026-10-24T11:52:54.886Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.6513452645739911,
      "accuracy_std_dev": 0.03893599144686849,
      "run_count": 9,
      "total_questions": 446,
      "avg_cost_per_task_usd": 0.012713934977578476,
      "avg_duration_per_task_ms": 951818.3878923766,
      "avg_output_tokens_per_task": 23060.636771300447,
      "last_run_timestamp": "2026-10-27T16:29:30.347Z"
    },
    {
      "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": "Crusoe",
      "accuracy": 0.6499997499999999,
      "accuracy_std_dev": 0.0378593669609979,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.015033419999999999,
      "avg_duration_per_task_ms": 89205.15,
      "avg_output_tokens_per_task": 19468.825,
      "last_run_timestamp": "2026-10-24T19:10:30.177Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Groq",
      "accuracy": 0.6495835,
      "accuracy_std_dev": 0.03429607808846371,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.034037475,
      "avg_duration_per_task_ms": 61916.2125,
      "avg_output_tokens_per_task": 14847.6825,
      "last_run_timestamp": "2026-10-24T18:54:16.139Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.6491228070175439,
      "accuracy_std_dev": 0.025681291234756373,
      "run_count": 8,
      "total_questions": 399,
      "avg_cost_per_task_usd": 0.01387597493734336,
      "avg_duration_per_task_ms": 805180.8822055138,
      "avg_output_tokens_per_task": 25965.736842105263,
      "last_run_timestamp": "2026-10-30T03:01:20.302Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "Venice",
      "accuracy": 0.6490788090452262,
      "accuracy_std_dev": 0.06101351971284725,
      "run_count": 8,
      "total_questions": 398,
      "avg_cost_per_task_usd": 0.02417512814070352,
      "avg_duration_per_task_ms": 427158.12060301506,
      "avg_output_tokens_per_task": 15670.065326633166,
      "last_run_timestamp": "2026-10-24T19:06:03.144Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.4-mini-20260317",
      "model_slug": "openai/gpt-5.4-mini",
      "model_name": "OpenAI: GPT-5.4 Mini",
      "provider_name": "OpenAI",
      "accuracy": 0.6488889999999999,
      "accuracy_std_dev": 0.019115994210782406,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.08501352666666667,
      "avg_duration_per_task_ms": 97755.09333333334,
      "avg_output_tokens_per_task": 12128.533333333333,
      "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": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "SiliconFlow",
      "accuracy": 0.64833325,
      "accuracy_std_dev": 0.021794475600195134,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.022383950000000007,
      "avg_duration_per_task_ms": 245046.31,
      "avg_output_tokens_per_task": 9447.4,
      "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": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "Venice",
      "accuracy": 0.647177495967742,
      "accuracy_std_dev": 0.07108568891563195,
      "run_count": 10,
      "total_questions": 496,
      "avg_cost_per_task_usd": 0.025456885080645163,
      "avg_duration_per_task_ms": 277301.2076612903,
      "avg_output_tokens_per_task": 25224.120967741936,
      "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": "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": 0.6466665,
      "accuracy_std_dev": 0.026666499999999982,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.0295618,
      "avg_duration_per_task_ms": 121397.75,
      "avg_output_tokens_per_task": 15953.93,
      "last_run_timestamp": "2026-10-01T19:04:28.503Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-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": "Inceptron",
      "accuracy": 0.6416665,
      "accuracy_std_dev": 0.02833350000000001,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.0220575,
      "avg_duration_per_task_ms": 236336.73,
      "avg_output_tokens_per_task": 9832.45,
      "last_run_timestamp": "2026-08-15T14:15:19.430Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.4-nano-20260317",
      "model_slug": "openai/gpt-5.4-nano",
      "model_name": "OpenAI: GPT-5.4 Nano",
      "provider_name": "Azure",
      "accuracy": 0.6413336000000001,
      "accuracy_std_dev": 0.014847377621654282,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.03101342,
      "avg_duration_per_task_ms": 232783.988,
      "avg_output_tokens_per_task": 15818.512,
      "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": "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": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "AkashML",
      "accuracy": 0.64,
      "accuracy_std_dev": 0.023726622023934812,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.016758953333333333,
      "avg_duration_per_task_ms": 469306.68666666665,
      "avg_output_tokens_per_task": 27813.52,
      "last_run_timestamp": "2026-10-30T01:16:49.428Z"
    },
    {
      "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "DeepInfra",
      "accuracy": 0.6393825126903554,
      "accuracy_std_dev": 0.042421817536747466,
      "run_count": 16,
      "total_questions": 788,
      "avg_cost_per_task_usd": 0.01991223350253807,
      "avg_duration_per_task_ms": 319706.6497461929,
      "avg_output_tokens_per_task": 16793.3730964467,
      "last_run_timestamp": "2026-10-24T19:19:43.152Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "inclusionai/ling-3.0-flash-fin-20260827",
      "model_slug": "inclusionai/ling-3.0-flash-fin",
      "model_name": "inclusionAI: Ling 3.0 Flash Fin",
      "provider_name": "DeepInfra",
      "accuracy": 0.638298,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 47,
      "avg_cost_per_task_usd": 0.020538510638297872,
      "avg_duration_per_task_ms": 686659.4680851063,
      "avg_output_tokens_per_task": 47094.31914893617,
      "last_run_timestamp": "2026-10-09T00:23:16.829Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Venice",
      "accuracy": 0.6366665,
      "accuracy_std_dev": 0.0033334999999999892,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.008583809999999999,
      "avg_duration_per_task_ms": 1454771.53,
      "avg_output_tokens_per_task": 17935.13,
      "last_run_timestamp": "2026-10-24T20:43:26.554Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Phala",
      "accuracy": 0.6353466979865772,
      "accuracy_std_dev": 0.036572065936844325,
      "run_count": 3,
      "total_questions": 149,
      "avg_cost_per_task_usd": 0.033807637583892616,
      "avg_duration_per_task_ms": 262695.20805369125,
      "avg_output_tokens_per_task": 17180.040268456374,
      "last_run_timestamp": "2026-10-24T18:33:47.845Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "AtlasCloud",
      "accuracy": 0.6333332500000001,
      "accuracy_std_dev": 0.12901321207123517,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.05091860250000001,
      "avg_duration_per_task_ms": 844509.295,
      "avg_output_tokens_per_task": 27160.0425,
      "last_run_timestamp": "2026-10-24T19:37:12.172Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Novita",
      "accuracy": 0.6333332500000001,
      "accuracy_std_dev": 0.016996502910525478,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.024200355,
      "avg_duration_per_task_ms": 140351.54,
      "avg_output_tokens_per_task": 10016.095,
      "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": 0.6333332500000001,
      "accuracy_std_dev": 0.03299835012826401,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.021382605000000002,
      "avg_duration_per_task_ms": 229847.125,
      "avg_output_tokens_per_task": 9805.265,
      "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": 0.6307017947368421,
      "accuracy_std_dev": 0.04099517109977711,
      "run_count": 4,
      "total_questions": 190,
      "avg_cost_per_task_usd": 0.02501245263157895,
      "avg_duration_per_task_ms": 66232.84736842105,
      "avg_output_tokens_per_task": 14146.842105263158,
      "last_run_timestamp": "2026-10-24T18:40:18.531Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "SambaNova",
      "accuracy": 0.6304712676767678,
      "accuracy_std_dev": 0.028764697857953853,
      "run_count": 8,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.041468260101010096,
      "avg_duration_per_task_ms": 81255.43686868687,
      "avg_output_tokens_per_task": 21574.416666666668,
      "last_run_timestamp": "2026-10-24T19:23:06.220Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "OpenInference",
      "accuracy": 0.6283335,
      "accuracy_std_dev": 0.18833350000000001,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.15963215,
      "avg_duration_per_task_ms": 1567590.01,
      "avg_output_tokens_per_task": 36431.88,
      "last_run_timestamp": "2026-10-17T19:19:25.280Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "AtlasCloud",
      "accuracy": 0.6283332500000001,
      "accuracy_std_dev": 0.023746444179866183,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.028451149999999998,
      "avg_duration_per_task_ms": 210873.285,
      "avg_output_tokens_per_task": 9700.365,
      "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": "Cerebras",
      "accuracy": 0.621666625,
      "accuracy_std_dev": 0.025331066237613754,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.06826625,
      "avg_duration_per_task_ms": 32002.2325,
      "avg_output_tokens_per_task": 21494.85,
      "last_run_timestamp": "2026-10-24T18:29:40.588Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.6155556666666666,
      "accuracy_std_dev": 0.011331263428624758,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.12338818000000001,
      "avg_duration_per_task_ms": 626122.66,
      "avg_output_tokens_per_task": 27741.31333333333,
      "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": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.615,
      "accuracy_std_dev": 0.03869391999397322,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0947713525,
      "avg_duration_per_task_ms": 55138.08,
      "avg_output_tokens_per_task": 3608.635,
      "last_run_timestamp": "2026-10-24T19:00:38.447Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "qwen/qwen3-coder-next-2025-02-03",
      "model_slug": "qwen/qwen3-coder-next",
      "model_name": "Qwen: Qwen3 Coder Next",
      "provider_name": "StreamLake",
      "accuracy": 0.6049998333333334,
      "accuracy_std_dev": 0.01979055610989468,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.04771943666666667,
      "avg_duration_per_task_ms": 93553.64666666667,
      "avg_output_tokens_per_task": 4645.643333333333,
      "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": "auto-routing",
      "accuracy": 0.6026666,
      "accuracy_std_dev": 0.023701878031919768,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.045193232,
      "avg_duration_per_task_ms": 105920.528,
      "avg_output_tokens_per_task": 4778.932,
      "last_run_timestamp": "2026-10-03T16:07:50.329Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Mancer 2",
      "accuracy": 0.6024903390804598,
      "accuracy_std_dev": 0.08896202865084331,
      "run_count": 7,
      "total_questions": 348,
      "avg_cost_per_task_usd": 0.031245040229885054,
      "avg_duration_per_task_ms": 1940985.2241379311,
      "avg_output_tokens_per_task": 61478.612068965514,
      "last_run_timestamp": "2026-10-24T19:11:22.177Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-2.5-pro",
      "model_slug": "google/gemini-2.5-pro",
      "model_name": "Google: Gemini 2.5 Pro",
      "provider_name": "Google",
      "accuracy": 0.6000001666666667,
      "accuracy_std_dev": 0.03771239112642186,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.19008634000000002,
      "avg_duration_per_task_ms": 186975.22,
      "avg_output_tokens_per_task": 11452.75,
      "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": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "SiliconFlow",
      "accuracy": 0.6000001666666667,
      "accuracy_std_dev": 0.13727506854659452,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.04416580999999999,
      "avg_duration_per_task_ms": 477149.00666666665,
      "avg_output_tokens_per_task": 8772.623333333333,
      "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": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "BaseTen",
      "accuracy": 0.593333,
      "accuracy_std_dev": 0.020000000000000018,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.04552234,
      "avg_duration_per_task_ms": 32685.39,
      "avg_output_tokens_per_task": 3125.98,
      "last_run_timestamp": "2026-09-26T08:22:32.737Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.6-20251208",
      "model_slug": "z-ai/glm-4.6v",
      "model_name": "Z.ai: GLM 4.6V",
      "provider_name": "Z.AI",
      "accuracy": 0.593333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.01896666,
      "avg_duration_per_task_ms": 306972.98,
      "avg_output_tokens_per_task": 7436.64,
      "last_run_timestamp": "2026-10-01T08:02:59.889Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-2.5-flash",
      "model_slug": "google/gemini-2.5-flash",
      "model_name": "Google: Gemini 2.5 Flash",
      "provider_name": "Google",
      "accuracy": 0.591111,
      "accuracy_std_dev": 0.006285295405415193,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.029060479999999996,
      "avg_duration_per_task_ms": 77755.87,
      "avg_output_tokens_per_task": 7763.98,
      "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": "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": 0.588889,
      "accuracy_std_dev": 0.008314764498569197,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.14882857333333332,
      "avg_duration_per_task_ms": 490464.7866666667,
      "avg_output_tokens_per_task": 24032.806666666667,
      "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": "qwen/qwen3-coder-next-2025-02-03",
      "model_slug": "qwen/qwen3-coder-next",
      "model_name": "Qwen: Qwen3 Coder Next",
      "provider_name": "Parasail",
      "accuracy": 0.5873334,
      "accuracy_std_dev": 0.015691608516656286,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.020115355999999997,
      "avg_duration_per_task_ms": 100935.536,
      "avg_output_tokens_per_task": 4366.388,
      "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": "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": "google/gemini-2.5-pro",
      "model_slug": "google/gemini-2.5-pro",
      "model_name": "Google: Gemini 2.5 Pro",
      "provider_name": "auto-routing",
      "accuracy": 0.58,
      "accuracy_std_dev": 0.026666999999999996,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.20638665,
      "avg_duration_per_task_ms": 179102.02,
      "avg_output_tokens_per_task": 12109.41,
      "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": "deepseek/deepseek-r1-0528",
      "model_slug": "deepseek/deepseek-r1-0528",
      "model_name": "DeepSeek: R1 0528",
      "provider_name": "Novita",
      "accuracy": 0.5789978786127168,
      "accuracy_std_dev": 0.05180400838836742,
      "run_count": 7,
      "total_questions": 346,
      "avg_cost_per_task_usd": 0.18764681502890174,
      "avg_duration_per_task_ms": 2044873.4393063583,
      "avg_output_tokens_per_task": 41085.294797687864,
      "last_run_timestamp": "2026-10-17T19:42:10.129Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-next-2025-02-03",
      "model_slug": "qwen/qwen3-coder-next",
      "model_name": "Qwen: Qwen3 Coder Next",
      "provider_name": "Ionstream",
      "accuracy": 0.57483,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 49,
      "avg_cost_per_task_usd": 0.017800938775510203,
      "avg_duration_per_task_ms": 142107.1224489796,
      "avg_output_tokens_per_task": 3993.1632653061224,
      "last_run_timestamp": "2026-08-08T16:03:49.463Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-next-2025-02-03",
      "model_slug": "qwen/qwen3-coder-next",
      "model_name": "Qwen: Qwen3 Coder Next",
      "provider_name": "Alibaba",
      "accuracy": 0.5733883127572016,
      "accuracy_std_dev": 0.039571612153158485,
      "run_count": 5,
      "total_questions": 243,
      "avg_cost_per_task_usd": 0.08109196296296296,
      "avg_duration_per_task_ms": 99593.68724279835,
      "avg_output_tokens_per_task": 6510.271604938272,
      "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/gemini-2.5-flash",
      "model_slug": "google/gemini-2.5-flash",
      "model_name": "Google: Gemini 2.5 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.5666665000000001,
      "accuracy_std_dev": 0.006666499999999964,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.028862480000000003,
      "avg_duration_per_task_ms": 75034.33,
      "avg_output_tokens_per_task": 7701.06,
      "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": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "Crusoe",
      "accuracy": 0.5666665,
      "accuracy_std_dev": 0.1466665,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.02785182,
      "avg_duration_per_task_ms": 139069.32,
      "avg_output_tokens_per_task": 11612.72,
      "last_run_timestamp": "2026-10-24T12:08:00.201Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Cerebras",
      "accuracy": 0.5648772953020134,
      "accuracy_std_dev": 0.0424873867714486,
      "run_count": 3,
      "total_questions": 149,
      "avg_cost_per_task_usd": 0.1486826308724832,
      "avg_duration_per_task_ms": 18002.88590604027,
      "avg_output_tokens_per_task": 2546.3288590604025,
      "last_run_timestamp": "2026-08-29T17:07:32.456Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "AkashML",
      "accuracy": 0.5633334999999999,
      "accuracy_std_dev": 0.09666650000000002,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.07923399999999998,
      "avg_duration_per_task_ms": 1068653.19,
      "avg_output_tokens_per_task": 62040.81,
      "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": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "AtlasCloud",
      "accuracy": 0.560133587973274,
      "accuracy_std_dev": 0.2552451920126611,
      "run_count": 9,
      "total_questions": 449,
      "avg_cost_per_task_usd": 0.015204305122494433,
      "avg_duration_per_task_ms": 265953.0779510022,
      "avg_output_tokens_per_task": 7739.80846325167,
      "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/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": 0.557500125,
      "accuracy_std_dev": 0.07859452244501124,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.032340572500000005,
      "avg_duration_per_task_ms": 393952.61,
      "avg_output_tokens_per_task": 16997.9425,
      "last_run_timestamp": "2026-10-24T19:54:36.551Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.55,
      "accuracy_std_dev": 0.036667000000000005,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.2373302,
      "avg_duration_per_task_ms": 164429.2,
      "avg_output_tokens_per_task": 12508.34,
      "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": "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": 0.546667,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.03543453,
      "avg_duration_per_task_ms": 61133.82,
      "avg_output_tokens_per_task": 6656.99,
      "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-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Novita",
      "accuracy": 0.5458451776504298,
      "accuracy_std_dev": 0.0501062696215529,
      "run_count": 7,
      "total_questions": 349,
      "avg_cost_per_task_usd": 0.010405719197707736,
      "avg_duration_per_task_ms": 192494.59598853867,
      "avg_output_tokens_per_task": 16106.47564469914,
      "last_run_timestamp": "2026-10-24T19:30:50.871Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-2.5-flash-lite",
      "model_slug": "google/gemini-2.5-flash-lite",
      "model_name": "Google: Gemini 2.5 Flash Lite",
      "provider_name": "Google AI Studio",
      "accuracy": 0.5370371414141414,
      "accuracy_std_dev": 0.020577999999999985,
      "run_count": 2,
      "total_questions": 99,
      "avg_cost_per_task_usd": 0.02047550505050505,
      "avg_duration_per_task_ms": 144559.1616161616,
      "avg_output_tokens_per_task": 38541.121212121216,
      "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": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "auto-routing",
      "accuracy": 0.535956005076142,
      "accuracy_std_dev": 0.0166977509787995,
      "run_count": 8,
      "total_questions": 394,
      "avg_cost_per_task_usd": 0.018755781725888324,
      "avg_duration_per_task_ms": 1279767.0253807106,
      "avg_output_tokens_per_task": 84685.69796954315,
      "last_run_timestamp": "2026-10-24T19:37:33.902Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.534014,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 49,
      "avg_cost_per_task_usd": 0.01161230612244898,
      "avg_duration_per_task_ms": 94666.5306122449,
      "avg_output_tokens_per_task": 16582.326530612245,
      "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": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Darkbloom",
      "accuracy": 0.5336213227665707,
      "accuracy_std_dev": 0.029334415082017253,
      "run_count": 7,
      "total_questions": 347,
      "avg_cost_per_task_usd": 0.002658521613832853,
      "avg_duration_per_task_ms": 386644.1469740634,
      "avg_output_tokens_per_task": 6857.3314121037465,
      "last_run_timestamp": "2026-10-30T01:32:41.347Z"
    },
    {
      "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": 0.5283335,
      "accuracy_std_dev": 0.01993049032261877,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.066619805,
      "avg_duration_per_task_ms": 287868.2675,
      "avg_output_tokens_per_task": 4057.665,
      "last_run_timestamp": "2026-10-24T18:33:38.931Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "auto-routing",
      "accuracy": 0.5258620229885057,
      "accuracy_std_dev": 0.09372402255384406,
      "run_count": 7,
      "total_questions": 348,
      "avg_cost_per_task_usd": 0.049525813218390814,
      "avg_duration_per_task_ms": 424260.5517241379,
      "avg_output_tokens_per_task": 7604.66091954023,
      "last_run_timestamp": "2026-10-24T18:46:40.201Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4.1-2025-04-14",
      "model_slug": "openai/gpt-4.1",
      "model_name": "OpenAI: GPT-4.1",
      "provider_name": "Azure",
      "accuracy": 0.522831205479452,
      "accuracy_std_dev": 0.027543964303555765,
      "run_count": 3,
      "total_questions": 146,
      "avg_cost_per_task_usd": 0.11947870547945207,
      "avg_duration_per_task_ms": 69921.19178082192,
      "avg_output_tokens_per_task": 2362.404109589041,
      "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": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "DeepInfra",
      "accuracy": 0.519759618556701,
      "accuracy_std_dev": 0.11063848893355334,
      "run_count": 4,
      "total_questions": 194,
      "avg_cost_per_task_usd": 0.05320747422680413,
      "avg_duration_per_task_ms": 383087.36597938143,
      "avg_output_tokens_per_task": 23944.39175257732,
      "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": "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": "deepseek/deepseek-v3.1-terminus",
      "model_slug": "deepseek/deepseek-v3.1-terminus",
      "model_name": "DeepSeek: DeepSeek V3.1 Terminus",
      "provider_name": "DeepInfra",
      "accuracy": 0.513889,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 48,
      "avg_cost_per_task_usd": 0.025741249999999997,
      "avg_duration_per_task_ms": 587337.8541666666,
      "avg_output_tokens_per_task": 10115.354166666666,
      "last_run_timestamp": "2026-08-08T16:03:43.735Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4.1-2025-04-14",
      "model_slug": "openai/gpt-4.1",
      "model_name": "OpenAI: GPT-4.1",
      "provider_name": "OpenAI",
      "accuracy": 0.5133333333333333,
      "accuracy_std_dev": 0.028803034820811636,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.12443134666666665,
      "avg_duration_per_task_ms": 52357.68,
      "avg_output_tokens_per_task": 2755.9866666666667,
      "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": "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Together",
      "accuracy": 0.512916625,
      "accuracy_std_dev": 0.03301683440662317,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.031192877499999997,
      "avg_duration_per_task_ms": 264931.19,
      "avg_output_tokens_per_task": 21003.91,
      "last_run_timestamp": "2026-10-24T19:23:13.562Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "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": "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": 0.4990475714285714,
      "accuracy_std_dev": 0.026047140242574162,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.025615628571428573,
      "avg_duration_per_task_ms": 596502.9742857143,
      "avg_output_tokens_per_task": 52103.73714285714,
      "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": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "Mancer 2",
      "accuracy": 0.493333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.14235776,
      "avg_duration_per_task_ms": 443941.74,
      "avg_output_tokens_per_task": 8833.66,
      "last_run_timestamp": "2026-08-06T00:00:02.908Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3-nano-30b-a3b",
      "model_slug": "nvidia/nemotron-3-nano-30b-a3b",
      "model_name": "NVIDIA: Nemotron 3 Nano 30B A3B",
      "provider_name": "Nebius",
      "accuracy": 0.48850570114942526,
      "accuracy_std_dev": 0.017439968302957548,
      "run_count": 7,
      "total_questions": 348,
      "avg_cost_per_task_usd": 0.031982068965517245,
      "avg_duration_per_task_ms": 642915.867816092,
      "avg_output_tokens_per_task": 59401.63505747126,
      "last_run_timestamp": "2026-10-23T09:26:25.528Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3-nano-30b-a3b",
      "model_slug": "nvidia/nemotron-3-nano-30b-a3b",
      "model_name": "NVIDIA: Nemotron 3 Nano 30B A3B",
      "provider_name": "Crusoe",
      "accuracy": 0.4857621557788945,
      "accuracy_std_dev": 0.02560762584874622,
      "run_count": 8,
      "total_questions": 398,
      "avg_cost_per_task_usd": 0.028534012562814074,
      "avg_duration_per_task_ms": 352717.891959799,
      "avg_output_tokens_per_task": 60736.11306532663,
      "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": "openai/gpt-5-nano-2025-08-07",
      "model_slug": "openai/gpt-5-nano",
      "model_name": "OpenAI: GPT-5 Nano",
      "provider_name": "auto-routing",
      "accuracy": 0.48333349999999997,
      "accuracy_std_dev": 0.03666650000000002,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.04059527,
      "avg_duration_per_task_ms": 669013.45,
      "avg_output_tokens_per_task": 93900.79,
      "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": "deepseek/deepseek-v3.1-terminus",
      "model_slug": "deepseek/deepseek-v3.1-terminus",
      "model_name": "DeepSeek: DeepSeek V3.1 Terminus",
      "provider_name": "SiliconFlow",
      "accuracy": 0.48000024999999996,
      "accuracy_std_dev": 0.01414201777284628,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.0673682,
      "avg_duration_per_task_ms": 287537.155,
      "avg_output_tokens_per_task": 4110.14,
      "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-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "auto-routing",
      "accuracy": 0.48,
      "accuracy_std_dev": 0.0312693389057076,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.050673248000000004,
      "avg_duration_per_task_ms": 91008.444,
      "avg_output_tokens_per_task": 2749.592,
      "last_run_timestamp": "2026-10-03T16:07:50.332Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-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": 0.47666650000000005,
      "accuracy_std_dev": 0.0166665,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.12559784000000002,
      "avg_duration_per_task_ms": 52928.96,
      "avg_output_tokens_per_task": 2668.55,
      "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": "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": 0.4754382631578947,
      "accuracy_std_dev": 0.054999999999999966,
      "run_count": 2,
      "total_questions": 95,
      "avg_cost_per_task_usd": 0.01757382105263158,
      "avg_duration_per_task_ms": 167109.45263157896,
      "avg_output_tokens_per_task": 28512.663157894738,
      "last_run_timestamp": "2026-10-01T19:06:10.500Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-32b-04-28",
      "model_slug": "qwen/qwen3-32b",
      "model_name": "Qwen: Qwen3 32B",
      "provider_name": "auto-routing",
      "accuracy": 0.475,
      "accuracy_std_dev": 0.00986024566123988,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.021282230000000003,
      "avg_duration_per_task_ms": 677068.455,
      "avg_output_tokens_per_task": 19681.38,
      "last_run_timestamp": "2026-08-29T16:16:50.456Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-32b-04-28",
      "model_slug": "qwen/qwen3-32b",
      "model_name": "Qwen: Qwen3 32B",
      "provider_name": "Nebius",
      "accuracy": 0.47333349999999996,
      "accuracy_std_dev": 0.008164965814380372,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.024006935,
      "avg_duration_per_task_ms": 869541.065,
      "avg_output_tokens_per_task": 22816.51,
      "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-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": "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": 0.472,
      "accuracy_std_dev": 0.023999999999999994,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.034465088000000005,
      "avg_duration_per_task_ms": 43484.756,
      "avg_output_tokens_per_task": 3636.224,
      "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": "auto-routing",
      "accuracy": 0.4706666,
      "accuracy_std_dev": 0.020483090046182005,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.023913492,
      "avg_duration_per_task_ms": 139319.672,
      "avg_output_tokens_per_task": 3726.148,
      "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/qwen3-30b-a3b-04-28",
      "model_slug": "qwen/qwen3-30b-a3b",
      "model_name": "Qwen: Qwen3 30B A3B",
      "provider_name": "DeepInfra",
      "accuracy": 0.47,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.03551402,
      "avg_duration_per_task_ms": 437816.58,
      "avg_output_tokens_per_task": 20922.04,
      "last_run_timestamp": "2026-10-08T23:09:09.727Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Parasail",
      "accuracy": 0.46820784971098267,
      "accuracy_std_dev": 0.04097451831844482,
      "run_count": 7,
      "total_questions": 346,
      "avg_cost_per_task_usd": 0.027368598265895955,
      "avg_duration_per_task_ms": 2298741.832369942,
      "avg_output_tokens_per_task": 136175.1647398844,
      "last_run_timestamp": "2026-10-24T19:05:04.624Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-next-80b-a3b-instruct-2509",
      "model_slug": "qwen/qwen3-next-80b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Next 80B A3B Instruct",
      "provider_name": "Parasail",
      "accuracy": 0.4680002,
      "accuracy_std_dev": 0.03798255037987838,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.019017924000000002,
      "avg_duration_per_task_ms": 96994.328,
      "avg_output_tokens_per_task": 3628.408,
      "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": "openai/gpt-5-nano-2025-08-07",
      "model_slug": "openai/gpt-5-nano",
      "model_name": "OpenAI: GPT-5 Nano",
      "provider_name": "OpenAI",
      "accuracy": 0.4644446666666666,
      "accuracy_std_dev": 0.029979285029648195,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.04325750666666667,
      "avg_duration_per_task_ms": 738509.7133333334,
      "avg_output_tokens_per_task": 100007.46,
      "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-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": 0.4640002,
      "accuracy_std_dev": 0.04144079793343752,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.024943207999999998,
      "avg_duration_per_task_ms": 59536.72,
      "avg_output_tokens_per_task": 3736.6,
      "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-nano-2025-08-07",
      "model_slug": "openai/gpt-5-nano",
      "model_name": "OpenAI: GPT-5 Nano",
      "provider_name": "Azure",
      "accuracy": 0.46166674999999996,
      "accuracy_std_dev": 0.04010416645047619,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.04448468,
      "avg_duration_per_task_ms": 738816.635,
      "avg_output_tokens_per_task": 95390.315,
      "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": "mistralai/mistral-small-2603",
      "model_slug": "mistralai/mistral-small-2603",
      "model_name": "Mistral: Mistral Small 4",
      "provider_name": "auto-routing",
      "accuracy": 0.46,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.011152899999999999,
      "avg_duration_per_task_ms": 87357.46,
      "avg_output_tokens_per_task": 9087.9,
      "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": "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": "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": 0.46,
      "accuracy_std_dev": 0.006667000000000006,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.03575541,
      "avg_duration_per_task_ms": 402821.3,
      "avg_output_tokens_per_task": 3754.72,
      "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": 0.4566668333333333,
      "accuracy_std_dev": 0.03590122251129556,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.021435916666666662,
      "avg_duration_per_task_ms": 222009.41666666666,
      "avg_output_tokens_per_task": 3511.05,
      "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": "qwen/qwen3-32b-04-28",
      "model_slug": "qwen/qwen3-32b",
      "model_name": "Qwen: Qwen3 32B",
      "provider_name": "Tenstorrent",
      "accuracy": 0.4541387986577181,
      "accuracy_std_dev": 0.02966298114934956,
      "run_count": 3,
      "total_questions": 149,
      "avg_cost_per_task_usd": 0.023203442953020136,
      "avg_duration_per_task_ms": 2013894.1275167784,
      "avg_output_tokens_per_task": 14181.58389261745,
      "last_run_timestamp": "2026-10-30T21:07:22.076Z"
    },
    {
      "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": 0.45333349999999994,
      "accuracy_std_dev": 0.00666650000000002,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.0063000800000000004,
      "avg_duration_per_task_ms": 38540.88,
      "avg_output_tokens_per_task": 1215.72,
      "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/qwen3-32b-04-28",
      "model_slug": "qwen/qwen3-32b",
      "model_name": "Qwen: Qwen3 32B",
      "provider_name": "DeepInfra",
      "accuracy": 0.4524997500000001,
      "accuracy_std_dev": 0.02520749723172652,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.02146402,
      "avg_duration_per_task_ms": 650786.745,
      "avg_output_tokens_per_task": 19908.01,
      "last_run_timestamp": "2026-10-09T00:21:42.483Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Parasail",
      "accuracy": 0.4521073362068965,
      "accuracy_std_dev": 0.045992926219636195,
      "run_count": 7,
      "total_questions": 348,
      "avg_cost_per_task_usd": 0.015361428160919543,
      "avg_duration_per_task_ms": 128943.27873563218,
      "avg_output_tokens_per_task": 3310.485632183908,
      "last_run_timestamp": "2026-10-24T19:23:50.567Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.45121959349593493,
      "accuracy_std_dev": 0.052473835908574476,
      "run_count": 5,
      "total_questions": 246,
      "avg_cost_per_task_usd": 0.05198310569105691,
      "avg_duration_per_task_ms": 74014.12195121951,
      "avg_output_tokens_per_task": 3228.670731707317,
      "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": "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": 0.45111100000000004,
      "accuracy_std_dev": 0.01571332689152746,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.023389446666666668,
      "avg_duration_per_task_ms": 54689.44666666666,
      "avg_output_tokens_per_task": 2398.8933333333334,
      "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/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Venice",
      "accuracy": 0.45095228571428575,
      "accuracy_std_dev": 0.024605712517417503,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.04007145714285715,
      "avg_duration_per_task_ms": 282805.1142857143,
      "avg_output_tokens_per_task": 3547.6828571428573,
      "last_run_timestamp": "2026-10-24T19:41:30.973Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Nebius",
      "accuracy": 0.45,
      "accuracy_std_dev": 0.041518865477893636,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.05302523142857142,
      "avg_duration_per_task_ms": 144621.62857142856,
      "avg_output_tokens_per_task": 3704.842857142857,
      "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-coder-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Coder 30B A3B Instruct",
      "provider_name": "Alibaba",
      "accuracy": 0.45,
      "accuracy_std_dev": 0.0033330000000000026,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.08659784,
      "avg_duration_per_task_ms": 80537.63,
      "avg_output_tokens_per_task": 4189.08,
      "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": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Groq",
      "accuracy": 0.4492010616438356,
      "accuracy_std_dev": 0.023681154664627318,
      "run_count": 6,
      "total_questions": 292,
      "avg_cost_per_task_usd": 0.05459431849315068,
      "avg_duration_per_task_ms": 157164.11643835617,
      "avg_output_tokens_per_task": 75683.87671232877,
      "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-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "Novita",
      "accuracy": 0.4466668,
      "accuracy_std_dev": 0.08033274135344816,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.20203069199999998,
      "avg_duration_per_task_ms": 115064.372,
      "avg_output_tokens_per_task": 3550.864,
      "last_run_timestamp": "2026-10-03T16:19:14.095Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4.1-mini-2025-04-14",
      "model_slug": "openai/gpt-4.1-mini",
      "model_name": "OpenAI: GPT-4.1 Mini",
      "provider_name": "auto-routing",
      "accuracy": 0.44333300000000003,
      "accuracy_std_dev": 0.009999999999999981,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.023192820000000003,
      "avg_duration_per_task_ms": 58714.73,
      "avg_output_tokens_per_task": 2330.08,
      "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": "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": "qwen/qwen3-14b-04-28",
      "model_slug": "qwen/qwen3-14b",
      "model_name": "Qwen: Qwen3 14B",
      "provider_name": "auto-routing",
      "accuracy": 0.44,
      "accuracy_std_dev": 0.0066669999999999785,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.03130282,
      "avg_duration_per_task_ms": 680376.23,
      "avg_output_tokens_per_task": 16688.33,
      "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": "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": "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": 0.4377778333333333,
      "accuracy_std_dev": 0.040490818956963014,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.053731056666666666,
      "avg_duration_per_task_ms": 129281.94,
      "avg_output_tokens_per_task": 3456.7733333333335,
      "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": "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": 0.43631422764227645,
      "accuracy_std_dev": 0.03840736243170051,
      "run_count": 5,
      "total_questions": 246,
      "avg_cost_per_task_usd": 0.057368723577235776,
      "avg_duration_per_task_ms": 251181.19512195123,
      "avg_output_tokens_per_task": 2298.849593495935,
      "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": "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": 0.4353741734693877,
      "accuracy_std_dev": 0.03760835138885378,
      "run_count": 4,
      "total_questions": 196,
      "avg_cost_per_task_usd": 0.04635305102040816,
      "avg_duration_per_task_ms": 63193.6581632653,
      "avg_output_tokens_per_task": 2275.091836734694,
      "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": "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": "openai/gpt-4o",
      "model_slug": "openai/gpt-4o",
      "model_name": "OpenAI: GPT-4o",
      "provider_name": "OpenAI",
      "accuracy": 0.4322221666666667,
      "accuracy_std_dev": 0.01651776073823432,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.21270556333333332,
      "avg_duration_per_task_ms": 46305.06,
      "avg_output_tokens_per_task": 2266.36,
      "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": "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": 0.431111,
      "accuracy_std_dev": 0.00314238253559302,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.24174756666666666,
      "avg_duration_per_task_ms": 46576.24,
      "avg_output_tokens_per_task": 2473.9733333333334,
      "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": "deepseek/deepseek-chat-v3",
      "model_slug": "deepseek/deepseek-chat",
      "model_name": "DeepSeek: DeepSeek V3",
      "provider_name": "DeepInfra",
      "accuracy": 0.4301802162162162,
      "accuracy_std_dev": 0.02412263108922131,
      "run_count": 3,
      "total_questions": 148,
      "avg_cost_per_task_usd": 0.06743897297297298,
      "avg_duration_per_task_ms": 360892.5135135135,
      "avg_output_tokens_per_task": 3318.9054054054054,
      "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": "qwen/qwen3-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "DeepInfra",
      "accuracy": 0.42916675,
      "accuracy_std_dev": 0.025967332577442378,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.027418319999999996,
      "avg_duration_per_task_ms": 84447.825,
      "avg_output_tokens_per_task": 2784.11,
      "last_run_timestamp": "2026-10-03T16:12:20.034Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "Venice",
      "accuracy": 0.4283335,
      "accuracy_std_dev": 0.02598098663927142,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.016813855,
      "avg_duration_per_task_ms": 118682.965,
      "avg_output_tokens_per_task": 2808.965,
      "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": "qwen/qwen3-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "Alibaba",
      "accuracy": 0.428,
      "accuracy_std_dev": 0.052391746594287164,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.49939273599999995,
      "avg_duration_per_task_ms": 111261.628,
      "avg_output_tokens_per_task": 3387.876,
      "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-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": 0.42799980000000004,
      "accuracy_std_dev": 0.012220332784339386,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.039161852,
      "avg_duration_per_task_ms": 146989.988,
      "avg_output_tokens_per_task": 3492.532,
      "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": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Parasail",
      "accuracy": 0.4276191428571428,
      "accuracy_std_dev": 0.03190820308513861,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.007563445714285715,
      "avg_duration_per_task_ms": 175165.4,
      "avg_output_tokens_per_task": 4466.028571428571,
      "last_run_timestamp": "2026-10-24T19:16:50.878Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "AkashML",
      "accuracy": 0.42666699999999996,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.0178204,
      "avg_duration_per_task_ms": 1378502.34,
      "avg_output_tokens_per_task": 126873.58,
      "last_run_timestamp": "2026-10-30T02:24:16.798Z"
    },
    {
      "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": 0.42666699999999996,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.02972828,
      "avg_duration_per_task_ms": 273042.46,
      "avg_output_tokens_per_task": 3546.89,
      "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": "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": 0.42666675,
      "accuracy_std_dev": 0.014142017772846282,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.050337635,
      "avg_duration_per_task_ms": 158009.49,
      "avg_output_tokens_per_task": 2226.425,
      "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": "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": 0.42666650000000006,
      "accuracy_std_dev": 0.00666650000000002,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.21223042999999997,
      "avg_duration_per_task_ms": 51685.1,
      "avg_output_tokens_per_task": 2189.48,
      "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": "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": 0.42592599999999997,
      "accuracy_std_dev": 0.03253776997070738,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.024220277777777778,
      "avg_duration_per_task_ms": 407107.03555555554,
      "avg_output_tokens_per_task": 3588.162222222222,
      "last_run_timestamp": "2026-10-24T19:25:31.387Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-chat-v3-0324",
      "model_slug": "deepseek/deepseek-chat-v3-0324",
      "model_name": "DeepSeek: DeepSeek V3 0324",
      "provider_name": "DeepInfra",
      "accuracy": 0.4256754864864865,
      "accuracy_std_dev": 0.024259060786069653,
      "run_count": 3,
      "total_questions": 148,
      "avg_cost_per_task_usd": 0.03375386486486487,
      "avg_duration_per_task_ms": 525936.777027027,
      "avg_output_tokens_per_task": 5145.756756756757,
      "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/qwen3-14b-04-28",
      "model_slug": "qwen/qwen3-14b",
      "model_name": "Qwen: Qwen3 14B",
      "provider_name": "DeepInfra",
      "accuracy": 0.4255553333333334,
      "accuracy_std_dev": 0.034462310778910675,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.029985173333333334,
      "avg_duration_per_task_ms": 822675.42,
      "avg_output_tokens_per_task": 18074.12,
      "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": "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": 0.42400019999999994,
      "accuracy_std_dev": 0.014514259655938366,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.024904308,
      "avg_duration_per_task_ms": 223225.116,
      "avg_output_tokens_per_task": 3537.724,
      "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/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "Io Net",
      "accuracy": 0.4236811784511785,
      "accuracy_std_dev": 0.029725846878880795,
      "run_count": 6,
      "total_questions": 297,
      "avg_cost_per_task_usd": 0.05290695286195286,
      "avg_duration_per_task_ms": 528585.1717171717,
      "avg_output_tokens_per_task": 22153.03367003367,
      "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": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "DeepInfra",
      "accuracy": 0.4204932602040816,
      "accuracy_std_dev": 0.025044575822109165,
      "run_count": 8,
      "total_questions": 392,
      "avg_cost_per_task_usd": 0.03310804591836735,
      "avg_duration_per_task_ms": 672143.8137755102,
      "avg_output_tokens_per_task": 6731.755102040816,
      "last_run_timestamp": "2026-10-24T19:18:23.891Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "mistralai/mistral-small-2603",
      "model_slug": "mistralai/mistral-small-2603",
      "model_name": "Mistral: Mistral Small 4",
      "provider_name": "Mistral",
      "accuracy": 0.41888899999999996,
      "accuracy_std_dev": 0.014614444065147779,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.01463076,
      "avg_duration_per_task_ms": 76549.69333333333,
      "avg_output_tokens_per_task": 8814.213333333333,
      "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-vl-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 VL 30B A3B Instruct",
      "provider_name": "Novita",
      "accuracy": 0.4175171377551021,
      "accuracy_std_dev": 0.02667947172423725,
      "run_count": 8,
      "total_questions": 392,
      "avg_cost_per_task_usd": 0.057072859693877556,
      "avg_duration_per_task_ms": 211166.0357142857,
      "avg_output_tokens_per_task": 4534.619897959184,
      "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": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Chutes",
      "accuracy": 0.41751264974619284,
      "accuracy_std_dev": 0.027965463072500252,
      "run_count": 8,
      "total_questions": 394,
      "avg_cost_per_task_usd": 0.004796167512690355,
      "avg_duration_per_task_ms": 327290.3121827411,
      "avg_output_tokens_per_task": 4759.779187817259,
      "last_run_timestamp": "2026-10-24T19:54:38.220Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4o",
      "model_slug": "openai/gpt-4o",
      "model_name": "OpenAI: GPT-4o",
      "provider_name": "auto-routing",
      "accuracy": 0.416667,
      "accuracy_std_dev": 0.03,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.19647352999999998,
      "avg_duration_per_task_ms": 53166.02,
      "avg_output_tokens_per_task": 2086.91,
      "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": "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": 0.411667,
      "accuracy_std_dev": 0.049749371855331,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.055448005,
      "avg_duration_per_task_ms": 142634.75,
      "avg_output_tokens_per_task": 3380.275,
      "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": "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": 0.40836190721649485,
      "accuracy_std_dev": 0.024997748289707123,
      "run_count": 6,
      "total_questions": 291,
      "avg_cost_per_task_usd": 0.011245838487972509,
      "avg_duration_per_task_ms": 117422.78006872852,
      "avg_output_tokens_per_task": 3161.364261168385,
      "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": "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": "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": 0.4066664000000001,
      "accuracy_std_dev": 0.025647109432448717,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.07750359200000001,
      "avg_duration_per_task_ms": 118177.072,
      "avg_output_tokens_per_task": 3117.652,
      "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": "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": 0.40533320000000006,
      "accuracy_std_dev": 0.01808612973966516,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.06744982,
      "avg_duration_per_task_ms": 383181.876,
      "avg_output_tokens_per_task": 10145.868,
      "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": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "Novita",
      "accuracy": 0.40389273722627733,
      "accuracy_std_dev": 0.01580206672840261,
      "run_count": 3,
      "total_questions": 137,
      "avg_cost_per_task_usd": 0.008183423357664232,
      "avg_duration_per_task_ms": 89294.24087591241,
      "avg_output_tokens_per_task": 1500.5474452554745,
      "last_run_timestamp": "2026-10-03T16:15:51.302Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-chat-v3",
      "model_slug": "deepseek/deepseek-chat",
      "model_name": "DeepSeek: DeepSeek V3",
      "provider_name": "StreamLake",
      "accuracy": 0.40380291275167784,
      "accuracy_std_dev": 0.008840018715415086,
      "run_count": 3,
      "total_questions": 149,
      "avg_cost_per_task_usd": 0.07104265771812081,
      "avg_duration_per_task_ms": 235544.13422818793,
      "avg_output_tokens_per_task": 8668.979865771813,
      "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": "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": 0.40333325000000003,
      "accuracy_std_dev": 0.01914841159698371,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.023807599999999998,
      "avg_duration_per_task_ms": 77145.375,
      "avg_output_tokens_per_task": 2775.135,
      "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/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "SiliconFlow",
      "accuracy": 0.4026668,
      "accuracy_std_dev": 0.027519660851107883,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.027686715999999997,
      "avg_duration_per_task_ms": 276072.088,
      "avg_output_tokens_per_task": 4225.576,
      "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": "Venice",
      "accuracy": 0.4,
      "accuracy_std_dev": 0.034769053930183376,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.076876536,
      "avg_duration_per_task_ms": 340404.576,
      "avg_output_tokens_per_task": 15764.94,
      "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-vl-235b-a22b-instruct",
      "model_slug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_name": "Qwen: Qwen3 VL 235B A22B Instruct",
      "provider_name": "Parasail",
      "accuracy": 0.39999980000000007,
      "accuracy_std_dev": 0.008432845839928543,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.036944032,
      "avg_duration_per_task_ms": 191861.204,
      "avg_output_tokens_per_task": 3680.784,
      "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": "deepseek/deepseek-chat-v3",
      "model_slug": "deepseek/deepseek-chat",
      "model_name": "DeepSeek: DeepSeek V3",
      "provider_name": "auto-routing",
      "accuracy": 0.39833349999999995,
      "accuracy_std_dev": 0.018333500000000003,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.06903712000000001,
      "avg_duration_per_task_ms": 315593.59,
      "avg_output_tokens_per_task": 7308.28,
      "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-vl-235b-a22b-instruct",
      "model_slug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_name": "Qwen: Qwen3 VL 235B A22B Instruct",
      "provider_name": "Alibaba",
      "accuracy": 0.39466680000000004,
      "accuracy_std_dev": 0.019505181079907976,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.0621838,
      "avg_duration_per_task_ms": 95587.668,
      "avg_output_tokens_per_task": 2912.132,
      "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": "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": 0.39166674999999995,
      "accuracy_std_dev": 0.00866020592639114,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.015426675,
      "avg_duration_per_task_ms": 99199.855,
      "avg_output_tokens_per_task": 615.135,
      "last_run_timestamp": "2026-08-29T17:06:47.617Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_slug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_name": "Qwen: Qwen3 VL 235B A22B Instruct",
      "provider_name": "DeepInfra",
      "accuracy": 0.39132881081081083,
      "accuracy_std_dev": 0.025676864524422683,
      "run_count": 6,
      "total_questions": 296,
      "avg_cost_per_task_usd": 0.058435516891891894,
      "avg_duration_per_task_ms": 376057.5371621622,
      "avg_output_tokens_per_task": 3843.489864864865,
      "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "CoreWeave",
      "accuracy": 0.3911292096774194,
      "accuracy_std_dev": 0.3156298317748815,
      "run_count": 5,
      "total_questions": 248,
      "avg_cost_per_task_usd": 0.005442725806451613,
      "avg_duration_per_task_ms": 343542.48387096776,
      "avg_output_tokens_per_task": 10219.729838709678,
      "last_run_timestamp": "2026-10-24T19:32:27.020Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "auto-routing",
      "accuracy": 0.387500125,
      "accuracy_std_dev": 0.12580578865203054,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0213678575,
      "avg_duration_per_task_ms": 206487.0225,
      "avg_output_tokens_per_task": 3317.835,
      "last_run_timestamp": "2026-10-24T19:36:27.834Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "Ambient",
      "accuracy": 0.386667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.014510460000000001,
      "avg_duration_per_task_ms": 64332.54,
      "avg_output_tokens_per_task": 4106.96,
      "last_run_timestamp": "2026-10-09T15:37:38.473Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 VL 30B A3B Instruct",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.386667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.0592903,
      "avg_duration_per_task_ms": 203536.16,
      "avg_output_tokens_per_task": 5170.96,
      "last_run_timestamp": "2026-10-04T03:00:16.402Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Coder 30B A3B Instruct",
      "provider_name": "SiliconFlow",
      "accuracy": 0.3866665,
      "accuracy_std_dev": 0.006666499999999992,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.020927889999999998,
      "avg_duration_per_task_ms": 183788.07,
      "avg_output_tokens_per_task": 3842.25,
      "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/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "CoreWeave",
      "accuracy": 0.38333325,
      "accuracy_std_dev": 0.007453597197830037,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.028690270000000004,
      "avg_duration_per_task_ms": 83258.84,
      "avg_output_tokens_per_task": 4293.12,
      "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": "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": 0.3826666,
      "accuracy_std_dev": 0.01083190350030869,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.026093336,
      "avg_duration_per_task_ms": 200624.988,
      "avg_output_tokens_per_task": 4302.312,
      "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/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Alibaba",
      "accuracy": 0.3773334,
      "accuracy_std_dev": 0.04552913987590804,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.05450766399999999,
      "avg_duration_per_task_ms": 92940.624,
      "avg_output_tokens_per_task": 3847.676,
      "last_run_timestamp": "2026-10-03T16:05:51.296Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "Alibaba",
      "accuracy": 0.37599980000000005,
      "accuracy_std_dev": 0.032276543943861154,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.055068684,
      "avg_duration_per_task_ms": 70747.116,
      "avg_output_tokens_per_task": 4411.06,
      "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": "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": 0.3738005794743429,
      "accuracy_std_dev": 0.022831389807948622,
      "run_count": 16,
      "total_questions": 799,
      "avg_cost_per_task_usd": 0.013108423028785983,
      "avg_duration_per_task_ms": 39230.8648310388,
      "avg_output_tokens_per_task": 749.793491864831,
      "last_run_timestamp": "2026-10-30T00:47:05.066Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "Nebius",
      "accuracy": 0.3688886666666667,
      "accuracy_std_dev": 0.01791611454963257,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.03368235333333333,
      "avg_duration_per_task_ms": 150468.61333333334,
      "avg_output_tokens_per_task": 4460.053333333333,
      "last_run_timestamp": "2026-10-23T08:14:36.871Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "GMICloud",
      "accuracy": 0.36833325,
      "accuracy_std_dev": 0.17853565236861657,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.015193219999999999,
      "avg_duration_per_task_ms": 127682.09,
      "avg_output_tokens_per_task": 2917.505,
      "last_run_timestamp": "2026-10-03T16:12:29.879Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.3681779823677581,
      "accuracy_std_dev": 0.14087050789486985,
      "run_count": 16,
      "total_questions": 794,
      "avg_cost_per_task_usd": 0.038621142317380355,
      "avg_duration_per_task_ms": 585118.7216624685,
      "avg_output_tokens_per_task": 145279.47858942064,
      "last_run_timestamp": "2026-10-24T19:41:04.164Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "CoreWeave",
      "accuracy": 0.364498756097561,
      "accuracy_std_dev": 0.08717599304625098,
      "run_count": 5,
      "total_questions": 246,
      "avg_cost_per_task_usd": 0.42022371138211384,
      "avg_duration_per_task_ms": 436503.325203252,
      "avg_output_tokens_per_task": 17041.613821138213,
      "last_run_timestamp": "2026-10-10T18:31:45.008Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.36,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.041000339999999996,
      "avg_duration_per_task_ms": 322345.3,
      "avg_output_tokens_per_task": 4212.74,
      "last_run_timestamp": "2026-10-04T03:00:10.237Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 VL 30B A3B Instruct",
      "provider_name": "DeepInfra",
      "accuracy": 0.35714285714285715,
      "accuracy_std_dev": 0.029979379745928993,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.07267795758928572,
      "avg_duration_per_task_ms": 400336.21875,
      "avg_output_tokens_per_task": 8020.354910714285,
      "last_run_timestamp": "2026-10-08T23:34:52.306Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "CoreWeave",
      "accuracy": 0.35333325000000004,
      "accuracy_std_dev": 0.03448008281874479,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.08867438999999999,
      "avg_duration_per_task_ms": 27089.57,
      "avg_output_tokens_per_task": 1029.7925,
      "last_run_timestamp": "2026-10-24T18:19:37.955Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 VL 30B A3B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.35175891959798994,
      "accuracy_std_dev": 0.022258587996757,
      "run_count": 8,
      "total_questions": 398,
      "avg_cost_per_task_usd": 0.0788611256281407,
      "avg_duration_per_task_ms": 211842.61055276383,
      "avg_output_tokens_per_task": 7349.442211055276,
      "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": "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": "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": "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": 0.32416675,
      "accuracy_std_dev": 0.025806488741932716,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.07527415750000001,
      "avg_duration_per_task_ms": 136836.5875,
      "avg_output_tokens_per_task": 7681.4025,
      "last_run_timestamp": "2026-10-03T08:23:09.856Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.3240741212121212,
      "accuracy_std_dev": 0.04532665948699506,
      "run_count": 8,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.04404032070707071,
      "avg_duration_per_task_ms": 93583.61111111111,
      "avg_output_tokens_per_task": 1344.4166666666667,
      "last_run_timestamp": "2026-10-24T18:56:17.284Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "BaseTen",
      "accuracy": 0.32207219594594594,
      "accuracy_std_dev": 0.32266469334989745,
      "run_count": 6,
      "total_questions": 296,
      "avg_cost_per_task_usd": 0.010810280405405406,
      "avg_duration_per_task_ms": 61078.11824324324,
      "avg_output_tokens_per_task": 11341.706081081082,
      "last_run_timestamp": "2026-10-24T19:09:36.304Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "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": 0.30927829896907216,
      "accuracy_std_dev": 0.041318422501349206,
      "run_count": 4,
      "total_questions": 194,
      "avg_cost_per_task_usd": 0.016999237113402062,
      "avg_duration_per_task_ms": 169063.57731958764,
      "avg_output_tokens_per_task": 1549.3092783505156,
      "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-4o-mini",
      "model_slug": "openai/gpt-4o-mini",
      "model_name": "OpenAI: GPT-4o-mini",
      "provider_name": "auto-routing",
      "accuracy": 0.30666649999999995,
      "accuracy_std_dev": 0.006666499999999992,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01779966,
      "avg_duration_per_task_ms": 55498.41,
      "avg_output_tokens_per_task": 3406.81,
      "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": "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": 0.2847618571428572,
      "accuracy_std_dev": 0.02910620321918911,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.1710796942857143,
      "avg_duration_per_task_ms": 139494.6,
      "avg_output_tokens_per_task": 1877.2057142857143,
      "last_run_timestamp": "2026-10-17T19:19:08.724Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4o-mini",
      "model_slug": "openai/gpt-4o-mini",
      "model_name": "OpenAI: GPT-4o-mini",
      "provider_name": "OpenAI",
      "accuracy": 0.2844446666666667,
      "accuracy_std_dev": 0.02997950740459592,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.025635820000000004,
      "avg_duration_per_task_ms": 67806.75333333333,
      "avg_output_tokens_per_task": 4107.166666666667,
      "last_run_timestamp": "2026-10-02T23:13:47.090Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen-2.5-72b-instruct",
      "model_slug": "qwen/qwen-2.5-72b-instruct",
      "model_name": "Qwen2.5 72B Instruct",
      "provider_name": "Novita",
      "accuracy": 0.2733335,
      "accuracy_std_dev": 0.013333499999999998,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.07945186,
      "avg_duration_per_task_ms": 632916.72,
      "avg_output_tokens_per_task": 5086.04,
      "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/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": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "AtlasCloud",
      "accuracy": 0.24333325000000003,
      "accuracy_std_dev": 0.08749614912204708,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.17476605,
      "avg_duration_per_task_ms": 545108.995,
      "avg_output_tokens_per_task": 11742.65,
      "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": "qwen/qwen3-coder-next-2025-02-03",
      "model_slug": "qwen/qwen3-coder-next",
      "model_name": "Qwen: Qwen3 Coder Next",
      "provider_name": "Novita",
      "accuracy": 0.24133339999999998,
      "accuracy_std_dev": 0.06591572936591082,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.013381287999999998,
      "avg_duration_per_task_ms": 21869.236,
      "avg_output_tokens_per_task": 1091.82,
      "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": "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": 0.24,
      "accuracy_std_dev": 0.033333,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.07327475,
      "avg_duration_per_task_ms": 662655.63,
      "avg_output_tokens_per_task": 5322.46,
      "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-vl-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 VL 30B A3B Instruct",
      "provider_name": "SiliconFlow",
      "accuracy": 0.23180078160919543,
      "accuracy_std_dev": 0.19038523699508184,
      "run_count": 7,
      "total_questions": 348,
      "avg_cost_per_task_usd": 0.19703842528735632,
      "avg_duration_per_task_ms": 1194250.8362068965,
      "avg_output_tokens_per_task": 7552.1781609195405,
      "last_run_timestamp": "2026-10-03T08:09:23.857Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Nebius",
      "accuracy": 0.2296295555555556,
      "accuracy_std_dev": 0.2407828888462297,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.07424728444444444,
      "avg_duration_per_task_ms": 161317.51555555555,
      "avg_output_tokens_per_task": 18679.246666666666,
      "last_run_timestamp": "2026-10-24T19:07:57.340Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta-llama/llama-3.1-8b-instruct",
      "model_slug": "meta-llama/llama-3.1-8b-instruct",
      "model_name": "Meta: Llama 3.1 8B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.22244883673469384,
      "accuracy_std_dev": 0.060049660254159636,
      "run_count": 5,
      "total_questions": 245,
      "avg_cost_per_task_usd": 0.004490742857142857,
      "avg_duration_per_task_ms": 16706.061224489797,
      "avg_output_tokens_per_task": 1551.4979591836734,
      "last_run_timestamp": "2026-10-24T19:30:38.210Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "CoreWeave",
      "accuracy": 0.21868997687861272,
      "accuracy_std_dev": 0.24138887954206054,
      "run_count": 7,
      "total_questions": 346,
      "avg_cost_per_task_usd": 0.009219916184971097,
      "avg_duration_per_task_ms": 788394.760115607,
      "avg_output_tokens_per_task": 61580.92485549133,
      "last_run_timestamp": "2026-10-24T19:33:50.345Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.1-terminus",
      "model_slug": "deepseek/deepseek-v3.1-terminus",
      "model_name": "DeepSeek: DeepSeek V3.1 Terminus",
      "provider_name": "AtlasCloud",
      "accuracy": 0.2116665,
      "accuracy_std_dev": 0.12643621800042107,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.018232810000000002,
      "avg_duration_per_task_ms": 80959.52,
      "avg_output_tokens_per_task": 1422.615,
      "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/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": "amazon/nova-micro-v1",
      "model_slug": "amazon/nova-micro-v1",
      "model_name": "Amazon: Nova Micro 1.0",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.17942213775510205,
      "accuracy_std_dev": 0.009739601557558703,
      "run_count": 4,
      "total_questions": 196,
      "avg_cost_per_task_usd": 0.010118459183673469,
      "avg_duration_per_task_ms": 155344.85714285713,
      "avg_output_tokens_per_task": 4996.790816326531,
      "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/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "Venice",
      "accuracy": 0.17133340000000002,
      "accuracy_std_dev": 0.261806292768986,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.006590812,
      "avg_duration_per_task_ms": 17587.43,
      "avg_output_tokens_per_task": 3014.246,
      "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": "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": 0.14,
      "accuracy_std_dev": 0.014142135623730954,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.0072234,
      "avg_duration_per_task_ms": 76455.195,
      "avg_output_tokens_per_task": 2322.375,
      "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": 0.111111,
      "accuracy_std_dev": 0.01571332689152746,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.007342933333333333,
      "avg_duration_per_task_ms": 57269.24,
      "avg_output_tokens_per_task": 2382.0866666666666,
      "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": 0.11,
      "accuracy_std_dev": 0.009999999999999995,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.006762440000000001,
      "avg_duration_per_task_ms": 64306.02,
      "avg_output_tokens_per_task": 2312.51,
      "last_run_timestamp": "2026-10-01T19:04:36.194Z"
    }
  ]
}
