{
  "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": "2026-11-01",
  "retrieved_at": "2026-11-01T01:45:38.316Z",
  "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": 2241,
  "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": "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": 0.9301345000000001,
      "accuracy_std_dev": 0.006832688313053163,
      "run_count": 18,
      "total_questions": 3564,
      "avg_cost_per_task_usd": 0.026123152076318747,
      "avg_duration_per_task_ms": 18695.293490460157,
      "avg_output_tokens_per_task": 2221.0524691358023,
      "last_run_timestamp": "2026-10-31T08:07:18.284Z"
    },
    {
      "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": "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": 0.929854,
      "accuracy_std_dev": 0.006348890821762996,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.025367877946127944,
      "avg_duration_per_task_ms": 20154.14898989899,
      "avg_output_tokens_per_task": 2304.1734006734005,
      "last_run_timestamp": "2026-10-31T08:33:04.486Z"
    },
    {
      "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": 0.9290121666666666,
      "accuracy_std_dev": 0.0044978590987515684,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.024844954545454545,
      "avg_duration_per_task_ms": 19502.118686868685,
      "avg_output_tokens_per_task": 2244.3215488215487,
      "last_run_timestamp": "2026-10-31T09:02:13.238Z"
    },
    {
      "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": "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": 0.9282826,
      "accuracy_std_dev": 0.00626296053316642,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.025195250505050507,
      "avg_duration_per_task_ms": 19633.72121212121,
      "avg_output_tokens_per_task": 2281.29898989899,
      "last_run_timestamp": "2026-10-31T08:09:30.514Z"
    },
    {
      "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": "Parasail",
      "accuracy": 0.9275846571787361,
      "accuracy_std_dev": 0.010407425048021134,
      "run_count": 13,
      "total_questions": 2389,
      "avg_cost_per_task_usd": 0.13203736626203433,
      "avg_duration_per_task_ms": 128986.54792800335,
      "avg_output_tokens_per_task": 8711.713269150272,
      "last_run_timestamp": "2026-10-31T08:39:06.851Z"
    },
    {
      "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": 0.92672599688196,
      "accuracy_std_dev": 0.007741573929745956,
      "run_count": 36,
      "total_questions": 6735,
      "avg_cost_per_task_usd": 0.17203167082405343,
      "avg_duration_per_task_ms": 164862.31818856718,
      "avg_output_tokens_per_task": 9318.532887899035,
      "last_run_timestamp": "2026-10-31T08:52:06.416Z"
    },
    {
      "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": 0.9260940999999999,
      "accuracy_std_dev": 0.00286181765491794,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.026339350505050502,
      "avg_duration_per_task_ms": 19433.036868686868,
      "avg_output_tokens_per_task": 2294.320202020202,
      "last_run_timestamp": "2026-10-31T08:17:29.639Z"
    },
    {
      "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": 0.9252448891402715,
      "accuracy_std_dev": 0.0069132275216017836,
      "run_count": 9,
      "total_questions": 1768,
      "avg_cost_per_task_usd": 0.10858752149321266,
      "avg_duration_per_task_ms": 188184.25339366516,
      "avg_output_tokens_per_task": 9458.231334841628,
      "last_run_timestamp": "2026-10-31T09:20:20.642Z"
    },
    {
      "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": "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": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Makora",
      "accuracy": 0.9240912983587339,
      "accuracy_std_dev": 0.015687164131543977,
      "run_count": 10,
      "total_questions": 1706,
      "avg_cost_per_task_usd": 0.1119299531066823,
      "avg_duration_per_task_ms": 198230.14243845252,
      "avg_output_tokens_per_task": 8354.249120750294,
      "last_run_timestamp": "2026-10-31T08:14:43.089Z"
    },
    {
      "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": 0.9237474745083064,
      "accuracy_std_dev": 0.02621443302538222,
      "run_count": 27,
      "total_questions": 5237,
      "avg_cost_per_task_usd": 0.12266296333778877,
      "avg_duration_per_task_ms": 159012.23353064733,
      "avg_output_tokens_per_task": 9966.579912163452,
      "last_run_timestamp": "2026-10-31T08:44:11.325Z"
    },
    {
      "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": "Near AI",
      "accuracy": 0.9234005000000001,
      "accuracy_std_dev": 0.005892500000000023,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.02620280303030303,
      "avg_duration_per_task_ms": 592141.2247474748,
      "avg_output_tokens_per_task": 21737.315656565657,
      "last_run_timestamp": "2026-10-31T10:06:11.748Z"
    },
    {
      "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": "Modal",
      "accuracy": 0.9231607898761836,
      "accuracy_std_dev": 0.009964755650903951,
      "run_count": 15,
      "total_questions": 2746,
      "avg_cost_per_task_usd": 0.14528911981063364,
      "avg_duration_per_task_ms": 123061.00109249818,
      "avg_output_tokens_per_task": 9541.6591405681,
      "last_run_timestamp": "2026-10-31T08:14:43.086Z"
    },
    {
      "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": 0.9228069789473685,
      "accuracy_std_dev": 0.011126928806578514,
      "run_count": 3,
      "total_questions": 570,
      "avg_cost_per_task_usd": 0.00580749649122807,
      "avg_duration_per_task_ms": 218614.69122807018,
      "avg_output_tokens_per_task": 19521.97543859649,
      "last_run_timestamp": "2026-10-31T08:40:34.158Z"
    },
    {
      "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": 0.92256925,
      "accuracy_std_dev": 0.005618592528382903,
      "run_count": 5,
      "total_questions": 960,
      "avg_cost_per_task_usd": 0.148641165625,
      "avg_duration_per_task_ms": 385822.7,
      "avg_output_tokens_per_task": 9697.144791666668,
      "last_run_timestamp": "2026-10-31T08:38:36.234Z"
    },
    {
      "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": "Sail Research",
      "accuracy": 0.9221626470588234,
      "accuracy_std_dev": 0.007429213654384402,
      "run_count": 17,
      "total_questions": 3366,
      "avg_cost_per_task_usd": 0.13078954872251933,
      "avg_duration_per_task_ms": 172274.93434343435,
      "avg_output_tokens_per_task": 11489.700237670826,
      "last_run_timestamp": "2026-10-31T09:01:05.876Z"
    },
    {
      "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": "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": 0.9208753333333334,
      "accuracy_std_dev": 0.003636621356271357,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.019656516835016836,
      "avg_duration_per_task_ms": 90437.9696969697,
      "avg_output_tokens_per_task": 20830.57239057239,
      "last_run_timestamp": "2026-10-31T08:09:30.885Z"
    },
    {
      "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": "Amazon Bedrock",
      "accuracy": 0.9205386,
      "accuracy_std_dev": 0.004178401589124734,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.14947429494949496,
      "avg_duration_per_task_ms": 120588.33434343434,
      "avg_output_tokens_per_task": 9732.809090909092,
      "last_run_timestamp": "2026-10-31T08:09:30.327Z"
    },
    {
      "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": 0.9199569999999999,
      "accuracy_std_dev": 0.006554291127331654,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.17177113406795225,
      "avg_duration_per_task_ms": 299167.57943067036,
      "avg_output_tokens_per_task": 9764.863636363636,
      "last_run_timestamp": "2026-10-31T08:33:15.782Z"
    },
    {
      "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": 0.9192646019184652,
      "accuracy_std_dev": 0.016279466419047314,
      "run_count": 3,
      "total_questions": 417,
      "avg_cost_per_task_usd": 0.01652008633093525,
      "avg_duration_per_task_ms": 447771.20383693045,
      "avg_output_tokens_per_task": 32878.398081534775,
      "last_run_timestamp": "2026-10-31T08:35:38.787Z"
    },
    {
      "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": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "BaseTen",
      "accuracy": 0.9190073210272871,
      "accuracy_std_dev": 0.01267182538981892,
      "run_count": 13,
      "total_questions": 2492,
      "avg_cost_per_task_usd": 0.1293751532905297,
      "avg_duration_per_task_ms": 116839.90609951846,
      "avg_output_tokens_per_task": 8587.111556982343,
      "last_run_timestamp": "2026-10-31T08:14:19.970Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "NextBit",
      "accuracy": 0.918981375,
      "accuracy_std_dev": 0.00679620392089402,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.12805398863636364,
      "avg_duration_per_task_ms": 541652.8712121212,
      "avg_output_tokens_per_task": 37116.541666666664,
      "last_run_timestamp": "2026-10-03T09:05:48.324Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "SiliconFlow",
      "accuracy": 0.9186303333333333,
      "accuracy_std_dev": 0.010408091158111372,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.025292042087542085,
      "avg_duration_per_task_ms": 160402.3451178451,
      "avg_output_tokens_per_task": 20957.713804713803,
      "last_run_timestamp": "2026-10-31T08:50:50.713Z"
    },
    {
      "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": 0.9185276396965866,
      "accuracy_std_dev": 0.01024254822724797,
      "run_count": 12,
      "total_questions": 2373,
      "avg_cost_per_task_usd": 0.13179984070796458,
      "avg_duration_per_task_ms": 262587.70375052677,
      "avg_output_tokens_per_task": 10414.297935103245,
      "last_run_timestamp": "2026-10-31T08:21:51.280Z"
    },
    {
      "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": 0.9177144177922683,
      "accuracy_std_dev": 0.013751154902061724,
      "run_count": 11,
      "total_questions": 2147,
      "avg_cost_per_task_usd": 0.2460820894271076,
      "avg_duration_per_task_ms": 409070.95482068,
      "avg_output_tokens_per_task": 37154.30647414998,
      "last_run_timestamp": "2026-10-31T08:34:22.445Z"
    },
    {
      "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": 0.9175082499999999,
      "accuracy_std_dev": 0.003571399938516188,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.025937152356902356,
      "avg_duration_per_task_ms": 19174.19191919192,
      "avg_output_tokens_per_task": 2235.470117845118,
      "last_run_timestamp": "2026-10-31T08:20:35.966Z"
    },
    {
      "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-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Baidu",
      "accuracy": 0.9162757342365815,
      "accuracy_std_dev": 0.00785699694603481,
      "run_count": 10,
      "total_questions": 1919,
      "avg_cost_per_task_usd": 0.09128974934861907,
      "avg_duration_per_task_ms": 645077.3731109953,
      "avg_output_tokens_per_task": 39631.607087024495,
      "last_run_timestamp": "2026-10-31T08:24:03.784Z"
    },
    {
      "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": 0.9159771681985294,
      "accuracy_std_dev": 0.02535263090623474,
      "run_count": 22,
      "total_questions": 4352,
      "avg_cost_per_task_usd": 0.14850530836397058,
      "avg_duration_per_task_ms": 399083.8046875,
      "avg_output_tokens_per_task": 9306.142463235294,
      "last_run_timestamp": "2026-10-31T09:12:10.294Z"
    },
    {
      "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": 0.9158379972375691,
      "accuracy_std_dev": 0.012441503983078046,
      "run_count": 13,
      "total_questions": 1810,
      "avg_cost_per_task_usd": 0.11735693314917128,
      "avg_duration_per_task_ms": 179712.15801104973,
      "avg_output_tokens_per_task": 8118.493370165746,
      "last_run_timestamp": "2026-10-31T08:07:17.446Z"
    },
    {
      "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": 0.9149830000000001,
      "accuracy_std_dev": 0.009335323856192691,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.15949898674242427,
      "avg_duration_per_task_ms": 579778.1357323233,
      "avg_output_tokens_per_task": 39960.98863636364,
      "last_run_timestamp": "2026-10-31T08:49:58.581Z"
    },
    {
      "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": 0.914552902173913,
      "accuracy_std_dev": 0.010001261731512806,
      "run_count": 3,
      "total_questions": 552,
      "avg_cost_per_task_usd": 0.01860343115942029,
      "avg_duration_per_task_ms": 136525.615942029,
      "avg_output_tokens_per_task": 19948.84963768116,
      "last_run_timestamp": "2026-10-31T08:12:46.334Z"
    },
    {
      "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": 0.9129758461538462,
      "accuracy_std_dev": 0.013360432120177481,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.033474385003885006,
      "avg_duration_per_task_ms": 223207.8550893551,
      "avg_output_tokens_per_task": 27728.331002331004,
      "last_run_timestamp": "2026-10-31T16:05:09.530Z"
    },
    {
      "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": 0.9129648851570964,
      "accuracy_std_dev": 0.008195106272424223,
      "run_count": 14,
      "total_questions": 2769,
      "avg_cost_per_task_usd": 0.0985225175153485,
      "avg_duration_per_task_ms": 283575.7085590466,
      "avg_output_tokens_per_task": 25018.89021307331,
      "last_run_timestamp": "2026-10-31T08:40:32.804Z"
    },
    {
      "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": 0.9126954247352496,
      "accuracy_std_dev": 0.015274129139141904,
      "run_count": 14,
      "total_questions": 2644,
      "avg_cost_per_task_usd": 0.16414042738275345,
      "avg_duration_per_task_ms": 497069.49243570346,
      "avg_output_tokens_per_task": 11341.725037821483,
      "last_run_timestamp": "2026-10-31T08:37:30.440Z"
    },
    {
      "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": 0.91245775,
      "accuracy_std_dev": 0.005709247427419856,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.027895760101010105,
      "avg_duration_per_task_ms": 177535.13888888888,
      "avg_output_tokens_per_task": 21414.281565656565,
      "last_run_timestamp": "2026-10-31T08:15:24.481Z"
    },
    {
      "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": 0.911926105263158,
      "accuracy_std_dev": 0.013578898854258508,
      "run_count": 19,
      "total_questions": 3762,
      "avg_cost_per_task_usd": 0.1411968139287613,
      "avg_duration_per_task_ms": 616719.149122807,
      "avg_output_tokens_per_task": 40116.86735778841,
      "last_run_timestamp": "2026-10-31T08:31:11.942Z"
    },
    {
      "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": 0.911616,
      "accuracy_std_dev": 0.004209000000000018,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.024379722222222223,
      "avg_duration_per_task_ms": 102982.82575757576,
      "avg_output_tokens_per_task": 20129.815656565657,
      "last_run_timestamp": "2026-10-31T08:34:49.804Z"
    },
    {
      "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-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "DekaLLM",
      "accuracy": 0.9106891772151898,
      "accuracy_std_dev": 0.008562817793226714,
      "run_count": 5,
      "total_questions": 948,
      "avg_cost_per_task_usd": 0.14278316033755276,
      "avg_duration_per_task_ms": 333972.91455696203,
      "avg_output_tokens_per_task": 37406.36181434599,
      "last_run_timestamp": "2026-10-31T08:31:11.854Z"
    },
    {
      "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": 0.9103534999999999,
      "accuracy_std_dev": 0.003644869990822727,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.019048060606060607,
      "avg_duration_per_task_ms": 135272.13636363635,
      "avg_output_tokens_per_task": 22610.929292929293,
      "last_run_timestamp": "2026-10-31T08:28:22.473Z"
    },
    {
      "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": 0.909632,
      "accuracy_std_dev": 0.0055275000000000185,
      "run_count": 2,
      "total_questions": 308,
      "avg_cost_per_task_usd": 0.05270355519480519,
      "avg_duration_per_task_ms": 55285.37662337662,
      "avg_output_tokens_per_task": 12143.461038961039,
      "last_run_timestamp": "2026-10-31T08:15:03.291Z"
    },
    {
      "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": "StreamLake",
      "accuracy": 0.9090910000000001,
      "accuracy_std_dev": 0.003367000000000009,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.018701613636363638,
      "avg_duration_per_task_ms": 137003.63383838383,
      "avg_output_tokens_per_task": 22059.484848484848,
      "last_run_timestamp": "2026-10-31T08:36:23.829Z"
    },
    {
      "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": "Together",
      "accuracy": 0.9086699166666665,
      "accuracy_std_dev": 0.015350820322588266,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.03558214772727273,
      "avg_duration_per_task_ms": 301721.86742424243,
      "avg_output_tokens_per_task": 29561.46254208754,
      "last_run_timestamp": "2026-10-31T16:17:15.690Z"
    },
    {
      "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": 0.9086667275553827,
      "accuracy_std_dev": 0.04751984064982348,
      "run_count": 13,
      "total_questions": 2573,
      "avg_cost_per_task_usd": 0.14066514496696458,
      "avg_duration_per_task_ms": 326890.27283326857,
      "avg_output_tokens_per_task": 10307.279440342014,
      "last_run_timestamp": "2026-10-31T09:37:23.390Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Phala",
      "accuracy": 0.9083779032258064,
      "accuracy_std_dev": 0.022205862232752862,
      "run_count": 12,
      "total_questions": 2232,
      "avg_cost_per_task_usd": 0.1474674775985663,
      "avg_duration_per_task_ms": 555743.3785842294,
      "avg_output_tokens_per_task": 37926.89829749104,
      "last_run_timestamp": "2026-10-31T08:44:11.309Z"
    },
    {
      "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": 0.9081554444444445,
      "accuracy_std_dev": 0.008876832231665518,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.033613874298540965,
      "avg_duration_per_task_ms": 210737.89562289562,
      "avg_output_tokens_per_task": 27019.509539842875,
      "last_run_timestamp": "2026-10-31T08:40:16.353Z"
    },
    {
      "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": 0.9078013574468085,
      "accuracy_std_dev": 0.012928926018815361,
      "run_count": 3,
      "total_questions": 470,
      "avg_cost_per_task_usd": 0.01807536170212766,
      "avg_duration_per_task_ms": 357786.7425531915,
      "avg_output_tokens_per_task": 17947.382978723403,
      "last_run_timestamp": "2026-10-31T09:05:48.501Z"
    },
    {
      "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": 0.9076662307692308,
      "accuracy_std_dev": 0.02122131075178129,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.15455806177156176,
      "avg_duration_per_task_ms": 600133.6076146077,
      "avg_output_tokens_per_task": 38964.08663558664,
      "last_run_timestamp": "2026-10-31T08:28:10.440Z"
    },
    {
      "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": 0.9076534680851064,
      "accuracy_std_dev": 0.005694785441280666,
      "run_count": 6,
      "total_questions": 1128,
      "avg_cost_per_task_usd": 0.025873300531914892,
      "avg_duration_per_task_ms": 176651.64184397162,
      "avg_output_tokens_per_task": 28451.522163120568,
      "last_run_timestamp": "2026-10-31T08:42:09.092Z"
    },
    {
      "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": 0.9074649821150856,
      "accuracy_std_dev": 0.007911648334003893,
      "run_count": 13,
      "total_questions": 2572,
      "avg_cost_per_task_usd": 0.030564498833592534,
      "avg_duration_per_task_ms": 392326.3246500778,
      "avg_output_tokens_per_task": 31615.309875583203,
      "last_run_timestamp": "2026-10-31T16:32:36.327Z"
    },
    {
      "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": 0.9074074999999998,
      "accuracy_std_dev": 0.008417500000000022,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.014508287878787877,
      "avg_duration_per_task_ms": 90086.95707070707,
      "avg_output_tokens_per_task": 21790.61111111111,
      "last_run_timestamp": "2026-10-31T08:13:11.036Z"
    },
    {
      "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": 0.9074072499999999,
      "accuracy_std_dev": 0.004610545920766882,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.025605553030303027,
      "avg_duration_per_task_ms": 96285.60858585859,
      "avg_output_tokens_per_task": 21160.45580808081,
      "last_run_timestamp": "2026-10-31T09:42:32.702Z"
    },
    {
      "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": 0.9071909310747666,
      "accuracy_std_dev": 0.03850936704324035,
      "run_count": 13,
      "total_questions": 2568,
      "avg_cost_per_task_usd": 0.1417193566978193,
      "avg_duration_per_task_ms": 214397.28777258567,
      "avg_output_tokens_per_task": 9503.626168224298,
      "last_run_timestamp": "2026-10-31T08:40:21.641Z"
    },
    {
      "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": 0.9071887349785409,
      "accuracy_std_dev": 0.014345789366918764,
      "run_count": 5,
      "total_questions": 932,
      "avg_cost_per_task_usd": 0.12532361909871245,
      "avg_duration_per_task_ms": 1379721.1190987125,
      "avg_output_tokens_per_task": 49889.66094420601,
      "last_run_timestamp": "2026-10-31T08:27:43.435Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "InferenceNet",
      "accuracy": 0.9071266666666665,
      "accuracy_std_dev": 0.007235753258799129,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.009577915824915826,
      "avg_duration_per_task_ms": 164204.16077441076,
      "avg_output_tokens_per_task": 18450.463804713803,
      "last_run_timestamp": "2026-10-31T08:24:03.698Z"
    },
    {
      "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": 0.9070474558880683,
      "accuracy_std_dev": 0.028977319190080154,
      "run_count": 13,
      "total_questions": 2573,
      "avg_cost_per_task_usd": 0.0985262642829382,
      "avg_duration_per_task_ms": 564503.555771473,
      "avg_output_tokens_per_task": 39384.385542168675,
      "last_run_timestamp": "2026-10-31T08:39:28.448Z"
    },
    {
      "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": 0.9070292653061225,
      "accuracy_std_dev": 0.005723067708842884,
      "run_count": 3,
      "total_questions": 588,
      "avg_cost_per_task_usd": 0.021270954081632654,
      "avg_duration_per_task_ms": 117754.6887755102,
      "avg_output_tokens_per_task": 21116.049319727892,
      "last_run_timestamp": "2026-10-31T08:51:31.997Z"
    },
    {
      "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-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Sail Research",
      "accuracy": 0.9065654999999998,
      "accuracy_std_dev": 0.006645834303030217,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.0029837954545454543,
      "avg_duration_per_task_ms": 222248.6893939394,
      "avg_output_tokens_per_task": 19714.793771043773,
      "last_run_timestamp": "2026-10-31T08:45:39.874Z"
    },
    {
      "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": 0.9062846666666666,
      "accuracy_std_dev": 0.0031744380430068155,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.011460047138047139,
      "avg_duration_per_task_ms": 97210.29461279462,
      "avg_output_tokens_per_task": 21369.54713804714,
      "last_run_timestamp": "2026-10-31T08:27:43.734Z"
    },
    {
      "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": 0.9062632644787644,
      "accuracy_std_dev": 0.007463126815174401,
      "run_count": 8,
      "total_questions": 1554,
      "avg_cost_per_task_usd": 0.024727387387387386,
      "avg_duration_per_task_ms": 89200.08815958817,
      "avg_output_tokens_per_task": 20548.897683397685,
      "last_run_timestamp": "2026-10-31T08:46:47.079Z"
    },
    {
      "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": 0.9061447499999998,
      "accuracy_std_dev": 0.005102889738912664,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.007892248737373738,
      "avg_duration_per_task_ms": 123987.29040404041,
      "avg_output_tokens_per_task": 16306.166666666666,
      "last_run_timestamp": "2026-10-31T08:53:36.415Z"
    },
    {
      "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": 0.9060041666666666,
      "accuracy_std_dev": 0.006179339377222198,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.032878036195286196,
      "avg_duration_per_task_ms": 809018.2592592592,
      "avg_output_tokens_per_task": 37448.21212121212,
      "last_run_timestamp": "2026-10-31T08:23:27.402Z"
    },
    {
      "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": 0.9059643571428573,
      "accuracy_std_dev": 0.02530737982319202,
      "run_count": 14,
      "total_questions": 2772,
      "avg_cost_per_task_usd": 0.039628380591630596,
      "avg_duration_per_task_ms": 275120.5137085137,
      "avg_output_tokens_per_task": 30935.2702020202,
      "last_run_timestamp": "2026-10-31T16:23:41.132Z"
    },
    {
      "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": 0.9059342499999998,
      "accuracy_std_dev": 0.010910172663874526,
      "run_count": 3,
      "total_questions": 528,
      "avg_cost_per_task_usd": 0.008819143939393941,
      "avg_duration_per_task_ms": 163195.54924242425,
      "avg_output_tokens_per_task": 22179.03787878788,
      "last_run_timestamp": "2026-10-31T08:49:09.016Z"
    },
    {
      "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": 0.9057238888888889,
      "accuracy_std_dev": 0.009356618102764875,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.03432713860830528,
      "avg_duration_per_task_ms": 196757.96408529743,
      "avg_output_tokens_per_task": 28444.87317620651,
      "last_run_timestamp": "2026-10-31T08:29:56.453Z"
    },
    {
      "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": 0.9054646923076923,
      "accuracy_std_dev": 0.020414937929725787,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.036716465034965035,
      "avg_duration_per_task_ms": 253595.11771561773,
      "avg_output_tokens_per_task": 30738.896658896658,
      "last_run_timestamp": "2026-10-31T16:09:19.241Z"
    },
    {
      "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": 0.9054373465045593,
      "accuracy_std_dev": 0.0037554479359991244,
      "run_count": 5,
      "total_questions": 987,
      "avg_cost_per_task_usd": 0.026207787234042552,
      "avg_duration_per_task_ms": 127141.36068895644,
      "avg_output_tokens_per_task": 21741.29989868288,
      "last_run_timestamp": "2026-10-31T08:31:11.848Z"
    },
    {
      "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": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "GMICloud",
      "accuracy": 0.9053629285714285,
      "accuracy_std_dev": 0.024694616907520228,
      "run_count": 14,
      "total_questions": 2772,
      "avg_cost_per_task_usd": 0.03438339466089466,
      "avg_duration_per_task_ms": 227033.76984126985,
      "avg_output_tokens_per_task": 29037.088383838385,
      "last_run_timestamp": "2026-10-31T16:04:48.154Z"
    },
    {
      "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": 0.9053496666666667,
      "accuracy_std_dev": 0.008612696635394351,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.02068693490460157,
      "avg_duration_per_task_ms": 208793.50953984287,
      "avg_output_tokens_per_task": 28854.00224466891,
      "last_run_timestamp": "2026-10-31T09:01:06.228Z"
    },
    {
      "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": 0.9052795516491482,
      "accuracy_std_dev": 0.020697851140189134,
      "run_count": 14,
      "total_questions": 2759,
      "avg_cost_per_task_usd": 0.1211796118158753,
      "avg_duration_per_task_ms": 328498.6843059079,
      "avg_output_tokens_per_task": 8296.681406306632,
      "last_run_timestamp": "2026-10-31T08:25:00.194Z"
    },
    {
      "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": 0.9051626666666668,
      "accuracy_std_dev": 0.007816314746875412,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.017028636363636364,
      "avg_duration_per_task_ms": 78357.59090909091,
      "avg_output_tokens_per_task": 16892.26430976431,
      "last_run_timestamp": "2026-10-31T08:58:04.691Z"
    },
    {
      "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": "Wafer",
      "accuracy": 0.9048818749999998,
      "accuracy_std_dev": 0.008666358814945001,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.008252799242424242,
      "avg_duration_per_task_ms": 82042.59659090909,
      "avg_output_tokens_per_task": 16383.162878787878,
      "last_run_timestamp": "2026-10-31T08:56:01.471Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta/muse-spark-open-20261015",
      "model_slug": "meta/muse-spark-open",
      "model_name": "Meta: Muse Spark Open",
      "provider_name": "Makora",
      "accuracy": 0.9048634068157615,
      "accuracy_std_dev": 0.012266915351464668,
      "run_count": 5,
      "total_questions": 939,
      "avg_cost_per_task_usd": 0.06399653567625134,
      "avg_duration_per_task_ms": 75580.11395101172,
      "avg_output_tokens_per_task": 14917.831735889244,
      "last_run_timestamp": "2026-10-31T08:13:57.643Z"
    },
    {
      "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": 0.9046013333333334,
      "accuracy_std_dev": 0.009152126540986068,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.02685950505050505,
      "avg_duration_per_task_ms": 136981.2441077441,
      "avg_output_tokens_per_task": 22236.84175084175,
      "last_run_timestamp": "2026-10-31T08:31:11.499Z"
    },
    {
      "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": 0.9046013333333334,
      "accuracy_std_dev": 0.003967811851498118,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.016546668350168353,
      "avg_duration_per_task_ms": 140240.16835016836,
      "avg_output_tokens_per_task": 21954.457912457914,
      "last_run_timestamp": "2026-10-31T08:27:53.304Z"
    },
    {
      "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": 0.9046013333333334,
      "accuracy_std_dev": 0.008285549120138134,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.01933321717171717,
      "avg_duration_per_task_ms": 139538.02861952863,
      "avg_output_tokens_per_task": 20900.63973063973,
      "last_run_timestamp": "2026-10-31T08:20:08.070Z"
    },
    {
      "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": 0.9045483013645222,
      "accuracy_std_dev": 0.011375036654859972,
      "run_count": 13,
      "total_questions": 2565,
      "avg_cost_per_task_usd": 0.027518346978557498,
      "avg_duration_per_task_ms": 307235.2007797271,
      "avg_output_tokens_per_task": 22778.734892787525,
      "last_run_timestamp": "2026-10-31T16:12:28.227Z"
    },
    {
      "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": 0.9044004522003033,
      "accuracy_std_dev": 0.00856689234728672,
      "run_count": 10,
      "total_questions": 1977,
      "avg_cost_per_task_usd": 0.014848743045017704,
      "avg_duration_per_task_ms": 370376.5558927668,
      "avg_output_tokens_per_task": 28855.582195245323,
      "last_run_timestamp": "2026-10-31T08:03:09.643Z"
    },
    {
      "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": 0.9042316703786192,
      "accuracy_std_dev": 0.0031656583896560934,
      "run_count": 5,
      "total_questions": 898,
      "avg_cost_per_task_usd": 0.22350682405345215,
      "avg_duration_per_task_ms": 441723.66035634745,
      "avg_output_tokens_per_task": 42286.14922048998,
      "last_run_timestamp": "2026-10-31T08:44:55.396Z"
    },
    {
      "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": 0.9041981924485772,
      "accuracy_std_dev": 0.01113877499359892,
      "run_count": 18,
      "total_questions": 3549,
      "avg_cost_per_task_usd": 0.027435362073823615,
      "avg_duration_per_task_ms": 185052.6246830093,
      "avg_output_tokens_per_task": 25932.813750352212,
      "last_run_timestamp": "2026-10-31T08:28:13.141Z"
    },
    {
      "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": "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-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Mistral",
      "accuracy": 0.9034793333333332,
      "accuracy_std_dev": 0.005722799567422317,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.025755799663299665,
      "avg_duration_per_task_ms": 89661.28282828283,
      "avg_output_tokens_per_task": 21298.74074074074,
      "last_run_timestamp": "2026-10-31T09:51:08.575Z"
    },
    {
      "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": 0.9033668,
      "accuracy_std_dev": 0.005061737504059227,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.026716864646464646,
      "avg_duration_per_task_ms": 106368.8292929293,
      "avg_output_tokens_per_task": 22085.938383838384,
      "last_run_timestamp": "2026-10-31T08:13:57.548Z"
    },
    {
      "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": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "auto-routing",
      "accuracy": 0.9031337692307693,
      "accuracy_std_dev": 0.017540530612769802,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.03434532362082362,
      "avg_duration_per_task_ms": 309648.5489510489,
      "avg_output_tokens_per_task": 30300.46114996115,
      "last_run_timestamp": "2026-10-31T16:18:48.433Z"
    },
    {
      "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": 0.9030760960809103,
      "accuracy_std_dev": 0.01701828861577222,
      "run_count": 12,
      "total_questions": 2373,
      "avg_cost_per_task_usd": 0.10602527054361567,
      "avg_duration_per_task_ms": 544239.2713864307,
      "avg_output_tokens_per_task": 37272.33965444585,
      "last_run_timestamp": "2026-10-31T08:11:08.576Z"
    },
    {
      "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": 0.9029179999999999,
      "accuracy_std_dev": 0.0028613344206273234,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.023974107744107742,
      "avg_duration_per_task_ms": 50011.877104377105,
      "avg_output_tokens_per_task": 19809.430976430976,
      "last_run_timestamp": "2026-10-31T08:51:04.442Z"
    },
    {
      "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": 0.9027699539584503,
      "accuracy_std_dev": 0.006295075818094274,
      "run_count": 9,
      "total_questions": 1781,
      "avg_cost_per_task_usd": 0.028251830432341377,
      "avg_duration_per_task_ms": 172539.25996631107,
      "avg_output_tokens_per_task": 28841.043234138124,
      "last_run_timestamp": "2026-10-31T08:31:10.643Z"
    },
    {
      "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": "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": "meta/muse-spark-open-20261015",
      "model_slug": "meta/muse-spark-open",
      "model_name": "Meta: Muse Spark Open",
      "provider_name": "Fireworks",
      "accuracy": 0.9023568,
      "accuracy_std_dev": 0.003838970664123373,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.0695745484848485,
      "avg_duration_per_task_ms": 344238.6484848485,
      "avg_output_tokens_per_task": 16158.315151515151,
      "last_run_timestamp": "2026-10-31T08:52:05.708Z"
    },
    {
      "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": 0.9020761666666667,
      "accuracy_std_dev": 0.005271695534223824,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.05407119865319866,
      "avg_duration_per_task_ms": 30482.292087542086,
      "avg_output_tokens_per_task": 2495.6936026936028,
      "last_run_timestamp": "2026-10-31T08:27:47.532Z"
    },
    {
      "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": 0.9015422819956616,
      "accuracy_std_dev": 0.037796971040504466,
      "run_count": 7,
      "total_questions": 1383,
      "avg_cost_per_task_usd": 0.09285763702096891,
      "avg_duration_per_task_ms": 766396.866955893,
      "avg_output_tokens_per_task": 36238.82212581345,
      "last_run_timestamp": "2026-10-31T16:24:40.574Z"
    },
    {
      "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": 0.9015150000000001,
      "accuracy_std_dev": 0.004208999999999963,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.0736977297979798,
      "avg_duration_per_task_ms": 138589.0126262626,
      "avg_output_tokens_per_task": 17125.45707070707,
      "last_run_timestamp": "2026-10-31T08:09:26.224Z"
    },
    {
      "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": "meta/muse-spark-open-20261015",
      "model_slug": "meta/muse-spark-open",
      "model_name": "Meta: Muse Spark Open",
      "provider_name": "BaseTen",
      "accuracy": 0.901515,
      "accuracy_std_dev": 0.0025249999999999995,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.0720575808080808,
      "avg_duration_per_task_ms": 100990.45707070707,
      "avg_output_tokens_per_task": 16730.29797979798,
      "last_run_timestamp": "2026-10-31T08:17:04.260Z"
    },
    {
      "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": "anthropic/claude-4.8-opus-20260528",
      "model_slug": "anthropic/claude-opus-4.8",
      "model_name": "Anthropic: Claude Opus 4.8",
      "provider_name": "Azure",
      "accuracy": 0.901320769230769,
      "accuracy_std_dev": 0.006035282463007334,
      "run_count": 26,
      "total_questions": 5148,
      "avg_cost_per_task_usd": 0.2067400357420358,
      "avg_duration_per_task_ms": 98162.20415695416,
      "avg_output_tokens_per_task": 7860.721445221445,
      "last_run_timestamp": "2026-10-31T08:46:46.997Z"
    },
    {
      "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": 0.9012343333333332,
      "accuracy_std_dev": 0.0037221826780652403,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.03380553956228956,
      "avg_duration_per_task_ms": 826187.4343434343,
      "avg_output_tokens_per_task": 38504.42255892256,
      "last_run_timestamp": "2026-10-31T08:14:20.398Z"
    },
    {
      "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": 0.9011541428571429,
      "accuracy_std_dev": 0.004473489479416358,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.053932944444444436,
      "avg_duration_per_task_ms": 26662.532467532466,
      "avg_output_tokens_per_task": 2480.9805194805194,
      "last_run_timestamp": "2026-10-31T08:25:18.221Z"
    },
    {
      "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": 0.9010944999999999,
      "accuracy_std_dev": 0.0012624999999999997,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.00798360101010101,
      "avg_duration_per_task_ms": 317918.4797979798,
      "avg_output_tokens_per_task": 28057.032828282827,
      "last_run_timestamp": "2026-10-31T16:08:33.700Z"
    },
    {
      "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": 0.901078703900709,
      "accuracy_std_dev": 0.008642828253125116,
      "run_count": 13,
      "total_questions": 2256,
      "avg_cost_per_task_usd": 0.025328547429078016,
      "avg_duration_per_task_ms": 170551.65514184398,
      "avg_output_tokens_per_task": 20949.003989361703,
      "last_run_timestamp": "2026-10-31T16:35:50.036Z"
    },
    {
      "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": 0.9009462380396732,
      "accuracy_std_dev": 0.019684329717769646,
      "run_count": 13,
      "total_questions": 2571,
      "avg_cost_per_task_usd": 0.12647589887203423,
      "avg_duration_per_task_ms": 518480.37456242705,
      "avg_output_tokens_per_task": 38463.91637495138,
      "last_run_timestamp": "2026-10-31T08:17:28.484Z"
    },
    {
      "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": 0.9006730000000001,
      "accuracy_std_dev": 0.006733999999999962,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.010972851010101009,
      "avg_duration_per_task_ms": 117479.04797979798,
      "avg_output_tokens_per_task": 19415.13888888889,
      "last_run_timestamp": "2026-10-31T08:11:46.712Z"
    },
    {
      "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": 0.9005650372881355,
      "accuracy_std_dev": 0.0005644881654108368,
      "run_count": 3,
      "total_questions": 590,
      "avg_cost_per_task_usd": 0.013123520338983051,
      "avg_duration_per_task_ms": 181977.62711864407,
      "avg_output_tokens_per_task": 21777.491525423728,
      "last_run_timestamp": "2026-10-31T08:31:11.333Z"
    },
    {
      "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": 0.9005527218243498,
      "accuracy_std_dev": 0.007008802337352753,
      "run_count": 14,
      "total_questions": 2653,
      "avg_cost_per_task_usd": 0.012335847719562759,
      "avg_duration_per_task_ms": 174702.59517527328,
      "avg_output_tokens_per_task": 26019.96117602714,
      "last_run_timestamp": "2026-10-31T08:07:19.235Z"
    },
    {
      "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": 0.9005006153846152,
      "accuracy_std_dev": 0.0058913432128904816,
      "run_count": 39,
      "total_questions": 7722,
      "avg_cost_per_task_usd": 0.21606604428904433,
      "avg_duration_per_task_ms": 93452.96775446775,
      "avg_output_tokens_per_task": 7881.437192437193,
      "last_run_timestamp": "2026-10-31T08:23:44.152Z"
    },
    {
      "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": 0.9002525,
      "accuracy_std_dev": 0.004014983343676526,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.014062290404040403,
      "avg_duration_per_task_ms": 87702.04797979798,
      "avg_output_tokens_per_task": 13895.555555555555,
      "last_run_timestamp": "2026-10-31T08:28:22.736Z"
    },
    {
      "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": 0.9000574515125321,
      "accuracy_std_dev": 0.11608222219996657,
      "run_count": 31,
      "total_questions": 5785,
      "avg_cost_per_task_usd": 0.16450815436473637,
      "avg_duration_per_task_ms": 193869.64321521175,
      "avg_output_tokens_per_task": 10791.89334485739,
      "last_run_timestamp": "2026-10-31T09:24:16.791Z"
    },
    {
      "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": 0.9000319047619048,
      "accuracy_std_dev": 0.005076517459315892,
      "run_count": 21,
      "total_questions": 4158,
      "avg_cost_per_task_usd": 0.05824434295334296,
      "avg_duration_per_task_ms": 35780.01707551708,
      "avg_output_tokens_per_task": 2517.5358345358345,
      "last_run_timestamp": "2026-10-31T09:01:04.740Z"
    },
    {
      "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": 0.8998963581973581,
      "accuracy_std_dev": 0.03142747825284958,
      "run_count": 9,
      "total_questions": 1287,
      "avg_cost_per_task_usd": 0.023476212121212123,
      "avg_duration_per_task_ms": 815676.9502719502,
      "avg_output_tokens_per_task": 21719.649572649574,
      "last_run_timestamp": "2026-10-31T16:03:21.470Z"
    },
    {
      "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": 0.8998314999999999,
      "accuracy_std_dev": 0.000841499999999995,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.11599434848484849,
      "avg_duration_per_task_ms": 1589231.9873737374,
      "avg_output_tokens_per_task": 46211.58585858586,
      "last_run_timestamp": "2026-10-31T08:31:10.881Z"
    },
    {
      "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": 0.8996625716694773,
      "accuracy_std_dev": 0.0020759481367959394,
      "run_count": 3,
      "total_questions": 593,
      "avg_cost_per_task_usd": 0.11324433389544686,
      "avg_duration_per_task_ms": 858665.3389544688,
      "avg_output_tokens_per_task": 45077.612141652615,
      "last_run_timestamp": "2026-10-31T08:28:18.045Z"
    },
    {
      "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": 0.8995687234401349,
      "accuracy_std_dev": 0.007447984730328704,
      "run_count": 9,
      "total_questions": 1779,
      "avg_cost_per_task_usd": 0.0197627712197864,
      "avg_duration_per_task_ms": 382016.3063518831,
      "avg_output_tokens_per_task": 22859.64811691962,
      "last_run_timestamp": "2026-10-31T08:33:02.133Z"
    },
    {
      "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": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "ModelRun",
      "accuracy": 0.8993857647058824,
      "accuracy_std_dev": 0.018671156495670343,
      "run_count": 17,
      "total_questions": 3366,
      "avg_cost_per_task_usd": 0.07737615121806299,
      "avg_duration_per_task_ms": 217682.72222222222,
      "avg_output_tokens_per_task": 30235.886809269163,
      "last_run_timestamp": "2026-10-31T16:10:09.223Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "NextBit",
      "accuracy": 0.8993638888888887,
      "accuracy_std_dev": 0.004029367042778795,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.030411617283950615,
      "avg_duration_per_task_ms": 237553.6509539843,
      "avg_output_tokens_per_task": 26633.1290684624,
      "last_run_timestamp": "2026-10-03T10:19:34.320Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-fable-5.1-20260831",
      "model_slug": "anthropic/claude-fable-5.1",
      "model_name": "Anthropic: Claude Fable 5.1",
      "provider_name": "auto-routing",
      "accuracy": 0.8993264,
      "accuracy_std_dev": 0.006511664598242117,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.13052546969696968,
      "avg_duration_per_task_ms": 30874.333333333332,
      "avg_output_tokens_per_task": 2403.130303030303,
      "last_run_timestamp": "2026-10-01T08:24:30.121Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-opus-5.5-20260921",
      "model_slug": "anthropic/claude-opus-5.5",
      "model_name": "Anthropic: Claude Opus 5.5",
      "provider_name": "Google",
      "accuracy": 0.8992302857142856,
      "accuracy_std_dev": 0.0048260090792136764,
      "run_count": 21,
      "total_questions": 4158,
      "avg_cost_per_task_usd": 0.05735562409812409,
      "avg_duration_per_task_ms": 25146.343674843676,
      "avg_output_tokens_per_task": 2479.3174603174602,
      "last_run_timestamp": "2026-10-31T09:06:56.339Z"
    },
    {
      "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": 0.8991100000000001,
      "accuracy_std_dev": 0.005600276243186554,
      "run_count": 14,
      "total_questions": 2772,
      "avg_cost_per_task_usd": 0.05768347799422799,
      "avg_duration_per_task_ms": 26053.75036075036,
      "avg_output_tokens_per_task": 2530.0901875901877,
      "last_run_timestamp": "2026-10-31T08:22:04.658Z"
    },
    {
      "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": 0.8989897750000003,
      "accuracy_std_dev": 0.005963928313148534,
      "run_count": 40,
      "total_questions": 7920,
      "avg_cost_per_task_usd": 0.22275342159090905,
      "avg_duration_per_task_ms": 94239.73522727273,
      "avg_output_tokens_per_task": 8122.755176767677,
      "last_run_timestamp": "2026-10-31T08:31:12.840Z"
    },
    {
      "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": 0.8987954230769233,
      "accuracy_std_dev": 0.0076484229655182515,
      "run_count": 26,
      "total_questions": 5148,
      "avg_cost_per_task_usd": 0.12695155264180263,
      "avg_duration_per_task_ms": 58394.58430458431,
      "avg_output_tokens_per_task": 4732.304195804196,
      "last_run_timestamp": "2026-10-31T08:31:13.199Z"
    },
    {
      "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": 0.8987492142857141,
      "accuracy_std_dev": 0.014339465397379814,
      "run_count": 14,
      "total_questions": 2772,
      "avg_cost_per_task_usd": 0.004243736652236653,
      "avg_duration_per_task_ms": 360908.05663780664,
      "avg_output_tokens_per_task": 23278.13023088023,
      "last_run_timestamp": "2026-10-31T08:39:08.723Z"
    },
    {
      "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": 0.8987081781963057,
      "accuracy_std_dev": 0.03249433928983099,
      "run_count": 14,
      "total_questions": 2761,
      "avg_cost_per_task_usd": 0.14709861970300617,
      "avg_duration_per_task_ms": 823480.3976819993,
      "avg_output_tokens_per_task": 36846.00507062658,
      "last_run_timestamp": "2026-10-31T08:15:03.301Z"
    },
    {
      "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": "BaseTen",
      "accuracy": 0.8986319552238805,
      "accuracy_std_dev": 0.009874782731787053,
      "run_count": 3,
      "total_questions": 536,
      "avg_cost_per_task_usd": 0.02413633582089552,
      "avg_duration_per_task_ms": 93980.86567164179,
      "avg_output_tokens_per_task": 19966.47014925373,
      "last_run_timestamp": "2026-10-31T08:27:51.477Z"
    },
    {
      "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": 0.8986254025773196,
      "accuracy_std_dev": 0.041870475079224964,
      "run_count": 10,
      "total_questions": 1940,
      "avg_cost_per_task_usd": 0.047417563402061845,
      "avg_duration_per_task_ms": 459421.87164948456,
      "avg_output_tokens_per_task": 15074.60618556701,
      "last_run_timestamp": "2026-10-31T16:08:56.804Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-luna",
      "model_slug": "openai/gpt-6-luna",
      "model_name": "GPT-6-Luna",
      "provider_name": "auto-routing",
      "accuracy": 0.898477,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 197,
      "avg_cost_per_task_usd": 0.003442197969543147,
      "avg_duration_per_task_ms": 73312.30456852792,
      "avg_output_tokens_per_task": 6831.9390862944165,
      "last_run_timestamp": "2026-10-08T13:29:57.437Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.8-opus-20260528",
      "model_slug": "anthropic/claude-opus-4.8",
      "model_name": "Anthropic: Claude Opus 4.8",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.8981480833333332,
      "accuracy_std_dev": 0.007606843601415013,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.19857798400673404,
      "avg_duration_per_task_ms": 95247.20033670033,
      "avg_output_tokens_per_task": 7731.943181818182,
      "last_run_timestamp": "2026-10-31T08:27:45.200Z"
    },
    {
      "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": 0.8981480000000002,
      "accuracy_std_dev": 0.016043822300187693,
      "run_count": 14,
      "total_questions": 2772,
      "avg_cost_per_task_usd": 0.019599273088023093,
      "avg_duration_per_task_ms": 202557.83225108226,
      "avg_output_tokens_per_task": 25548.025613275615,
      "last_run_timestamp": "2026-10-31T09:18:59.078Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "SiliconFlow",
      "accuracy": 0.898148,
      "accuracy_std_dev": 0.012012972737836369,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.03667170117845118,
      "avg_duration_per_task_ms": 157797.89814814815,
      "avg_output_tokens_per_task": 27519.53787878788,
      "last_run_timestamp": "2026-10-03T09:28:24.400Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8981478333333334,
      "accuracy_std_dev": 0.007386487717823356,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.12543263804713806,
      "avg_duration_per_task_ms": 965889.989057239,
      "avg_output_tokens_per_task": 49917.90404040404,
      "last_run_timestamp": "2026-10-31T08:45:01.505Z"
    },
    {
      "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": 0.89814775,
      "accuracy_std_dev": 0.0055197238778311765,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.11915268055555557,
      "avg_duration_per_task_ms": 825703.6161616162,
      "avg_output_tokens_per_task": 47450.94823232323,
      "last_run_timestamp": "2026-10-31T08:07:17.850Z"
    },
    {
      "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": 0.8980075833333333,
      "accuracy_std_dev": 0.004784274352123693,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.23032522264309765,
      "avg_duration_per_task_ms": 110445.48653198653,
      "avg_output_tokens_per_task": 8998.862373737375,
      "last_run_timestamp": "2026-10-31T08:24:37.252Z"
    },
    {
      "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": 0.8978673333333335,
      "accuracy_std_dev": 0.007570651065499923,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.03724634175084175,
      "avg_duration_per_task_ms": 745311.5892255892,
      "avg_output_tokens_per_task": 37370.69696969697,
      "last_run_timestamp": "2026-10-31T08:12:46.423Z"
    },
    {
      "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": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "GMICloud",
      "accuracy": 0.8976429,
      "accuracy_std_dev": 0.019280368027867086,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.024536105555555552,
      "avg_duration_per_task_ms": 164306.0883838384,
      "avg_output_tokens_per_task": 28203.131818181817,
      "last_run_timestamp": "2026-10-31T08:31:11.869Z"
    },
    {
      "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": 0.8973062307692307,
      "accuracy_std_dev": 0.017706017439337668,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.030415017871017868,
      "avg_duration_per_task_ms": 251068.9052059052,
      "avg_output_tokens_per_task": 27243.771173271172,
      "last_run_timestamp": "2026-10-31T08:36:32.971Z"
    },
    {
      "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": 0.897306,
      "accuracy_std_dev": 0.004761657064510171,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.026144708754208754,
      "avg_duration_per_task_ms": 176560.02693602693,
      "avg_output_tokens_per_task": 21638.170033670034,
      "last_run_timestamp": "2026-10-31T08:57:38.005Z"
    },
    {
      "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": 0.8973059999999999,
      "accuracy_std_dev": 0.0033669999999999534,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.07477493686868687,
      "avg_duration_per_task_ms": 142684.398989899,
      "avg_output_tokens_per_task": 17437.89393939394,
      "last_run_timestamp": "2026-10-31T08:49:18.270Z"
    },
    {
      "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": 0.8970256666666665,
      "accuracy_std_dev": 0.00667522615513675,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.06776180639730639,
      "avg_duration_per_task_ms": 222686.2255892256,
      "avg_output_tokens_per_task": 15778.796296296296,
      "last_run_timestamp": "2026-10-31T08:27:47.368Z"
    },
    {
      "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": 0.8970215358404846,
      "accuracy_std_dev": 0.034154544866046785,
      "run_count": 21,
      "total_questions": 3962,
      "avg_cost_per_task_usd": 0.37220477612317004,
      "avg_duration_per_task_ms": 491484.80312973243,
      "avg_output_tokens_per_task": 56849.085562847045,
      "last_run_timestamp": "2026-10-31T16:18:08.626Z"
    },
    {
      "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": 0.8965292307692307,
      "accuracy_std_dev": 0.011041742442998501,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.20059593162393163,
      "avg_duration_per_task_ms": 96185.55477855478,
      "avg_output_tokens_per_task": 7804.96891996892,
      "last_run_timestamp": "2026-10-31T08:32:06.814Z"
    },
    {
      "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": 0.8965206666666669,
      "accuracy_std_dev": 0.0064727425889047965,
      "run_count": 15,
      "total_questions": 2970,
      "avg_cost_per_task_usd": 0.20788648787878788,
      "avg_duration_per_task_ms": 100622.61919191919,
      "avg_output_tokens_per_task": 8101.048148148148,
      "last_run_timestamp": "2026-10-31T08:27:47.283Z"
    },
    {
      "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": 0.89642245,
      "accuracy_std_dev": 0.008143539230426761,
      "run_count": 40,
      "total_questions": 7920,
      "avg_cost_per_task_usd": 0.20629075454545454,
      "avg_duration_per_task_ms": 70772.6904040404,
      "avg_output_tokens_per_task": 7507.606691919192,
      "last_run_timestamp": "2026-10-31T08:51:35.523Z"
    },
    {
      "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": 0.896184,
      "accuracy_std_dev": 0.007570749896806769,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0237254696969697,
      "avg_duration_per_task_ms": 319850.3367003367,
      "avg_output_tokens_per_task": 19569.10606060606,
      "last_run_timestamp": "2026-10-31T08:20:35.720Z"
    },
    {
      "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": 0.896157644859813,
      "accuracy_std_dev": 0.009978064213845336,
      "run_count": 13,
      "total_questions": 2568,
      "avg_cost_per_task_usd": 0.01666022313084112,
      "avg_duration_per_task_ms": 348918.9147196262,
      "avg_output_tokens_per_task": 28403.781542056076,
      "last_run_timestamp": "2026-10-31T09:16:30.257Z"
    },
    {
      "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": 0.8961037142857144,
      "accuracy_std_dev": 0.004737329618927418,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.054530834776334774,
      "avg_duration_per_task_ms": 30476.945887445887,
      "avg_output_tokens_per_task": 2521.178932178932,
      "last_run_timestamp": "2026-10-31T08:40:20.874Z"
    },
    {
      "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-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "AtlasCloud",
      "accuracy": 0.895412375,
      "accuracy_std_dev": 0.020051295631065218,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.02414115151515151,
      "avg_duration_per_task_ms": 210274.94823232322,
      "avg_output_tokens_per_task": 27810.135101010103,
      "last_run_timestamp": "2026-10-31T08:27:49.932Z"
    },
    {
      "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": 0.8952905937499999,
      "accuracy_std_dev": 0.016969367452854572,
      "run_count": 8,
      "total_questions": 1536,
      "avg_cost_per_task_usd": 0.032397834635416664,
      "avg_duration_per_task_ms": 236218.794921875,
      "avg_output_tokens_per_task": 32195.716796875,
      "last_run_timestamp": "2026-10-31T09:17:17.877Z"
    },
    {
      "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": 0.8952341538461538,
      "accuracy_std_dev": 0.0074752677939800664,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.12034864141414141,
      "avg_duration_per_task_ms": 58356.32323232323,
      "avg_output_tokens_per_task": 4592.7913752913755,
      "last_run_timestamp": "2026-10-31T08:31:11.765Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "BaseTen",
      "accuracy": 0.8951904072144289,
      "accuracy_std_dev": 0.019820245214898276,
      "run_count": 13,
      "total_questions": 2495,
      "avg_cost_per_task_usd": 0.159686151503006,
      "avg_duration_per_task_ms": 368440.1851703407,
      "avg_output_tokens_per_task": 40060.66533066132,
      "last_run_timestamp": "2026-10-31T08:09:30.320Z"
    },
    {
      "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": "Cohere",
      "accuracy": 0.8951417142857143,
      "accuracy_std_dev": 0.009679824699036726,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.008688588023088023,
      "avg_duration_per_task_ms": 171107.75108225108,
      "avg_output_tokens_per_task": 30552.396103896102,
      "last_run_timestamp": "2026-10-31T09:09:17.234Z"
    },
    {
      "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": 0.8951046923076924,
      "accuracy_std_dev": 0.009277198784651573,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.23170269114219116,
      "avg_duration_per_task_ms": 110619.57381507382,
      "avg_output_tokens_per_task": 9043.770007770008,
      "last_run_timestamp": "2026-10-31T08:13:10.867Z"
    },
    {
      "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": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "Tencent",
      "accuracy": 0.894276,
      "accuracy_std_dev": 0.008869991905295046,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.11497575656565656,
      "avg_duration_per_task_ms": 654340.1272727273,
      "avg_output_tokens_per_task": 45820.55656565657,
      "last_run_timestamp": "2026-10-31T08:45:35.702Z"
    },
    {
      "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": 0.8941497500000002,
      "accuracy_std_dev": 0.012326495474687834,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.010706986111111112,
      "avg_duration_per_task_ms": 265829.2607323232,
      "avg_output_tokens_per_task": 20748.16351010101,
      "last_run_timestamp": "2026-10-31T08:33:46.211Z"
    },
    {
      "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": 0.893939358490566,
      "accuracy_std_dev": 0.004245539018258487,
      "run_count": 3,
      "total_questions": 583,
      "avg_cost_per_task_usd": 0.023210332761578047,
      "avg_duration_per_task_ms": 116069.80960548884,
      "avg_output_tokens_per_task": 13628.193825042881,
      "last_run_timestamp": "2026-10-31T08:17:29.099Z"
    },
    {
      "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": 0.8939393333333333,
      "accuracy_std_dev": 0.012217692780917704,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.02448665488215488,
      "avg_duration_per_task_ms": 139886.7744107744,
      "avg_output_tokens_per_task": 20249.15824915825,
      "last_run_timestamp": "2026-10-31T09:02:19.878Z"
    },
    {
      "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": 0.8939391999999999,
      "accuracy_std_dev": 0.01075335131761256,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.0050906424242424244,
      "avg_duration_per_task_ms": 221838.88434343433,
      "avg_output_tokens_per_task": 27900.379292929294,
      "last_run_timestamp": "2026-10-31T16:31:13.440Z"
    },
    {
      "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": 0.893627599657094,
      "accuracy_std_dev": 0.01880407170058742,
      "run_count": 12,
      "total_questions": 2333,
      "avg_cost_per_task_usd": 0.118091234462066,
      "avg_duration_per_task_ms": 525572.4204886412,
      "avg_output_tokens_per_task": 43317.68838405486,
      "last_run_timestamp": "2026-10-31T08:55:34.589Z"
    },
    {
      "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": "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": 0.8933780000000001,
      "accuracy_std_dev": 0.0028613344206273234,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.03135734848484849,
      "avg_duration_per_task_ms": 31339.205387205388,
      "avg_output_tokens_per_task": 13928.784511784512,
      "last_run_timestamp": "2026-10-31T08:28:17.755Z"
    },
    {
      "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": 0.8933779999999999,
      "accuracy_std_dev": 0.0034590473640392213,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.01648433501683502,
      "avg_duration_per_task_ms": 93266.93602693603,
      "avg_output_tokens_per_task": 13659.838383838383,
      "last_run_timestamp": "2026-10-31T08:40:23.037Z"
    },
    {
      "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": "PrimeIntellect",
      "accuracy": 0.8931621743589744,
      "accuracy_std_dev": 0.0023230000000000195,
      "run_count": 2,
      "total_questions": 390,
      "avg_cost_per_task_usd": 0.020242015384615383,
      "avg_duration_per_task_ms": 141496.24102564104,
      "avg_output_tokens_per_task": 40323.36666666667,
      "last_run_timestamp": "2026-10-31T08:52:13.905Z"
    },
    {
      "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": "AtlasCloud",
      "accuracy": 0.8930975,
      "accuracy_std_dev": 0.005892499999999967,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.11303334343434343,
      "avg_duration_per_task_ms": 814053.6843434343,
      "avg_output_tokens_per_task": 45048.43686868687,
      "last_run_timestamp": "2026-10-31T08:27:45.248Z"
    },
    {
      "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": 0.89301948861912,
      "accuracy_std_dev": 0.023661754900260477,
      "run_count": 10,
      "total_questions": 1977,
      "avg_cost_per_task_usd": 0.06508963581183612,
      "avg_duration_per_task_ms": 539075.6282245826,
      "avg_output_tokens_per_task": 36246.24228629236,
      "last_run_timestamp": "2026-10-31T16:05:14.632Z"
    },
    {
      "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": 0.8928929372738239,
      "accuracy_std_dev": 0.015585617718297187,
      "run_count": 17,
      "total_questions": 3316,
      "avg_cost_per_task_usd": 0.014895325693606755,
      "avg_duration_per_task_ms": 187175.30186972255,
      "avg_output_tokens_per_task": 21447.96381182147,
      "last_run_timestamp": "2026-10-31T08:57:07.688Z"
    },
    {
      "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": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Novita",
      "accuracy": 0.892817,
      "accuracy_std_dev": 0.00915257617650169,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.007166784511784513,
      "avg_duration_per_task_ms": 121588.36531986532,
      "avg_output_tokens_per_task": 11764.878787878788,
      "last_run_timestamp": "2026-10-31T08:24:44.390Z"
    },
    {
      "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": 0.892817,
      "accuracy_std_dev": 0.0067805750493597225,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.11382627104377104,
      "avg_duration_per_task_ms": 836209.1372053872,
      "avg_output_tokens_per_task": 45359.85101010101,
      "last_run_timestamp": "2026-10-31T09:36:23.668Z"
    },
    {
      "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": 0.8923205,
      "accuracy_std_dev": 0.00915553451470747,
      "run_count": 26,
      "total_questions": 5148,
      "avg_cost_per_task_usd": 0.24410277389277388,
      "avg_duration_per_task_ms": 85496.22163947164,
      "avg_output_tokens_per_task": 9073.098096348096,
      "last_run_timestamp": "2026-10-31T08:12:47.061Z"
    },
    {
      "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": "anthropic/claude-4.7-opus-20260416",
      "model_slug": "anthropic/claude-opus-4.7",
      "model_name": "Anthropic: Claude Opus 4.7",
      "provider_name": "Anthropic",
      "accuracy": 0.8918068666666666,
      "accuracy_std_dev": 0.011242368859907686,
      "run_count": 15,
      "total_questions": 2970,
      "avg_cost_per_task_usd": 0.20911976700336699,
      "avg_duration_per_task_ms": 100253.8824915825,
      "avg_output_tokens_per_task": 8136.723905723906,
      "last_run_timestamp": "2026-10-31T08:39:17.193Z"
    },
    {
      "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": 0.891800684137931,
      "accuracy_std_dev": 0.016199095521633556,
      "run_count": 11,
      "total_questions": 2175,
      "avg_cost_per_task_usd": 0.14719310068965516,
      "avg_duration_per_task_ms": 655158.9301149425,
      "avg_output_tokens_per_task": 37074.59586206896,
      "last_run_timestamp": "2026-10-31T08:49:49.384Z"
    },
    {
      "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-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "OpenInference",
      "accuracy": 0.8916946666666665,
      "accuracy_std_dev": 0.009944030347678707,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.011122356902356903,
      "avg_duration_per_task_ms": 302424.50673400675,
      "avg_output_tokens_per_task": 25444.905723905726,
      "last_run_timestamp": "2026-10-31T09:01:06.311Z"
    },
    {
      "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": 0.8913204444444445,
      "accuracy_std_dev": 0.028427403794035355,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.10673178563411896,
      "avg_duration_per_task_ms": 440642.3956228956,
      "avg_output_tokens_per_task": 36504.574635241304,
      "last_run_timestamp": "2026-10-31T16:10:36.257Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Parasail",
      "accuracy": 0.891312813400857,
      "accuracy_std_dev": 0.01013985402366827,
      "run_count": 13,
      "total_questions": 2567,
      "avg_cost_per_task_usd": 0.007902175301908844,
      "avg_duration_per_task_ms": 334153.24970783014,
      "avg_output_tokens_per_task": 27876.923646279705,
      "last_run_timestamp": "2026-10-31T08:16:11.302Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.8910532857142858,
      "accuracy_std_dev": 0.008331061289180467,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.02299151948051948,
      "avg_duration_per_task_ms": 246585.4624819625,
      "avg_output_tokens_per_task": 30738.57647907648,
      "last_run_timestamp": "2026-10-31T08:09:30.604Z"
    },
    {
      "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": "anthropic/claude-opus-5-20260723",
      "model_slug": "anthropic/claude-opus-5",
      "model_name": "Anthropic: Claude Opus 5",
      "provider_name": "Google",
      "accuracy": 0.8906571072375901,
      "accuracy_std_dev": 0.009177998565046731,
      "run_count": 40,
      "total_questions": 7917,
      "avg_cost_per_task_usd": 0.13327076518883413,
      "avg_duration_per_task_ms": 59327.010610079575,
      "avg_output_tokens_per_task": 4760.107869142352,
      "last_run_timestamp": "2026-10-31T08:46:09.151Z"
    },
    {
      "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-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "GMICloud",
      "accuracy": 0.8899710714285712,
      "accuracy_std_dev": 0.03484950356372031,
      "run_count": 14,
      "total_questions": 2772,
      "avg_cost_per_task_usd": 0.11476333874458873,
      "avg_duration_per_task_ms": 499182.5613275613,
      "avg_output_tokens_per_task": 36700.38997113997,
      "last_run_timestamp": "2026-10-31T08:09:42.543Z"
    },
    {
      "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": 0.889876224111283,
      "accuracy_std_dev": 0.005998909101508626,
      "run_count": 7,
      "total_questions": 1294,
      "avg_cost_per_task_usd": 0.030587522411128287,
      "avg_duration_per_task_ms": 772887.7256568779,
      "avg_output_tokens_per_task": 36034.0115919629,
      "last_run_timestamp": "2026-10-31T08:09:29.858Z"
    },
    {
      "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": 0.8895201249999999,
      "accuracy_std_dev": 0.013120057749849085,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.10712794823232324,
      "avg_duration_per_task_ms": 473198.4646464646,
      "avg_output_tokens_per_task": 37189.18434343435,
      "last_run_timestamp": "2026-10-31T08:45:16.520Z"
    },
    {
      "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": 0.8894587960168979,
      "accuracy_std_dev": 0.02138212178317713,
      "run_count": 9,
      "total_questions": 1657,
      "avg_cost_per_task_usd": 0.15853517501508751,
      "avg_duration_per_task_ms": 684130.0814725407,
      "avg_output_tokens_per_task": 50418.83343391672,
      "last_run_timestamp": "2026-10-31T16:34:03.928Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-fable-5.1-20260831",
      "model_slug": "anthropic/claude-fable-5.1",
      "model_name": "Anthropic: Claude Fable 5.1",
      "provider_name": "Anthropic",
      "accuracy": 0.8894498333333334,
      "accuracy_std_dev": 0.02853141346198763,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.136280962962963,
      "avg_duration_per_task_ms": 40331.19781144781,
      "avg_output_tokens_per_task": 2516.2390572390573,
      "last_run_timestamp": "2026-10-01T08:24:17.840Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-luna-pro-20260922",
      "model_slug": "openai/gpt-6-luna-pro",
      "model_name": "OpenAI: GPT-6 Luna Pro",
      "provider_name": "Azure",
      "accuracy": 0.8894498333333334,
      "accuracy_std_dev": 0.009896244159893304,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.008765856902356902,
      "avg_duration_per_task_ms": 59664.589225589225,
      "avg_output_tokens_per_task": 12901.691919191919,
      "last_run_timestamp": "2026-10-01T08:29:53.989Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Reka",
      "accuracy": 0.8891708512988952,
      "accuracy_std_dev": 0.008107170901595804,
      "run_count": 17,
      "total_questions": 3349,
      "avg_cost_per_task_usd": 0.015164630934607346,
      "avg_duration_per_task_ms": 200146.0895789788,
      "avg_output_tokens_per_task": 27111.550313526426,
      "last_run_timestamp": "2026-10-31T09:08:32.372Z"
    },
    {
      "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": 0.8891478461538462,
      "accuracy_std_dev": 0.008503682775454532,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.20683341996891996,
      "avg_duration_per_task_ms": 100040.29953379954,
      "avg_output_tokens_per_task": 8045.402097902098,
      "last_run_timestamp": "2026-10-31T09:23:04.665Z"
    },
    {
      "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-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "InferenceNet",
      "accuracy": 0.8888027820710974,
      "accuracy_std_dev": 0.021558691528940226,
      "run_count": 10,
      "total_questions": 1941,
      "avg_cost_per_task_usd": 0.019635562596599692,
      "avg_duration_per_task_ms": 258062.31633178773,
      "avg_output_tokens_per_task": 30486.462132921173,
      "last_run_timestamp": "2026-10-31T08:42:01.224Z"
    },
    {
      "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": 0.8887267478134111,
      "accuracy_std_dev": 0.05987639223737212,
      "run_count": 22,
      "total_questions": 4116,
      "avg_cost_per_task_usd": 0.008590749028182704,
      "avg_duration_per_task_ms": 154650.31656948494,
      "avg_output_tokens_per_task": 24587.18537414966,
      "last_run_timestamp": "2026-10-31T09:24:59.899Z"
    },
    {
      "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": 0.8884226074592075,
      "accuracy_std_dev": 0.027668835320627423,
      "run_count": 11,
      "total_questions": 2145,
      "avg_cost_per_task_usd": 0.008902693706293706,
      "avg_duration_per_task_ms": 299088.97715617716,
      "avg_output_tokens_per_task": 27126.232167832168,
      "last_run_timestamp": "2026-10-31T08:51:08.284Z"
    },
    {
      "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": 0.8881691813471503,
      "accuracy_std_dev": 0.0030120000000000147,
      "run_count": 2,
      "total_questions": 386,
      "avg_cost_per_task_usd": 0.1504010103626943,
      "avg_duration_per_task_ms": 651195.4274611399,
      "avg_output_tokens_per_task": 42839.41709844559,
      "last_run_timestamp": "2026-10-31T16:08:08.539Z"
    },
    {
      "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": 0.8880469999999999,
      "accuracy_std_dev": 0.011198752698403515,
      "run_count": 14,
      "total_questions": 2772,
      "avg_cost_per_task_usd": 0.025298165945165943,
      "avg_duration_per_task_ms": 331181.7139249639,
      "avg_output_tokens_per_task": 26439.20562770563,
      "last_run_timestamp": "2026-10-31T09:11:16.308Z"
    },
    {
      "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-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8878064999999999,
      "accuracy_std_dev": 0.018774829060473484,
      "run_count": 14,
      "total_questions": 2772,
      "avg_cost_per_task_usd": 0.028150706349206345,
      "avg_duration_per_task_ms": 209828.97294372294,
      "avg_output_tokens_per_task": 25892.321789321788,
      "last_run_timestamp": "2026-10-31T08:53:51.020Z"
    },
    {
      "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": 0.8876718896377422,
      "accuracy_std_dev": 0.0038308239938977078,
      "run_count": 6,
      "total_questions": 1187,
      "avg_cost_per_task_usd": 0.032938989890480205,
      "avg_duration_per_task_ms": 750389.2257792755,
      "avg_output_tokens_per_task": 38026.47177759057,
      "last_run_timestamp": "2026-10-31T08:03:09.435Z"
    },
    {
      "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": 0.8875500348763478,
      "accuracy_std_dev": 0.013802209238356553,
      "run_count": 8,
      "total_questions": 1577,
      "avg_cost_per_task_usd": 0.1294895707038681,
      "avg_duration_per_task_ms": 1720933.2999365884,
      "avg_output_tokens_per_task": 32195.95688015219,
      "last_run_timestamp": "2026-10-31T16:35:08.639Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "NextBit",
      "accuracy": 0.887537924012158,
      "accuracy_std_dev": 0.020250680116973837,
      "run_count": 5,
      "total_questions": 987,
      "avg_cost_per_task_usd": 0.12033997061803446,
      "avg_duration_per_task_ms": 580755.4113475177,
      "avg_output_tokens_per_task": 34357.5663627153,
      "last_run_timestamp": "2026-10-03T16:44:56.199Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-fable-5.1-20260831",
      "model_slug": "anthropic/claude-fable-5.1",
      "model_name": "Anthropic: Claude Fable 5.1",
      "provider_name": "Azure",
      "accuracy": 0.8872051666666666,
      "accuracy_std_dev": 0.024761493125527695,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.1434989882154882,
      "avg_duration_per_task_ms": 42121.643939393936,
      "avg_output_tokens_per_task": 2660.465488215488,
      "last_run_timestamp": "2026-10-01T08:24:34.535Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Relace",
      "accuracy": 0.8871015424486148,
      "accuracy_std_dev": 0.007433711649565692,
      "run_count": 6,
      "total_questions": 1119,
      "avg_cost_per_task_usd": 0.100128890974084,
      "avg_duration_per_task_ms": 379563.4110813226,
      "avg_output_tokens_per_task": 36832.25379803396,
      "last_run_timestamp": "2026-10-31T16:50:08.743Z"
    },
    {
      "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-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-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Nebius",
      "accuracy": 0.8865318000000001,
      "accuracy_std_dev": 0.007110736653821441,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.030907244444444443,
      "avg_duration_per_task_ms": 199105.52626262626,
      "avg_output_tokens_per_task": 25534.58484848485,
      "last_run_timestamp": "2026-10-31T08:28:17.984Z"
    },
    {
      "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": 0.8864282307692305,
      "accuracy_std_dev": 0.014063708976564978,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.11727622494172495,
      "avg_duration_per_task_ms": 585593.1717171717,
      "avg_output_tokens_per_task": 30597.257575757576,
      "last_run_timestamp": "2026-10-31T16:23:41.134Z"
    },
    {
      "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": "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": 0.886083,
      "accuracy_std_dev": 0.012320026406898109,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.022550329966329965,
      "avg_duration_per_task_ms": 149365.32154882155,
      "avg_output_tokens_per_task": 44803.96127946128,
      "last_run_timestamp": "2026-10-31T08:13:10.777Z"
    },
    {
      "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": 0.8856841481205807,
      "accuracy_std_dev": 0.008667673595090004,
      "run_count": 14,
      "total_questions": 2687,
      "avg_cost_per_task_usd": 0.09660396911053219,
      "avg_duration_per_task_ms": 451439.94491998514,
      "avg_output_tokens_per_task": 28785.028656494233,
      "last_run_timestamp": "2026-10-31T16:12:28.231Z"
    },
    {
      "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": "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": "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": 0.8855216666666669,
      "accuracy_std_dev": 0.0023805928299947096,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.016851747474747474,
      "avg_duration_per_task_ms": 62951.132996632994,
      "avg_output_tokens_per_task": 13874.079124579124,
      "last_run_timestamp": "2026-10-31T08:03:10.178Z"
    },
    {
      "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": "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-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Nebius",
      "accuracy": 0.8846799999999999,
      "accuracy_std_dev": 0.021044000000000007,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.08896683838383838,
      "avg_duration_per_task_ms": 416558.4267676768,
      "avg_output_tokens_per_task": 22204.20202020202,
      "last_run_timestamp": "2026-10-31T08:57:00.457Z"
    },
    {
      "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": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8842123333333334,
      "accuracy_std_dev": 0.018675407929502712,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.008158511784511784,
      "avg_duration_per_task_ms": 397047.721661055,
      "avg_output_tokens_per_task": 28750.622895622895,
      "last_run_timestamp": "2026-10-31T16:48:29.809Z"
    },
    {
      "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": 0.8840456666666667,
      "accuracy_std_dev": 0.005144110505120299,
      "run_count": 3,
      "total_questions": 585,
      "avg_cost_per_task_usd": 0.02067068717948718,
      "avg_duration_per_task_ms": 91684.32991452991,
      "avg_output_tokens_per_task": 41136.99487179487,
      "last_run_timestamp": "2026-10-31T08:35:13.435Z"
    },
    {
      "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": 0.8840252222222222,
      "accuracy_std_dev": 0.02805224637777707,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.19950497250280583,
      "avg_duration_per_task_ms": 1012777.083613917,
      "avg_output_tokens_per_task": 58973.21717171717,
      "last_run_timestamp": "2026-10-31T16:24:42.659Z"
    },
    {
      "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": 0.8839239383667181,
      "accuracy_std_dev": 0.011423232412500406,
      "run_count": 8,
      "total_questions": 1298,
      "avg_cost_per_task_usd": 0.00878981278890601,
      "avg_duration_per_task_ms": 844366.6594761171,
      "avg_output_tokens_per_task": 29702.687981510015,
      "last_run_timestamp": "2026-10-31T08:34:49.723Z"
    },
    {
      "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": 0,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.014708931818181817,
      "avg_duration_per_task_ms": 71351.09343434343,
      "avg_output_tokens_per_task": 13550.636363636364,
      "last_run_timestamp": "2026-10-31T08:27:52.629Z"
    },
    {
      "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": 0.8836312915601023,
      "accuracy_std_dev": 0.003620000000000012,
      "run_count": 2,
      "total_questions": 391,
      "avg_cost_per_task_usd": 0.006277416879795397,
      "avg_duration_per_task_ms": 104811.05115089515,
      "avg_output_tokens_per_task": 10905.734015345268,
      "last_run_timestamp": "2026-10-31T08:31:24.328Z"
    },
    {
      "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": 0.8832547529880477,
      "accuracy_std_dev": 0.048459311911738724,
      "run_count": 14,
      "total_questions": 2761,
      "avg_cost_per_task_usd": 0.07450339333574792,
      "avg_duration_per_task_ms": 577739.1611734879,
      "avg_output_tokens_per_task": 39363.77290836653,
      "last_run_timestamp": "2026-10-31T16:51:33.999Z"
    },
    {
      "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": 0.8831221012658229,
      "accuracy_std_dev": 0.01342904913771097,
      "run_count": 12,
      "total_questions": 2370,
      "avg_cost_per_task_usd": 0.1860454797468354,
      "avg_duration_per_task_ms": 320097.15611814347,
      "avg_output_tokens_per_task": 46697.94894514768,
      "last_run_timestamp": "2026-10-31T08:25:00.286Z"
    },
    {
      "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": 0.8829965,
      "accuracy_std_dev": 0.002525500000000014,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.006819977272727273,
      "avg_duration_per_task_ms": 114306.47727272728,
      "avg_output_tokens_per_task": 11182.35606060606,
      "last_run_timestamp": "2026-10-31T08:07:19.817Z"
    },
    {
      "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": 0.8829290000000001,
      "accuracy_std_dev": 0.012699110785799135,
      "run_count": 4,
      "total_questions": 783,
      "avg_cost_per_task_usd": 0.027103435504469985,
      "avg_duration_per_task_ms": 200899.3601532567,
      "avg_output_tokens_per_task": 30539.469987228607,
      "last_run_timestamp": "2026-10-31T08:15:03.390Z"
    },
    {
      "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": 0.8827922637511884,
      "accuracy_std_dev": 0.009496956336599638,
      "run_count": 38,
      "total_questions": 7363,
      "avg_cost_per_task_usd": 0.1404521245416271,
      "avg_duration_per_task_ms": 63191.223414369146,
      "avg_output_tokens_per_task": 5059.606817873149,
      "last_run_timestamp": "2026-10-31T08:31:11.082Z"
    },
    {
      "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": 0.8825948758169935,
      "accuracy_std_dev": 0.010166708210464918,
      "run_count": 14,
      "total_questions": 2754,
      "avg_cost_per_task_usd": 0.007344207697893974,
      "avg_duration_per_task_ms": 375236.5925925926,
      "avg_output_tokens_per_task": 20279.856209150326,
      "last_run_timestamp": "2026-10-31T08:07:18.580Z"
    },
    {
      "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": 0.882365125,
      "accuracy_std_dev": 0.0043279540904883765,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.029606650252525252,
      "avg_duration_per_task_ms": 78746.81123737374,
      "avg_output_tokens_per_task": 24519.371843434343,
      "last_run_timestamp": "2026-10-31T09:02:59.407Z"
    },
    {
      "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": 0.8821789256756756,
      "accuracy_std_dev": 0.01706691482241463,
      "run_count": 6,
      "total_questions": 1184,
      "avg_cost_per_task_usd": 0.04157157263513513,
      "avg_duration_per_task_ms": 164106.9873310811,
      "avg_output_tokens_per_task": 34476.61317567567,
      "last_run_timestamp": "2026-10-31T08:31:11.691Z"
    },
    {
      "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-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Reka",
      "accuracy": 0.881818,
      "accuracy_std_dev": 0.03925680632450885,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.010723755555555554,
      "avg_duration_per_task_ms": 145330.13333333333,
      "avg_output_tokens_per_task": 16217.381818181819,
      "last_run_timestamp": "2026-10-31T09:01:13.864Z"
    },
    {
      "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": 0.8817657344384302,
      "accuracy_std_dev": 0.02198561693228856,
      "run_count": 15,
      "total_questions": 2956,
      "avg_cost_per_task_usd": 0.09460041407307172,
      "avg_duration_per_task_ms": 453932.54093369417,
      "avg_output_tokens_per_task": 33735.36434370771,
      "last_run_timestamp": "2026-10-31T16:08:08.538Z"
    },
    {
      "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": 0.8817339999999999,
      "accuracy_std_dev": 0.016681744617994862,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.008588069444444443,
      "avg_duration_per_task_ms": 150888.15656565657,
      "avg_output_tokens_per_task": 16172.003787878788,
      "last_run_timestamp": "2026-10-31T08:07:19.698Z"
    },
    {
      "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": 0.881655907348243,
      "accuracy_std_dev": 0.03261050414332772,
      "run_count": 19,
      "total_questions": 3756,
      "avg_cost_per_task_usd": 0.14437988578274757,
      "avg_duration_per_task_ms": 367673.1634717785,
      "avg_output_tokens_per_task": 10849.925452609159,
      "last_run_timestamp": "2026-10-31T09:14:35.378Z"
    },
    {
      "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": "StreamLake",
      "accuracy": 0.8814812,
      "accuracy_std_dev": 0.011112816535874249,
      "run_count": 15,
      "total_questions": 2970,
      "avg_cost_per_task_usd": 0.006134084511784513,
      "avg_duration_per_task_ms": 402598.8686868687,
      "avg_output_tokens_per_task": 29715.115824915825,
      "last_run_timestamp": "2026-10-31T08:19:56.454Z"
    },
    {
      "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": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Crusoe",
      "accuracy": 0.8810325,
      "accuracy_std_dev": 0.016020429392393534,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.03953295538720539,
      "avg_duration_per_task_ms": 203574.52356902356,
      "avg_output_tokens_per_task": 32800.2398989899,
      "last_run_timestamp": "2026-10-31T08:47:44.999Z"
    },
    {
      "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": 0.8809893076923078,
      "accuracy_std_dev": 0.018410802353236562,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.12893885664335664,
      "avg_duration_per_task_ms": 63425.74864024864,
      "avg_output_tokens_per_task": 4924.540792540793,
      "last_run_timestamp": "2026-10-31T08:31:10.506Z"
    },
    {
      "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": 0.8808676325443787,
      "accuracy_std_dev": 0.011642004324486309,
      "run_count": 9,
      "total_questions": 1690,
      "avg_cost_per_task_usd": 0.06728608106508875,
      "avg_duration_per_task_ms": 317431.9201183432,
      "avg_output_tokens_per_task": 18592.419526627218,
      "last_run_timestamp": "2026-10-31T16:03:21.450Z"
    },
    {
      "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": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Cloudflare",
      "accuracy": 0.880186085003036,
      "accuracy_std_dev": 0.03105021938412785,
      "run_count": 9,
      "total_questions": 1647,
      "avg_cost_per_task_usd": 0.204991802671524,
      "avg_duration_per_task_ms": 1379722.3017607771,
      "avg_output_tokens_per_task": 51102.45476624165,
      "last_run_timestamp": "2026-10-31T16:04:29.120Z"
    },
    {
      "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": 0.8800631670886074,
      "accuracy_std_dev": 0.006743341790240229,
      "run_count": 8,
      "total_questions": 1580,
      "avg_cost_per_task_usd": 0.033778830379746834,
      "avg_duration_per_task_ms": 152045.89430379748,
      "avg_output_tokens_per_task": 27961.480379746834,
      "last_run_timestamp": "2026-10-31T09:42:46.430Z"
    },
    {
      "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": 0.8800035555555555,
      "accuracy_std_dev": 0.060993454403113545,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.13239759820426486,
      "avg_duration_per_task_ms": 1665151.1414141415,
      "avg_output_tokens_per_task": 55809.35241301908,
      "last_run_timestamp": "2026-10-31T16:13:09.866Z"
    },
    {
      "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": 0.8799533076923076,
      "accuracy_std_dev": 0.01785792442666288,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.1271698951048951,
      "avg_duration_per_task_ms": 62325.81235431235,
      "avg_output_tokens_per_task": 4863.158508158508,
      "last_run_timestamp": "2026-10-31T08:12:45.681Z"
    },
    {
      "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": 0.879872690530519,
      "accuracy_std_dev": 0.053065810239372686,
      "run_count": 18,
      "total_questions": 3506,
      "avg_cost_per_task_usd": 0.00976165202509983,
      "avg_duration_per_task_ms": 414485.4175698802,
      "avg_output_tokens_per_task": 26831.222760981174,
      "last_run_timestamp": "2026-10-31T08:47:06.377Z"
    },
    {
      "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": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Venice",
      "accuracy": 0.87984,
      "accuracy_std_dev": 0.018052164925570553,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.036141941919191914,
      "avg_duration_per_task_ms": 139656.41224747474,
      "avg_output_tokens_per_task": 28081.721590909092,
      "last_run_timestamp": "2026-10-31T09:14:35.460Z"
    },
    {
      "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": 0.879749096574958,
      "accuracy_std_dev": 0.006147652748894318,
      "run_count": 9,
      "total_questions": 1781,
      "avg_cost_per_task_usd": 0.017198152723189218,
      "avg_duration_per_task_ms": 289110.672094329,
      "avg_output_tokens_per_task": 31088.430656934306,
      "last_run_timestamp": "2026-10-31T08:46:18.716Z"
    },
    {
      "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": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Mancer 2",
      "accuracy": 0.879564846153846,
      "accuracy_std_dev": 0.03143742775944268,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.012390225330225331,
      "avg_duration_per_task_ms": 350248.7618492618,
      "avg_output_tokens_per_task": 22401.122377622378,
      "last_run_timestamp": "2026-10-31T08:24:31.558Z"
    },
    {
      "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": 0.8793102638453499,
      "accuracy_std_dev": 0.020773557275777312,
      "run_count": 10,
      "total_questions": 1914,
      "avg_cost_per_task_usd": 0.09333590334378267,
      "avg_duration_per_task_ms": 520471.578892372,
      "avg_output_tokens_per_task": 28040.05590386625,
      "last_run_timestamp": "2026-10-31T08:31:12.870Z"
    },
    {
      "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": 0.8791213851308232,
      "accuracy_std_dev": 0.03905943528044792,
      "run_count": 20,
      "total_questions": 3134,
      "avg_cost_per_task_usd": 0.02867042214422464,
      "avg_duration_per_task_ms": 357853.9288449266,
      "avg_output_tokens_per_task": 25473.93746011487,
      "last_run_timestamp": "2026-10-31T08:10:35.431Z"
    },
    {
      "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": "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": "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": 0.8782263333333332,
      "accuracy_std_dev": 0.004199386807089298,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.020176612794612796,
      "avg_duration_per_task_ms": 120195.73063973064,
      "avg_output_tokens_per_task": 42971.59090909091,
      "last_run_timestamp": "2026-10-31T08:09:30.418Z"
    },
    {
      "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": "Parasail",
      "accuracy": 0.8782088436213992,
      "accuracy_std_dev": 0.01624543331919451,
      "run_count": 9,
      "total_questions": 1701,
      "avg_cost_per_task_usd": 0.18331693709582594,
      "avg_duration_per_task_ms": 766864.758377425,
      "avg_output_tokens_per_task": 52771.66372721928,
      "last_run_timestamp": "2026-10-31T16:35:50.033Z"
    },
    {
      "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": 0.8781144000000001,
      "accuracy_std_dev": 0.016133498487308946,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.023055602525252527,
      "avg_duration_per_task_ms": 719094.7823232323,
      "avg_output_tokens_per_task": 35705.98131313131,
      "last_run_timestamp": "2026-10-31T16:46:00.416Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Baidu",
      "accuracy": 0.8778525555555555,
      "accuracy_std_dev": 0.02507239829024519,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.13925247306397306,
      "avg_duration_per_task_ms": 898484.416386083,
      "avg_output_tokens_per_task": 61943.26487093154,
      "last_run_timestamp": "2026-10-31T16:04:29.123Z"
    },
    {
      "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": 0.8777055714285714,
      "accuracy_std_dev": 0.04371368642773152,
      "run_count": 14,
      "total_questions": 2772,
      "avg_cost_per_task_usd": 0.09160316883116884,
      "avg_duration_per_task_ms": 609338.455988456,
      "avg_output_tokens_per_task": 40699.61507936508,
      "last_run_timestamp": "2026-10-31T16:16:10.979Z"
    },
    {
      "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": 0.8774956609137057,
      "accuracy_std_dev": 0.014904254367126186,
      "run_count": 15,
      "total_questions": 2955,
      "avg_cost_per_task_usd": 0.10885685989847717,
      "avg_duration_per_task_ms": 463191.3489001692,
      "avg_output_tokens_per_task": 31188.99323181049,
      "last_run_timestamp": "2026-10-31T16:02:14.362Z"
    },
    {
      "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": 0.8774795484247374,
      "accuracy_std_dev": 0.0070708188238871586,
      "run_count": 13,
      "total_questions": 2571,
      "avg_cost_per_task_usd": 0.11807617502917153,
      "avg_duration_per_task_ms": 759991.2372617659,
      "avg_output_tokens_per_task": 30946.45468689226,
      "last_run_timestamp": "2026-10-31T16:19:28.502Z"
    },
    {
      "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": "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-k2.7-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-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": "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": 0.8763216813793104,
      "accuracy_std_dev": 0.01096372812565248,
      "run_count": 11,
      "total_questions": 2175,
      "avg_cost_per_task_usd": 0.0099918183908046,
      "avg_duration_per_task_ms": 425649.0579310345,
      "avg_output_tokens_per_task": 22880.83264367816,
      "last_run_timestamp": "2026-10-31T08:32:56.980Z"
    },
    {
      "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": 0.8762625,
      "accuracy_std_dev": 0.009259500000000032,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.014013469696969696,
      "avg_duration_per_task_ms": 171317.10353535353,
      "avg_output_tokens_per_task": 26009.79292929293,
      "last_run_timestamp": "2026-10-31T08:09:25.873Z"
    },
    {
      "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": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Decart",
      "accuracy": 0.8759833172804533,
      "accuracy_std_dev": 0.02383166285579381,
      "run_count": 6,
      "total_questions": 1059,
      "avg_cost_per_task_usd": 0.006660402266288952,
      "avg_duration_per_task_ms": 451840.41265344666,
      "avg_output_tokens_per_task": 36610.579792256845,
      "last_run_timestamp": "2026-10-31T08:15:03.294Z"
    },
    {
      "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": 0.8759819999999999,
      "accuracy_std_dev": 0.04446362050785635,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.019776964646464647,
      "avg_duration_per_task_ms": 142117.47643097644,
      "avg_output_tokens_per_task": 20178.936026936026,
      "last_run_timestamp": "2026-10-31T09:01:06.393Z"
    },
    {
      "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": 0.875981888888889,
      "accuracy_std_dev": 0.0056117777821777124,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.08140571212121211,
      "avg_duration_per_task_ms": 407227.8625140292,
      "avg_output_tokens_per_task": 23146.814814814814,
      "last_run_timestamp": "2026-10-31T16:15:54.664Z"
    },
    {
      "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": 0.8756311250000001,
      "accuracy_std_dev": 0.012722348362600958,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.03608339141414141,
      "avg_duration_per_task_ms": 154494.60795454544,
      "avg_output_tokens_per_task": 29914.79419191919,
      "last_run_timestamp": "2026-10-31T08:55:44.046Z"
    },
    {
      "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": 0.8755709520547945,
      "accuracy_std_dev": 0.007533567739715949,
      "run_count": 3,
      "total_questions": 584,
      "avg_cost_per_task_usd": 0.031521625,
      "avg_duration_per_task_ms": 247947.07705479453,
      "avg_output_tokens_per_task": 62802.890410958906,
      "last_run_timestamp": "2026-10-31T08:14:43.030Z"
    },
    {
      "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": 0.875421,
      "accuracy_std_dev": 0.010101000000000027,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.05678013888888889,
      "avg_duration_per_task_ms": 453449.72222222225,
      "avg_output_tokens_per_task": 62821.7904040404,
      "last_run_timestamp": "2026-10-31T08:23:57.761Z"
    },
    {
      "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": 0.8754206666666668,
      "accuracy_std_dev": 0.004955863351178644,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.01787970538720539,
      "avg_duration_per_task_ms": 181946.53367003368,
      "avg_output_tokens_per_task": 29504.771043771045,
      "last_run_timestamp": "2026-10-31T08:09:31.246Z"
    },
    {
      "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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Venice",
      "accuracy": 0.8753238600311043,
      "accuracy_std_dev": 0.02970275849959406,
      "run_count": 13,
      "total_questions": 2572,
      "avg_cost_per_task_usd": 0.17156172589424573,
      "avg_duration_per_task_ms": 364836.6096423017,
      "avg_output_tokens_per_task": 39108.15202177294,
      "last_run_timestamp": "2026-10-31T16:32:36.327Z"
    },
    {
      "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": 0.8753084,
      "accuracy_std_dev": 0.016963206932652818,
      "run_count": 15,
      "total_questions": 2970,
      "avg_cost_per_task_usd": 0.008280871043771043,
      "avg_duration_per_task_ms": 217247.1521885522,
      "avg_output_tokens_per_task": 29293.272727272728,
      "last_run_timestamp": "2026-10-31T08:31:11.360Z"
    },
    {
      "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": 0.8751679521169355,
      "accuracy_std_dev": 0.029185533960542576,
      "run_count": 11,
      "total_questions": 1984,
      "avg_cost_per_task_usd": 0.027124963205645166,
      "avg_duration_per_task_ms": 364845.5393145161,
      "avg_output_tokens_per_task": 25176.949596774193,
      "last_run_timestamp": "2026-10-31T16:15:26.636Z"
    },
    {
      "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": "CoreWeave",
      "accuracy": 0.8746724444444444,
      "accuracy_std_dev": 0.019889553484240645,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.20093631930415262,
      "avg_duration_per_task_ms": 573135.2710437711,
      "avg_output_tokens_per_task": 58795.35746352413,
      "last_run_timestamp": "2026-10-31T16:09:16.278Z"
    },
    {
      "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": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Cloudflare",
      "accuracy": 0.874560281359906,
      "accuracy_std_dev": 0.03491023586403909,
      "run_count": 13,
      "total_questions": 2559,
      "avg_cost_per_task_usd": 0.10398778819851505,
      "avg_duration_per_task_ms": 633649.7999218445,
      "avg_output_tokens_per_task": 25844.287612348573,
      "last_run_timestamp": "2026-10-31T16:08:08.536Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Makora",
      "accuracy": 0.8745317312013828,
      "accuracy_std_dev": 0.011186512454241005,
      "run_count": 6,
      "total_questions": 1157,
      "avg_cost_per_task_usd": 0.0043274416594641305,
      "avg_duration_per_task_ms": 185028.52031114954,
      "avg_output_tokens_per_task": 21870.51080380294,
      "last_run_timestamp": "2026-09-26T08:53:59.525Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "auto-routing",
      "accuracy": 0.8744105999999999,
      "accuracy_std_dev": 0.014149500063253112,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.09050255303030302,
      "avg_duration_per_task_ms": 196310.64444444445,
      "avg_output_tokens_per_task": 14775.43686868687,
      "last_run_timestamp": "2026-10-31T08:07:27.863Z"
    },
    {
      "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": 0.8743250351006383,
      "accuracy_std_dev": 0.026296453097608783,
      "run_count": 21,
      "total_questions": 4074,
      "avg_cost_per_task_usd": 0.03975532817869414,
      "avg_duration_per_task_ms": 80921.18360333824,
      "avg_output_tokens_per_task": 25880.549828178693,
      "last_run_timestamp": "2026-10-31T09:03:20.938Z"
    },
    {
      "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": 0.8742983333333334,
      "accuracy_std_dev": 0.0055553002519115975,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.02925406565656566,
      "avg_duration_per_task_ms": 153006.37373737374,
      "avg_output_tokens_per_task": 28825.781144781144,
      "last_run_timestamp": "2026-10-31T08:50:42.011Z"
    },
    {
      "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": 0.8740494840344654,
      "accuracy_std_dev": 0.0038017229265163467,
      "run_count": 10,
      "total_questions": 1973,
      "avg_cost_per_task_usd": 0.04434169640141917,
      "avg_duration_per_task_ms": 202197.7790167258,
      "avg_output_tokens_per_task": 38209.482513938165,
      "last_run_timestamp": "2026-10-31T08:10:37.345Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Together",
      "accuracy": 0.8740044607508533,
      "accuracy_std_dev": 0.007605951280996289,
      "run_count": 3,
      "total_questions": 586,
      "avg_cost_per_task_usd": 0.10401818771331058,
      "avg_duration_per_task_ms": 459122.319112628,
      "avg_output_tokens_per_task": 30022.856655290103,
      "last_run_timestamp": "2026-08-22T16:49:15.579Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Together",
      "accuracy": 0.8737693295358652,
      "accuracy_std_dev": 0.014113612302391693,
      "run_count": 12,
      "total_questions": 2370,
      "avg_cost_per_task_usd": 0.09247935569620254,
      "avg_duration_per_task_ms": 203634.65358649788,
      "avg_output_tokens_per_task": 23372.77805907173,
      "last_run_timestamp": "2026-10-31T08:40:26.156Z"
    },
    {
      "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": 0.873737,
      "accuracy_std_dev": 0.007273551310513108,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.01830383333333333,
      "avg_duration_per_task_ms": 381305.08585858584,
      "avg_output_tokens_per_task": 30305.045454545456,
      "last_run_timestamp": "2026-10-31T08:27:46.342Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-luna-20260922",
      "model_slug": "openai/gpt-6-luna",
      "model_name": "OpenAI: GPT-6 Luna",
      "provider_name": "auto-routing",
      "accuracy": 0.873737,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0029526111111111116,
      "avg_duration_per_task_ms": 49722.71717171717,
      "avg_output_tokens_per_task": 5746.707070707071,
      "last_run_timestamp": "2026-10-01T08:23:37.659Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.7-plus-20260602",
      "model_slug": "qwen/qwen3.7-plus",
      "model_name": "Qwen: Qwen3.7 Plus",
      "provider_name": "Alibaba",
      "accuracy": 0.873737,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.045554454545454544,
      "avg_duration_per_task_ms": 605480.6616161616,
      "avg_output_tokens_per_task": 35385.22727272727,
      "last_run_timestamp": "2026-10-01T08:40:06.225Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Venice",
      "accuracy": 0.873677142857143,
      "accuracy_std_dev": 0.013379294172377723,
      "run_count": 14,
      "total_questions": 2772,
      "avg_cost_per_task_usd": 0.00609786507936508,
      "avg_duration_per_task_ms": 261742.29906204907,
      "avg_output_tokens_per_task": 18465.902236652237,
      "last_run_timestamp": "2026-10-31T08:36:26.194Z"
    },
    {
      "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": 0.8728018888888889,
      "accuracy_std_dev": 0.022448311515045904,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.11960105274971941,
      "avg_duration_per_task_ms": 614007.7817059484,
      "avg_output_tokens_per_task": 35342.742424242424,
      "last_run_timestamp": "2026-10-31T16:25:53.821Z"
    },
    {
      "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": 0.8726149999999999,
      "accuracy_std_dev": 0.0020998715833752102,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.017651676767676767,
      "avg_duration_per_task_ms": 95564.90235690236,
      "avg_output_tokens_per_task": 29293.20202020202,
      "last_run_timestamp": "2026-10-31T08:33:05.096Z"
    },
    {
      "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": 0.8722639999999998,
      "accuracy_std_dev": 0.01391911622553672,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.01943295707070707,
      "avg_duration_per_task_ms": 314686.08207070705,
      "avg_output_tokens_per_task": 35123.857954545456,
      "last_run_timestamp": "2026-10-31T08:09:31.159Z"
    },
    {
      "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": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8720325587863464,
      "accuracy_std_dev": 0.09552669457201043,
      "run_count": 12,
      "total_questions": 2373,
      "avg_cost_per_task_usd": 0.026620932996207335,
      "avg_duration_per_task_ms": 196837.85040033713,
      "avg_output_tokens_per_task": 22844.949431099874,
      "last_run_timestamp": "2026-10-31T16:04:29.094Z"
    },
    {
      "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": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Baidu",
      "accuracy": 0.8718665555555556,
      "accuracy_std_dev": 0.016936558486245815,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.03607248540965208,
      "avg_duration_per_task_ms": 502797.54545454547,
      "avg_output_tokens_per_task": 35101.41638608305,
      "last_run_timestamp": "2026-10-31T16:06:00.097Z"
    },
    {
      "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": 0.8714925000000001,
      "accuracy_std_dev": 0.007251768353305299,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.007901208754208754,
      "avg_duration_per_task_ms": 584978.2028619528,
      "avg_output_tokens_per_task": 44690.53282828283,
      "last_run_timestamp": "2026-10-31T16:48:05.771Z"
    },
    {
      "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": "anthropic/claude-5-fable-20260609",
      "model_slug": "anthropic/claude-fable-5",
      "model_name": "Anthropic: Claude Fable 5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.8712119166666668,
      "accuracy_std_dev": 0.038220154043075064,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.1817093308080808,
      "avg_duration_per_task_ms": 45516.91035353536,
      "avg_output_tokens_per_task": 3430.526515151515,
      "last_run_timestamp": "2026-10-31T08:20:29.734Z"
    },
    {
      "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": 0.8711184444444444,
      "accuracy_std_dev": 0.016592189722953073,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.04206737822671156,
      "avg_duration_per_task_ms": 668364.7727272727,
      "avg_output_tokens_per_task": 35229.93041526375,
      "last_run_timestamp": "2026-10-31T16:16:55.624Z"
    },
    {
      "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": "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": 0.8708191333333333,
      "accuracy_std_dev": 0.005011630059061523,
      "run_count": 15,
      "total_questions": 2970,
      "avg_cost_per_task_usd": 0.004255378451178452,
      "avg_duration_per_task_ms": 39486.23198653199,
      "avg_output_tokens_per_task": 7745.265656565656,
      "last_run_timestamp": "2026-10-31T08:49:45.152Z"
    },
    {
      "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": 0.8708100545730167,
      "accuracy_std_dev": 0.007904713862626544,
      "run_count": 10,
      "total_questions": 1979,
      "avg_cost_per_task_usd": 0.049469633148054575,
      "avg_duration_per_task_ms": 293502.65942395147,
      "avg_output_tokens_per_task": 21011.941384537644,
      "last_run_timestamp": "2026-10-31T16:14:42.903Z"
    },
    {
      "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": 0.8703704,
      "accuracy_std_dev": 0.00583193054142453,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.004018268686868687,
      "avg_duration_per_task_ms": 37454.91717171717,
      "avg_output_tokens_per_task": 7796.7444444444445,
      "last_run_timestamp": "2026-10-31T08:14:20.331Z"
    },
    {
      "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": "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": 0.8703605355249205,
      "accuracy_std_dev": 0.03387763198941155,
      "run_count": 10,
      "total_questions": 1886,
      "avg_cost_per_task_usd": 0.014264545599151641,
      "avg_duration_per_task_ms": 348340.7725344645,
      "avg_output_tokens_per_task": 28651.488865323437,
      "last_run_timestamp": "2026-10-31T08:28:24.326Z"
    },
    {
      "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": 0.870112333932584,
      "accuracy_std_dev": 0.07630745338957075,
      "run_count": 23,
      "total_questions": 4450,
      "avg_cost_per_task_usd": 0.21229808741573034,
      "avg_duration_per_task_ms": 491047.8352808989,
      "avg_output_tokens_per_task": 61310.42764044944,
      "last_run_timestamp": "2026-10-31T09:02:18.616Z"
    },
    {
      "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": 0.8700336,
      "accuracy_std_dev": 0.008343089106559986,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.07480730858585859,
      "avg_duration_per_task_ms": 106906.41464646465,
      "avg_output_tokens_per_task": 12161.163636363637,
      "last_run_timestamp": "2026-10-31T08:03:09.289Z"
    },
    {
      "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": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "DeepInfra",
      "accuracy": 0.8699400591341078,
      "accuracy_std_dev": 0.01681928135474285,
      "run_count": 10,
      "total_questions": 1894,
      "avg_cost_per_task_usd": 0.07171590390707498,
      "avg_duration_per_task_ms": 408524.5543822598,
      "avg_output_tokens_per_task": 23628.092397043296,
      "last_run_timestamp": "2026-10-31T08:45:16.438Z"
    },
    {
      "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": "Fireworks",
      "accuracy": 0.8697338870630417,
      "accuracy_std_dev": 0.011260986943785035,
      "run_count": 17,
      "total_questions": 3347,
      "avg_cost_per_task_usd": 0.04833613654018524,
      "avg_duration_per_task_ms": 200583.7881685091,
      "avg_output_tokens_per_task": 37254.27158649537,
      "last_run_timestamp": "2026-10-31T08:35:39.353Z"
    },
    {
      "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": 0.8697226153846153,
      "accuracy_std_dev": 0.03509376134681856,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.18790192113442114,
      "avg_duration_per_task_ms": 46125.006604506605,
      "avg_output_tokens_per_task": 3550.8018648018647,
      "last_run_timestamp": "2026-10-31T08:26:18.906Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Z.AI",
      "accuracy": 0.8695517762430939,
      "accuracy_std_dev": 0.027759107412948937,
      "run_count": 11,
      "total_questions": 2172,
      "avg_cost_per_task_usd": 0.10362995395948434,
      "avg_duration_per_task_ms": 356439.69521178637,
      "avg_output_tokens_per_task": 23309.89779005525,
      "last_run_timestamp": "2026-10-31T08:27:45.609Z"
    },
    {
      "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": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "ModelRun",
      "accuracy": 0.8695284999999998,
      "accuracy_std_dev": 0.01157224020015143,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.018078708333333332,
      "avg_duration_per_task_ms": 141813.27146464647,
      "avg_output_tokens_per_task": 29935.99873737374,
      "last_run_timestamp": "2026-10-31T08:24:37.341Z"
    },
    {
      "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": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Parasail",
      "accuracy": 0.869060888888889,
      "accuracy_std_dev": 0.00603234318660567,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.008162227272727272,
      "avg_duration_per_task_ms": 360559.57575757575,
      "avg_output_tokens_per_task": 28803.3265993266,
      "last_run_timestamp": "2026-10-31T16:18:48.433Z"
    },
    {
      "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": 0.8690236000000001,
      "accuracy_std_dev": 0.0056741455250989255,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.004051724242424242,
      "avg_duration_per_task_ms": 38666.534343434345,
      "avg_output_tokens_per_task": 7872.538383838384,
      "last_run_timestamp": "2026-10-31T08:35:55.398Z"
    },
    {
      "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": "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": "anthropic/claude-5-fable-20260609",
      "model_slug": "anthropic/claude-fable-5",
      "model_name": "Anthropic: Claude Fable 5",
      "provider_name": "auto-routing",
      "accuracy": 0.8684277692307691,
      "accuracy_std_dev": 0.04024631421662945,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.1798952365967366,
      "avg_duration_per_task_ms": 44033.45998445999,
      "avg_output_tokens_per_task": 3389.598679098679,
      "last_run_timestamp": "2026-10-31T08:13:57.565Z"
    },
    {
      "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": "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": 0.8681253333333332,
      "accuracy_std_dev": 0.00965467639137749,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.021851893939393942,
      "avg_duration_per_task_ms": 524499.1397306398,
      "avg_output_tokens_per_task": 46276.513468013465,
      "last_run_timestamp": "2026-10-31T08:03:09.464Z"
    },
    {
      "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": 0.8679098076923076,
      "accuracy_std_dev": 0.028085900492293595,
      "run_count": 26,
      "total_questions": 5148,
      "avg_cost_per_task_usd": 0.1592424807692308,
      "avg_duration_per_task_ms": 153716.81565656565,
      "avg_output_tokens_per_task": 15358.370240870241,
      "last_run_timestamp": "2026-10-31T08:57:30.925Z"
    },
    {
      "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": "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": 0.8676153636363637,
      "accuracy_std_dev": 0.004014278793432891,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.0702307419651056,
      "avg_duration_per_task_ms": 231760.75160697888,
      "avg_output_tokens_per_task": 11437.757575757576,
      "last_run_timestamp": "2026-10-31T08:17:29.729Z"
    },
    {
      "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": 0.8675645333333333,
      "accuracy_std_dev": 0.00826267276262484,
      "run_count": 15,
      "total_questions": 2970,
      "avg_cost_per_task_usd": 0.004299042424242424,
      "avg_duration_per_task_ms": 36217.24208754209,
      "avg_output_tokens_per_task": 7829.657575757576,
      "last_run_timestamp": "2026-10-31T08:38:41.553Z"
    },
    {
      "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": 0.8675644166666667,
      "accuracy_std_dev": 0.02533032544684445,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.0076437706228956225,
      "avg_duration_per_task_ms": 424115.41877104377,
      "avg_output_tokens_per_task": 25925.149831649833,
      "last_run_timestamp": "2026-10-31T08:17:29.008Z"
    },
    {
      "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-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "auto-routing",
      "accuracy": 0.8674240000000001,
      "accuracy_std_dev": 0.0048169925783625705,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.004064253787878788,
      "avg_duration_per_task_ms": 38624.37373737374,
      "avg_output_tokens_per_task": 7893.444444444444,
      "last_run_timestamp": "2026-10-31T08:27:45.642Z"
    },
    {
      "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": 0.8673413018549746,
      "accuracy_std_dev": 0.05933095903135433,
      "run_count": 9,
      "total_questions": 1779,
      "avg_cost_per_task_usd": 0.25496933951658235,
      "avg_duration_per_task_ms": 1215292.3445756042,
      "avg_output_tokens_per_task": 56493.202360876894,
      "last_run_timestamp": "2026-10-31T16:10:09.220Z"
    },
    {
      "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": 0.8671716000000002,
      "accuracy_std_dev": 0.006408346388890055,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.0041782994949494945,
      "avg_duration_per_task_ms": 35977.948484848486,
      "avg_output_tokens_per_task": 7724.74494949495,
      "last_run_timestamp": "2026-10-31T09:34:38.477Z"
    },
    {
      "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": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8668163333333333,
      "accuracy_std_dev": 0.017267918242927966,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.07368100841750841,
      "avg_duration_per_task_ms": 292389.96184062853,
      "avg_output_tokens_per_task": 20963.185746352414,
      "last_run_timestamp": "2026-10-31T16:09:19.242Z"
    },
    {
      "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": 0.866566745508982,
      "accuracy_std_dev": 0.010605013521914997,
      "run_count": 10,
      "total_questions": 1670,
      "avg_cost_per_task_usd": 0.05591402215568862,
      "avg_duration_per_task_ms": 429070.1862275449,
      "avg_output_tokens_per_task": 18535.95988023952,
      "last_run_timestamp": "2026-10-31T16:05:14.702Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "DeepInfra",
      "accuracy": 0.8659487366153845,
      "accuracy_std_dev": 0.016174915476337644,
      "run_count": 9,
      "total_questions": 1625,
      "avg_cost_per_task_usd": 0.005155188923076923,
      "avg_duration_per_task_ms": 688915.9618461538,
      "avg_output_tokens_per_task": 28341.659692307694,
      "last_run_timestamp": "2026-10-31T16:22:14.129Z"
    },
    {
      "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": 0.8658810000000001,
      "accuracy_std_dev": 0.04916542176842041,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.11255429335016834,
      "avg_duration_per_task_ms": 478692.68855218857,
      "avg_output_tokens_per_task": 28354.992845117846,
      "last_run_timestamp": "2026-10-31T16:01:53.651Z"
    },
    {
      "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": 0.8657292702702702,
      "accuracy_std_dev": 0.027335158307768007,
      "run_count": 37,
      "total_questions": 7326,
      "avg_cost_per_task_usd": 0.17380805050505052,
      "avg_duration_per_task_ms": 159066.58531258532,
      "avg_output_tokens_per_task": 16179.980343980344,
      "last_run_timestamp": "2026-10-31T08:33:01.888Z"
    },
    {
      "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": "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-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "InferenceNet",
      "accuracy": 0.8650982536651893,
      "accuracy_std_dev": 0.03856212210008964,
      "run_count": 15,
      "total_questions": 2933,
      "avg_cost_per_task_usd": 0.016069654619843162,
      "avg_duration_per_task_ms": 435995.1789976134,
      "avg_output_tokens_per_task": 41080.64609614729,
      "last_run_timestamp": "2026-10-31T08:51:00.880Z"
    },
    {
      "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": 0.8650607692307695,
      "accuracy_std_dev": 0.013738118503771485,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.7501272863247864,
      "avg_duration_per_task_ms": 502836.99456099456,
      "avg_output_tokens_per_task": 29840.98679098679,
      "last_run_timestamp": "2026-10-31T08:23:25.857Z"
    },
    {
      "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": 0.8649456666666666,
      "accuracy_std_dev": 0.026980251460816472,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.13166756565656565,
      "avg_duration_per_task_ms": 468303.6307519641,
      "avg_output_tokens_per_task": 34458.837261503926,
      "last_run_timestamp": "2026-10-31T16:30:53.502Z"
    },
    {
      "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": 0.8647586666666667,
      "accuracy_std_dev": 0.008943734131906087,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.05048964983164983,
      "avg_duration_per_task_ms": 169537.7323232323,
      "avg_output_tokens_per_task": 41902.75589225589,
      "last_run_timestamp": "2026-10-31T08:15:03.392Z"
    },
    {
      "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": 0.8645427692307693,
      "accuracy_std_dev": 0.039793064181421686,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.18345110683760685,
      "avg_duration_per_task_ms": 45870.56643356643,
      "avg_output_tokens_per_task": 3459.1363636363635,
      "last_run_timestamp": "2026-10-31T08:09:25.606Z"
    },
    {
      "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": 0.8644132307692307,
      "accuracy_std_dev": 0.07186191609570917,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.006970269619269618,
      "avg_duration_per_task_ms": 569923.0031080031,
      "avg_output_tokens_per_task": 28743.998834498834,
      "last_run_timestamp": "2026-10-31T08:07:18.761Z"
    },
    {
      "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": 0.8641975,
      "accuracy_std_dev": 0.018071121259715286,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.09928920664983164,
      "avg_duration_per_task_ms": 478656.78282828286,
      "avg_output_tokens_per_task": 26358.21085858586,
      "last_run_timestamp": "2026-10-31T16:03:21.454Z"
    },
    {
      "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": 0.8638953076923078,
      "accuracy_std_dev": 0.009790647408346209,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.8417052486402485,
      "avg_duration_per_task_ms": 411336.81507381506,
      "avg_output_tokens_per_task": 33477.120435120436,
      "last_run_timestamp": "2026-10-31T08:13:57.628Z"
    },
    {
      "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": 0.863852179419525,
      "accuracy_std_dev": 0.02754414592685712,
      "run_count": 10,
      "total_questions": 1895,
      "avg_cost_per_task_usd": 0.012207302902374668,
      "avg_duration_per_task_ms": 440151.53984168865,
      "avg_output_tokens_per_task": 28225.067018469657,
      "last_run_timestamp": "2026-10-31T08:36:30.903Z"
    },
    {
      "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": 0.8637468444693992,
      "accuracy_std_dev": 0.038710484413858166,
      "run_count": 9,
      "total_questions": 1781,
      "avg_cost_per_task_usd": 0.16561091353172375,
      "avg_duration_per_task_ms": 965884.440763616,
      "avg_output_tokens_per_task": 55799.21560920831,
      "last_run_timestamp": "2026-10-31T16:11:41.349Z"
    },
    {
      "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": 0.863649682795699,
      "accuracy_std_dev": 0.01354727385860341,
      "run_count": 6,
      "total_questions": 1116,
      "avg_cost_per_task_usd": 0.01399078853046595,
      "avg_duration_per_task_ms": 437830.958781362,
      "avg_output_tokens_per_task": 25123.625448028673,
      "last_run_timestamp": "2026-10-31T08:16:12.491Z"
    },
    {
      "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-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Relace",
      "accuracy": 0.8636363333333335,
      "accuracy_std_dev": 0.008473463964374614,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.056405106902356894,
      "avg_duration_per_task_ms": 429778.9991582492,
      "avg_output_tokens_per_task": 55743.61616161616,
      "last_run_timestamp": "2026-10-31T16:10:09.223Z"
    },
    {
      "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": 0.8635067692307692,
      "accuracy_std_dev": 0.008442363578577946,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.7928760839160839,
      "avg_duration_per_task_ms": 453047.2128982129,
      "avg_output_tokens_per_task": 31542.781274281275,
      "last_run_timestamp": "2026-10-31T08:15:24.290Z"
    },
    {
      "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": 0.8633403579175706,
      "accuracy_std_dev": 0.03134350405895737,
      "run_count": 7,
      "total_questions": 1383,
      "avg_cost_per_task_usd": 0.03313610050614606,
      "avg_duration_per_task_ms": 634565.6637744035,
      "avg_output_tokens_per_task": 38353.89515545915,
      "last_run_timestamp": "2026-10-31T08:09:30.864Z"
    },
    {
      "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": "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": 0.8630751666666665,
      "accuracy_std_dev": 0.009048483766478349,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.0646409191919192,
      "avg_duration_per_task_ms": 213655.1308922559,
      "avg_output_tokens_per_task": 10505.142255892257,
      "last_run_timestamp": "2026-10-31T08:13:57.551Z"
    },
    {
      "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": 0.8629839648725213,
      "accuracy_std_dev": 0.012680294246303573,
      "run_count": 9,
      "total_questions": 1765,
      "avg_cost_per_task_usd": 0.07565230934844192,
      "avg_duration_per_task_ms": 287267.9807365439,
      "avg_output_tokens_per_task": 20888.441359773373,
      "last_run_timestamp": "2026-10-31T16:05:36.103Z"
    },
    {
      "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": 0.8624118181818182,
      "accuracy_std_dev": 0.010959488075454655,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.0702179696969697,
      "avg_duration_per_task_ms": 231211.1772268136,
      "avg_output_tokens_per_task": 11433.751147842057,
      "last_run_timestamp": "2026-10-31T08:01:54.465Z"
    },
    {
      "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": 0.8623781294346978,
      "accuracy_std_dev": 0.016503860237700806,
      "run_count": 13,
      "total_questions": 2565,
      "avg_cost_per_task_usd": 0.08223287719298245,
      "avg_duration_per_task_ms": 172990.00506822614,
      "avg_output_tokens_per_task": 24199.74580896686,
      "last_run_timestamp": "2026-10-31T08:36:29.328Z"
    },
    {
      "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": 0.8622376508158508,
      "accuracy_std_dev": 0.05418494311168491,
      "run_count": 11,
      "total_questions": 2145,
      "avg_cost_per_task_usd": 0.07540476596736596,
      "avg_duration_per_task_ms": 485054.5156177156,
      "avg_output_tokens_per_task": 20286.058275058276,
      "last_run_timestamp": "2026-10-31T08:40:20.956Z"
    },
    {
      "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": 0.8621209999999999,
      "accuracy_std_dev": 0.016724375791042256,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.09178991565656566,
      "avg_duration_per_task_ms": 180823.8116161616,
      "avg_output_tokens_per_task": 25366.796464646464,
      "last_run_timestamp": "2026-10-31T16:04:29.122Z"
    },
    {
      "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": 0.8618672673926194,
      "accuracy_std_dev": 0.052319300007587534,
      "run_count": 9,
      "total_questions": 1653,
      "avg_cost_per_task_usd": 0.18415829764065333,
      "avg_duration_per_task_ms": 1613453.7084089534,
      "avg_output_tokens_per_task": 53479.07320024198,
      "last_run_timestamp": "2026-10-31T16:04:29.123Z"
    },
    {
      "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": 0.8617656666666667,
      "accuracy_std_dev": 0.018346628633681503,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.15345084287317617,
      "avg_duration_per_task_ms": 965626.3894500561,
      "avg_output_tokens_per_task": 48791.631313131315,
      "last_run_timestamp": "2026-10-31T16:05:14.701Z"
    },
    {
      "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": 0.8617121428571427,
      "accuracy_std_dev": 0.029687962026136223,
      "run_count": 14,
      "total_questions": 2772,
      "avg_cost_per_task_usd": 0.15537036976911978,
      "avg_duration_per_task_ms": 152704.13347763347,
      "avg_output_tokens_per_task": 15319.186507936507,
      "last_run_timestamp": "2026-10-31T08:27:53.385Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Ambient",
      "accuracy": 0.8613915000000001,
      "accuracy_std_dev": 0.04603966955336814,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.004525512626262627,
      "avg_duration_per_task_ms": 565963.2584175084,
      "avg_output_tokens_per_task": 18978.082491582492,
      "last_run_timestamp": "2026-08-29T10:57:26.955Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Morph",
      "accuracy": 0.8612792,
      "accuracy_std_dev": 0.03136197207702347,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.005120404040404041,
      "avg_duration_per_task_ms": 530889.794949495,
      "avg_output_tokens_per_task": 17542.636363636364,
      "last_run_timestamp": "2026-09-05T09:22:54.724Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.6-opus-20260205",
      "model_slug": "anthropic/claude-opus-4.6",
      "model_name": "Anthropic: Claude Opus 4.6",
      "provider_name": "Azure",
      "accuracy": 0.8611756923076923,
      "accuracy_std_dev": 0.018869779487458927,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.7569548310023311,
      "avg_duration_per_task_ms": 510623.6328671329,
      "avg_output_tokens_per_task": 30145.944444444445,
      "last_run_timestamp": "2026-10-31T08:30:25.667Z"
    },
    {
      "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": 0.861111076923077,
      "accuracy_std_dev": 0.007633099173604557,
      "run_count": 26,
      "total_questions": 5148,
      "avg_cost_per_task_usd": 0.8831994118104116,
      "avg_duration_per_task_ms": 414231.13616938615,
      "avg_output_tokens_per_task": 33454.88888888889,
      "last_run_timestamp": "2026-10-31T08:24:08.173Z"
    },
    {
      "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": 0.8607304246575341,
      "accuracy_std_dev": 0.01148253802030522,
      "run_count": 7,
      "total_questions": 1314,
      "avg_cost_per_task_usd": 0.02474931582952816,
      "avg_duration_per_task_ms": 139190.8691019787,
      "avg_output_tokens_per_task": 37895.824200913245,
      "last_run_timestamp": "2026-10-31T08:41:38.334Z"
    },
    {
      "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": "Chutes",
      "accuracy": 0.8606433333333333,
      "accuracy_std_dev": 0.02111334608988141,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.14206713636363638,
      "avg_duration_per_task_ms": 1276209.3136924803,
      "avg_output_tokens_per_task": 52392.67283950617,
      "last_run_timestamp": "2026-10-31T16:08:31.067Z"
    },
    {
      "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": 0.8606058666666668,
      "accuracy_std_dev": 0.011601831394894306,
      "run_count": 15,
      "total_questions": 2970,
      "avg_cost_per_task_usd": 0.8426976885521886,
      "avg_duration_per_task_ms": 457770.45521885523,
      "avg_output_tokens_per_task": 33527.7531986532,
      "last_run_timestamp": "2026-10-31T08:07:51.358Z"
    },
    {
      "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": 0.860546054237288,
      "accuracy_std_dev": 0.01522905267434664,
      "run_count": 9,
      "total_questions": 1770,
      "avg_cost_per_task_usd": 0.1431118,
      "avg_duration_per_task_ms": 301601.77118644066,
      "avg_output_tokens_per_task": 30086.04463276836,
      "last_run_timestamp": "2026-10-31T09:06:29.131Z"
    },
    {
      "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": 0.8604851819757365,
      "accuracy_std_dev": 0.010061527125579174,
      "run_count": 9,
      "total_questions": 1731,
      "avg_cost_per_task_usd": 0.004011181398035817,
      "avg_duration_per_task_ms": 248472.44656268053,
      "avg_output_tokens_per_task": 28672.775852108607,
      "last_run_timestamp": "2026-10-31T16:05:36.103Z"
    },
    {
      "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": "anthropic/claude-4.5-opus-20251124",
      "model_slug": "anthropic/claude-opus-4.5",
      "model_name": "Anthropic: Claude Opus 4.5",
      "provider_name": "Google",
      "accuracy": 0.8601396923076924,
      "accuracy_std_dev": 0.009255270194157684,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.8530733566433567,
      "avg_duration_per_task_ms": 482253.25835275836,
      "avg_output_tokens_per_task": 33932.094017094016,
      "last_run_timestamp": "2026-10-31T08:48:00.920Z"
    },
    {
      "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-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "DigitalOcean",
      "accuracy": 0.859429902017291,
      "accuracy_std_dev": 0.05098775585122179,
      "run_count": 11,
      "total_questions": 2082,
      "avg_cost_per_task_usd": 0.011825536023054757,
      "avg_duration_per_task_ms": 442610.81796349667,
      "avg_output_tokens_per_task": 25201.657060518733,
      "last_run_timestamp": "2026-10-31T08:27:46.434Z"
    },
    {
      "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": 0.8590389615384615,
      "accuracy_std_dev": 0.017598304950020197,
      "run_count": 26,
      "total_questions": 5148,
      "avg_cost_per_task_usd": 0.8044641944444444,
      "avg_duration_per_task_ms": 487156.2365967366,
      "avg_output_tokens_per_task": 30492.046425796427,
      "last_run_timestamp": "2026-10-31T09:25:22.406Z"
    },
    {
      "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": 0.8590346,
      "accuracy_std_dev": 0.016996610739791626,
      "run_count": 15,
      "total_questions": 2970,
      "avg_cost_per_task_usd": 0.75415548989899,
      "avg_duration_per_task_ms": 507155.8878787879,
      "avg_output_tokens_per_task": 30052.987878787877,
      "last_run_timestamp": "2026-10-31T08:09:30.224Z"
    },
    {
      "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": "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": "anthropic/claude-5-fable-20260609",
      "model_slug": "anthropic/claude-fable-5",
      "model_name": "Anthropic: Claude Fable 5",
      "provider_name": "Google",
      "accuracy": 0.858959888888889,
      "accuracy_std_dev": 0.04917250355133622,
      "run_count": 27,
      "total_questions": 5346,
      "avg_cost_per_task_usd": 0.18915268649457537,
      "avg_duration_per_task_ms": 41450.0987654321,
      "avg_output_tokens_per_task": 3406.6427235316123,
      "last_run_timestamp": "2026-10-31T08:27:45.697Z"
    },
    {
      "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": 0.8588278767123287,
      "accuracy_std_dev": 0.014585318196353787,
      "run_count": 3,
      "total_questions": 438,
      "avg_cost_per_task_usd": 0.005061143835616439,
      "avg_duration_per_task_ms": 414471.07305936073,
      "avg_output_tokens_per_task": 13714.753424657534,
      "last_run_timestamp": "2026-10-31T16:45:39.994Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.858586,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.08651329797979797,
      "avg_duration_per_task_ms": 499241.5707070707,
      "avg_output_tokens_per_task": 35858.23737373737,
      "last_run_timestamp": "2026-10-04T03:06:10.317Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.5-opus-20251124",
      "model_slug": "anthropic/claude-opus-4.5",
      "model_name": "Anthropic: Claude Opus 4.5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.8581648333333333,
      "accuracy_std_dev": 0.025909089369798823,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.8336664709595959,
      "avg_duration_per_task_ms": 439138.7053872054,
      "avg_output_tokens_per_task": 33233.371632996634,
      "last_run_timestamp": "2026-10-31T08:35:35.377Z"
    },
    {
      "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": 0.857990281303602,
      "accuracy_std_dev": 0.030501594473235694,
      "run_count": 12,
      "total_questions": 2332,
      "avg_cost_per_task_usd": 0.13072562135506002,
      "avg_duration_per_task_ms": 510944.2062607204,
      "avg_output_tokens_per_task": 43400.17238421956,
      "last_run_timestamp": "2026-10-31T16:16:35.161Z"
    },
    {
      "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": 0.8579124,
      "accuracy_std_dev": 0.017384854495796048,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.02057177575757576,
      "avg_duration_per_task_ms": 445221.4888888889,
      "avg_output_tokens_per_task": 40903.725252525255,
      "last_run_timestamp": "2026-10-31T08:25:16.899Z"
    },
    {
      "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": 0.8575497692307692,
      "accuracy_std_dev": 0.016422560561869862,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.8395061091686092,
      "avg_duration_per_task_ms": 459222.6048951049,
      "avg_output_tokens_per_task": 33403.42074592075,
      "last_run_timestamp": "2026-10-31T08:40:44.982Z"
    },
    {
      "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": "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": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Azure",
      "accuracy": 0.8570892222222223,
      "accuracy_std_dev": 0.031137791974732713,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.015986119528619527,
      "avg_duration_per_task_ms": 333449.29124579125,
      "avg_output_tokens_per_task": 28587.83613916947,
      "last_run_timestamp": "2026-10-31T16:04:48.120Z"
    },
    {
      "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": "auto-routing",
      "accuracy": 0.8569022857142858,
      "accuracy_std_dev": 0.009351683113815608,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.005601821789321789,
      "avg_duration_per_task_ms": 514036.772005772,
      "avg_output_tokens_per_task": 30934.699134199134,
      "last_run_timestamp": "2026-10-31T16:16:55.696Z"
    },
    {
      "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": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "Minimax",
      "accuracy": 0.8568555555555556,
      "accuracy_std_dev": 0.007583374638878124,
      "run_count": 18,
      "total_questions": 3564,
      "avg_cost_per_task_usd": 0.06372034118967451,
      "avg_duration_per_task_ms": 605513.0182379349,
      "avg_output_tokens_per_task": 35143.53956228956,
      "last_run_timestamp": "2026-10-31T16:08:56.835Z"
    },
    {
      "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": 0.8568247201068804,
      "accuracy_std_dev": 0.015059304005690374,
      "run_count": 9,
      "total_questions": 1497,
      "avg_cost_per_task_usd": 0.004656902471609886,
      "avg_duration_per_task_ms": 413401.8376753507,
      "avg_output_tokens_per_task": 9906.163660654642,
      "last_run_timestamp": "2026-10-31T09:01:06.018Z"
    },
    {
      "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": 0.856687139561058,
      "accuracy_std_dev": 0.016735505421958285,
      "run_count": 9,
      "total_questions": 1777,
      "avg_cost_per_task_usd": 0.03746025323579066,
      "avg_duration_per_task_ms": 607320.918964547,
      "avg_output_tokens_per_task": 34838.97186268993,
      "last_run_timestamp": "2026-10-31T16:11:17.542Z"
    },
    {
      "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": 0.8565023144979739,
      "accuracy_std_dev": 0.02045561379357996,
      "run_count": 45,
      "total_questions": 8884,
      "avg_cost_per_task_usd": 0.2887550784556506,
      "avg_duration_per_task_ms": 342908.4242458352,
      "avg_output_tokens_per_task": 51958.11143628996,
      "last_run_timestamp": "2026-10-31T16:26:37.892Z"
    },
    {
      "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": 0.8558309545454545,
      "accuracy_std_dev": 0.010665991197502755,
      "run_count": 22,
      "total_questions": 4356,
      "avg_cost_per_task_usd": 0.13619998117539026,
      "avg_duration_per_task_ms": 429202.50367309456,
      "avg_output_tokens_per_task": 44305.976124885216,
      "last_run_timestamp": "2026-10-31T16:26:59.503Z"
    },
    {
      "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": 0.8557935365853659,
      "accuracy_std_dev": 0.035827629886346836,
      "run_count": 41,
      "total_questions": 8118,
      "avg_cost_per_task_usd": 0.16502105962059618,
      "avg_duration_per_task_ms": 151492.22037447648,
      "avg_output_tokens_per_task": 15265.38950480414,
      "last_run_timestamp": "2026-10-31T08:51:08.192Z"
    },
    {
      "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": 0.8556823130287649,
      "accuracy_std_dev": 0.021169598748854452,
      "run_count": 12,
      "total_questions": 2364,
      "avg_cost_per_task_usd": 0.08525841285956007,
      "avg_duration_per_task_ms": 295845.3159898477,
      "avg_output_tokens_per_task": 25006.826565143823,
      "last_run_timestamp": "2026-10-31T08:34:08.441Z"
    },
    {
      "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": 0.8554993333333332,
      "accuracy_std_dev": 0.021190911552938067,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.043522018518518515,
      "avg_duration_per_task_ms": 663758.2626262626,
      "avg_output_tokens_per_task": 36261.223905723906,
      "last_run_timestamp": "2026-10-31T16:02:14.360Z"
    },
    {
      "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": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Fireworks",
      "accuracy": 0.8548688033707866,
      "accuracy_std_dev": 0.012107174401981674,
      "run_count": 3,
      "total_questions": 356,
      "avg_cost_per_task_usd": 0.00593769382022472,
      "avg_duration_per_task_ms": 221303.48314606742,
      "avg_output_tokens_per_task": 23969.63202247191,
      "last_run_timestamp": "2026-10-31T16:10:36.259Z"
    },
    {
      "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": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Inceptron",
      "accuracy": 0.8545531474226805,
      "accuracy_std_dev": 0.018117679996070137,
      "run_count": 10,
      "total_questions": 1940,
      "avg_cost_per_task_usd": 0.08358154742268041,
      "avg_duration_per_task_ms": 325785.14793814434,
      "avg_output_tokens_per_task": 28959.811340206186,
      "last_run_timestamp": "2026-10-31T08:56:36.753Z"
    },
    {
      "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": 0.8540966666666665,
      "accuracy_std_dev": 0.004199386807089248,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.013766887205387206,
      "avg_duration_per_task_ms": 205590.59595959596,
      "avg_output_tokens_per_task": 29236.930976430976,
      "last_run_timestamp": "2026-10-31T08:17:28.743Z"
    },
    {
      "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": 0.8540819492179129,
      "accuracy_std_dev": 0.036199470332124874,
      "run_count": 24,
      "total_questions": 4667,
      "avg_cost_per_task_usd": 0.004028577030212127,
      "avg_duration_per_task_ms": 78708.61088493679,
      "avg_output_tokens_per_task": 15107.985001071353,
      "last_run_timestamp": "2026-10-31T08:12:47.827Z"
    },
    {
      "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": 0.8539094444444444,
      "accuracy_std_dev": 0.018180043149387204,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.013447226711560045,
      "avg_duration_per_task_ms": 780881.4635241302,
      "avg_output_tokens_per_task": 28188.217171717173,
      "last_run_timestamp": "2026-10-31T16:17:15.719Z"
    },
    {
      "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": 0.8537224444444443,
      "accuracy_std_dev": 0.027781459489343335,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.07644378900112234,
      "avg_duration_per_task_ms": 480473.94051627384,
      "avg_output_tokens_per_task": 22998.121212121212,
      "last_run_timestamp": "2026-10-31T16:12:07.201Z"
    },
    {
      "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": 0.8537037,
      "accuracy_std_dev": 0.012630753865466617,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.10325877878787879,
      "avg_duration_per_task_ms": 455294.6515151515,
      "avg_output_tokens_per_task": 31795.09494949495,
      "last_run_timestamp": "2026-10-31T08:16:11.390Z"
    },
    {
      "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": 0.8536755833333335,
      "accuracy_std_dev": 0.03297161670148616,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.7094076557239056,
      "avg_duration_per_task_ms": 485859.81018518517,
      "avg_output_tokens_per_task": 28361.97053872054,
      "last_run_timestamp": "2026-10-31T08:09:42.551Z"
    },
    {
      "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-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Sail Research",
      "accuracy": 0.8530808309576838,
      "accuracy_std_dev": 0.0923187678677938,
      "run_count": 23,
      "total_questions": 4490,
      "avg_cost_per_task_usd": 0.007865783073496658,
      "avg_duration_per_task_ms": 559605.6879732739,
      "avg_output_tokens_per_task": 27305.323830734967,
      "last_run_timestamp": "2026-10-31T09:04:13.487Z"
    },
    {
      "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": "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": 0.8523164263322883,
      "accuracy_std_dev": 0.010034068893524688,
      "run_count": 5,
      "total_questions": 957,
      "avg_cost_per_task_usd": 0.08246928213166145,
      "avg_duration_per_task_ms": 272317.5339602926,
      "avg_output_tokens_per_task": 27908.60710553814,
      "last_run_timestamp": "2026-10-31T08:44:39.028Z"
    },
    {
      "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": 0.8521884,
      "accuracy_std_dev": 0.01923598183197312,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.030990156565656565,
      "avg_duration_per_task_ms": 619926.6106060606,
      "avg_output_tokens_per_task": 34097.76565656566,
      "last_run_timestamp": "2026-10-31T16:02:14.339Z"
    },
    {
      "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": 0.8521639072512648,
      "accuracy_std_dev": 0.012873132633608224,
      "run_count": 9,
      "total_questions": 1779,
      "avg_cost_per_task_usd": 0.08519417144463183,
      "avg_duration_per_task_ms": 288650.1905564924,
      "avg_output_tokens_per_task": 29819.736368746486,
      "last_run_timestamp": "2026-10-31T08:22:52.355Z"
    },
    {
      "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": 0.852137797527048,
      "accuracy_std_dev": 0.006816033761653469,
      "run_count": 10,
      "total_questions": 1941,
      "avg_cost_per_task_usd": 0.09748740082431735,
      "avg_duration_per_task_ms": 655650.1257083977,
      "avg_output_tokens_per_task": 41096.26429675425,
      "last_run_timestamp": "2026-10-31T16:11:17.543Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Venice",
      "accuracy": 0.8521062280219781,
      "accuracy_std_dev": 0.006055500000000047,
      "run_count": 2,
      "total_questions": 364,
      "avg_cost_per_task_usd": 0.005501717032967033,
      "avg_duration_per_task_ms": 539158.467032967,
      "avg_output_tokens_per_task": 15491.851648351649,
      "last_run_timestamp": "2026-10-17T08:17:58.236Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Ionstream",
      "accuracy": 0.851852,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.16585097979797978,
      "avg_duration_per_task_ms": 389107.60101010103,
      "avg_output_tokens_per_task": 37536.631313131315,
      "last_run_timestamp": "2026-08-04T03:12:10.622Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5-2025-08-07",
      "model_slug": "openai/gpt-5",
      "model_name": "OpenAI: GPT-5",
      "provider_name": "Azure",
      "accuracy": 0.8518517499999999,
      "accuracy_std_dev": 0.0020620621201845155,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.21063902146464644,
      "avg_duration_per_task_ms": 251655.29545454544,
      "avg_output_tokens_per_task": 19971.159090909092,
      "last_run_timestamp": "2026-10-01T16:27:58.636Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Parasail",
      "accuracy": 0.8512906666666668,
      "accuracy_std_dev": 0.00555530025191155,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.017986696969696972,
      "avg_duration_per_task_ms": 129036.91750841751,
      "avg_output_tokens_per_task": 29778.104377104377,
      "last_run_timestamp": "2026-10-31T08:56:30.449Z"
    },
    {
      "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": 0.8510099999999999,
      "accuracy_std_dev": 0.009995248508666484,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.013658645833333335,
      "avg_duration_per_task_ms": 646317.4116161616,
      "avg_output_tokens_per_task": 19001.599116161615,
      "last_run_timestamp": "2026-10-31T16:05:36.294Z"
    },
    {
      "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": "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": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "auto-routing",
      "accuracy": 0.8509165555555555,
      "accuracy_std_dev": 0.023084461599916624,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.046726610549943884,
      "avg_duration_per_task_ms": 634694.9528619528,
      "avg_output_tokens_per_task": 38211.56285072952,
      "last_run_timestamp": "2026-10-31T16:03:21.441Z"
    },
    {
      "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": 0.8508805384615384,
      "accuracy_std_dev": 0.0098378318641346,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.09800575252525251,
      "avg_duration_per_task_ms": 565289.1045066045,
      "avg_output_tokens_per_task": 30479.255633255634,
      "last_run_timestamp": "2026-10-31T16:01:53.623Z"
    },
    {
      "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": 0.8507273166273585,
      "accuracy_std_dev": 0.011165344551781631,
      "run_count": 10,
      "total_questions": 1696,
      "avg_cost_per_task_usd": 0.006636630896226415,
      "avg_duration_per_task_ms": 404585.84021226416,
      "avg_output_tokens_per_task": 16352.901533018869,
      "last_run_timestamp": "2026-10-31T16:08:33.701Z"
    },
    {
      "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": "AtlasCloud",
      "accuracy": 0.8505423333333332,
      "accuracy_std_dev": 0.03289190627224608,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.010776331649831649,
      "avg_duration_per_task_ms": 434634.2154882155,
      "avg_output_tokens_per_task": 38163.07968574635,
      "last_run_timestamp": "2026-10-31T16:04:29.124Z"
    },
    {
      "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": 0.8504273076923077,
      "accuracy_std_dev": 0.019136243013573914,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.12441163325563323,
      "avg_duration_per_task_ms": 380633.7595182595,
      "avg_output_tokens_per_task": 38591.44055944056,
      "last_run_timestamp": "2026-10-31T16:08:08.543Z"
    },
    {
      "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": 0.8502976923076923,
      "accuracy_std_dev": 0.04563619826137448,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.16715858896658897,
      "avg_duration_per_task_ms": 165688.25796425797,
      "avg_output_tokens_per_task": 16495.934343434343,
      "last_run_timestamp": "2026-10-31T08:12:45.468Z"
    },
    {
      "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": 0.8502042599087539,
      "accuracy_std_dev": 0.01590746948888509,
      "run_count": 37,
      "total_questions": 7014,
      "avg_cost_per_task_usd": 0.16371538508696884,
      "avg_duration_per_task_ms": 188635.26946107784,
      "avg_output_tokens_per_task": 29341.152266894784,
      "last_run_timestamp": "2026-10-31T09:24:22.125Z"
    },
    {
      "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": 0.8501748923435677,
      "accuracy_std_dev": 0.00914970442953917,
      "run_count": 13,
      "total_questions": 2573,
      "avg_cost_per_task_usd": 0.03854331947143411,
      "avg_duration_per_task_ms": 385001.11737271666,
      "avg_output_tokens_per_task": 39135.41857753595,
      "last_run_timestamp": "2026-10-31T16:26:59.494Z"
    },
    {
      "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": "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": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Fireworks",
      "accuracy": 0.8490562452830188,
      "accuracy_std_dev": 0.004459999999999964,
      "run_count": 2,
      "total_questions": 265,
      "avg_cost_per_task_usd": 0.017200528301886793,
      "avg_duration_per_task_ms": 165535.98490566039,
      "avg_output_tokens_per_task": 28515.603773584906,
      "last_run_timestamp": "2026-10-31T08:20:07.444Z"
    },
    {
      "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": 0.84890575,
      "accuracy_std_dev": 0.05035112340045156,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.1629483771043771,
      "avg_duration_per_task_ms": 162891.44065656565,
      "avg_output_tokens_per_task": 16117.493265993266,
      "last_run_timestamp": "2026-10-31T08:45:37.246Z"
    },
    {
      "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": 0.8487436923076923,
      "accuracy_std_dev": 0.015982537328089897,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.05245943783993783,
      "avg_duration_per_task_ms": 270756.14801864803,
      "avg_output_tokens_per_task": 45704.256410256414,
      "last_run_timestamp": "2026-10-31T16:53:23.847Z"
    },
    {
      "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": "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": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8477888928571429,
      "accuracy_std_dev": 0.029215151613674717,
      "run_count": 10,
      "total_questions": 1960,
      "avg_cost_per_task_usd": 0.1592395025510204,
      "avg_duration_per_task_ms": 438119.2229591837,
      "avg_output_tokens_per_task": 36131.57244897959,
      "last_run_timestamp": "2026-10-31T08:13:34.182Z"
    },
    {
      "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": 0.8476458282485876,
      "accuracy_std_dev": 0.0101614457790704,
      "run_count": 9,
      "total_questions": 1770,
      "avg_cost_per_task_usd": 0.007194387005649717,
      "avg_duration_per_task_ms": 581669.6435028248,
      "avg_output_tokens_per_task": 17793.33220338983,
      "last_run_timestamp": "2026-10-31T16:12:28.235Z"
    },
    {
      "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": 0.8475903301204819,
      "accuracy_std_dev": 0.009100393388336854,
      "run_count": 13,
      "total_questions": 2490,
      "avg_cost_per_task_usd": 0.06626451124497991,
      "avg_duration_per_task_ms": 191280.34939759035,
      "avg_output_tokens_per_task": 20610.90562248996,
      "last_run_timestamp": "2026-10-31T16:08:31.065Z"
    },
    {
      "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": 0.8473217418810629,
      "accuracy_std_dev": 0.031381320315484754,
      "run_count": 12,
      "total_questions": 2371,
      "avg_cost_per_task_usd": 0.009025258118937158,
      "avg_duration_per_task_ms": 479873.75579924084,
      "avg_output_tokens_per_task": 32127.526360185577,
      "last_run_timestamp": "2026-10-31T08:31:37.876Z"
    },
    {
      "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": "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": 0.8469361045003815,
      "accuracy_std_dev": 0.04231320024956013,
      "run_count": 8,
      "total_questions": 1311,
      "avg_cost_per_task_usd": 0.02445259496567506,
      "avg_duration_per_task_ms": 437957.52021357743,
      "avg_output_tokens_per_task": 38826.28070175439,
      "last_run_timestamp": "2026-10-31T08:55:46.529Z"
    },
    {
      "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": 0.8468660769230769,
      "accuracy_std_dev": 0.017453184397394857,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.12967881662781666,
      "avg_duration_per_task_ms": 476170.62509712507,
      "avg_output_tokens_per_task": 48636.332556332556,
      "last_run_timestamp": "2026-10-31T16:09:19.244Z"
    },
    {
      "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": 0.8468582586644126,
      "accuracy_std_dev": 0.013598361278927057,
      "run_count": 6,
      "total_questions": 1183,
      "avg_cost_per_task_usd": 0.26757311665257816,
      "avg_duration_per_task_ms": 498008.1420118343,
      "avg_output_tokens_per_task": 60614.17836010144,
      "last_run_timestamp": "2026-10-31T16:45:45.621Z"
    },
    {
      "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": 0.8468012941176469,
      "accuracy_std_dev": 0.0195477464979672,
      "run_count": 34,
      "total_questions": 6732,
      "avg_cost_per_task_usd": 0.23950519132501485,
      "avg_duration_per_task_ms": 493273.4841057635,
      "avg_output_tokens_per_task": 47308.403149138445,
      "last_run_timestamp": "2026-10-31T17:01:17.198Z"
    },
    {
      "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": 0.8466142222222222,
      "accuracy_std_dev": 0.010836379374616662,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.14184849326599328,
      "avg_duration_per_task_ms": 144926.00224466893,
      "avg_output_tokens_per_task": 32057.991582491584,
      "last_run_timestamp": "2026-10-31T08:40:49.272Z"
    },
    {
      "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": 0.8465387902641119,
      "accuracy_std_dev": 0.025032600265254102,
      "run_count": 10,
      "total_questions": 1931,
      "avg_cost_per_task_usd": 0.10684122164681513,
      "avg_duration_per_task_ms": 489966.58933195233,
      "avg_output_tokens_per_task": 34902.65044018643,
      "last_run_timestamp": "2026-10-31T09:15:06.724Z"
    },
    {
      "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": 0.8463805000000001,
      "accuracy_std_dev": 0.04340338597102762,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.09990382196969698,
      "avg_duration_per_task_ms": 1192685.5157828282,
      "avg_output_tokens_per_task": 39566.316287878784,
      "last_run_timestamp": "2026-10-31T16:01:53.650Z"
    },
    {
      "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": 0.8459293152073732,
      "accuracy_std_dev": 0.017246738162321286,
      "run_count": 11,
      "total_questions": 2170,
      "avg_cost_per_task_usd": 0.010452895391705068,
      "avg_duration_per_task_ms": 454110.96359447006,
      "avg_output_tokens_per_task": 34047.99170506912,
      "last_run_timestamp": "2026-10-31T08:27:51.327Z"
    },
    {
      "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": 0.8456636089524402,
      "accuracy_std_dev": 0.018418318454142355,
      "run_count": 17,
      "total_questions": 3217,
      "avg_cost_per_task_usd": 0.14255490332608023,
      "avg_duration_per_task_ms": 266408.3546782717,
      "avg_output_tokens_per_task": 28803.044140503574,
      "last_run_timestamp": "2026-10-31T08:33:10.900Z"
    },
    {
      "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": 0.8454543999999999,
      "accuracy_std_dev": 0.04169381837443052,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.0038799494949494947,
      "avg_duration_per_task_ms": 558146.4727272728,
      "avg_output_tokens_per_task": 40172.065151515155,
      "last_run_timestamp": "2026-10-31T16:18:02.140Z"
    },
    {
      "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": 0.8452858999999998,
      "accuracy_std_dev": 0.01791875201820708,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.14104453232323233,
      "avg_duration_per_task_ms": 915068.2414141414,
      "avg_output_tokens_per_task": 55465.26919191919,
      "last_run_timestamp": "2026-10-31T16:10:36.259Z"
    },
    {
      "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": 0.8449505654205607,
      "accuracy_std_dev": 0.010833165164405542,
      "run_count": 13,
      "total_questions": 2568,
      "avg_cost_per_task_usd": 0.01864547196261682,
      "avg_duration_per_task_ms": 133412.9898753894,
      "avg_output_tokens_per_task": 18649.78894080997,
      "last_run_timestamp": "2026-10-31T16:05:14.635Z"
    },
    {
      "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": 0.8446468343465044,
      "accuracy_std_dev": 0.01783018725336332,
      "run_count": 10,
      "total_questions": 1974,
      "avg_cost_per_task_usd": 0.0660811013171226,
      "avg_duration_per_task_ms": 245353.75430597772,
      "avg_output_tokens_per_task": 20336.286220871327,
      "last_run_timestamp": "2026-10-31T08:32:53.731Z"
    },
    {
      "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": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "DigitalOcean",
      "accuracy": 0.8441557142857142,
      "accuracy_std_dev": 0.035688144801137224,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.004898924242424242,
      "avg_duration_per_task_ms": 1008847.8311688312,
      "avg_output_tokens_per_task": 27765.935786435788,
      "last_run_timestamp": "2026-10-31T16:46:54.774Z"
    },
    {
      "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": "PrimeIntellect",
      "accuracy": 0.8439077604166666,
      "accuracy_std_dev": 0.009070260174817954,
      "run_count": 12,
      "total_questions": 2112,
      "avg_cost_per_task_usd": 0.12450989630681818,
      "avg_duration_per_task_ms": 153233.97348484848,
      "avg_output_tokens_per_task": 31747.522727272728,
      "last_run_timestamp": "2026-10-31T08:45:40.383Z"
    },
    {
      "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": 0.8438831999999998,
      "accuracy_std_dev": 0.010796604899689534,
      "run_count": 30,
      "total_questions": 5940,
      "avg_cost_per_task_usd": 0.1486385685185185,
      "avg_duration_per_task_ms": 173235.29764309764,
      "avg_output_tokens_per_task": 32642.098316498315,
      "last_run_timestamp": "2026-10-31T08:24:42.872Z"
    },
    {
      "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": 0.8435638461538462,
      "accuracy_std_dev": 0.032311682699906065,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.05297190248640248,
      "avg_duration_per_task_ms": 206855.22494172494,
      "avg_output_tokens_per_task": 21970.065268065267,
      "last_run_timestamp": "2026-10-31T16:28:22.255Z"
    },
    {
      "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": "Novita",
      "accuracy": 0.843457793224299,
      "accuracy_std_dev": 0.05231442632207949,
      "run_count": 13,
      "total_questions": 2568,
      "avg_cost_per_task_usd": 0.07964885591900311,
      "avg_duration_per_task_ms": 557670.449376947,
      "avg_output_tokens_per_task": 43630.647196261685,
      "last_run_timestamp": "2026-10-31T16:30:10.027Z"
    },
    {
      "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": "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-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Relace",
      "accuracy": 0.8433784397283531,
      "accuracy_std_dev": 0.003767029717016947,
      "run_count": 6,
      "total_questions": 1178,
      "avg_cost_per_task_usd": 0.1209133650254669,
      "avg_duration_per_task_ms": 239041.5050933786,
      "avg_output_tokens_per_task": 32072.49235993209,
      "last_run_timestamp": "2026-10-31T09:12:37.593Z"
    },
    {
      "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": 0.8432182184684685,
      "accuracy_std_dev": 0.019959068726883127,
      "run_count": 9,
      "total_questions": 1332,
      "avg_cost_per_task_usd": 0.01802810960960961,
      "avg_duration_per_task_ms": 98985.72822822822,
      "avg_output_tokens_per_task": 15473.123123123123,
      "last_run_timestamp": "2026-10-31T16:06:00.067Z"
    },
    {
      "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": 0.8431583644859812,
      "accuracy_std_dev": 0.027490521207033807,
      "run_count": 8,
      "total_questions": 1391,
      "avg_cost_per_task_usd": 0.08865105607476635,
      "avg_duration_per_task_ms": 1400795.4744787922,
      "avg_output_tokens_per_task": 40690.26815240834,
      "last_run_timestamp": "2026-10-31T16:04:23.355Z"
    },
    {
      "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": 0.8426947486187845,
      "accuracy_std_dev": 0.027181441039645704,
      "run_count": 11,
      "total_questions": 2172,
      "avg_cost_per_task_usd": 0.01345692311233886,
      "avg_duration_per_task_ms": 682984.7329650092,
      "avg_output_tokens_per_task": 42267.71869244936,
      "last_run_timestamp": "2026-10-31T08:57:00.536Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "thinkingmachines/inkling-20260715",
      "model_slug": "thinkingmachines/inkling",
      "model_name": "Thinking Machines: Inkling",
      "provider_name": "Together",
      "accuracy": 0.8425437868421053,
      "accuracy_std_dev": 0.00573439306339873,
      "run_count": 6,
      "total_questions": 1140,
      "avg_cost_per_task_usd": 0.08756267807017544,
      "avg_duration_per_task_ms": 282086.36140350875,
      "avg_output_tokens_per_task": 21480.616666666665,
      "last_run_timestamp": "2026-10-01T16:06:12.801Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Decart",
      "accuracy": 0.8422337612565445,
      "accuracy_std_dev": 0.023212455418589382,
      "run_count": 10,
      "total_questions": 1910,
      "avg_cost_per_task_usd": 0.10300291204188483,
      "avg_duration_per_task_ms": 203070.60157068062,
      "avg_output_tokens_per_task": 30819.346073298428,
      "last_run_timestamp": "2026-10-31T08:52:06.333Z"
    },
    {
      "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": 0.842080958563536,
      "accuracy_std_dev": 0.01924201603413352,
      "run_count": 11,
      "total_questions": 2172,
      "avg_cost_per_task_usd": 0.02133896546961326,
      "avg_duration_per_task_ms": 337269.2900552486,
      "avg_output_tokens_per_task": 42459.25598526704,
      "last_run_timestamp": "2026-10-31T08:44:51.872Z"
    },
    {
      "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": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Makora",
      "accuracy": 0.8416364037363778,
      "accuracy_std_dev": 0.015589563439410502,
      "run_count": 11,
      "total_questions": 1927,
      "avg_cost_per_task_usd": 0.1375403404255319,
      "avg_duration_per_task_ms": 272709.34509600414,
      "avg_output_tokens_per_task": 32850.36170212766,
      "last_run_timestamp": "2026-10-31T09:42:32.959Z"
    },
    {
      "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": 0.8414135533290239,
      "accuracy_std_dev": 0.01226653991581053,
      "run_count": 9,
      "total_questions": 1547,
      "avg_cost_per_task_usd": 0.006907548157724628,
      "avg_duration_per_task_ms": 661601.5623787977,
      "avg_output_tokens_per_task": 17049.550096961862,
      "last_run_timestamp": "2026-10-31T16:34:27.227Z"
    },
    {
      "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": "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": 0.840891889212828,
      "accuracy_std_dev": 0.01539818649719325,
      "run_count": 16,
      "total_questions": 3087,
      "avg_cost_per_task_usd": 0.022577127632005183,
      "avg_duration_per_task_ms": 250189.5425979916,
      "avg_output_tokens_per_task": 38449.29964366699,
      "last_run_timestamp": "2026-10-31T08:54:15.638Z"
    },
    {
      "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": 0.8408359820854386,
      "accuracy_std_dev": 0.011558364030339736,
      "run_count": 11,
      "total_questions": 2177,
      "avg_cost_per_task_usd": 0.14478948920532844,
      "avg_duration_per_task_ms": 203069.29581993568,
      "avg_output_tokens_per_task": 32673.29122645843,
      "last_run_timestamp": "2026-10-31T09:21:48.528Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "Venice",
      "accuracy": 0.8408155555555555,
      "accuracy_std_dev": 0.0506301697894998,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.0956665359147026,
      "avg_duration_per_task_ms": 384651.6548821549,
      "avg_output_tokens_per_task": 21452.003367003366,
      "last_run_timestamp": "2026-10-31T16:15:29.279Z"
    },
    {
      "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": 0.8407407,
      "accuracy_std_dev": 0.024165205569371837,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.013536882323232325,
      "avg_duration_per_task_ms": 528967.1176767676,
      "avg_output_tokens_per_task": 40146.823737373736,
      "last_run_timestamp": "2026-10-31T08:45:25.206Z"
    },
    {
      "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": 0.8406767200863932,
      "accuracy_std_dev": 0.03362626844282443,
      "run_count": 12,
      "total_questions": 2315,
      "avg_cost_per_task_usd": 0.030011868250539957,
      "avg_duration_per_task_ms": 286135.6241900648,
      "avg_output_tokens_per_task": 9270.458315334774,
      "last_run_timestamp": "2026-10-31T08:16:12.311Z"
    },
    {
      "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": 0.8406685714285715,
      "accuracy_std_dev": 0.01850043854968342,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.02123720418470418,
      "avg_duration_per_task_ms": 436896.4862914863,
      "avg_output_tokens_per_task": 42244.7886002886,
      "last_run_timestamp": "2026-10-31T08:09:30.600Z"
    },
    {
      "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": 0.8406284444444444,
      "accuracy_std_dev": 0.012648005025924159,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.08990638720538721,
      "avg_duration_per_task_ms": 1092332.543771044,
      "avg_output_tokens_per_task": 45444.60101010101,
      "last_run_timestamp": "2026-10-31T16:33:40.856Z"
    },
    {
      "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": "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": 0.8404879999999999,
      "accuracy_std_dev": 0.020003728864889173,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.037026556818181816,
      "avg_duration_per_task_ms": 113840.68308080808,
      "avg_output_tokens_per_task": 34919.68308080808,
      "last_run_timestamp": "2026-10-31T09:18:03.023Z"
    },
    {
      "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": 0.8404308818757921,
      "accuracy_std_dev": 0.01259237381870473,
      "run_count": 20,
      "total_questions": 3945,
      "avg_cost_per_task_usd": 0.10827099493029152,
      "avg_duration_per_task_ms": 326193.6968314322,
      "avg_output_tokens_per_task": 34498.35411913815,
      "last_run_timestamp": "2026-10-31T08:48:28.167Z"
    },
    {
      "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": 0.8401781603650587,
      "accuracy_std_dev": 0.014388231606073072,
      "run_count": 16,
      "total_questions": 3068,
      "avg_cost_per_task_usd": 0.07506749576271186,
      "avg_duration_per_task_ms": 388300.5521512386,
      "avg_output_tokens_per_task": 22304.916558018253,
      "last_run_timestamp": "2026-10-31T08:43:48.577Z"
    },
    {
      "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": 0.8400751575262545,
      "accuracy_std_dev": 0.014386829490645235,
      "run_count": 13,
      "total_questions": 2571,
      "avg_cost_per_task_usd": 0.23145275106962276,
      "avg_duration_per_task_ms": 284144.37495138077,
      "avg_output_tokens_per_task": 52451.106184364064,
      "last_run_timestamp": "2026-10-31T16:57:18.670Z"
    },
    {
      "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": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "StreamLake",
      "accuracy": 0.8397306000000001,
      "accuracy_std_dev": 0.024613795376576923,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.16074421262626265,
      "avg_duration_per_task_ms": 1000006.3535353536,
      "avg_output_tokens_per_task": 52830.63737373737,
      "last_run_timestamp": "2026-10-31T16:39:29.347Z"
    },
    {
      "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": 0.839693111111111,
      "accuracy_std_dev": 0.009704287579832906,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.0060121301907968576,
      "avg_duration_per_task_ms": 439670.8260381594,
      "avg_output_tokens_per_task": 25980.03591470258,
      "last_run_timestamp": "2026-10-31T16:45:40.019Z"
    },
    {
      "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": 0.8391730814863104,
      "accuracy_std_dev": 0.011927781960856079,
      "run_count": 31,
      "total_questions": 6136,
      "avg_cost_per_task_usd": 0.26252152737940027,
      "avg_duration_per_task_ms": 275082.67649934813,
      "avg_output_tokens_per_task": 57033.21740547588,
      "last_run_timestamp": "2026-10-31T16:16:55.693Z"
    },
    {
      "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": 0.8387440275454993,
      "accuracy_std_dev": 0.019690853286246784,
      "run_count": 11,
      "total_questions": 2033,
      "avg_cost_per_task_usd": 0.07400631874077719,
      "avg_duration_per_task_ms": 403244.14658140676,
      "avg_output_tokens_per_task": 24425.756025577964,
      "last_run_timestamp": "2026-10-31T16:18:04.837Z"
    },
    {
      "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": 0.8386898181818183,
      "accuracy_std_dev": 0.022320469186247312,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.1030642401285583,
      "avg_duration_per_task_ms": 384608.90220385673,
      "avg_output_tokens_per_task": 35835.99357208448,
      "last_run_timestamp": "2026-10-31T08:36:48.545Z"
    },
    {
      "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": 0.8384273393416928,
      "accuracy_std_dev": 0.02801308482872427,
      "run_count": 13,
      "total_questions": 2552,
      "avg_cost_per_task_usd": 0.06636876214733543,
      "avg_duration_per_task_ms": 1061039.8240595611,
      "avg_output_tokens_per_task": 33105.52742946708,
      "last_run_timestamp": "2026-10-31T16:05:36.103Z"
    },
    {
      "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": 0.8383837142857142,
      "accuracy_std_dev": 0.012139811903865117,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.1527254054834055,
      "avg_duration_per_task_ms": 220532.05050505052,
      "avg_output_tokens_per_task": 34470.08658008658,
      "last_run_timestamp": "2026-10-31T08:45:52.318Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Z.AI",
      "accuracy": 0.8383836923076923,
      "accuracy_std_dev": 0.06521496654955188,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.20231615501165504,
      "avg_duration_per_task_ms": 532695.9992229992,
      "avg_output_tokens_per_task": 45956.61693861694,
      "last_run_timestamp": "2026-10-31T16:18:48.433Z"
    },
    {
      "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": 0.8379036028306059,
      "accuracy_std_dev": 0.019624931352535135,
      "run_count": 13,
      "total_questions": 2261,
      "avg_cost_per_task_usd": 0.09598150818222025,
      "avg_duration_per_task_ms": 610315.1362229103,
      "avg_output_tokens_per_task": 29918.706324635117,
      "last_run_timestamp": "2026-10-31T16:13:36.316Z"
    },
    {
      "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": "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": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8374772307692306,
      "accuracy_std_dev": 0.012060975124207245,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.08510472455322457,
      "avg_duration_per_task_ms": 722897.0310800311,
      "avg_output_tokens_per_task": 49108.018648018646,
      "last_run_timestamp": "2026-10-31T16:46:10.481Z"
    },
    {
      "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": 0.8371593636363637,
      "accuracy_std_dev": 0.013027195241792026,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.141865238292011,
      "avg_duration_per_task_ms": 181084.47015610652,
      "avg_output_tokens_per_task": 32015.05968778696,
      "last_run_timestamp": "2026-10-31T08:16:11.213Z"
    },
    {
      "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": "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": 0.8361277244542866,
      "accuracy_std_dev": 0.026614847390527968,
      "run_count": 16,
      "total_questions": 3161,
      "avg_cost_per_task_usd": 0.02298093546346093,
      "avg_duration_per_task_ms": 307170.6646630813,
      "avg_output_tokens_per_task": 18491.486871243276,
      "last_run_timestamp": "2026-10-31T08:31:11.445Z"
    },
    {
      "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": "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": 0.8356642307692309,
      "accuracy_std_dev": 0.03305920742722029,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.05416566394716394,
      "avg_duration_per_task_ms": 248261.90637140637,
      "avg_output_tokens_per_task": 22972.331002331004,
      "last_run_timestamp": "2026-10-31T16:08:08.536Z"
    },
    {
      "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": 0.8348581735005084,
      "accuracy_std_dev": 0.011926219294758369,
      "run_count": 18,
      "total_questions": 2951,
      "avg_cost_per_task_usd": 0.1190305462555066,
      "avg_duration_per_task_ms": 291312.74483226024,
      "avg_output_tokens_per_task": 28021.837004405286,
      "last_run_timestamp": "2026-10-31T08:09:30.690Z"
    },
    {
      "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": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Mara",
      "accuracy": 0.8346120933503837,
      "accuracy_std_dev": 0.029860364128538167,
      "run_count": 6,
      "total_questions": 782,
      "avg_cost_per_task_usd": 0.09530405242966751,
      "avg_duration_per_task_ms": 330065.4283887468,
      "avg_output_tokens_per_task": 39472.66240409207,
      "last_run_timestamp": "2026-10-31T16:08:31.063Z"
    },
    {
      "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": 0.8344576188870152,
      "accuracy_std_dev": 0.007843600547934322,
      "run_count": 3,
      "total_questions": 593,
      "avg_cost_per_task_usd": 0.033416617200674535,
      "avg_duration_per_task_ms": 208747.37942664418,
      "avg_output_tokens_per_task": 14517.00168634064,
      "last_run_timestamp": "2026-10-31T08:47:23.899Z"
    },
    {
      "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": 0.8344555555555555,
      "accuracy_std_dev": 0.01969686689756132,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.09693281088664421,
      "avg_duration_per_task_ms": 1118597.827721661,
      "avg_output_tokens_per_task": 44356.06565656565,
      "last_run_timestamp": "2026-10-31T16:08:56.903Z"
    },
    {
      "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": 0.8341543388067869,
      "accuracy_std_dev": 0.008301927538228698,
      "run_count": 10,
      "total_questions": 1827,
      "avg_cost_per_task_usd": 0.0431522140120416,
      "avg_duration_per_task_ms": 94743.94636015326,
      "avg_output_tokens_per_task": 15061.821018062397,
      "last_run_timestamp": "2026-10-31T08:22:53.415Z"
    },
    {
      "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": "Alibaba",
      "accuracy": 0.8341254705882353,
      "accuracy_std_dev": 0.010649185432078013,
      "run_count": 17,
      "total_questions": 3366,
      "avg_cost_per_task_usd": 0.25025737195484254,
      "avg_duration_per_task_ms": 384709.65567439096,
      "avg_output_tokens_per_task": 40253.002970885325,
      "last_run_timestamp": "2026-10-31T08:41:02.403Z"
    },
    {
      "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-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Cloudflare",
      "accuracy": 0.8338420961538461,
      "accuracy_std_dev": 0.01798642514244685,
      "run_count": 17,
      "total_questions": 3276,
      "avg_cost_per_task_usd": 0.2751184893162393,
      "avg_duration_per_task_ms": 391372.9496336996,
      "avg_output_tokens_per_task": 46721.76800976801,
      "last_run_timestamp": "2026-10-31T16:56:08.871Z"
    },
    {
      "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": 0.8337216923076923,
      "accuracy_std_dev": 0.02540450493782736,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.13457357070707068,
      "avg_duration_per_task_ms": 539843.0198135198,
      "avg_output_tokens_per_task": 43011.44755244755,
      "last_run_timestamp": "2026-10-31T16:18:28.926Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-flash-20260826",
      "model_slug": "qwen/qwen3.8-flash",
      "model_name": "Qwen: Qwen3.8 Flash",
      "provider_name": "Makora",
      "accuracy": 0.8333329999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 156,
      "avg_cost_per_task_usd": 0.003982064102564103,
      "avg_duration_per_task_ms": 34178.44871794872,
      "avg_output_tokens_per_task": 8341.692307692309,
      "last_run_timestamp": "2026-09-09T15:03:02.824Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "CoreWeave",
      "accuracy": 0.833146111111111,
      "accuracy_std_dev": 0.01582137855662708,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.13870717059483728,
      "avg_duration_per_task_ms": 298196.682379349,
      "avg_output_tokens_per_task": 57171.39337822671,
      "last_run_timestamp": "2026-10-31T16:45:48.945Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Cloudflare",
      "accuracy": 0.8330725023474179,
      "accuracy_std_dev": 0.010168177287347969,
      "run_count": 7,
      "total_questions": 1278,
      "avg_cost_per_task_usd": 0.029710525039123633,
      "avg_duration_per_task_ms": 655555.5469483568,
      "avg_output_tokens_per_task": 42323.18544600939,
      "last_run_timestamp": "2026-10-03T09:40:53.626Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Parasail",
      "accuracy": 0.8328014100518549,
      "accuracy_std_dev": 0.010819971665955098,
      "run_count": 13,
      "total_questions": 2507,
      "avg_cost_per_task_usd": 0.23187096051057043,
      "avg_duration_per_task_ms": 302340.8938970881,
      "avg_output_tokens_per_task": 52526.70203430395,
      "last_run_timestamp": "2026-10-31T16:28:00.196Z"
    },
    {
      "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": 0.832772,
      "accuracy_std_dev": 0.0698509274145066,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.024015700336700335,
      "avg_duration_per_task_ms": 613738.0185185185,
      "avg_output_tokens_per_task": 59848.74074074074,
      "last_run_timestamp": "2026-10-31T08:32:40.133Z"
    },
    {
      "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": 0.8327310906791908,
      "accuracy_std_dev": 0.020914023298509846,
      "run_count": 14,
      "total_questions": 2768,
      "avg_cost_per_task_usd": 0.02508389053468208,
      "avg_duration_per_task_ms": 393221.2268786127,
      "avg_output_tokens_per_task": 46919.84176300578,
      "last_run_timestamp": "2026-10-31T08:25:10.696Z"
    },
    {
      "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": 0.8326857692307692,
      "accuracy_std_dev": 0.011020358329744859,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.03860720590520591,
      "avg_duration_per_task_ms": 230762.9052059052,
      "avg_output_tokens_per_task": 14476.024864024865,
      "last_run_timestamp": "2026-10-31T08:25:00.370Z"
    },
    {
      "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": 0.8325688873598369,
      "accuracy_std_dev": 0.019237670701256954,
      "run_count": 10,
      "total_questions": 1962,
      "avg_cost_per_task_usd": 0.013576182466870541,
      "avg_duration_per_task_ms": 438297.6029561672,
      "avg_output_tokens_per_task": 39848.997961264016,
      "last_run_timestamp": "2026-10-31T08:26:26.747Z"
    },
    {
      "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": 0.8325147922396857,
      "accuracy_std_dev": 0.029981271466154875,
      "run_count": 11,
      "total_questions": 2036,
      "avg_cost_per_task_usd": 0.011545964145383104,
      "avg_duration_per_task_ms": 399504.8664047151,
      "avg_output_tokens_per_task": 22873.502946954814,
      "last_run_timestamp": "2026-10-31T08:40:04.563Z"
    },
    {
      "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": 0.8323464262877442,
      "accuracy_std_dev": 0.027721810831950897,
      "run_count": 9,
      "total_questions": 1689,
      "avg_cost_per_task_usd": 0.006784689757252812,
      "avg_duration_per_task_ms": 749921.5624629959,
      "avg_output_tokens_per_task": 18211.90763765542,
      "last_run_timestamp": "2026-10-31T16:04:29.121Z"
    },
    {
      "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": 0.8323081814454126,
      "accuracy_std_dev": 0.01351251744901741,
      "run_count": 10,
      "total_questions": 1951,
      "avg_cost_per_task_usd": 0.15155813992824194,
      "avg_duration_per_task_ms": 272697.51101998973,
      "avg_output_tokens_per_task": 34260.651973347005,
      "last_run_timestamp": "2026-10-31T08:37:13.096Z"
    },
    {
      "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": 0.8320106428571429,
      "accuracy_std_dev": 0.007771735884483182,
      "run_count": 14,
      "total_questions": 2772,
      "avg_cost_per_task_usd": 0.07774830699855699,
      "avg_duration_per_task_ms": 706637.7438672439,
      "avg_output_tokens_per_task": 57999.62698412698,
      "last_run_timestamp": "2026-10-31T16:05:36.105Z"
    },
    {
      "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": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "AkashML",
      "accuracy": 0.8319222756474104,
      "accuracy_std_dev": 0.02047048229572281,
      "run_count": 21,
      "total_questions": 4016,
      "avg_cost_per_task_usd": 0.04150572758964145,
      "avg_duration_per_task_ms": 442057.90986055776,
      "avg_output_tokens_per_task": 45303.60109561753,
      "last_run_timestamp": "2026-10-31T16:27:38.037Z"
    },
    {
      "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": 0.8318478292335116,
      "accuracy_std_dev": 0.010811745465322452,
      "run_count": 17,
      "total_questions": 2805,
      "avg_cost_per_task_usd": 0.020570678431372547,
      "avg_duration_per_task_ms": 255956.05668449198,
      "avg_output_tokens_per_task": 40944.02673796791,
      "last_run_timestamp": "2026-10-31T08:20:07.217Z"
    },
    {
      "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": 0.831769061433447,
      "accuracy_std_dev": 0.006801018573632123,
      "run_count": 12,
      "total_questions": 2344,
      "avg_cost_per_task_usd": 0.03448967320819113,
      "avg_duration_per_task_ms": 208520.74530716724,
      "avg_output_tokens_per_task": 14194.120733788395,
      "last_run_timestamp": "2026-10-31T08:37:28.112Z"
    },
    {
      "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": 0.8316496,
      "accuracy_std_dev": 0.007062535765573171,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.022192562626262628,
      "avg_duration_per_task_ms": 11557.427272727273,
      "avg_output_tokens_per_task": 14346.162626262627,
      "last_run_timestamp": "2026-10-31T08:15:46.337Z"
    },
    {
      "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": "PrimeIntellect",
      "accuracy": 0.831532,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 185,
      "avg_cost_per_task_usd": 0.01692656216216216,
      "avg_duration_per_task_ms": 560352.7027027027,
      "avg_output_tokens_per_task": 13578.437837837839,
      "last_run_timestamp": "2026-10-31T18:41:15.616Z"
    },
    {
      "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": 0.8314562888513515,
      "accuracy_std_dev": 0.012434701966673737,
      "run_count": 10,
      "total_questions": 1776,
      "avg_cost_per_task_usd": 0.21311484403153153,
      "avg_duration_per_task_ms": 464295.1722972973,
      "avg_output_tokens_per_task": 48323.44707207207,
      "last_run_timestamp": "2026-10-31T16:35:08.642Z"
    },
    {
      "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": 0.8314150997442455,
      "accuracy_std_dev": 0.012107134010057875,
      "run_count": 8,
      "total_questions": 1564,
      "avg_cost_per_task_usd": 0.02770090409207161,
      "avg_duration_per_task_ms": 583377.9558823529,
      "avg_output_tokens_per_task": 14708.68989769821,
      "last_run_timestamp": "2026-10-31T08:48:50.088Z"
    },
    {
      "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": 0.8313539815914488,
      "accuracy_std_dev": 0.013525888763386546,
      "run_count": 9,
      "total_questions": 1684,
      "avg_cost_per_task_usd": 0.02395000890736342,
      "avg_duration_per_task_ms": 287524.2090261283,
      "avg_output_tokens_per_task": 47679.917458432305,
      "last_run_timestamp": "2026-10-31T08:55:40.953Z"
    },
    {
      "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/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "CoreWeave",
      "accuracy": 0.8311777307159355,
      "accuracy_std_dev": 0.014826610062963942,
      "run_count": 14,
      "total_questions": 2165,
      "avg_cost_per_task_usd": 0.039417619399538105,
      "avg_duration_per_task_ms": 289131.9612009238,
      "avg_output_tokens_per_task": 32076.019399538105,
      "last_run_timestamp": "2026-10-31T16:19:08.399Z"
    },
    {
      "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": 0.8310883333333333,
      "accuracy_std_dev": 0.009654676391377534,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.20155932323232326,
      "avg_duration_per_task_ms": 266665.86195286195,
      "avg_output_tokens_per_task": 37910.23063973064,
      "last_run_timestamp": "2026-10-31T08:09:31.329Z"
    },
    {
      "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": 0.8309762,
      "accuracy_std_dev": 0.004338052761320463,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.04722357777777777,
      "avg_duration_per_task_ms": 258256.24646464645,
      "avg_output_tokens_per_task": 15298.06111111111,
      "last_run_timestamp": "2026-10-31T08:20:09.369Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Alibaba",
      "accuracy": 0.8304553494159472,
      "accuracy_std_dev": 0.047802944677603286,
      "run_count": 10,
      "total_questions": 1969,
      "avg_cost_per_task_usd": 0.194895594718131,
      "avg_duration_per_task_ms": 603493.3336719146,
      "avg_output_tokens_per_task": 46389.83697308278,
      "last_run_timestamp": "2026-10-31T16:25:21.093Z"
    },
    {
      "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": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Cloudflare",
      "accuracy": 0.8302074978687126,
      "accuracy_std_dev": 0.018798024809619108,
      "run_count": 12,
      "total_questions": 2346,
      "avg_cost_per_task_usd": 0.14303728772378516,
      "avg_duration_per_task_ms": 291115.02642796247,
      "avg_output_tokens_per_task": 32292.15643648764,
      "last_run_timestamp": "2026-10-31T08:14:43.196Z"
    },
    {
      "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-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": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "Venice",
      "accuracy": 0.8295141534054742,
      "accuracy_std_dev": 0.11811955384270635,
      "run_count": 8,
      "total_questions": 1571,
      "avg_cost_per_task_usd": 0.12611730617441116,
      "avg_duration_per_task_ms": 777712.6670910248,
      "avg_output_tokens_per_task": 38502.726288987906,
      "last_run_timestamp": "2026-10-31T16:02:14.362Z"
    },
    {
      "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": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8292178888888889,
      "accuracy_std_dev": 0.011827528589828504,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.00797142255892256,
      "avg_duration_per_task_ms": 508240.4511784512,
      "avg_output_tokens_per_task": 20513.609988776654,
      "last_run_timestamp": "2026-10-31T16:01:53.649Z"
    },
    {
      "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": 0.8288813468715698,
      "accuracy_std_dev": 0.048221988340541345,
      "run_count": 14,
      "total_questions": 2733,
      "avg_cost_per_task_usd": 0.09670949213318698,
      "avg_duration_per_task_ms": 542922.2052689352,
      "avg_output_tokens_per_task": 29030.10171972192,
      "last_run_timestamp": "2026-10-31T08:36:47.227Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "CoreWeave",
      "accuracy": 0.828497570984456,
      "accuracy_std_dev": 0.015230029187102692,
      "run_count": 5,
      "total_questions": 965,
      "avg_cost_per_task_usd": 0.11257541554404144,
      "avg_duration_per_task_ms": 438124.0186528497,
      "avg_output_tokens_per_task": 31261.672538860104,
      "last_run_timestamp": "2026-08-29T10:53:09.444Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Decart",
      "accuracy": 0.82828275,
      "accuracy_std_dev": 0.004610363346581274,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.0223316047979798,
      "avg_duration_per_task_ms": 615131.3409090909,
      "avg_output_tokens_per_task": 56159.15404040404,
      "last_run_timestamp": "2026-10-17T08:18:07.611Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Friendli",
      "accuracy": 0.8280963467299964,
      "accuracy_std_dev": 0.019705580037059044,
      "run_count": 14,
      "total_questions": 2737,
      "avg_cost_per_task_usd": 0.2455083120204604,
      "avg_duration_per_task_ms": 398401.90317866276,
      "avg_output_tokens_per_task": 55634.735842162954,
      "last_run_timestamp": "2026-10-31T16:06:46.433Z"
    },
    {
      "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": 0.8280958888888889,
      "accuracy_std_dev": 0.00754749308555785,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.047590010101010105,
      "avg_duration_per_task_ms": 453658.93883277214,
      "avg_output_tokens_per_task": 26356.35521885522,
      "last_run_timestamp": "2026-10-31T16:10:31.145Z"
    },
    {
      "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": 0.8280885384615385,
      "accuracy_std_dev": 0.053874268721241306,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.05164174397824397,
      "avg_duration_per_task_ms": 528859.5505050505,
      "avg_output_tokens_per_task": 24724.757575757576,
      "last_run_timestamp": "2026-10-31T16:01:53.625Z"
    },
    {
      "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": 0.8278389154691463,
      "accuracy_std_dev": 0.01931102964125829,
      "run_count": 6,
      "total_questions": 1183,
      "avg_cost_per_task_usd": 0.20861228233305154,
      "avg_duration_per_task_ms": 311401.96280642436,
      "avg_output_tokens_per_task": 47465.20540997464,
      "last_run_timestamp": "2026-09-12T08:53:53.147Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "Phala",
      "accuracy": 0.8272467692307692,
      "accuracy_std_dev": 0.013669555786794399,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.06361650932400933,
      "avg_duration_per_task_ms": 543058.7459207459,
      "avg_output_tokens_per_task": 50077.30147630148,
      "last_run_timestamp": "2026-10-31T16:08:56.901Z"
    },
    {
      "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": 0.8270583636363636,
      "accuracy_std_dev": 0.00941702237510474,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.04288225068870524,
      "avg_duration_per_task_ms": 229942.66437098256,
      "avg_output_tokens_per_task": 14178.259871441689,
      "last_run_timestamp": "2026-10-31T08:36:47.145Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "BaseTen",
      "accuracy": 0.8269358,
      "accuracy_std_dev": 0.006767606146932625,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.04578402121212121,
      "avg_duration_per_task_ms": 60291.652525252524,
      "avg_output_tokens_per_task": 11285.754545454545,
      "last_run_timestamp": "2026-09-26T09:19:03.901Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "Google",
      "accuracy": 0.8268397142857142,
      "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/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "auto-routing",
      "accuracy": 0.8266562135355892,
      "accuracy_std_dev": 0.04220692579963176,
      "run_count": 13,
      "total_questions": 2571,
      "avg_cost_per_task_usd": 0.08716451886425515,
      "avg_duration_per_task_ms": 756672.2042007002,
      "avg_output_tokens_per_task": 31524.158304161803,
      "last_run_timestamp": "2026-10-31T16:01:53.624Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Friendli",
      "accuracy": 0.8266521780172412,
      "accuracy_std_dev": 0.022372414615840108,
      "run_count": 12,
      "total_questions": 2320,
      "avg_cost_per_task_usd": 0.02234062025862069,
      "avg_duration_per_task_ms": 235753.16637931034,
      "avg_output_tokens_per_task": 44456.286637931036,
      "last_run_timestamp": "2026-10-31T08:28:04.955Z"
    },
    {
      "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-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Friendli",
      "accuracy": 0.826459,
      "accuracy_std_dev": 0.009410002267091476,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.14383780639730645,
      "avg_duration_per_task_ms": 198190.3564814815,
      "avg_output_tokens_per_task": 35703.72769360269,
      "last_run_timestamp": "2026-10-31T08:32:01.077Z"
    },
    {
      "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": 0.826417643153527,
      "accuracy_std_dev": 0.01028202813164795,
      "run_count": 4,
      "total_questions": 723,
      "avg_cost_per_task_usd": 0.010974759336099586,
      "avg_duration_per_task_ms": 1541635.907330567,
      "avg_output_tokens_per_task": 31233.395573997233,
      "last_run_timestamp": "2026-10-31T08:24:48.465Z"
    },
    {
      "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": 0.8264159143968872,
      "accuracy_std_dev": 0.0040531378816294926,
      "run_count": 4,
      "total_questions": 771,
      "avg_cost_per_task_usd": 0.13822659403372245,
      "avg_duration_per_task_ms": 272716.0363164721,
      "avg_output_tokens_per_task": 34382.090791180286,
      "last_run_timestamp": "2026-10-31T08:38:15.143Z"
    },
    {
      "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": 0.8264111325878595,
      "accuracy_std_dev": 0.01425677691801343,
      "run_count": 10,
      "total_questions": 1878,
      "avg_cost_per_task_usd": 0.02454355484558041,
      "avg_duration_per_task_ms": 264967.17305644305,
      "avg_output_tokens_per_task": 50263.019169329076,
      "last_run_timestamp": "2026-10-31T08:17:29.105Z"
    },
    {
      "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": 0.8263836795903439,
      "accuracy_std_dev": 0.029166866420929173,
      "run_count": 7,
      "total_questions": 1367,
      "avg_cost_per_task_usd": 0.07172207900512069,
      "avg_duration_per_task_ms": 248028.12801755668,
      "avg_output_tokens_per_task": 21803.532553035846,
      "last_run_timestamp": "2026-10-31T09:17:12.929Z"
    },
    {
      "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": 0.826225,
      "accuracy_std_dev": 0.028929060222090415,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.09852584455667786,
      "avg_duration_per_task_ms": 1409812.7817059483,
      "avg_output_tokens_per_task": 45348.74074074074,
      "last_run_timestamp": "2026-10-31T16:08:33.697Z"
    },
    {
      "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": 0.8259871818181819,
      "accuracy_std_dev": 0.010774204132752242,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.03615010422405877,
      "avg_duration_per_task_ms": 305458.8122130395,
      "avg_output_tokens_per_task": 15557.78971533517,
      "last_run_timestamp": "2026-10-31T08:32:33.985Z"
    },
    {
      "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": 0.8258563532084044,
      "accuracy_std_dev": 0.010360732650818109,
      "run_count": 9,
      "total_questions": 1761,
      "avg_cost_per_task_usd": 0.008003823395797843,
      "avg_duration_per_task_ms": 338944.8080636002,
      "avg_output_tokens_per_task": 19222.395797842135,
      "last_run_timestamp": "2026-10-31T16:08:31.066Z"
    },
    {
      "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": 0.8258509999999999,
      "accuracy_std_dev": 0.01267571617086958,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.194836923681257,
      "avg_duration_per_task_ms": 926868.8086419753,
      "avg_output_tokens_per_task": 51862.713243546576,
      "last_run_timestamp": "2026-10-31T16:50:02.353Z"
    },
    {
      "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": 0.82575775,
      "accuracy_std_dev": 0.010065865980505628,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.14380135795454543,
      "avg_duration_per_task_ms": 418612.6521464646,
      "avg_output_tokens_per_task": 28784.420454545456,
      "last_run_timestamp": "2026-10-31T08:40:11.483Z"
    },
    {
      "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": "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": 0.825389385046729,
      "accuracy_std_dev": 0.014184335849948446,
      "run_count": 11,
      "total_questions": 2140,
      "avg_cost_per_task_usd": 0.03383954813084113,
      "avg_duration_per_task_ms": 407090.0654205608,
      "avg_output_tokens_per_task": 47991.5476635514,
      "last_run_timestamp": "2026-10-31T08:48:01.125Z"
    },
    {
      "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": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "AkashML",
      "accuracy": 0.8242303292875989,
      "accuracy_std_dev": 0.016434035763013298,
      "run_count": 20,
      "total_questions": 3790,
      "avg_cost_per_task_usd": 0.030382969920844324,
      "avg_duration_per_task_ms": 266171.0530343008,
      "avg_output_tokens_per_task": 14053.306068601583,
      "last_run_timestamp": "2026-10-31T08:10:37.346Z"
    },
    {
      "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": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Alibaba",
      "accuracy": 0.8238164069389231,
      "accuracy_std_dev": 0.03209164605620812,
      "run_count": 14,
      "total_questions": 2767,
      "avg_cost_per_task_usd": 0.08185260065052402,
      "avg_duration_per_task_ms": 342308.28370075894,
      "avg_output_tokens_per_task": 20334.039392844235,
      "last_run_timestamp": "2026-10-31T16:02:14.359Z"
    },
    {
      "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": "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": "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": 0.8225450489150423,
      "accuracy_std_dev": 0.023532795309234442,
      "run_count": 14,
      "total_questions": 2719,
      "avg_cost_per_task_usd": 0.026525477749172497,
      "avg_duration_per_task_ms": 850874.7859507172,
      "avg_output_tokens_per_task": 49891.477013607946,
      "last_run_timestamp": "2026-10-31T08:42:18.327Z"
    },
    {
      "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": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Chutes",
      "accuracy": 0.8223905,
      "accuracy_std_dev": 0.019262412208322532,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.03192426473063974,
      "avg_duration_per_task_ms": 260830.34259259258,
      "avg_output_tokens_per_task": 14583.287878787878,
      "last_run_timestamp": "2026-10-31T08:17:29.187Z"
    },
    {
      "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/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "DeepInfra",
      "accuracy": 0.8221762091333679,
      "accuracy_std_dev": 0.01082457395235492,
      "run_count": 10,
      "total_questions": 1927,
      "avg_cost_per_task_usd": 0.024801258432797096,
      "avg_duration_per_task_ms": 452703.2304099637,
      "avg_output_tokens_per_task": 24752.417747794498,
      "last_run_timestamp": "2026-10-31T16:42:08.154Z"
    },
    {
      "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": 0.8216674158517757,
      "accuracy_std_dev": 0.01474539844154779,
      "run_count": 10,
      "total_questions": 1943,
      "avg_cost_per_task_usd": 0.018114979413278436,
      "avg_duration_per_task_ms": 598455.0849202265,
      "avg_output_tokens_per_task": 20142.77406073083,
      "last_run_timestamp": "2026-10-31T16:13:07.320Z"
    },
    {
      "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": 0.8216317349778621,
      "accuracy_std_dev": 0.09289881485376644,
      "run_count": 8,
      "total_questions": 1581,
      "avg_cost_per_task_usd": 0.09027165528146741,
      "avg_duration_per_task_ms": 645885.798228969,
      "avg_output_tokens_per_task": 25622.579380139152,
      "last_run_timestamp": "2026-10-31T16:16:30.410Z"
    },
    {
      "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": 0.8215486666666668,
      "accuracy_std_dev": 0.015611992961253275,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.02833054377104377,
      "avg_duration_per_task_ms": 228888.5606060606,
      "avg_output_tokens_per_task": 56421.28451178451,
      "last_run_timestamp": "2026-10-31T08:03:09.999Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "DeepInfra",
      "accuracy": 0.8214198161925602,
      "accuracy_std_dev": 0.04954945097072217,
      "run_count": 8,
      "total_questions": 1371,
      "avg_cost_per_task_usd": 0.04645164551422319,
      "avg_duration_per_task_ms": 608936.6484318016,
      "avg_output_tokens_per_task": 28638.388767323122,
      "last_run_timestamp": "2026-09-26T09:56:00.145Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Makora",
      "accuracy": 0.8213506405228759,
      "accuracy_std_dev": 0.02248216924587128,
      "run_count": 3,
      "total_questions": 459,
      "avg_cost_per_task_usd": 0.08758356862745097,
      "avg_duration_per_task_ms": 304618.9760348584,
      "avg_output_tokens_per_task": 34720.03050108933,
      "last_run_timestamp": "2026-08-29T10:57:24.797Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta/muse-glimmer-30b-20260810",
      "model_slug": "meta/muse-glimmer-30b",
      "model_name": "Meta: Muse Glimmer 30B",
      "provider_name": "Parasail",
      "accuracy": 0.8212070710659898,
      "accuracy_std_dev": 0.011946281848340899,
      "run_count": 3,
      "total_questions": 591,
      "avg_cost_per_task_usd": 0.01807649915397631,
      "avg_duration_per_task_ms": 302940.0592216582,
      "avg_output_tokens_per_task": 16413.382402707277,
      "last_run_timestamp": "2026-08-29T09:59:35.603Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Io Net",
      "accuracy": 0.8210525603238865,
      "accuracy_std_dev": 0.07599126782917703,
      "run_count": 7,
      "total_questions": 1235,
      "avg_cost_per_task_usd": 0.007044106882591094,
      "avg_duration_per_task_ms": 525854.9595141701,
      "avg_output_tokens_per_task": 14201.662348178137,
      "last_run_timestamp": "2026-10-03T10:22:02.091Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "CoreWeave",
      "accuracy": 0.8203833076923077,
      "accuracy_std_dev": 0.018076297388844354,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.042996285159285164,
      "avg_duration_per_task_ms": 241748.8038073038,
      "avg_output_tokens_per_task": 14318.661227661227,
      "last_run_timestamp": "2026-10-31T08:09:30.241Z"
    },
    {
      "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": 0.8202748856394587,
      "accuracy_std_dev": 0.060149001089135595,
      "run_count": 24,
      "total_questions": 4582,
      "avg_cost_per_task_usd": 0.0076442278481012645,
      "avg_duration_per_task_ms": 496763.3814927979,
      "avg_output_tokens_per_task": 17412.37341772152,
      "last_run_timestamp": "2026-10-31T16:45:45.611Z"
    },
    {
      "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": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Mancer 2",
      "accuracy": 0.8200275393258427,
      "accuracy_std_dev": 0.011794769394173345,
      "run_count": 9,
      "total_questions": 1691,
      "avg_cost_per_task_usd": 0.03818077646363098,
      "avg_duration_per_task_ms": 215799.96629213484,
      "avg_output_tokens_per_task": 14692.68894145476,
      "last_run_timestamp": "2026-10-31T08:55:11.136Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "StreamLake",
      "accuracy": 0.819865,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.13854771717171718,
      "avg_duration_per_task_ms": 540824.1212121212,
      "avg_output_tokens_per_task": 31240.651515151516,
      "last_run_timestamp": "2026-09-07T06:00:57.184Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "auto-routing",
      "accuracy": 0.8194060909090909,
      "accuracy_std_dev": 0.013884218840464785,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.033594433884297524,
      "avg_duration_per_task_ms": 211994.46005509642,
      "avg_output_tokens_per_task": 13897.44214876033,
      "last_run_timestamp": "2026-10-31T08:09:30.951Z"
    },
    {
      "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": 0.819304,
      "accuracy_std_dev": 0.02899661507831561,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.08773566049382715,
      "avg_duration_per_task_ms": 1029441.7283950617,
      "avg_output_tokens_per_task": 43316.07407407407,
      "last_run_timestamp": "2026-10-31T16:08:08.535Z"
    },
    {
      "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": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "DigitalOcean",
      "accuracy": 0.8191759532044761,
      "accuracy_std_dev": 0.09354392281067757,
      "run_count": 10,
      "total_questions": 1966,
      "avg_cost_per_task_usd": 0.06251466022380467,
      "avg_duration_per_task_ms": 1731771.3148524924,
      "avg_output_tokens_per_task": 19617.884028484234,
      "last_run_timestamp": "2026-10-31T16:13:09.868Z"
    },
    {
      "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": "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": 0.8184407692307691,
      "accuracy_std_dev": 0.012119050255716884,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.3141125089355089,
      "avg_duration_per_task_ms": 289618.7264957265,
      "avg_output_tokens_per_task": 20750.078088578088,
      "last_run_timestamp": "2026-10-31T08:15:24.404Z"
    },
    {
      "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/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "Parasail",
      "accuracy": 0.8179862262773724,
      "accuracy_std_dev": 0.005388830217320935,
      "run_count": 9,
      "total_questions": 1781,
      "avg_cost_per_task_usd": 0.026104914093206064,
      "avg_duration_per_task_ms": 325164.1538461539,
      "avg_output_tokens_per_task": 26027.919708029196,
      "last_run_timestamp": "2026-10-31T16:04:48.178Z"
    },
    {
      "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": 0.8178727706422019,
      "accuracy_std_dev": 0.03174962566960436,
      "run_count": 5,
      "total_questions": 981,
      "avg_cost_per_task_usd": 0.023857419979612642,
      "avg_duration_per_task_ms": 166236.06523955148,
      "avg_output_tokens_per_task": 14523.781855249745,
      "last_run_timestamp": "2026-10-31T08:07:51.399Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta/muse-glimmer-30b-20260810",
      "model_slug": "meta/muse-glimmer-30b",
      "model_name": "Meta: Muse Glimmer 30B",
      "provider_name": "Together",
      "accuracy": 0.8178077777777777,
      "accuracy_std_dev": 0.009999885885768644,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.024910318742985414,
      "avg_duration_per_task_ms": 190389.1571268238,
      "avg_output_tokens_per_task": 16514.1734006734,
      "last_run_timestamp": "2026-10-03T10:10:35.343Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "StreamLake",
      "accuracy": 0.8177376144306652,
      "accuracy_std_dev": 0.03151916867911054,
      "run_count": 9,
      "total_questions": 1774,
      "avg_cost_per_task_usd": 0.013423886696730551,
      "avg_duration_per_task_ms": 855848.4718151072,
      "avg_output_tokens_per_task": 39679.556933483655,
      "last_run_timestamp": "2026-10-17T08:19:03.202Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "CoreWeave",
      "accuracy": 0.8176205555555557,
      "accuracy_std_dev": 0.01985605337372578,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.006378395061728395,
      "avg_duration_per_task_ms": 440890.9601571268,
      "avg_output_tokens_per_task": 18699.455667789,
      "last_run_timestamp": "2026-10-31T16:12:28.233Z"
    },
    {
      "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": 0.8174803333333334,
      "accuracy_std_dev": 0.015915170704568507,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.01320759595959596,
      "avg_duration_per_task_ms": 508912.26851851854,
      "avg_output_tokens_per_task": 12721.0867003367,
      "last_run_timestamp": "2026-10-31T08:36:26.277Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Phala",
      "accuracy": 0.8173077377622378,
      "accuracy_std_dev": 0.04183386114142466,
      "run_count": 3,
      "total_questions": 572,
      "avg_cost_per_task_usd": 0.00918055944055944,
      "avg_duration_per_task_ms": 556054.2482517483,
      "avg_output_tokens_per_task": 20397.216783216783,
      "last_run_timestamp": "2026-08-29T19:08:46.030Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Reka",
      "accuracy": 0.8172424786585366,
      "accuracy_std_dev": 0.032381379178472315,
      "run_count": 8,
      "total_questions": 984,
      "avg_cost_per_task_usd": 0.0567359319105691,
      "avg_duration_per_task_ms": 328244.918699187,
      "avg_output_tokens_per_task": 21710.94918699187,
      "last_run_timestamp": "2026-10-31T16:26:17.224Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "OpenInference",
      "accuracy": 0.8171817336142323,
      "accuracy_std_dev": 0.04982830145510509,
      "run_count": 11,
      "total_questions": 2136,
      "avg_cost_per_task_usd": 0.018343408239700373,
      "avg_duration_per_task_ms": 793559.7439138577,
      "avg_output_tokens_per_task": 48776.84925093633,
      "last_run_timestamp": "2026-10-31T08:31:53.887Z"
    },
    {
      "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": 0.8171795172255397,
      "accuracy_std_dev": 0.008683223783407119,
      "run_count": 11,
      "total_questions": 2177,
      "avg_cost_per_task_usd": 0.03297541662838769,
      "avg_duration_per_task_ms": 545292.2503445108,
      "avg_output_tokens_per_task": 15758.780431786863,
      "last_run_timestamp": "2026-10-31T08:29:48.679Z"
    },
    {
      "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": 0.8170594444444444,
      "accuracy_std_dev": 0.01614747979036502,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.007438095959595959,
      "avg_duration_per_task_ms": 496093.94051627384,
      "avg_output_tokens_per_task": 18405.01571268238,
      "last_run_timestamp": "2026-10-31T16:13:09.863Z"
    },
    {
      "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": "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": 0.8169574545454545,
      "accuracy_std_dev": 0.03032911546941362,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.029461516069788796,
      "avg_duration_per_task_ms": 587955.8774104683,
      "avg_output_tokens_per_task": 53445.840679522495,
      "last_run_timestamp": "2026-10-31T08:57:28.755Z"
    },
    {
      "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": 0.8168723333333333,
      "accuracy_std_dev": 0.007519404837559345,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.026859203703703705,
      "avg_duration_per_task_ms": 171179.37822671156,
      "avg_output_tokens_per_task": 23402.503928170594,
      "last_run_timestamp": "2026-10-31T16:09:44.821Z"
    },
    {
      "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": 0.8165543999999997,
      "accuracy_std_dev": 0.010562628756138299,
      "run_count": 30,
      "total_questions": 5940,
      "avg_cost_per_task_usd": 0.33263593888888887,
      "avg_duration_per_task_ms": 295154.08097643097,
      "avg_output_tokens_per_task": 21001.168855218853,
      "last_run_timestamp": "2026-10-31T08:27:57.306Z"
    },
    {
      "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": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "SiliconFlow",
      "accuracy": 0.815937111111111,
      "accuracy_std_dev": 0.004627552783646224,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.011142754769921435,
      "avg_duration_per_task_ms": 816705.024691358,
      "avg_output_tokens_per_task": 18593.037598204264,
      "last_run_timestamp": "2026-10-31T16:13:09.867Z"
    },
    {
      "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": 0.815563,
      "accuracy_std_dev": 0.04477470762482865,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.20474397923681256,
      "avg_duration_per_task_ms": 873869.0656565657,
      "avg_output_tokens_per_task": 51153.14253647587,
      "last_run_timestamp": "2026-10-31T16:08:08.537Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "BaseTen",
      "accuracy": 0.815286480254777,
      "accuracy_std_dev": 0.006665997576319691,
      "run_count": 4,
      "total_questions": 785,
      "avg_cost_per_task_usd": 0.08723442165605096,
      "avg_duration_per_task_ms": 185900.4152866242,
      "avg_output_tokens_per_task": 24915.47388535032,
      "last_run_timestamp": "2026-08-29T18:58:14.477Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Relace",
      "accuracy": 0.8152402988505746,
      "accuracy_std_dev": 0.029604949040447602,
      "run_count": 4,
      "total_questions": 783,
      "avg_cost_per_task_usd": 0.28207722477650066,
      "avg_duration_per_task_ms": 913648.006385696,
      "avg_output_tokens_per_task": 58474.00127713921,
      "last_run_timestamp": "2026-10-31T16:12:07.198Z"
    },
    {
      "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": 0.81523575,
      "accuracy_std_dev": 0.04077706865920479,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.14845152777777776,
      "avg_duration_per_task_ms": 1663278.9867424243,
      "avg_output_tokens_per_task": 48562.03977272727,
      "last_run_timestamp": "2026-10-31T16:08:56.833Z"
    },
    {
      "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": 0.8151218824415977,
      "accuracy_std_dev": 0.008978671073432733,
      "run_count": 7,
      "total_questions": 1327,
      "avg_cost_per_task_usd": 0.04307039035418237,
      "avg_duration_per_task_ms": 180891.98040693294,
      "avg_output_tokens_per_task": 17051.95177091183,
      "last_run_timestamp": "2026-10-31T08:27:47.450Z"
    },
    {
      "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": 0.8148975767263426,
      "accuracy_std_dev": 0.10918305406308805,
      "run_count": 9,
      "total_questions": 1564,
      "avg_cost_per_task_usd": 0.041120762148337583,
      "avg_duration_per_task_ms": 136841.38746803068,
      "avg_output_tokens_per_task": 28739.403452685423,
      "last_run_timestamp": "2026-10-31T16:08:08.538Z"
    },
    {
      "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": 0.8147523703703704,
      "accuracy_std_dev": 0.01034211180579509,
      "run_count": 27,
      "total_questions": 5346,
      "avg_cost_per_task_usd": 0.3634217265245044,
      "avg_duration_per_task_ms": 358529.47362514026,
      "avg_output_tokens_per_task": 22930.12757201646,
      "last_run_timestamp": "2026-10-31T08:33:07.626Z"
    },
    {
      "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": "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": "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": "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": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Phala",
      "accuracy": 0.8125701666666667,
      "accuracy_std_dev": 0.050032071865343405,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.20077214393939394,
      "avg_duration_per_task_ms": 1584443.372053872,
      "avg_output_tokens_per_task": 47804.85858585859,
      "last_run_timestamp": "2026-10-31T16:31:54.109Z"
    },
    {
      "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": "anthropic/claude-4.5-sonnet-20250929",
      "model_slug": "anthropic/claude-sonnet-4.5",
      "model_name": "Anthropic: Claude Sonnet 4.5",
      "provider_name": "Anthropic",
      "accuracy": 0.8120790625,
      "accuracy_std_dev": 0.01643967442830283,
      "run_count": 16,
      "total_questions": 3168,
      "avg_cost_per_task_usd": 0.3404591770833333,
      "avg_duration_per_task_ms": 353918.67329545453,
      "avg_output_tokens_per_task": 22524.086174242424,
      "last_run_timestamp": "2026-10-31T08:46:10.648Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "DigitalOcean",
      "accuracy": 0.812079,
      "accuracy_std_dev": 0.1279844797709863,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.15244413194444442,
      "avg_duration_per_task_ms": 1980931.956439394,
      "avg_output_tokens_per_task": 55523.080808080806,
      "last_run_timestamp": "2026-10-24T16:14:24.978Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3-ultra-550b-a55b-20260604",
      "model_slug": "nvidia/nemotron-3-ultra-550b-a55b",
      "model_name": "NVIDIA: Nemotron 3 Ultra",
      "provider_name": "Together",
      "accuracy": 0.812016,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 172,
      "avg_cost_per_task_usd": 0.14135672674418603,
      "avg_duration_per_task_ms": 305120.88372093026,
      "avg_output_tokens_per_task": 39843.74418604651,
      "last_run_timestamp": "2026-08-04T03:11:59.879Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Ionstream",
      "accuracy": 0.8118685,
      "accuracy_std_dev": 0.0618254994258033,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.042041156565656566,
      "avg_duration_per_task_ms": 1127766.4457070706,
      "avg_output_tokens_per_task": 24710.378787878788,
      "last_run_timestamp": "2026-08-29T19:26:41.546Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "auto-routing",
      "accuracy": 0.8116347777777777,
      "accuracy_std_dev": 0.0244922962662765,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.008957320426487093,
      "avg_duration_per_task_ms": 563804.7289562289,
      "avg_output_tokens_per_task": 19164.750841750843,
      "last_run_timestamp": "2026-10-31T16:08:08.535Z"
    },
    {
      "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": 0.810606,
      "accuracy_std_dev": 0.010546997511140292,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.044063582702020196,
      "avg_duration_per_task_ms": 62743.707702020205,
      "avg_output_tokens_per_task": 15119.022095959595,
      "last_run_timestamp": "2026-10-31T08:43:59.854Z"
    },
    {
      "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": 0.8105124444444445,
      "accuracy_std_dev": 0.0452643157786968,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.23235617452300789,
      "avg_duration_per_task_ms": 1158864.2744107745,
      "avg_output_tokens_per_task": 53109.03591470258,
      "last_run_timestamp": "2026-10-31T16:30:07.021Z"
    },
    {
      "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": 0.8103538311386522,
      "accuracy_std_dev": 0.10032075765294651,
      "run_count": 14,
      "total_questions": 2582,
      "avg_cost_per_task_usd": 0.041116779628195196,
      "avg_duration_per_task_ms": 153458.49147947328,
      "avg_output_tokens_per_task": 19676.14329976762,
      "last_run_timestamp": "2026-10-31T16:17:39.770Z"
    },
    {
      "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": 0.8102452857142856,
      "accuracy_std_dev": 0.09184708177837814,
      "run_count": 14,
      "total_questions": 2772,
      "avg_cost_per_task_usd": 0.09330289826839828,
      "avg_duration_per_task_ms": 515911.21789321786,
      "avg_output_tokens_per_task": 34419.12373737374,
      "last_run_timestamp": "2026-10-31T16:05:36.297Z"
    },
    {
      "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": 0.8102317777777777,
      "accuracy_std_dev": 0.01704457789821722,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.22739129012345682,
      "avg_duration_per_task_ms": 1030818.7098765433,
      "avg_output_tokens_per_task": 61856.86588103255,
      "last_run_timestamp": "2026-10-31T16:12:28.230Z"
    },
    {
      "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": "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": 0.809764,
      "accuracy_std_dev": 0.007273551310513057,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.027726367003367003,
      "avg_duration_per_task_ms": 248023.59595959596,
      "avg_output_tokens_per_task": 55228.13468013468,
      "last_run_timestamp": "2026-10-31T08:07:51.513Z"
    },
    {
      "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": 0.8096608426103648,
      "accuracy_std_dev": 0.012817931849127972,
      "run_count": 3,
      "total_questions": 521,
      "avg_cost_per_task_usd": 0.0079749443378119,
      "avg_duration_per_task_ms": 289711.4932821497,
      "avg_output_tokens_per_task": 31380.54894433781,
      "last_run_timestamp": "2026-10-31T16:04:23.357Z"
    },
    {
      "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": 0.8095958000000001,
      "accuracy_std_dev": 0.023247585009200426,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.11788618585858586,
      "avg_duration_per_task_ms": 776373.9363636364,
      "avg_output_tokens_per_task": 61313.756565656564,
      "last_run_timestamp": "2026-10-31T16:10:09.226Z"
    },
    {
      "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": "anthropic/claude-4.5-sonnet-20250929",
      "model_slug": "anthropic/claude-sonnet-4.5",
      "model_name": "Anthropic: Claude Sonnet 4.5",
      "provider_name": "Azure",
      "accuracy": 0.8091630714285715,
      "accuracy_std_dev": 0.01971105683716007,
      "run_count": 14,
      "total_questions": 2772,
      "avg_cost_per_task_usd": 0.34238485173160177,
      "avg_duration_per_task_ms": 356759.47510822513,
      "avg_output_tokens_per_task": 22659.93614718615,
      "last_run_timestamp": "2026-10-31T08:17:29.306Z"
    },
    {
      "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": "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/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "DekaLLM",
      "accuracy": 0.807532574733096,
      "accuracy_std_dev": 0.023704551349195916,
      "run_count": 6,
      "total_questions": 1124,
      "avg_cost_per_task_usd": 0.039211227758007124,
      "avg_duration_per_task_ms": 252786.0587188612,
      "avg_output_tokens_per_task": 39158.109430604985,
      "last_run_timestamp": "2026-10-31T16:09:44.822Z"
    },
    {
      "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": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Novita",
      "accuracy": 0.8063973333333333,
      "accuracy_std_dev": 0.024015690667191354,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.2794591790123457,
      "avg_duration_per_task_ms": 997160.2497194164,
      "avg_output_tokens_per_task": 63521.387205387204,
      "last_run_timestamp": "2026-10-31T16:12:28.234Z"
    },
    {
      "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": 0.8063972,
      "accuracy_std_dev": 0.10239770597117887,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.05410009898989898,
      "avg_duration_per_task_ms": 1231927.9055555556,
      "avg_output_tokens_per_task": 29021.94191919192,
      "last_run_timestamp": "2026-10-31T16:05:14.716Z"
    },
    {
      "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": "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": 0.8047136666666667,
      "accuracy_std_dev": 0.0407068763642377,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.02657082996632996,
      "avg_duration_per_task_ms": 316645.5987654321,
      "avg_output_tokens_per_task": 25739.34006734007,
      "last_run_timestamp": "2026-10-31T16:05:36.303Z"
    },
    {
      "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": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "DeepInfra",
      "accuracy": 0.8040673640873017,
      "accuracy_std_dev": 0.02506085284963421,
      "run_count": 7,
      "total_questions": 1008,
      "avg_cost_per_task_usd": 0.0705810753968254,
      "avg_duration_per_task_ms": 1609832.9444444445,
      "avg_output_tokens_per_task": 40293.83829365079,
      "last_run_timestamp": "2026-10-31T16:05:14.633Z"
    },
    {
      "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": 0.8040581192136969,
      "accuracy_std_dev": 0.06940046297396005,
      "run_count": 8,
      "total_questions": 1577,
      "avg_cost_per_task_usd": 0.19658499112238428,
      "avg_duration_per_task_ms": 1692732.626506024,
      "avg_output_tokens_per_task": 45873.13633481294,
      "last_run_timestamp": "2026-10-31T16:22:14.129Z"
    },
    {
      "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": 0.8038719166666666,
      "accuracy_std_dev": 0.01983021612530708,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.3387011611952862,
      "avg_duration_per_task_ms": 355211.85016835015,
      "avg_output_tokens_per_task": 22440.643097643097,
      "last_run_timestamp": "2026-10-31T08:53:23.022Z"
    },
    {
      "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": "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": 0.8028182609489051,
      "accuracy_std_dev": 0.07654548992882891,
      "run_count": 9,
      "total_questions": 1644,
      "avg_cost_per_task_usd": 0.019102108272506083,
      "avg_duration_per_task_ms": 309069.2311435523,
      "avg_output_tokens_per_task": 40420.16605839416,
      "last_run_timestamp": "2026-10-31T08:33:53.031Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Mistral",
      "accuracy": 0.8006073724696356,
      "accuracy_std_dev": 0.018676167604730915,
      "run_count": 5,
      "total_questions": 988,
      "avg_cost_per_task_usd": 0.026427639676113363,
      "avg_duration_per_task_ms": 228773.67307692306,
      "avg_output_tokens_per_task": 52620.16093117409,
      "last_run_timestamp": "2026-10-31T08:22:47.424Z"
    },
    {
      "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": 0.7990709066202091,
      "accuracy_std_dev": 0.036071343734006646,
      "run_count": 11,
      "total_questions": 1435,
      "avg_cost_per_task_usd": 0.00675192613240418,
      "avg_duration_per_task_ms": 310615.30383275263,
      "avg_output_tokens_per_task": 26833.981184668988,
      "last_run_timestamp": "2026-10-31T16:04:29.125Z"
    },
    {
      "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": 0.7981739230769231,
      "accuracy_std_dev": 0.03994574297037949,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.043110703574203574,
      "avg_duration_per_task_ms": 487860.4627039627,
      "avg_output_tokens_per_task": 42588.88306138306,
      "last_run_timestamp": "2026-10-31T16:13:09.867Z"
    },
    {
      "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": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7955478888888888,
      "accuracy_std_dev": 0.030696558538356792,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.039506446689113364,
      "avg_duration_per_task_ms": 192749.99270482603,
      "avg_output_tokens_per_task": 21847.699214365883,
      "last_run_timestamp": "2026-10-31T16:08:31.063Z"
    },
    {
      "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": 0.7955151838886546,
      "accuracy_std_dev": 0.021167952217358864,
      "run_count": 12,
      "total_questions": 2371,
      "avg_cost_per_task_usd": 0.0645802066638549,
      "avg_duration_per_task_ms": 194462.45592576973,
      "avg_output_tokens_per_task": 20212.99072121468,
      "last_run_timestamp": "2026-10-31T08:22:42.983Z"
    },
    {
      "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/qwen3.5-9b-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": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.792742111111111,
      "accuracy_std_dev": 0.011968486588114497,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.017912929292929293,
      "avg_duration_per_task_ms": 62394.14758698092,
      "avg_output_tokens_per_task": 5816.072390572391,
      "last_run_timestamp": "2026-10-31T16:10:36.258Z"
    },
    {
      "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": 0.7925423333333335,
      "accuracy_std_dev": 0.014817180897974928,
      "run_count": 13,
      "total_questions": 2427,
      "avg_cost_per_task_usd": 0.004677955088586733,
      "avg_duration_per_task_ms": 1415043.368355995,
      "avg_output_tokens_per_task": 30720.006180469714,
      "last_run_timestamp": "2026-10-31T16:08:08.537Z"
    },
    {
      "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": 0.7923681111111112,
      "accuracy_std_dev": 0.042678175464605506,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.1784357115600449,
      "avg_duration_per_task_ms": 555057.9292929292,
      "avg_output_tokens_per_task": 55554.39898989899,
      "last_run_timestamp": "2026-10-31T16:15:54.668Z"
    },
    {
      "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": "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": 0.7912072841391009,
      "accuracy_std_dev": 0.007680330288036654,
      "run_count": 6,
      "total_questions": 1179,
      "avg_cost_per_task_usd": 0.012160957591178965,
      "avg_duration_per_task_ms": 774485.8049194232,
      "avg_output_tokens_per_task": 40209.361323155215,
      "last_run_timestamp": "2026-10-31T16:11:17.542Z"
    },
    {
      "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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "GMICloud",
      "accuracy": 0.7888407857142856,
      "accuracy_std_dev": 0.03143577289644615,
      "run_count": 14,
      "total_questions": 2772,
      "avg_cost_per_task_usd": 0.10358753607503608,
      "avg_duration_per_task_ms": 373041.1764069264,
      "avg_output_tokens_per_task": 27977.449855699855,
      "last_run_timestamp": "2026-10-31T16:09:41.959Z"
    },
    {
      "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": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "Together",
      "accuracy": 0.7874268851655088,
      "accuracy_std_dev": 0.015558291548790305,
      "run_count": 13,
      "total_questions": 2447,
      "avg_cost_per_task_usd": 0.008662264814058032,
      "avg_duration_per_task_ms": 377262.1029832448,
      "avg_output_tokens_per_task": 34212.00081732734,
      "last_run_timestamp": "2026-10-31T16:15:26.662Z"
    },
    {
      "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": 0.7872267663934426,
      "accuracy_std_dev": 0.013963506030602306,
      "run_count": 3,
      "total_questions": 488,
      "avg_cost_per_task_usd": 0.008632290983606556,
      "avg_duration_per_task_ms": 586832.4200819673,
      "avg_output_tokens_per_task": 37091.375,
      "last_run_timestamp": "2026-10-31T16:20:33.127Z"
    },
    {
      "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": 0.7865693333333335,
      "accuracy_std_dev": 0.009968547225036253,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.02101957912457912,
      "avg_duration_per_task_ms": 246528.65207631874,
      "avg_output_tokens_per_task": 12493.957351290685,
      "last_run_timestamp": "2026-10-31T16:10:36.260Z"
    },
    {
      "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": "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": 0.78493275,
      "accuracy_std_dev": 0.010934561842501992,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.03407076388888889,
      "avg_duration_per_task_ms": 483005.77398989897,
      "avg_output_tokens_per_task": 67906.58964646465,
      "last_run_timestamp": "2026-10-31T08:31:13.285Z"
    },
    {
      "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": 0.7848860000000001,
      "accuracy_std_dev": 0.020676163312697362,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.31222273961840624,
      "avg_duration_per_task_ms": 978633.652637486,
      "avg_output_tokens_per_task": 78687.65488215488,
      "last_run_timestamp": "2026-10-31T16:04:02.366Z"
    },
    {
      "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": 0.7837654143365532,
      "accuracy_std_dev": 0.02906493883169204,
      "run_count": 10,
      "total_questions": 1967,
      "avg_cost_per_task_usd": 0.011243548551093035,
      "avg_duration_per_task_ms": 599323.9059481444,
      "avg_output_tokens_per_task": 37234.770716827654,
      "last_run_timestamp": "2026-10-31T16:03:21.448Z"
    },
    {
      "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": "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-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "DeepInfra",
      "accuracy": 0.7803449165517241,
      "accuracy_std_dev": 0.016592666104035245,
      "run_count": 10,
      "total_questions": 1450,
      "avg_cost_per_task_usd": 0.04745455310344828,
      "avg_duration_per_task_ms": 981137.0944827587,
      "avg_output_tokens_per_task": 24061.69103448276,
      "last_run_timestamp": "2026-10-31T16:04:29.123Z"
    },
    {
      "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-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Sail Research",
      "accuracy": 0.779785,
      "accuracy_std_dev": 0.12136469148745921,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.017734758352758354,
      "avg_duration_per_task_ms": 640150.5850815851,
      "avg_output_tokens_per_task": 32813.10295260295,
      "last_run_timestamp": "2026-10-31T08:48:05.270Z"
    },
    {
      "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": 0.7792768229342326,
      "accuracy_std_dev": 0.05369266686739757,
      "run_count": 9,
      "total_questions": 1779,
      "avg_cost_per_task_usd": 0.159092970207982,
      "avg_duration_per_task_ms": 948891.7240022485,
      "avg_output_tokens_per_task": 66199.51714446319,
      "last_run_timestamp": "2026-10-31T16:04:51.396Z"
    },
    {
      "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": 0.77925075,
      "accuracy_std_dev": 0.01980242038710168,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.02103724936868687,
      "avg_duration_per_task_ms": 475607.264520202,
      "avg_output_tokens_per_task": 27969.91224747475,
      "last_run_timestamp": "2026-10-31T16:16:35.158Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Phala",
      "accuracy": 0.7786195454545455,
      "accuracy_std_dev": 0.04555204390657037,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.21966554178145087,
      "avg_duration_per_task_ms": 867448.9150596878,
      "avg_output_tokens_per_task": 73058.02662993572,
      "last_run_timestamp": "2026-10-17T16:06:46.266Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "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": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "auto-routing",
      "accuracy": 0.7763531538461537,
      "accuracy_std_dev": 0.018975981644198205,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.005880460761460762,
      "avg_duration_per_task_ms": 664693.4401709401,
      "avg_output_tokens_per_task": 36578.56293706294,
      "last_run_timestamp": "2026-10-31T16:03:21.447Z"
    },
    {
      "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": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "GMICloud",
      "accuracy": 0.7759257999999999,
      "accuracy_std_dev": 0.12807761483865945,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.007387053030303032,
      "avg_duration_per_task_ms": 370638.1904040404,
      "avg_output_tokens_per_task": 23199.052525252526,
      "last_run_timestamp": "2026-10-31T16:19:28.505Z"
    },
    {
      "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": 0.7754896779268858,
      "accuracy_std_dev": 0.1073162914081702,
      "run_count": 11,
      "total_questions": 2161,
      "avg_cost_per_task_usd": 0.1401641702915317,
      "avg_duration_per_task_ms": 835967.7473391949,
      "avg_output_tokens_per_task": 38050.62887552059,
      "last_run_timestamp": "2026-10-31T17:00:44.290Z"
    },
    {
      "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": 0.7752525,
      "accuracy_std_dev": 0.02576620138055538,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.011368148148148148,
      "avg_duration_per_task_ms": 910071.1422558923,
      "avg_output_tokens_per_task": 43361.55303030303,
      "last_run_timestamp": "2026-10-31T09:00:40.586Z"
    },
    {
      "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": 0.7744107142857142,
      "accuracy_std_dev": 0.007947545348170354,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.005179176046176046,
      "avg_duration_per_task_ms": 801836.0988455989,
      "avg_output_tokens_per_task": 28610.891053391053,
      "last_run_timestamp": "2026-10-31T16:17:18.501Z"
    },
    {
      "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": 0.7741524627118646,
      "accuracy_std_dev": 0.024920321732780762,
      "run_count": 7,
      "total_questions": 1180,
      "avg_cost_per_task_usd": 0.010626239830508475,
      "avg_duration_per_task_ms": 1250036.8076271187,
      "avg_output_tokens_per_task": 40411.0593220339,
      "last_run_timestamp": "2026-10-31T08:17:43.227Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "inclusionai/ling-3.0-flash-20260723",
      "model_slug": "inclusionai/ling-3.0-flash",
      "model_name": "inclusionAI: Ling 3.0 Flash",
      "provider_name": "DeepInfra",
      "accuracy": 0.7741139703903096,
      "accuracy_std_dev": 0.09718478875928833,
      "run_count": 4,
      "total_questions": 743,
      "avg_cost_per_task_usd": 0.004475604306864065,
      "avg_duration_per_task_ms": 312870.95827725437,
      "avg_output_tokens_per_task": 22926.07267833109,
      "last_run_timestamp": "2026-10-08T22:27:01.555Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "Phala",
      "accuracy": 0.7736031459537571,
      "accuracy_std_dev": 0.07146005981537049,
      "run_count": 7,
      "total_questions": 1384,
      "avg_cost_per_task_usd": 0.01837689523121387,
      "avg_duration_per_task_ms": 1270075.3273121386,
      "avg_output_tokens_per_task": 17747.31430635838,
      "last_run_timestamp": "2026-10-31T16:45:39.993Z"
    },
    {
      "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": "Venice",
      "accuracy": 0.7732982293193716,
      "accuracy_std_dev": 0.013698592432801259,
      "run_count": 5,
      "total_questions": 955,
      "avg_cost_per_task_usd": 0.00232973612565445,
      "avg_duration_per_task_ms": 792784.167539267,
      "avg_output_tokens_per_task": 15348.367539267016,
      "last_run_timestamp": "2026-10-31T16:10:36.260Z"
    },
    {
      "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": 0.7732351365494685,
      "accuracy_std_dev": 0.025060485373108846,
      "run_count": 9,
      "total_questions": 1223,
      "avg_cost_per_task_usd": 0.013341370400654133,
      "avg_duration_per_task_ms": 54198.84873262469,
      "avg_output_tokens_per_task": 17681.87653311529,
      "last_run_timestamp": "2026-10-31T16:39:49.876Z"
    },
    {
      "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": 0.7732082857142858,
      "accuracy_std_dev": 0.007069241982283623,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.020289174603174608,
      "avg_duration_per_task_ms": 740766.9199134199,
      "avg_output_tokens_per_task": 50513.87373737374,
      "last_run_timestamp": "2026-10-31T16:23:19.027Z"
    },
    {
      "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": 0.7722054594731511,
      "accuracy_std_dev": 0.007712250974261645,
      "run_count": 10,
      "total_questions": 1974,
      "avg_cost_per_task_usd": 0.011376429078014186,
      "avg_duration_per_task_ms": 917720.4442755826,
      "avg_output_tokens_per_task": 11827.906788247214,
      "last_run_timestamp": "2026-10-31T16:02:14.358Z"
    },
    {
      "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": 0.7714860192771085,
      "accuracy_std_dev": 0.016470310033005864,
      "run_count": 6,
      "total_questions": 830,
      "avg_cost_per_task_usd": 0.009284253012048193,
      "avg_duration_per_task_ms": 720483.7698795181,
      "avg_output_tokens_per_task": 30820.519277108433,
      "last_run_timestamp": "2026-10-31T16:24:40.574Z"
    },
    {
      "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": 0.7713804000000001,
      "accuracy_std_dev": 0.01747606822600553,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.014499158585858585,
      "avg_duration_per_task_ms": 1080917.518181818,
      "avg_output_tokens_per_task": 51478.64141414141,
      "last_run_timestamp": "2026-10-31T08:09:31.070Z"
    },
    {
      "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": "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Parasail",
      "accuracy": 0.7704203757292881,
      "accuracy_std_dev": 0.015785967573985243,
      "run_count": 9,
      "total_questions": 1714,
      "avg_cost_per_task_usd": 0.02183358868144691,
      "avg_duration_per_task_ms": 237188.9049008168,
      "avg_output_tokens_per_task": 29025.875729288215,
      "last_run_timestamp": "2026-10-31T16:18:04.835Z"
    },
    {
      "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": 0.7702020714285713,
      "accuracy_std_dev": 0.010195756445448158,
      "run_count": 14,
      "total_questions": 2772,
      "avg_cost_per_task_usd": 0.01475207287157287,
      "avg_duration_per_task_ms": 141322.34632034632,
      "avg_output_tokens_per_task": 29381.38961038961,
      "last_run_timestamp": "2026-10-31T16:51:23.396Z"
    },
    {
      "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": 0.7698409107142858,
      "accuracy_std_dev": 0.0035135000000000027,
      "run_count": 2,
      "total_questions": 336,
      "avg_cost_per_task_usd": 0.012509422619047618,
      "avg_duration_per_task_ms": 211684.1994047619,
      "avg_output_tokens_per_task": 31185.72619047619,
      "last_run_timestamp": "2026-10-31T16:53:46.692Z"
    },
    {
      "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": 0.7694625413724936,
      "accuracy_std_dev": 0.015728262890591427,
      "run_count": 18,
      "total_questions": 3541,
      "avg_cost_per_task_usd": 0.006914428410053658,
      "avg_duration_per_task_ms": 433264.68483479245,
      "avg_output_tokens_per_task": 28218.570742728043,
      "last_run_timestamp": "2026-10-31T16:02:14.375Z"
    },
    {
      "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": 0.7692009872315728,
      "accuracy_std_dev": 0.009952898819619341,
      "run_count": 9,
      "total_questions": 1723,
      "avg_cost_per_task_usd": 0.024908178757980265,
      "avg_duration_per_task_ms": 73252.19210679048,
      "avg_output_tokens_per_task": 26085.17933836332,
      "last_run_timestamp": "2026-10-31T16:06:00.111Z"
    },
    {
      "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": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "DeepInfra",
      "accuracy": 0.7683982261904762,
      "accuracy_std_dev": 0.037461158230359067,
      "run_count": 10,
      "total_questions": 1848,
      "avg_cost_per_task_usd": 0.0074100124458874456,
      "avg_duration_per_task_ms": 907022.0817099568,
      "avg_output_tokens_per_task": 19100.03896103896,
      "last_run_timestamp": "2026-10-31T16:04:29.109Z"
    },
    {
      "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": 0.7669798106936416,
      "accuracy_std_dev": 0.011686708686296891,
      "run_count": 7,
      "total_questions": 1384,
      "avg_cost_per_task_usd": 0.019905706647398845,
      "avg_duration_per_task_ms": 1154492.3439306358,
      "avg_output_tokens_per_task": 60181.387283236996,
      "last_run_timestamp": "2026-10-31T16:03:21.445Z"
    },
    {
      "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": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "Venice",
      "accuracy": 0.7657183855579869,
      "accuracy_std_dev": 0.13627692230884958,
      "run_count": 12,
      "total_questions": 2285,
      "avg_cost_per_task_usd": 0.10443865426695843,
      "avg_duration_per_task_ms": 680524.7737417943,
      "avg_output_tokens_per_task": 29927.162800875274,
      "last_run_timestamp": "2026-10-31T16:40:50.323Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Nebius",
      "accuracy": 0.7656566,
      "accuracy_std_dev": 0.0131009148474448,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.01855774343434343,
      "avg_duration_per_task_ms": 128486.40656565657,
      "avg_output_tokens_per_task": 30699.099494949496,
      "last_run_timestamp": "2026-10-31T16:52:10.938Z"
    },
    {
      "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": 0.7653769053571428,
      "accuracy_std_dev": 0.011846747415441054,
      "run_count": 17,
      "total_questions": 3360,
      "avg_cost_per_task_usd": 0.016902593452380953,
      "avg_duration_per_task_ms": 110624.59017857142,
      "avg_output_tokens_per_task": 5988.1654761904765,
      "last_run_timestamp": "2026-10-31T09:06:38.813Z"
    },
    {
      "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": 0.76536225,
      "accuracy_std_dev": 0.00536920838555365,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.004458252525252525,
      "avg_duration_per_task_ms": 487032.047979798,
      "avg_output_tokens_per_task": 29595.234848484848,
      "last_run_timestamp": "2026-10-31T16:16:55.698Z"
    },
    {
      "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": "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": 0.7643096666666668,
      "accuracy_std_dev": 0.011195306675368749,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.021048479236812575,
      "avg_duration_per_task_ms": 1070472.4147025815,
      "avg_output_tokens_per_task": 52516.912457912455,
      "last_run_timestamp": "2026-10-31T16:51:21.148Z"
    },
    {
      "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": 0.7636362999999998,
      "accuracy_std_dev": 0.129902922031069,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.029436740404040405,
      "avg_duration_per_task_ms": 136141.02676767675,
      "avg_output_tokens_per_task": 22545.73888888889,
      "last_run_timestamp": "2026-10-31T16:08:08.538Z"
    },
    {
      "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Crusoe",
      "accuracy": 0.7632998,
      "accuracy_std_dev": 0.007030464092789305,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.007469111111111111,
      "avg_duration_per_task_ms": 123178.52121212121,
      "avg_output_tokens_per_task": 29681.70909090909,
      "last_run_timestamp": "2026-10-31T16:38:46.120Z"
    },
    {
      "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": "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": "qwen/qwen3-coder-next-2025-02-03",
      "model_slug": "qwen/qwen3-coder-next",
      "model_name": "Qwen: Qwen3 Coder Next",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.7626260000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.022027883838383836,
      "avg_duration_per_task_ms": 90904.42929292929,
      "avg_output_tokens_per_task": 14522.767676767677,
      "last_run_timestamp": "2026-10-04T03:02:47.748Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "InferenceNet",
      "accuracy": 0.7626259999999999,
      "accuracy_std_dev": 0.01346799999999998,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.01717197474747475,
      "avg_duration_per_task_ms": 622307.4974747475,
      "avg_output_tokens_per_task": 61038.34090909091,
      "last_run_timestamp": "2026-10-03T10:05:09.589Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Groq",
      "accuracy": 0.7622521111111111,
      "accuracy_std_dev": 0.011328776902153444,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.01744060213243547,
      "avg_duration_per_task_ms": 72362.74859708193,
      "avg_output_tokens_per_task": 28828.154320987655,
      "last_run_timestamp": "2026-10-31T16:23:59.761Z"
    },
    {
      "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": 0.7620651111111112,
      "accuracy_std_dev": 0.013326932508974654,
      "run_count": 18,
      "total_questions": 3564,
      "avg_cost_per_task_usd": 0.018258098765432097,
      "avg_duration_per_task_ms": 148659.74102132436,
      "avg_output_tokens_per_task": 30202.418630751963,
      "last_run_timestamp": "2026-10-31T16:08:33.703Z"
    },
    {
      "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": "Together",
      "accuracy": 0.7585687491525424,
      "accuracy_std_dev": 0.0142815403075283,
      "run_count": 9,
      "total_questions": 1770,
      "avg_cost_per_task_usd": 0.016968081355932207,
      "avg_duration_per_task_ms": 256703.14350282485,
      "avg_output_tokens_per_task": 28488.8406779661,
      "last_run_timestamp": "2026-10-31T16:22:37.563Z"
    },
    {
      "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": 0.7584460168918918,
      "accuracy_std_dev": 0.03053004550509561,
      "run_count": 9,
      "total_questions": 1776,
      "avg_cost_per_task_usd": 0.00475450731981982,
      "avg_duration_per_task_ms": 646469.7742117117,
      "avg_output_tokens_per_task": 28382.398085585584,
      "last_run_timestamp": "2026-10-31T16:20:12.041Z"
    },
    {
      "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/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Phala",
      "accuracy": 0.7574711434482759,
      "accuracy_std_dev": 0.05835625158448523,
      "run_count": 11,
      "total_questions": 2175,
      "avg_cost_per_task_usd": 0.014400611954022986,
      "avg_duration_per_task_ms": 105760.68091954023,
      "avg_output_tokens_per_task": 6123.435402298851,
      "last_run_timestamp": "2026-10-31T08:19:36.628Z"
    },
    {
      "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": "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": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Novita",
      "accuracy": 0.7563452081218275,
      "accuracy_std_dev": 0.0358022938821275,
      "run_count": 6,
      "total_questions": 1182,
      "avg_cost_per_task_usd": 0.014391308798646361,
      "avg_duration_per_task_ms": 1040556.6945854484,
      "avg_output_tokens_per_task": 51071.41201353638,
      "last_run_timestamp": "2026-10-31T08:41:02.325Z"
    },
    {
      "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": 0.7558922222222222,
      "accuracy_std_dev": 0.040404060185904256,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.05619586251402918,
      "avg_duration_per_task_ms": 1048664.5718294051,
      "avg_output_tokens_per_task": 28923.416386083052,
      "last_run_timestamp": "2026-10-31T16:08:31.065Z"
    },
    {
      "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": 0.7539042317380352,
      "accuracy_std_dev": 0.03122595132646972,
      "run_count": 12,
      "total_questions": 1985,
      "avg_cost_per_task_usd": 0.045778590428211585,
      "avg_duration_per_task_ms": 263292.8277078086,
      "avg_output_tokens_per_task": 17254.131486146096,
      "last_run_timestamp": "2026-10-31T08:39:08.315Z"
    },
    {
      "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": "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": 0.7517599999999999,
      "accuracy_std_dev": 0.013833343768524714,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.01652851698806244,
      "avg_duration_per_task_ms": 1314711.4719926538,
      "avg_output_tokens_per_task": 54869.09182736455,
      "last_run_timestamp": "2026-10-31T16:03:21.448Z"
    },
    {
      "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": 0.7514029999999999,
      "accuracy_std_dev": 0.013279676389129373,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.015289352693602695,
      "avg_duration_per_task_ms": 350945.7154882155,
      "avg_output_tokens_per_task": 62534.38636363636,
      "last_run_timestamp": "2026-10-31T16:38:47.420Z"
    },
    {
      "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": 0.7510100999999999,
      "accuracy_std_dev": 0.032456701001334076,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.3156345363636364,
      "avg_duration_per_task_ms": 917451.5242424243,
      "avg_output_tokens_per_task": 80246.15656565657,
      "last_run_timestamp": "2026-10-31T16:39:49.878Z"
    },
    {
      "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": 0.7495323333333332,
      "accuracy_std_dev": 0.02683967111903158,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.1896214567901235,
      "avg_duration_per_task_ms": 1077261.9915824917,
      "avg_output_tokens_per_task": 85826.17845117845,
      "last_run_timestamp": "2026-10-31T16:01:53.649Z"
    },
    {
      "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": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "GMICloud",
      "accuracy": 0.7469697,
      "accuracy_std_dev": 0.09552459944438396,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.13509163080808081,
      "avg_duration_per_task_ms": 945862.1585858585,
      "avg_output_tokens_per_task": 71261.06515151515,
      "last_run_timestamp": "2026-10-31T16:04:29.124Z"
    },
    {
      "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": "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": "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": 0.7450601862113403,
      "accuracy_std_dev": 0.02284683246884292,
      "run_count": 8,
      "total_questions": 1552,
      "avg_cost_per_task_usd": 0.016163423324742267,
      "avg_duration_per_task_ms": 377921.4826030928,
      "avg_output_tokens_per_task": 53593.9587628866,
      "last_run_timestamp": "2026-10-31T08:07:17.099Z"
    },
    {
      "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": 0.7437335555555555,
      "accuracy_std_dev": 0.04255083833646293,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.30186382940516276,
      "avg_duration_per_task_ms": 1012019.6812570146,
      "avg_output_tokens_per_task": 89968.04769921437,
      "last_run_timestamp": "2026-10-31T16:11:15.423Z"
    },
    {
      "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": "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": "SiliconFlow",
      "accuracy": 0.7413018888888888,
      "accuracy_std_dev": 0.01318452125658845,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.11432362065095399,
      "avg_duration_per_task_ms": 753653.6470258137,
      "avg_output_tokens_per_task": 44622.36812570146,
      "last_run_timestamp": "2026-10-31T16:04:48.154Z"
    },
    {
      "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": 0.7410596900662253,
      "accuracy_std_dev": 0.04968796474861267,
      "run_count": 8,
      "total_questions": 1510,
      "avg_cost_per_task_usd": 0.017916279470198675,
      "avg_duration_per_task_ms": 948164.3721854305,
      "avg_output_tokens_per_task": 45200.233774834436,
      "last_run_timestamp": "2026-10-31T16:47:37.740Z"
    },
    {
      "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": "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": 0.7374669124010552,
      "accuracy_std_dev": 0.0902272932043027,
      "run_count": 11,
      "total_questions": 1895,
      "avg_cost_per_task_usd": 0.004063712401055408,
      "avg_duration_per_task_ms": 556408.5973614776,
      "avg_output_tokens_per_task": 25344.54142480211,
      "last_run_timestamp": "2026-10-31T08:04:26.691Z"
    },
    {
      "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": "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": 0.7366819680365297,
      "accuracy_std_dev": 0.008315102174527942,
      "run_count": 9,
      "total_questions": 1752,
      "avg_cost_per_task_usd": 0.006853743150684932,
      "avg_duration_per_task_ms": 245825.2191780822,
      "avg_output_tokens_per_task": 8475.880707762557,
      "last_run_timestamp": "2026-10-31T16:02:14.362Z"
    },
    {
      "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": 0.736486429054054,
      "accuracy_std_dev": 0.05162874176700847,
      "run_count": 9,
      "total_questions": 1776,
      "avg_cost_per_task_usd": 0.3068156143018018,
      "avg_duration_per_task_ms": 1317780.9831081082,
      "avg_output_tokens_per_task": 97153.50563063064,
      "last_run_timestamp": "2026-10-31T16:18:48.427Z"
    },
    {
      "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": 0.7352164576271186,
      "accuracy_std_dev": 0.012339799164982321,
      "run_count": 9,
      "total_questions": 1770,
      "avg_cost_per_task_usd": 0.006587220903954802,
      "avg_duration_per_task_ms": 332459.9661016949,
      "avg_output_tokens_per_task": 26156.45536723164,
      "last_run_timestamp": "2026-10-31T16:16:13.798Z"
    },
    {
      "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": 0.7348357476056337,
      "accuracy_std_dev": 0.05024499959642196,
      "run_count": 9,
      "total_questions": 1775,
      "avg_cost_per_task_usd": 0.3023959695774648,
      "avg_duration_per_task_ms": 1292289.6828169015,
      "avg_output_tokens_per_task": 95338.63492957747,
      "last_run_timestamp": "2026-10-31T16:18:28.925Z"
    },
    {
      "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": 0.732440575,
      "accuracy_std_dev": 0.057246596407316816,
      "run_count": 3,
      "total_questions": 560,
      "avg_cost_per_task_usd": 0.014524144642857144,
      "avg_duration_per_task_ms": 428029.5660714286,
      "avg_output_tokens_per_task": 58878.205357142855,
      "last_run_timestamp": "2026-10-31T08:03:10.090Z"
    },
    {
      "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": 0.732155,
      "accuracy_std_dev": 0.02424727451075689,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.016461675757575758,
      "avg_duration_per_task_ms": 673272.2666666667,
      "avg_output_tokens_per_task": 60883.92525252525,
      "last_run_timestamp": "2026-10-31T08:31:11.846Z"
    },
    {
      "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "auto-routing",
      "accuracy": 0.731949111111111,
      "accuracy_std_dev": 0.02440615873296666,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.006272946689113356,
      "avg_duration_per_task_ms": 384042.4797979798,
      "avg_output_tokens_per_task": 21207.913580246914,
      "last_run_timestamp": "2026-10-31T16:08:31.072Z"
    },
    {
      "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": 0.7319023750000001,
      "accuracy_std_dev": 0.006382992557129826,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.0060275726010100995,
      "avg_duration_per_task_ms": 414095.6856060606,
      "avg_output_tokens_per_task": 7948.38952020202,
      "last_run_timestamp": "2026-10-31T16:08:56.903Z"
    },
    {
      "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": 0.7306920882647944,
      "accuracy_std_dev": 0.018533662569629808,
      "run_count": 7,
      "total_questions": 997,
      "avg_cost_per_task_usd": 0.016062558676028082,
      "avg_duration_per_task_ms": 67044.65897693079,
      "avg_output_tokens_per_task": 9235.0962888666,
      "last_run_timestamp": "2026-10-31T16:30:53.500Z"
    },
    {
      "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": 0.730226631893972,
      "accuracy_std_dev": 0.11726978279845851,
      "run_count": 13,
      "total_questions": 2339,
      "avg_cost_per_task_usd": 0.03321617400598546,
      "avg_duration_per_task_ms": 1275412.009833262,
      "avg_output_tokens_per_task": 34879.14194100043,
      "last_run_timestamp": "2026-10-31T16:04:48.182Z"
    },
    {
      "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": 0.7300354102564102,
      "accuracy_std_dev": 0.01294792675952992,
      "run_count": 39,
      "total_questions": 7722,
      "avg_cost_per_task_usd": 0.20437522118622112,
      "avg_duration_per_task_ms": 238565.11292411294,
      "avg_output_tokens_per_task": 38135.92411292411,
      "last_run_timestamp": "2026-10-31T08:33:11.343Z"
    },
    {
      "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": 0.7293447692307693,
      "accuracy_std_dev": 0.008838355156352458,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.19052701903651906,
      "avg_duration_per_task_ms": 238729.40753690753,
      "avg_output_tokens_per_task": 37914.3142968143,
      "last_run_timestamp": "2026-10-31T08:20:09.279Z"
    },
    {
      "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": "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": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Cohere",
      "accuracy": 0.7267116666666668,
      "accuracy_std_dev": 0.01049864519301841,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.021460314814814814,
      "avg_duration_per_task_ms": 244130.04882154881,
      "avg_output_tokens_per_task": 76264.67003367003,
      "last_run_timestamp": "2026-10-31T08:19:36.711Z"
    },
    {
      "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": "anthropic/claude-4.5-haiku-20251001",
      "model_slug": "anthropic/claude-haiku-4.5",
      "model_name": "Anthropic: Claude Haiku 4.5",
      "provider_name": "Azure",
      "accuracy": 0.7263041800146951,
      "accuracy_std_dev": 0.01201198916070684,
      "run_count": 14,
      "total_questions": 2722,
      "avg_cost_per_task_usd": 0.1868326627479794,
      "avg_duration_per_task_ms": 344051.7843497428,
      "avg_output_tokens_per_task": 37176.1304188097,
      "last_run_timestamp": "2026-10-31T08:16:11.695Z"
    },
    {
      "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": 0.7257187692307693,
      "accuracy_std_dev": 0.010132498721014512,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.18636140015540018,
      "avg_duration_per_task_ms": 303150.1456876457,
      "avg_output_tokens_per_task": 37101.67715617715,
      "last_run_timestamp": "2026-10-31T08:09:30.774Z"
    },
    {
      "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": "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": 0.7250282222222222,
      "accuracy_std_dev": 0.015036749624967174,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.005513952300785634,
      "avg_duration_per_task_ms": 307362.290684624,
      "avg_output_tokens_per_task": 8054.221099887766,
      "last_run_timestamp": "2026-10-31T16:22:55.771Z"
    },
    {
      "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": 0.7236893527788727,
      "accuracy_std_dev": 0.03777209883376474,
      "run_count": 13,
      "total_questions": 2537,
      "avg_cost_per_task_usd": 0.010590881750098544,
      "avg_duration_per_task_ms": 263440.8253843122,
      "avg_output_tokens_per_task": 48345.98265668112,
      "last_run_timestamp": "2026-10-31T08:46:43.680Z"
    },
    {
      "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": 0.7223064,
      "accuracy_std_dev": 0.010291729001484623,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.00442305606060606,
      "avg_duration_per_task_ms": 663677.5994949494,
      "avg_output_tokens_per_task": 7916.334343434343,
      "last_run_timestamp": "2026-10-31T16:02:18.363Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "ModelRun",
      "accuracy": 0.7219408075949366,
      "accuracy_std_dev": 0.025598999999999983,
      "run_count": 2,
      "total_questions": 395,
      "avg_cost_per_task_usd": 0.27939689620253166,
      "avg_duration_per_task_ms": 450249.90379746835,
      "avg_output_tokens_per_task": 45941.00253164557,
      "last_run_timestamp": "2026-08-15T08:16:15.061Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.5-haiku-20251001",
      "model_slug": "anthropic/claude-haiku-4.5",
      "model_name": "Anthropic: Claude Haiku 4.5",
      "provider_name": "Google",
      "accuracy": 0.7212367804878049,
      "accuracy_std_dev": 0.01324537121083399,
      "run_count": 41,
      "total_questions": 8118,
      "avg_cost_per_task_usd": 0.19729785008622816,
      "avg_duration_per_task_ms": 270228.84429662477,
      "avg_output_tokens_per_task": 36826.08869179601,
      "last_run_timestamp": "2026-10-31T08:29:35.324Z"
    },
    {
      "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": 0.7183810211428571,
      "accuracy_std_dev": 0.051458865728389994,
      "run_count": 18,
      "total_questions": 3500,
      "avg_cost_per_task_usd": 0.008799481142857142,
      "avg_duration_per_task_ms": 536929.9885714286,
      "avg_output_tokens_per_task": 26082.768,
      "last_run_timestamp": "2026-10-31T16:52:55.644Z"
    },
    {
      "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": 0.7183742857142856,
      "accuracy_std_dev": 0.012394494661678109,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.01898589466089466,
      "avg_duration_per_task_ms": 501749.6746031746,
      "avg_output_tokens_per_task": 67465.82683982684,
      "last_run_timestamp": "2026-10-31T08:06:04.074Z"
    },
    {
      "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": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Venice",
      "accuracy": 0.7140853333333333,
      "accuracy_std_dev": 0.0635853157908508,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.1300334169472503,
      "avg_duration_per_task_ms": 644759.9270482603,
      "avg_output_tokens_per_task": 39066.33557800225,
      "last_run_timestamp": "2026-10-31T16:56:31.817Z"
    },
    {
      "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": "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": 0.71001675,
      "accuracy_std_dev": 0.018948683824675006,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.015486457070707072,
      "avg_duration_per_task_ms": 1105788.4797979798,
      "avg_output_tokens_per_task": 64321.568181818184,
      "last_run_timestamp": "2026-10-31T16:02:14.361Z"
    },
    {
      "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": 0.7060420000000001,
      "accuracy_std_dev": 0.04619957164303584,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.015228148709315378,
      "avg_duration_per_task_ms": 964326.9915824916,
      "avg_output_tokens_per_task": 49378.965768799106,
      "last_run_timestamp": "2026-10-31T16:11:46.431Z"
    },
    {
      "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": "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": 0.7038606610169492,
      "accuracy_std_dev": 0.025828434379814764,
      "run_count": 11,
      "total_questions": 2124,
      "avg_cost_per_task_usd": 0.009122728342749528,
      "avg_duration_per_task_ms": 365582.73399246705,
      "avg_output_tokens_per_task": 45350.33474576271,
      "last_run_timestamp": "2026-10-31T08:03:09.796Z"
    },
    {
      "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": 0.703844,
      "accuracy_std_dev": 0.13064484273849722,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.611304741161616,
      "avg_duration_per_task_ms": 637806.0361952862,
      "avg_output_tokens_per_task": 40663.66961279461,
      "last_run_timestamp": "2026-10-31T08:38:33.095Z"
    },
    {
      "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": "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-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "DeepInfra",
      "accuracy": 0.6970787019475021,
      "accuracy_std_dev": 0.046119854611155975,
      "run_count": 12,
      "total_questions": 2362,
      "avg_cost_per_task_usd": 0.007317635478408128,
      "avg_duration_per_task_ms": 292695.22861981374,
      "avg_output_tokens_per_task": 40004.958086367486,
      "last_run_timestamp": "2026-10-31T08:15:24.314Z"
    },
    {
      "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": 0.6950760566414041,
      "accuracy_std_dev": 0.14052251643685132,
      "run_count": 38,
      "total_questions": 7521,
      "avg_cost_per_task_usd": 0.6626253082036964,
      "avg_duration_per_task_ms": 603719.9066613482,
      "avg_output_tokens_per_task": 41294.83499534636,
      "last_run_timestamp": "2026-10-31T08:51:32.081Z"
    },
    {
      "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": "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": 0.6937554455696204,
      "accuracy_std_dev": 0.09367172425871108,
      "run_count": 10,
      "total_questions": 1975,
      "avg_cost_per_task_usd": 0.011443888607594935,
      "avg_duration_per_task_ms": 875181.9529113924,
      "avg_output_tokens_per_task": 33496.65417721519,
      "last_run_timestamp": "2026-10-31T16:08:08.535Z"
    },
    {
      "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": "Amazon Bedrock",
      "accuracy": 0.693300564102564,
      "accuracy_std_dev": 0.1359986435897281,
      "run_count": 39,
      "total_questions": 7722,
      "avg_cost_per_task_usd": 0.6583114172494173,
      "avg_duration_per_task_ms": 444898.01476301474,
      "avg_output_tokens_per_task": 40970.9775964776,
      "last_run_timestamp": "2026-10-31T08:31:32.893Z"
    },
    {
      "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": 0.6929553846153845,
      "accuracy_std_dev": 0.13884156644759268,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.616928526029526,
      "avg_duration_per_task_ms": 639057.0835275835,
      "avg_output_tokens_per_task": 40970.16472416472,
      "last_run_timestamp": "2026-10-31T08:09:30.979Z"
    },
    {
      "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": 0.6917897692307693,
      "accuracy_std_dev": 0.14551018900679394,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.6185969393939393,
      "avg_duration_per_task_ms": 638891.4533799534,
      "avg_output_tokens_per_task": 41138.08430458431,
      "last_run_timestamp": "2026-10-31T08:28:17.671Z"
    },
    {
      "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": 0.6912458,
      "accuracy_std_dev": 0.010837313198390075,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.0064883984848484845,
      "avg_duration_per_task_ms": 106966.29242424242,
      "avg_output_tokens_per_task": 15944.740404040404,
      "last_run_timestamp": "2026-10-31T16:08:56.904Z"
    },
    {
      "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": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "DeepInfra",
      "accuracy": 0.6898380904522613,
      "accuracy_std_dev": 0.12509647859063733,
      "run_count": 9,
      "total_questions": 1194,
      "avg_cost_per_task_usd": 0.0919431105527638,
      "avg_duration_per_task_ms": 1244619.9221105527,
      "avg_output_tokens_per_task": 26175.759631490786,
      "last_run_timestamp": "2026-10-31T16:05:14.699Z"
    },
    {
      "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": 0.687366775928297,
      "accuracy_std_dev": 0.029137221034831948,
      "run_count": 7,
      "total_questions": 781,
      "avg_cost_per_task_usd": 0.008982548015364916,
      "avg_duration_per_task_ms": 54182.23815620998,
      "avg_output_tokens_per_task": 5818.932138284251,
      "last_run_timestamp": "2026-10-31T16:13:09.867Z"
    },
    {
      "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": 0.6866834096080909,
      "accuracy_std_dev": 0.0076575271261354015,
      "run_count": 4,
      "total_questions": 791,
      "avg_cost_per_task_usd": 0.007103366624525917,
      "avg_duration_per_task_ms": 1621893.5347661187,
      "avg_output_tokens_per_task": 50543.86852085967,
      "last_run_timestamp": "2026-10-31T16:36:31.734Z"
    },
    {
      "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": "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": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "Venice",
      "accuracy": 0.6842701805792164,
      "accuracy_std_dev": 0.044485547973956784,
      "run_count": 9,
      "total_questions": 1761,
      "avg_cost_per_task_usd": 0.028337780806360024,
      "avg_duration_per_task_ms": 842767.0045428734,
      "avg_output_tokens_per_task": 18259.638273708122,
      "last_run_timestamp": "2026-10-31T16:23:19.099Z"
    },
    {
      "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": 0.681502625,
      "accuracy_std_dev": 0.06815065878246794,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.007994105429292929,
      "avg_duration_per_task_ms": 1440482.744949495,
      "avg_output_tokens_per_task": 62597.993055555555,
      "last_run_timestamp": "2026-10-31T16:08:08.531Z"
    },
    {
      "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": 0.67979451369863,
      "accuracy_std_dev": 0.019550333432791698,
      "run_count": 9,
      "total_questions": 1752,
      "avg_cost_per_task_usd": 0.0022822819634703197,
      "avg_duration_per_task_ms": 207690.84988584474,
      "avg_output_tokens_per_task": 4088.747716894977,
      "last_run_timestamp": "2026-10-31T16:18:48.428Z"
    },
    {
      "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": 0.6785352626262626,
      "accuracy_std_dev": 0.014156861524116758,
      "run_count": 13,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.005069473232323233,
      "avg_duration_per_task_ms": 756595.5121212121,
      "avg_output_tokens_per_task": 50528.773232323234,
      "last_run_timestamp": "2026-10-31T16:04:29.122Z"
    },
    {
      "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": 0.677399,
      "accuracy_std_dev": 0.12205074947844442,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.0072654311868686865,
      "avg_duration_per_task_ms": 728833.754419192,
      "avg_output_tokens_per_task": 25199.926136363636,
      "last_run_timestamp": "2026-10-31T16:45:39.993Z"
    },
    {
      "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": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "DeepInfra",
      "accuracy": 0.6692905358600583,
      "accuracy_std_dev": 0.017641172479174967,
      "run_count": 10,
      "total_questions": 1715,
      "avg_cost_per_task_usd": 0.00673951195335277,
      "avg_duration_per_task_ms": 477949.7562682216,
      "avg_output_tokens_per_task": 47964.45364431487,
      "last_run_timestamp": "2026-10-31T16:03:21.483Z"
    },
    {
      "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": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Venice",
      "accuracy": 0.6689113333333333,
      "accuracy_std_dev": 0.016128109050832822,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.005779179573512907,
      "avg_duration_per_task_ms": 407135.462962963,
      "avg_output_tokens_per_task": 15838.005050505051,
      "last_run_timestamp": "2026-10-31T16:17:15.718Z"
    },
    {
      "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": 0.6687991999999999,
      "accuracy_std_dev": 0.14462356096164505,
      "run_count": 15,
      "total_questions": 2970,
      "avg_cost_per_task_usd": 0.6179110515151515,
      "avg_duration_per_task_ms": 635772.7828282828,
      "avg_output_tokens_per_task": 41062.00437710438,
      "last_run_timestamp": "2026-10-31T08:22:47.508Z"
    },
    {
      "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": 0.6687465155984118,
      "accuracy_std_dev": 0.0639810001554281,
      "run_count": 9,
      "total_questions": 1763,
      "avg_cost_per_task_usd": 0.05512392682926829,
      "avg_duration_per_task_ms": 1146484.0079410097,
      "avg_output_tokens_per_task": 24612.394781622235,
      "last_run_timestamp": "2026-10-31T16:02:14.362Z"
    },
    {
      "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": "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": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "CoreWeave",
      "accuracy": 0.664235,
      "accuracy_std_dev": 0.019262151293260175,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.007772474747474748,
      "avg_duration_per_task_ms": 551813.6610549944,
      "avg_output_tokens_per_task": 59957.01627384961,
      "last_run_timestamp": "2026-10-31T16:08:31.061Z"
    },
    {
      "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": "Amazon Bedrock",
      "accuracy": 0.6623645555555556,
      "accuracy_std_dev": 0.013998065156863738,
      "run_count": 18,
      "total_questions": 3564,
      "avg_cost_per_task_usd": 0.00905068686868687,
      "avg_duration_per_task_ms": 323979.22923681256,
      "avg_output_tokens_per_task": 59978.55218855219,
      "last_run_timestamp": "2026-10-31T16:08:08.537Z"
    },
    {
      "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": 0.6603534999999999,
      "accuracy_std_dev": 0.021669912920914115,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.008152684974747474,
      "avg_duration_per_task_ms": 795790.2424242424,
      "avg_output_tokens_per_task": 55529.520833333336,
      "last_run_timestamp": "2026-10-31T16:08:56.902Z"
    },
    {
      "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": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "auto-routing",
      "accuracy": 0.6595585555555555,
      "accuracy_std_dev": 0.13091400954155713,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.0148099898989899,
      "avg_duration_per_task_ms": 436249.1038159371,
      "avg_output_tokens_per_task": 12881.502805836139,
      "last_run_timestamp": "2026-10-31T16:01:53.629Z"
    },
    {
      "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": "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Phala",
      "accuracy": 0.6491281532846714,
      "accuracy_std_dev": 0.20940255131760213,
      "run_count": 9,
      "total_questions": 1644,
      "avg_cost_per_task_usd": 0.00955052798053528,
      "avg_duration_per_task_ms": 180102.42518248176,
      "avg_output_tokens_per_task": 15744.736618004867,
      "last_run_timestamp": "2026-10-31T16:54:17.849Z"
    },
    {
      "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": 0.6490834444444443,
      "accuracy_std_dev": 0.02275484053564142,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.00663854152637486,
      "avg_duration_per_task_ms": 622965.2239057239,
      "avg_output_tokens_per_task": 43822.446127946125,
      "last_run_timestamp": "2026-10-31T16:01:53.622Z"
    },
    {
      "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": "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": "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-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Darkbloom",
      "accuracy": 0.64025675,
      "accuracy_std_dev": 0.01324879906585875,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.001088048611111111,
      "avg_duration_per_task_ms": 331146.75505050505,
      "avg_output_tokens_per_task": 11572.473484848484,
      "last_run_timestamp": "2026-10-31T16:31:13.438Z"
    },
    {
      "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-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-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "Darkbloom",
      "accuracy": 0.6351663037907099,
      "accuracy_std_dev": 0.019472921033065377,
      "run_count": 10,
      "total_questions": 1873,
      "avg_cost_per_task_usd": 0.005367699412706888,
      "avg_duration_per_task_ms": 807188.147357181,
      "avg_output_tokens_per_task": 35119.06513614522,
      "last_run_timestamp": "2026-10-31T08:20:09.538Z"
    },
    {
      "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-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Google",
      "accuracy": 0.6303372741573033,
      "accuracy_std_dev": 0.1138422900195217,
      "run_count": 9,
      "total_questions": 1780,
      "avg_cost_per_task_usd": 0.009750982584269664,
      "avg_duration_per_task_ms": 363946.58370786515,
      "avg_output_tokens_per_task": 27697.321348314606,
      "last_run_timestamp": "2026-10-31T16:12:07.129Z"
    },
    {
      "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": 0.6302757276308385,
      "accuracy_std_dev": 0.007550033035656602,
      "run_count": 9,
      "total_questions": 1777,
      "avg_cost_per_task_usd": 0.001856440067529544,
      "avg_duration_per_task_ms": 236824.7703995498,
      "avg_output_tokens_per_task": 10141.058525604953,
      "last_run_timestamp": "2026-10-31T16:10:09.221Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "Darkbloom",
      "accuracy": 0.6295420719476744,
      "accuracy_std_dev": 0.1792338661514529,
      "run_count": 14,
      "total_questions": 2752,
      "avg_cost_per_task_usd": 0.02676043095930232,
      "avg_duration_per_task_ms": 725458.0534156977,
      "avg_output_tokens_per_task": 37903.851744186046,
      "last_run_timestamp": "2026-10-31T16:08:33.702Z"
    },
    {
      "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": "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": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Groq",
      "accuracy": 0.6272831084474886,
      "accuracy_std_dev": 0.013804623233441915,
      "run_count": 9,
      "total_questions": 1752,
      "avg_cost_per_task_usd": 0.016875459474885845,
      "avg_duration_per_task_ms": 80669.67808219178,
      "avg_output_tokens_per_task": 56020.17294520548,
      "last_run_timestamp": "2026-10-31T16:52:47.514Z"
    },
    {
      "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": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Fireworks",
      "accuracy": 0.6182949425287356,
      "accuracy_std_dev": 0.0013475000000000015,
      "run_count": 2,
      "total_questions": 348,
      "avg_cost_per_task_usd": 0.011105244252873564,
      "avg_duration_per_task_ms": 345775.4454022989,
      "avg_output_tokens_per_task": 38606.68103448276,
      "last_run_timestamp": "2026-08-22T17:46:40.285Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 VL 30B A3B Instruct",
      "provider_name": "SiliconFlow",
      "accuracy": 0.6170940030769231,
      "accuracy_std_dev": 0.014320214565431623,
      "run_count": 5,
      "total_questions": 975,
      "avg_cost_per_task_usd": 0.015717948717948717,
      "avg_duration_per_task_ms": 780521.6543589743,
      "avg_output_tokens_per_task": 16712.84717948718,
      "last_run_timestamp": "2026-10-03T11:06:38.068Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-14b-04-28",
      "model_slug": "qwen/qwen3-14b",
      "model_name": "Qwen: Qwen3 14B",
      "provider_name": "DeepInfra",
      "accuracy": 0.6123375670628184,
      "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": "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": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "Mara",
      "accuracy": 0.6059481003717473,
      "accuracy_std_dev": 0.024067533945591992,
      "run_count": 3,
      "total_questions": 269,
      "avg_cost_per_task_usd": 0.03276318587360595,
      "avg_duration_per_task_ms": 68520.27137546468,
      "avg_output_tokens_per_task": 7032.501858736059,
      "last_run_timestamp": "2026-10-31T16:08:33.699Z"
    },
    {
      "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": "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/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": 0.5972300452232899,
      "accuracy_std_dev": 0.06739364692023929,
      "run_count": 9,
      "total_questions": 1769,
      "avg_cost_per_task_usd": 0.012534433578292821,
      "avg_duration_per_task_ms": 1310890.6302996043,
      "avg_output_tokens_per_task": 57642.5850763143,
      "last_run_timestamp": "2026-10-31T16:05:14.631Z"
    },
    {
      "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": "qwen/qwen3-14b-04-28",
      "model_slug": "qwen/qwen3-14b",
      "model_name": "Qwen: Qwen3 14B",
      "provider_name": "auto-routing",
      "accuracy": 0.591751,
      "accuracy_std_dev": 0.0025249999999999995,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.0029208080808080807,
      "avg_duration_per_task_ms": 521963.51515151514,
      "avg_output_tokens_per_task": 11688.588383838383,
      "last_run_timestamp": "2026-10-01T16:32:22.405Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Cerebras",
      "accuracy": 0.5905352222222223,
      "accuracy_std_dev": 0.008575970638654383,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.01031239337822671,
      "avg_duration_per_task_ms": 11512.907968574635,
      "avg_output_tokens_per_task": 13295.387205387206,
      "last_run_timestamp": "2026-10-31T16:21:16.340Z"
    },
    {
      "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": 0.5896185734720416,
      "accuracy_std_dev": 0.2364552374805523,
      "run_count": 8,
      "total_questions": 1538,
      "avg_cost_per_task_usd": 0.08412247919375812,
      "avg_duration_per_task_ms": 1774600.33289987,
      "avg_output_tokens_per_task": 39146.597529258775,
      "last_run_timestamp": "2026-10-31T16:02:18.348Z"
    },
    {
      "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": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "SambaNova",
      "accuracy": 0.5833334540682414,
      "accuracy_std_dev": 0.047090518334207944,
      "run_count": 8,
      "total_questions": 762,
      "avg_cost_per_task_usd": 0.030894883202099738,
      "avg_duration_per_task_ms": 61623.28215223097,
      "avg_output_tokens_per_task": 6621.586614173229,
      "last_run_timestamp": "2026-10-31T16:41:30.138Z"
    },
    {
      "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": 0.5830470075601375,
      "accuracy_std_dev": 0.0830182052186651,
      "run_count": 9,
      "total_questions": 1455,
      "avg_cost_per_task_usd": 0.05815560893470791,
      "avg_duration_per_task_ms": 352334.5237113402,
      "avg_output_tokens_per_task": 22047.301030927836,
      "last_run_timestamp": "2026-10-31T16:21:16.340Z"
    },
    {
      "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": 0.5810130708860759,
      "accuracy_std_dev": 0.012019500000000016,
      "run_count": 2,
      "total_questions": 395,
      "avg_cost_per_task_usd": 0.01962629113924051,
      "avg_duration_per_task_ms": 376708.9949367089,
      "avg_output_tokens_per_task": 97953.7417721519,
      "last_run_timestamp": "2026-10-31T08:09:25.783Z"
    },
    {
      "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": "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": "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "DigitalOcean",
      "accuracy": 0.539057,
      "accuracy_std_dev": 0.13614692394027858,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.003152918181818182,
      "avg_duration_per_task_ms": 212250.40606060607,
      "avg_output_tokens_per_task": 7608.965656565657,
      "last_run_timestamp": "2026-10-31T16:10:36.257Z"
    },
    {
      "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": 0.49700711111111107,
      "accuracy_std_dev": 0.01216009606545061,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.002385578002244669,
      "avg_duration_per_task_ms": 9967.221099887767,
      "avg_output_tokens_per_task": 2372.858024691358,
      "last_run_timestamp": "2026-10-31T16:04:29.124Z"
    },
    {
      "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": "CoreWeave",
      "accuracy": 0.4835387777777777,
      "accuracy_std_dev": 0.011018803965925972,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.0025987850729517394,
      "avg_duration_per_task_ms": 48627.889450056115,
      "avg_output_tokens_per_task": 2852.319865319865,
      "last_run_timestamp": "2026-10-31T16:04:29.124Z"
    },
    {
      "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": 0.482768822853688,
      "accuracy_std_dev": 0.014790995023742912,
      "run_count": 17,
      "total_questions": 3308,
      "avg_cost_per_task_usd": 0.001309269044740024,
      "avg_duration_per_task_ms": 60513.06045949214,
      "avg_output_tokens_per_task": 2352.4238210399035,
      "last_run_timestamp": "2026-10-31T16:01:53.654Z"
    },
    {
      "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": 0.48148122222222217,
      "accuracy_std_dev": 0.009848511007341592,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.002354208754208754,
      "avg_duration_per_task_ms": 15824.776655443322,
      "avg_output_tokens_per_task": 2212.813131313131,
      "last_run_timestamp": "2026-10-31T16:05:14.701Z"
    },
    {
      "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": "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": "Parasail",
      "accuracy": 0.4762437777777777,
      "accuracy_std_dev": 0.013290307736406826,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.0013282469135802468,
      "avg_duration_per_task_ms": 46260.15375982042,
      "avg_output_tokens_per_task": 2383.4377104377104,
      "last_run_timestamp": "2026-10-31T16:36:11.519Z"
    },
    {
      "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": 0.47568244444444446,
      "accuracy_std_dev": 0.009052379136154844,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.0013990594837261505,
      "avg_duration_per_task_ms": 79190.52525252526,
      "avg_output_tokens_per_task": 2352.858024691358,
      "last_run_timestamp": "2026-10-31T16:23:41.139Z"
    },
    {
      "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": "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": 0.46944120359955005,
      "accuracy_std_dev": 0.013089737311683874,
      "run_count": 9,
      "total_questions": 1778,
      "avg_cost_per_task_usd": 0.0012956473565804273,
      "avg_duration_per_task_ms": 85305.89820022497,
      "avg_output_tokens_per_task": 2966.7919010123733,
      "last_run_timestamp": "2026-10-31T16:31:54.106Z"
    },
    {
      "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": 0.4675925,
      "accuracy_std_dev": 0.012321014262226957,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.0007978666666666667,
      "avg_duration_per_task_ms": 159410.78737373737,
      "avg_output_tokens_per_task": 2242.5338383838384,
      "last_run_timestamp": "2026-10-31T16:02:14.361Z"
    },
    {
      "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": "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": 0.4320892331288344,
      "accuracy_std_dev": 0.09307635724371684,
      "run_count": 10,
      "total_questions": 1956,
      "avg_cost_per_task_usd": 0.004473661042944785,
      "avg_duration_per_task_ms": 43451.16871165644,
      "avg_output_tokens_per_task": 1885.2627811860941,
      "last_run_timestamp": "2026-10-31T16:05:36.136Z"
    },
    {
      "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": "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": 0.3325851111111111,
      "accuracy_std_dev": 0.09545144872126171,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.008632653759820426,
      "avg_duration_per_task_ms": 536841.7272727273,
      "avg_output_tokens_per_task": 21346.39674523008,
      "last_run_timestamp": "2026-10-31T16:45:48.945Z"
    },
    {
      "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": 0.2882121979695431,
      "accuracy_std_dev": 0.01622631095200357,
      "run_count": 9,
      "total_questions": 1773,
      "avg_cost_per_task_usd": 0.0028333936830231246,
      "avg_duration_per_task_ms": 49259.41624365482,
      "avg_output_tokens_per_task": 35365.48110547095,
      "last_run_timestamp": "2026-10-31T16:04:29.122Z"
    },
    {
      "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": 0.2800224444444444,
      "accuracy_std_dev": 0.02023328750193112,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.0012389450056116721,
      "avg_duration_per_task_ms": 144968.38888888888,
      "avg_output_tokens_per_task": 17614.97474747475,
      "last_run_timestamp": "2026-10-31T16:08:31.064Z"
    },
    {
      "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": "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": 0.2713588193548387,
      "accuracy_std_dev": 0.013427628284002107,
      "run_count": 9,
      "total_questions": 1705,
      "avg_cost_per_task_usd": 0.009860867448680352,
      "avg_duration_per_task_ms": 405997.0574780059,
      "avg_output_tokens_per_task": 44105.84926686217,
      "last_run_timestamp": "2026-10-31T16:04:29.124Z"
    },
    {
      "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": "DeepInfra",
      "accuracy": 0.2656565000000001,
      "accuracy_std_dev": 0.010207038417190359,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.00039605808080808073,
      "avg_duration_per_task_ms": 384073.6873737374,
      "avg_output_tokens_per_task": 9498.437373737373,
      "last_run_timestamp": "2026-10-31T16:08:31.068Z"
    },
    {
      "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": 0.2568276666666667,
      "accuracy_std_dev": 0.01356372386592676,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.00042733389450056114,
      "avg_duration_per_task_ms": 74390.75308641975,
      "avg_output_tokens_per_task": 8223.042087542088,
      "last_run_timestamp": "2026-10-31T16:55:07.715Z"
    },
    {
      "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": 0.0016840000000000002,
      "accuracy_std_dev": 0,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.003431577441077441,
      "avg_duration_per_task_ms": 155214.41582491584,
      "avg_output_tokens_per_task": 3231.1026936026938,
      "last_run_timestamp": "2026-10-31T16:04:43.943Z"
    },
    {
      "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": "Azure",
      "accuracy": 0.8005555833333333,
      "accuracy_std_dev": 0.022353925693482164,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.8806157300000002,
      "avg_duration_per_task_ms": 125606.79666666666,
      "avg_output_tokens_per_task": 5738.843333333333,
      "last_run_timestamp": "2026-10-31T11:20:38.480Z"
    },
    {
      "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": 0.8000000909090909,
      "accuracy_std_dev": 0.020695063581156234,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.8742483381818181,
      "avg_duration_per_task_ms": 136193.65272727274,
      "avg_output_tokens_per_task": 5706.489090909091,
      "last_run_timestamp": "2026-10-31T12:24:55.848Z"
    },
    {
      "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": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "BaseTen",
      "accuracy": 0.7981221830985916,
      "accuracy_std_dev": 0.05639978873679898,
      "run_count": 3,
      "total_questions": 142,
      "avg_cost_per_task_usd": 0.007624964788732395,
      "avg_duration_per_task_ms": 30939.802816901407,
      "avg_output_tokens_per_task": 8761.732394366198,
      "last_run_timestamp": "2026-10-31T11:56:20.980Z"
    },
    {
      "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": 0.795,
      "accuracy_std_dev": 0.02696008022985094,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.8826118166666667,
      "avg_duration_per_task_ms": 119564.475,
      "avg_output_tokens_per_task": 5739.836666666667,
      "last_run_timestamp": "2026-10-31T11:38:36.784Z"
    },
    {
      "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": 0.7946430446428572,
      "accuracy_std_dev": 0.027357193240535482,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.20150866741071427,
      "avg_duration_per_task_ms": 823434.4174107143,
      "avg_output_tokens_per_task": 34488.85267857143,
      "last_run_timestamp": "2026-10-31T19:01:09.612Z"
    },
    {
      "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": 0.7938889166666666,
      "accuracy_std_dev": 0.016433554015379277,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.8273622516666667,
      "avg_duration_per_task_ms": 126448.29,
      "avg_output_tokens_per_task": 5727.116666666667,
      "last_run_timestamp": "2026-10-31T11:18:52.340Z"
    },
    {
      "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": 0.7936329325842696,
      "accuracy_std_dev": 0.028642456621828596,
      "run_count": 9,
      "total_questions": 445,
      "avg_cost_per_task_usd": 0.12726180674157303,
      "avg_duration_per_task_ms": 2869756.1258426965,
      "avg_output_tokens_per_task": 27340.368539325842,
      "last_run_timestamp": "2026-10-31T20:31:41.691Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "inclusionai/ling-3.1-flash-20261002",
      "model_slug": "inclusionai/ling-3.1-flash",
      "model_name": "inclusionAI: Ling 3.1 Flash",
      "provider_name": "Novita",
      "accuracy": 0.793333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.00828808,
      "avg_duration_per_task_ms": 196026.56,
      "avg_output_tokens_per_task": 16303,
      "last_run_timestamp": "2026-10-14T19:21:35.384Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "inclusionai/ling-3.1-flash-20261002",
      "model_slug": "inclusionai/ling-3.1-flash",
      "model_name": "inclusionAI: Ling 3.1 Flash",
      "provider_name": "DeepInfra",
      "accuracy": 0.793333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.03146312,
      "avg_duration_per_task_ms": 636210.72,
      "avg_output_tokens_per_task": 16745.44,
      "last_run_timestamp": "2026-10-14T19:17:59.742Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-5-fable-20260609",
      "model_slug": "anthropic/claude-fable-5",
      "model_name": "Anthropic: Claude Fable 5",
      "provider_name": "Google",
      "accuracy": 0.7916667916666666,
      "accuracy_std_dev": 0.021709424691016815,
      "run_count": 24,
      "total_questions": 1200,
      "avg_cost_per_task_usd": 0.9256366683333332,
      "avg_duration_per_task_ms": 93473.74,
      "avg_output_tokens_per_task": 5715.59,
      "last_run_timestamp": "2026-10-31T12:30:34.973Z"
    },
    {
      "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": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "Alibaba",
      "accuracy": 0.7899033875278396,
      "accuracy_std_dev": 0.023747050173666146,
      "run_count": 9,
      "total_questions": 449,
      "avg_cost_per_task_usd": 0.18324271046770602,
      "avg_duration_per_task_ms": 436487.29844097997,
      "avg_output_tokens_per_task": 30283.42316258352,
      "last_run_timestamp": "2026-10-31T19:12:41.306Z"
    },
    {
      "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": 0.788889,
      "accuracy_std_dev": 0.015234869040023505,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.057788776666666666,
      "avg_duration_per_task_ms": 893038.1233333333,
      "avg_output_tokens_per_task": 40447.416666666664,
      "last_run_timestamp": "2026-10-31T13:13:33.031Z"
    },
    {
      "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": 0.7883333333333333,
      "accuracy_std_dev": 0.020570601102598386,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.4772300233333334,
      "avg_duration_per_task_ms": 119792.23833333333,
      "avg_output_tokens_per_task": 7037.026666666667,
      "last_run_timestamp": "2026-10-31T12:17:12.454Z"
    },
    {
      "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": 0.7882883714036617,
      "accuracy_std_dev": 0.015477690995371508,
      "run_count": 23,
      "total_questions": 1147,
      "avg_cost_per_task_usd": 0.0821072258064516,
      "avg_duration_per_task_ms": 149668.3635571055,
      "avg_output_tokens_per_task": 12209.048823016565,
      "last_run_timestamp": "2026-10-31T12:32:50.229Z"
    },
    {
      "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": 0.788,
      "accuracy_std_dev": 0.02246951515275753,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.15925035199999998,
      "avg_duration_per_task_ms": 71120.884,
      "avg_output_tokens_per_task": 6245.168,
      "last_run_timestamp": "2026-10-31T11:10:36.361Z"
    },
    {
      "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": 0.7875705677966102,
      "accuracy_std_dev": 0.021829507772196534,
      "run_count": 12,
      "total_questions": 590,
      "avg_cost_per_task_usd": 0.07927353559322034,
      "avg_duration_per_task_ms": 1138885.6372881357,
      "avg_output_tokens_per_task": 26180.43559322034,
      "last_run_timestamp": "2026-10-31T20:22:41.827Z"
    },
    {
      "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": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "DeepInfra",
      "accuracy": 0.786127239884393,
      "accuracy_std_dev": 0.023660230134130145,
      "run_count": 7,
      "total_questions": 346,
      "avg_cost_per_task_usd": 0.09665253179190753,
      "avg_duration_per_task_ms": 380894.4450867052,
      "avg_output_tokens_per_task": 16737.754335260117,
      "last_run_timestamp": "2026-10-31T12:56:48.075Z"
    },
    {
      "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": 0.7858794166666667,
      "accuracy_std_dev": 0.021093449409710124,
      "run_count": 3,
      "total_questions": 144,
      "avg_cost_per_task_usd": 0.07943493055555556,
      "avg_duration_per_task_ms": 334937.25694444444,
      "avg_output_tokens_per_task": 18674.54861111111,
      "last_run_timestamp": "2026-10-31T19:43:30.408Z"
    },
    {
      "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": 0.7844446666666666,
      "accuracy_std_dev": 0.03478329384760578,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.04264756333333333,
      "avg_duration_per_task_ms": 115094.22333333333,
      "avg_output_tokens_per_task": 4914.21,
      "last_run_timestamp": "2026-10-31T12:26:47.613Z"
    },
    {
      "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": 0.7844444444444445,
      "accuracy_std_dev": 0.02266234163693451,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.19777408000000002,
      "avg_duration_per_task_ms": 762878.1888888889,
      "avg_output_tokens_per_task": 31234.38888888889,
      "last_run_timestamp": "2026-10-31T19:47:27.263Z"
    },
    {
      "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": 0.7844443333333334,
      "accuracy_std_dev": 0.01632984089872824,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.0785912688888889,
      "avg_duration_per_task_ms": 47362.52222222222,
      "avg_output_tokens_per_task": 4720.888888888889,
      "last_run_timestamp": "2026-10-31T11:58:26.584Z"
    },
    {
      "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": 0.7833333333333333,
      "accuracy_std_dev": 0.011706202852429232,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.1729425,
      "avg_duration_per_task_ms": 66541.35,
      "avg_output_tokens_per_task": 6203.445,
      "last_run_timestamp": "2026-10-31T12:38:55.501Z"
    },
    {
      "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": "google/gemini-3.7-flash-20260813",
      "model_slug": "google/gemini-3.7-flash",
      "model_name": "Google: Gemini 3.7 Flash",
      "provider_name": "Google",
      "accuracy": 0.7831052054794521,
      "accuracy_std_dev": 0.017155537965592597,
      "run_count": 3,
      "total_questions": 146,
      "avg_cost_per_task_usd": 0.14992966438356164,
      "avg_duration_per_task_ms": 183129.3904109589,
      "avg_output_tokens_per_task": 16313.164383561643,
      "last_run_timestamp": "2026-10-15T21:26:28.195Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Novita",
      "accuracy": 0.7825,
      "accuracy_std_dev": 0.023790007240436065,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.08953825,
      "avg_duration_per_task_ms": 274829.5175,
      "avg_output_tokens_per_task": 19992.6725,
      "last_run_timestamp": "2026-10-31T12:57:09.932Z"
    },
    {
      "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": 0.7822223333333334,
      "accuracy_std_dev": 0.01571327975171178,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.011317603333333334,
      "avg_duration_per_task_ms": 599501.3033333333,
      "avg_output_tokens_per_task": 21464.33,
      "last_run_timestamp": "2026-10-31T14:08:11.994Z"
    },
    {
      "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": 0.781333,
      "accuracy_std_dev": 0.027129319932501044,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.0587014,
      "avg_duration_per_task_ms": 871009.496,
      "avg_output_tokens_per_task": 40629.384,
      "last_run_timestamp": "2026-10-31T13:41:45.309Z"
    },
    {
      "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": 0.7811976407035177,
      "accuracy_std_dev": 0.02110393286871771,
      "run_count": 16,
      "total_questions": 796,
      "avg_cost_per_task_usd": 0.09273568467336683,
      "avg_duration_per_task_ms": 66515.5364321608,
      "avg_output_tokens_per_task": 4679.810301507538,
      "last_run_timestamp": "2026-10-31T13:12:01.731Z"
    },
    {
      "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": 0.7808333749999999,
      "accuracy_std_dev": 0.010769114145758466,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0823930725,
      "avg_duration_per_task_ms": 41435.8875,
      "avg_output_tokens_per_task": 4744.8375,
      "last_run_timestamp": "2026-10-31T12:40:46.044Z"
    },
    {
      "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": "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": 0.7800001,
      "accuracy_std_dev": 0.02108170349592273,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.174296886,
      "avg_duration_per_task_ms": 72081.318,
      "avg_output_tokens_per_task": 6231.684,
      "last_run_timestamp": "2026-10-31T12:10:32.453Z"
    },
    {
      "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": 0.78,
      "accuracy_std_dev": 0.021081938910409114,
      "run_count": 18,
      "total_questions": 900,
      "avg_cost_per_task_usd": 0.18010515333333332,
      "avg_duration_per_task_ms": 64530.58666666667,
      "avg_output_tokens_per_task": 6325.823333333334,
      "last_run_timestamp": "2026-10-31T12:07:29.560Z"
    },
    {
      "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": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.78,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.0042799399999999994,
      "avg_duration_per_task_ms": 627315.22,
      "avg_output_tokens_per_task": 12675.1,
      "last_run_timestamp": "2026-10-31T21:10:18.835Z"
    },
    {
      "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": 0.7799999000000001,
      "accuracy_std_dev": 0.034896621932359,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.080800072,
      "avg_duration_per_task_ms": 206899.728,
      "avg_output_tokens_per_task": 25416.206,
      "last_run_timestamp": "2026-10-31T19:22:16.313Z"
    },
    {
      "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": "auto-routing",
      "accuracy": 0.7786668,
      "accuracy_std_dev": 0.01484714413481597,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.014840148000000001,
      "avg_duration_per_task_ms": 379677.016,
      "avg_output_tokens_per_task": 21736.348,
      "last_run_timestamp": "2026-10-31T12:28:58.923Z"
    },
    {
      "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": 0.7786666,
      "accuracy_std_dev": 0.010666575000439441,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.031286299999999996,
      "avg_duration_per_task_ms": 221994.964,
      "avg_output_tokens_per_task": 12740.492,
      "last_run_timestamp": "2026-10-31T12:48:16.034Z"
    },
    {
      "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": 0.7784256831875607,
      "accuracy_std_dev": 0.029691712716069543,
      "run_count": 21,
      "total_questions": 1029,
      "avg_cost_per_task_usd": 0.07680297376093294,
      "avg_duration_per_task_ms": 74568.79203109816,
      "avg_output_tokens_per_task": 4283.289601554908,
      "last_run_timestamp": "2026-10-31T13:24:01.335Z"
    },
    {
      "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": "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": 0.7779999000000001,
      "accuracy_std_dev": 0.020011142453393298,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.081736628,
      "avg_duration_per_task_ms": 135707.96,
      "avg_output_tokens_per_task": 4671.7,
      "last_run_timestamp": "2026-10-31T12:38:49.692Z"
    },
    {
      "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-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "AtlasCloud",
      "accuracy": 0.777778,
      "accuracy_std_dev": 0.01257094435593444,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.016507473333333335,
      "avg_duration_per_task_ms": 92510.76666666666,
      "avg_output_tokens_per_task": 10861.266666666666,
      "last_run_timestamp": "2026-10-31T11:43:38.949Z"
    },
    {
      "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": 0.777459306590258,
      "accuracy_std_dev": 0.02712463143584833,
      "run_count": 7,
      "total_questions": 349,
      "avg_cost_per_task_usd": 0.17213766475644698,
      "avg_duration_per_task_ms": 55665.0888252149,
      "avg_output_tokens_per_task": 4643.535816618911,
      "last_run_timestamp": "2026-10-31T12:26:01.294Z"
    },
    {
      "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": 0.777333342857143,
      "accuracy_std_dev": 0.021153949387347253,
      "run_count": 35,
      "total_questions": 1750,
      "avg_cost_per_task_usd": 0.4968248080000001,
      "avg_duration_per_task_ms": 118772.41142857143,
      "avg_output_tokens_per_task": 7078.027428571429,
      "last_run_timestamp": "2026-10-31T11:36:20.802Z"
    },
    {
      "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": 0.7771427857142857,
      "accuracy_std_dev": 0.017405512982051608,
      "run_count": 14,
      "total_questions": 700,
      "avg_cost_per_task_usd": 0.4396931257142857,
      "avg_duration_per_task_ms": 90015.33428571429,
      "avg_output_tokens_per_task": 5705.564285714286,
      "last_run_timestamp": "2026-10-31T12:26:17.134Z"
    },
    {
      "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": 0.7766666666666666,
      "accuracy_std_dev": 0.021430304782143333,
      "run_count": 24,
      "total_questions": 1200,
      "avg_cost_per_task_usd": 0.4784604191666666,
      "avg_duration_per_task_ms": 102044.3925,
      "avg_output_tokens_per_task": 7128.966666666666,
      "last_run_timestamp": "2026-10-31T12:54:18.844Z"
    },
    {
      "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": 0.7766666666666666,
      "accuracy_std_dev": 0.014782246611692315,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.17363205,
      "avg_duration_per_task_ms": 73619.57333333333,
      "avg_output_tokens_per_task": 6340.086666666667,
      "last_run_timestamp": "2026-10-31T12:24:45.929Z"
    },
    {
      "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": 0.7766666388888891,
      "accuracy_std_dev": 0.023280401603963375,
      "run_count": 36,
      "total_questions": 1800,
      "avg_cost_per_task_usd": 0.5182516222222223,
      "avg_duration_per_task_ms": 108585.71333333333,
      "avg_output_tokens_per_task": 7098.225555555556,
      "last_run_timestamp": "2026-10-31T12:26:08.030Z"
    },
    {
      "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": 0.7766665833333334,
      "accuracy_std_dev": 0.0379571517689318,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.10135336166666668,
      "avg_duration_per_task_ms": 172908.82,
      "avg_output_tokens_per_task": 6639.301666666666,
      "last_run_timestamp": "2026-10-31T11:53:06.149Z"
    },
    {
      "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": 0.7764815,
      "accuracy_std_dev": 0.02583109973193382,
      "run_count": 18,
      "total_questions": 900,
      "avg_cost_per_task_usd": 0.13240519666666667,
      "avg_duration_per_task_ms": 319498.1977777778,
      "avg_output_tokens_per_task": 19319.296666666665,
      "last_run_timestamp": "2026-10-31T13:25:38.758Z"
    },
    {
      "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": 0.7760940000000001,
      "accuracy_std_dev": 0.020936325423770052,
      "run_count": 4,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.06970348989898989,
      "avg_duration_per_task_ms": 95095.22222222222,
      "avg_output_tokens_per_task": 4292.19696969697,
      "last_run_timestamp": "2026-10-31T12:44:06.053Z"
    },
    {
      "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": 0.7760001999999999,
      "accuracy_std_dev": 0.028472283599318148,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.06820797999999999,
      "avg_duration_per_task_ms": 658315.768,
      "avg_output_tokens_per_task": 13133.756,
      "last_run_timestamp": "2026-10-31T13:30:16.905Z"
    },
    {
      "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": 0.7758335,
      "accuracy_std_dev": 0.016308646311082982,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.15398643500000003,
      "avg_duration_per_task_ms": 330861.565,
      "avg_output_tokens_per_task": 19948.0325,
      "last_run_timestamp": "2026-10-31T13:41:05.689Z"
    },
    {
      "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": 0.7757695461847389,
      "accuracy_std_dev": 0.016169521055368324,
      "run_count": 10,
      "total_questions": 498,
      "avg_cost_per_task_usd": 0.08735203012048193,
      "avg_duration_per_task_ms": 139136.04417670684,
      "avg_output_tokens_per_task": 4443.712851405622,
      "last_run_timestamp": "2026-10-31T12:29:23.419Z"
    },
    {
      "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": 0.7757574545454545,
      "accuracy_std_dev": 0.02074823590031709,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.08103076363636365,
      "avg_duration_per_task_ms": 336159.44,
      "avg_output_tokens_per_task": 20208.643636363635,
      "last_run_timestamp": "2026-10-31T12:24:53.587Z"
    },
    {
      "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": "qwen/qwen3.5-122b-a10b-20260224",
      "model_slug": "qwen/qwen3.5-122b-a10b",
      "model_name": "Qwen: Qwen3.5-122B-A10B",
      "provider_name": "Novita",
      "accuracy": 0.7753500890756303,
      "accuracy_std_dev": 0.023224689913460387,
      "run_count": 12,
      "total_questions": 595,
      "avg_cost_per_task_usd": 0.20088185714285717,
      "avg_duration_per_task_ms": 515898.5899159664,
      "avg_output_tokens_per_task": 28073.515966386556,
      "last_run_timestamp": "2026-10-31T18:53:29.091Z"
    },
    {
      "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": 0.7753334,
      "accuracy_std_dev": 0.026339740994170767,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.065886822,
      "avg_duration_per_task_ms": 257998.344,
      "avg_output_tokens_per_task": 14164.592,
      "last_run_timestamp": "2026-10-31T12:39:11.628Z"
    },
    {
      "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": 0.7751514545454545,
      "accuracy_std_dev": 0.026375843918816165,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.10104016181818182,
      "avg_duration_per_task_ms": 294074.4890909091,
      "avg_output_tokens_per_task": 21520.51818181818,
      "last_run_timestamp": "2026-10-31T11:29:11.154Z"
    },
    {
      "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": 0.775,
      "accuracy_std_dev": 0.01787294665828779,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.09322250500000001,
      "avg_duration_per_task_ms": 167383.47,
      "avg_output_tokens_per_task": 5123.47,
      "last_run_timestamp": "2026-10-31T12:32:51.992Z"
    },
    {
      "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-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": 0.7745454545454545,
      "accuracy_std_dev": 0.02425759601131023,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.48932630545454553,
      "avg_duration_per_task_ms": 124123.69272727273,
      "avg_output_tokens_per_task": 7977.827272727272,
      "last_run_timestamp": "2026-10-31T11:58:26.584Z"
    },
    {
      "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": 0.7744445,
      "accuracy_std_dev": 0.018425810536219746,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.45357258333333333,
      "avg_duration_per_task_ms": 120535.45,
      "avg_output_tokens_per_task": 7078.328333333333,
      "last_run_timestamp": "2026-10-31T11:40:34.382Z"
    },
    {
      "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": 0.7744444166666666,
      "accuracy_std_dev": 0.018823368390816466,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.4382412116666667,
      "avg_duration_per_task_ms": 94200.57833333334,
      "avg_output_tokens_per_task": 5678.8533333333335,
      "last_run_timestamp": "2026-10-31T12:52:03.935Z"
    },
    {
      "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": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Friendli",
      "accuracy": 0.7739394545454545,
      "accuracy_std_dev": 0.019165412105066386,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.06798592909090909,
      "avg_duration_per_task_ms": 45561.432727272724,
      "avg_output_tokens_per_task": 4731.818181818182,
      "last_run_timestamp": "2026-10-31T12:41:41.331Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Reka",
      "accuracy": 0.7733335,
      "accuracy_std_dev": 0.00666650000000002,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.11407636,
      "avg_duration_per_task_ms": 360681.75,
      "avg_output_tokens_per_task": 19997.86,
      "last_run_timestamp": "2026-10-01T22:40:52.402Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Io Net",
      "accuracy": 0.7733335,
      "accuracy_std_dev": 0.013333499999999998,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.10867763,
      "avg_duration_per_task_ms": 521628.03,
      "avg_output_tokens_per_task": 12092.47,
      "last_run_timestamp": "2026-09-19T08:11:59.798Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "DekaLLM",
      "accuracy": 0.7733334,
      "accuracy_std_dev": 0.02108173511929225,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.05645720600000001,
      "avg_duration_per_task_ms": 156313.234,
      "avg_output_tokens_per_task": 16426.692,
      "last_run_timestamp": "2026-10-31T11:34:45.444Z"
    },
    {
      "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": 0.773333375,
      "accuracy_std_dev": 0.02236071704428047,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.04385152750000001,
      "avg_duration_per_task_ms": 107910.09,
      "avg_output_tokens_per_task": 4940.6725,
      "last_run_timestamp": "2026-10-31T12:00:41.768Z"
    },
    {
      "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": 0.7733333571428571,
      "accuracy_std_dev": 0.023771504885372875,
      "run_count": 14,
      "total_questions": 700,
      "avg_cost_per_task_usd": 0.5080915685714286,
      "avg_duration_per_task_ms": 123754.58285714286,
      "avg_output_tokens_per_task": 8039.411428571429,
      "last_run_timestamp": "2026-10-31T11:50:39.610Z"
    },
    {
      "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": 0.7733333333333333,
      "accuracy_std_dev": 0.02309381831909905,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.16708652333333332,
      "avg_duration_per_task_ms": 79748.64333333333,
      "avg_output_tokens_per_task": 6314.02,
      "last_run_timestamp": "2026-10-31T11:48:36.920Z"
    },
    {
      "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": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "ModelRun",
      "accuracy": 0.77333325,
      "accuracy_std_dev": 0.029814193116492343,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.06547900499999999,
      "avg_duration_per_task_ms": 79239.345,
      "avg_output_tokens_per_task": 12411.3875,
      "last_run_timestamp": "2026-10-31T11:09:23.381Z"
    },
    {
      "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": 0.7733331666666667,
      "accuracy_std_dev": 0.021081851068448003,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.05726335,
      "avg_duration_per_task_ms": 158508.54333333333,
      "avg_output_tokens_per_task": 4588.7266666666665,
      "last_run_timestamp": "2026-10-31T13:11:06.716Z"
    },
    {
      "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": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "GMICloud",
      "accuracy": 0.773333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.26629643999999997,
      "avg_duration_per_task_ms": 438741.58,
      "avg_output_tokens_per_task": 27876.64,
      "last_run_timestamp": "2026-08-29T16:58:30.455Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "Makora",
      "accuracy": 0.773333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.11485508,
      "avg_duration_per_task_ms": 381375.48,
      "avg_output_tokens_per_task": 14199.1,
      "last_run_timestamp": "2026-10-13T03:50:19.636Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "StreamLake",
      "accuracy": 0.773333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.057549260000000005,
      "avg_duration_per_task_ms": 140710.34,
      "avg_output_tokens_per_task": 4743.94,
      "last_run_timestamp": "2026-09-11T04:09:34.110Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.7-opus-20260416",
      "model_slug": "anthropic/claude-opus-4.7",
      "model_name": "Anthropic: Claude Opus 4.7",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7730554583333333,
      "accuracy_std_dev": 0.031238603972727888,
      "run_count": 24,
      "total_questions": 1200,
      "avg_cost_per_task_usd": 0.4483734791666666,
      "avg_duration_per_task_ms": 79019.83166666667,
      "avg_output_tokens_per_task": 5657.6883333333335,
      "last_run_timestamp": "2026-10-31T12:41:58.029Z"
    },
    {
      "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": 0.7725926666666666,
      "accuracy_std_dev": 0.033176625054396364,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.05704026,
      "avg_duration_per_task_ms": 115075.10666666667,
      "avg_output_tokens_per_task": 6558.98,
      "last_run_timestamp": "2026-10-31T12:50:03.839Z"
    },
    {
      "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": 0.7724602745591939,
      "accuracy_std_dev": 0.036002860879163164,
      "run_count": 8,
      "total_questions": 397,
      "avg_cost_per_task_usd": 0.18971839546599495,
      "avg_duration_per_task_ms": 615216.5390428212,
      "avg_output_tokens_per_task": 29565.539042821158,
      "last_run_timestamp": "2026-10-31T19:27:20.607Z"
    },
    {
      "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": 0.77222225,
      "accuracy_std_dev": 0.02746491926417226,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.1697550733333333,
      "avg_duration_per_task_ms": 565928.75,
      "avg_output_tokens_per_task": 35366.556666666664,
      "last_run_timestamp": "2026-10-31T19:08:11.310Z"
    },
    {
      "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-4.8-opus-20260528",
      "model_slug": "anthropic/claude-opus-4.8",
      "model_name": "Anthropic: Claude Opus 4.8",
      "provider_name": "Azure",
      "accuracy": 0.7722222083333333,
      "accuracy_std_dev": 0.029606428127997207,
      "run_count": 24,
      "total_questions": 1200,
      "avg_cost_per_task_usd": 0.5155781208333333,
      "avg_duration_per_task_ms": 125395.55583333333,
      "avg_output_tokens_per_task": 7961.098333333333,
      "last_run_timestamp": "2026-10-31T13:09:20.558Z"
    },
    {
      "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": 0.7720001999999999,
      "accuracy_std_dev": 0.024367767418456705,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.14564374000000002,
      "avg_duration_per_task_ms": 314489.068,
      "avg_output_tokens_per_task": 20798.24,
      "last_run_timestamp": "2026-10-31T13:16:57.928Z"
    },
    {
      "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": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Parasail",
      "accuracy": 0.7718182727272727,
      "accuracy_std_dev": 0.031151820244762325,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.06332707454545455,
      "avg_duration_per_task_ms": 243359.18363636362,
      "avg_output_tokens_per_task": 11661.423636363636,
      "last_run_timestamp": "2026-10-31T12:57:09.934Z"
    },
    {
      "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": 0.7716665,
      "accuracy_std_dev": 0.0016665000000000152,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.050993250000000004,
      "avg_duration_per_task_ms": 677393.46,
      "avg_output_tokens_per_task": 30421.97,
      "last_run_timestamp": "2026-10-31T13:28:05.897Z"
    },
    {
      "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": 0.7715774062499999,
      "accuracy_std_dev": 0.022644299042712767,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.14963372544642858,
      "avg_duration_per_task_ms": 341867.6473214286,
      "avg_output_tokens_per_task": 20209.140625,
      "last_run_timestamp": "2026-10-31T14:04:49.550Z"
    },
    {
      "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": 0.7714602976588629,
      "accuracy_std_dev": 0.028416062527791094,
      "run_count": 6,
      "total_questions": 299,
      "avg_cost_per_task_usd": 0.08344332775919731,
      "avg_duration_per_task_ms": 249520.2474916388,
      "avg_output_tokens_per_task": 12314.311036789297,
      "last_run_timestamp": "2026-10-31T19:22:52.905Z"
    },
    {
      "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": 0.7713334,
      "accuracy_std_dev": 0.0305578329670152,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.08027632600000001,
      "avg_duration_per_task_ms": 358319.328,
      "avg_output_tokens_per_task": 18206.802,
      "last_run_timestamp": "2026-10-31T20:27:36.016Z"
    },
    {
      "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": 0.7713334,
      "accuracy_std_dev": 0.020011225740568715,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.051776854,
      "avg_duration_per_task_ms": 108112.036,
      "avg_output_tokens_per_task": 4665.94,
      "last_run_timestamp": "2026-10-31T13:28:16.093Z"
    },
    {
      "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": 0.77120534375,
      "accuracy_std_dev": 0.030014310742266483,
      "run_count": 18,
      "total_questions": 896,
      "avg_cost_per_task_usd": 0.03428859486607143,
      "avg_duration_per_task_ms": 485944.55691964284,
      "avg_output_tokens_per_task": 12331.224330357143,
      "last_run_timestamp": "2026-10-31T19:33:25.394Z"
    },
    {
      "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": 0.770833375,
      "accuracy_std_dev": 0.02283704805758343,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.2110248925,
      "avg_duration_per_task_ms": 381381.2675,
      "avg_output_tokens_per_task": 17629.1275,
      "last_run_timestamp": "2026-10-31T12:14:30.525Z"
    },
    {
      "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": "Parasail",
      "accuracy": 0.7707407777777777,
      "accuracy_std_dev": 0.023188894510460925,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.11877632888888891,
      "avg_duration_per_task_ms": 486710.77777777775,
      "avg_output_tokens_per_task": 21345.64,
      "last_run_timestamp": "2026-10-17T11:37:24.257Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta/muse-glimmer-30b-20260810",
      "model_slug": "meta/muse-glimmer-30b",
      "model_name": "Meta: Muse Glimmer 30B",
      "provider_name": "Parasail",
      "accuracy": 0.7706935637583894,
      "accuracy_std_dev": 0.013644816256244217,
      "run_count": 3,
      "total_questions": 149,
      "avg_cost_per_task_usd": 0.09587820805369127,
      "avg_duration_per_task_ms": 288021.6644295302,
      "avg_output_tokens_per_task": 13114.979865771813,
      "last_run_timestamp": "2026-08-29T08:48:12.957Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Phala",
      "accuracy": 0.7706668,
      "accuracy_std_dev": 0.027194791985966728,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.13800210599999999,
      "avg_duration_per_task_ms": 324537.226,
      "avg_output_tokens_per_task": 20086.83,
      "last_run_timestamp": "2026-10-31T13:20:51.002Z"
    },
    {
      "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": "anthropic/claude-4.5-opus-20251124",
      "model_slug": "anthropic/claude-opus-4.5",
      "model_name": "Anthropic: Claude Opus 4.5",
      "provider_name": "Anthropic",
      "accuracy": 0.7704762857142857,
      "accuracy_std_dev": 0.028139468010578456,
      "run_count": 14,
      "total_questions": 700,
      "avg_cost_per_task_usd": 0.5353783828571429,
      "avg_duration_per_task_ms": 182998.61857142858,
      "avg_output_tokens_per_task": 10560.904285714285,
      "last_run_timestamp": "2026-10-31T12:38:33.924Z"
    },
    {
      "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": 0.7700000714285714,
      "accuracy_std_dev": 0.014638549891415803,
      "run_count": 14,
      "total_questions": 700,
      "avg_cost_per_task_usd": 0.14041841571428573,
      "avg_duration_per_task_ms": 235377.64428571428,
      "avg_output_tokens_per_task": 25220.13142857143,
      "last_run_timestamp": "2026-10-31T13:20:49.855Z"
    },
    {
      "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": 0.77,
      "accuracy_std_dev": 0.02472072904467018,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.47639071666666666,
      "avg_duration_per_task_ms": 109574.44833333333,
      "avg_output_tokens_per_task": 7160.583333333333,
      "last_run_timestamp": "2026-10-31T11:27:21.008Z"
    },
    {
      "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": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "DeepInfra",
      "accuracy": 0.769999875,
      "accuracy_std_dev": 0.03382961802340332,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0406579,
      "avg_duration_per_task_ms": 682767.975,
      "avg_output_tokens_per_task": 12586.08,
      "last_run_timestamp": "2026-10-31T12:34:48.818Z"
    },
    {
      "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": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "DeepInfra",
      "accuracy": 0.7699214781144781,
      "accuracy_std_dev": 0.02849514533750542,
      "run_count": 6,
      "total_questions": 297,
      "avg_cost_per_task_usd": 0.010507353535353536,
      "avg_duration_per_task_ms": 977066.3771043771,
      "avg_output_tokens_per_task": 18208.95286195286,
      "last_run_timestamp": "2026-10-31T12:24:58.436Z"
    },
    {
      "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": 0.7696296666666667,
      "accuracy_std_dev": 0.024966483982420205,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.040881691111111104,
      "avg_duration_per_task_ms": 198238.98888888888,
      "avg_output_tokens_per_task": 13013.271111111111,
      "last_run_timestamp": "2026-10-31T20:06:41.224Z"
    },
    {
      "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": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Modal",
      "accuracy": 0.7693333,
      "accuracy_std_dev": 0.027680754646685506,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.080204858,
      "avg_duration_per_task_ms": 55518.588,
      "avg_output_tokens_per_task": 4833.54,
      "last_run_timestamp": "2026-10-31T12:18:44.291Z"
    },
    {
      "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": 0.7690024251781471,
      "accuracy_std_dev": 0.02344176596675554,
      "run_count": 17,
      "total_questions": 842,
      "avg_cost_per_task_usd": 0.05268543230403801,
      "avg_duration_per_task_ms": 269011.41448931117,
      "avg_output_tokens_per_task": 11083.264845605701,
      "last_run_timestamp": "2026-10-31T12:29:57.959Z"
    },
    {
      "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": 0.7685184999999999,
      "accuracy_std_dev": 0.027604418545769074,
      "run_count": 36,
      "total_questions": 1800,
      "avg_cost_per_task_usd": 0.5604697755555554,
      "avg_duration_per_task_ms": 113686.57388888889,
      "avg_output_tokens_per_task": 8067.973333333333,
      "last_run_timestamp": "2026-10-31T12:31:31.431Z"
    },
    {
      "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": 0.7685184444444445,
      "accuracy_std_dev": 0.02067440250976561,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.07816885555555556,
      "avg_duration_per_task_ms": 93704.89777777778,
      "avg_output_tokens_per_task": 4401.093333333333,
      "last_run_timestamp": "2026-10-31T12:34:51.325Z"
    },
    {
      "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": 0.7684848181818182,
      "accuracy_std_dev": 0.034709847551636586,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.05697716181818182,
      "avg_duration_per_task_ms": 301468.96545454545,
      "avg_output_tokens_per_task": 12590.405454545455,
      "last_run_timestamp": "2026-10-31T11:18:52.425Z"
    },
    {
      "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": 0.768333375,
      "accuracy_std_dev": 0.027638527357375157,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.12531909,
      "avg_duration_per_task_ms": 166385.7025,
      "avg_output_tokens_per_task": 7956.0025,
      "last_run_timestamp": "2026-10-31T13:24:47.637Z"
    },
    {
      "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": 0.76833325,
      "accuracy_std_dev": 0.0414661475053748,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.11548195250000001,
      "avg_duration_per_task_ms": 265843.155,
      "avg_output_tokens_per_task": 11738.8375,
      "last_run_timestamp": "2026-10-31T21:36:17.452Z"
    },
    {
      "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": 0.7681482222222222,
      "accuracy_std_dev": 0.017435059205697385,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.052077991111111104,
      "avg_duration_per_task_ms": 111760.7,
      "avg_output_tokens_per_task": 4431.0888888888885,
      "last_run_timestamp": "2026-10-31T13:22:10.288Z"
    },
    {
      "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": 0.7681481111111111,
      "accuracy_std_dev": 0.022394386178705306,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.05412603333333333,
      "avg_duration_per_task_ms": 221687.28666666665,
      "avg_output_tokens_per_task": 12553.397777777778,
      "last_run_timestamp": "2026-10-31T20:15:14.788Z"
    },
    {
      "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": 0.768,
      "accuracy_std_dev": 0.020612843894038495,
      "run_count": 30,
      "total_questions": 1500,
      "avg_cost_per_task_usd": 0.06597447600000002,
      "avg_duration_per_task_ms": 69126.944,
      "avg_output_tokens_per_task": 4654.312,
      "last_run_timestamp": "2026-10-31T12:41:59.417Z"
    },
    {
      "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": 0.7677776666666667,
      "accuracy_std_dev": 0.043104839177934014,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.06842919666666666,
      "avg_duration_per_task_ms": 974829.46,
      "avg_output_tokens_per_task": 44158.48333333333,
      "last_run_timestamp": "2026-10-31T13:18:29.756Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Decart",
      "accuracy": 0.7675,
      "accuracy_std_dev": 0.019561171859068173,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0578507225,
      "avg_duration_per_task_ms": 50974.42,
      "avg_output_tokens_per_task": 4677.1725,
      "last_run_timestamp": "2026-10-24T12:36:03.010Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Alibaba",
      "accuracy": 0.7674509411764706,
      "accuracy_std_dev": 0.021678839985437112,
      "run_count": 17,
      "total_questions": 850,
      "avg_cost_per_task_usd": 0.12603030588235295,
      "avg_duration_per_task_ms": 92367.94588235294,
      "avg_output_tokens_per_task": 4470.871764705882,
      "last_run_timestamp": "2026-10-31T12:50:48.724Z"
    },
    {
      "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": 0.7672726363636364,
      "accuracy_std_dev": 0.030278820027065213,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.43524433636363635,
      "avg_duration_per_task_ms": 90898.93818181819,
      "avg_output_tokens_per_task": 5582.549090909091,
      "last_run_timestamp": "2026-10-31T12:33:27.768Z"
    },
    {
      "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": 0.7672223333333333,
      "accuracy_std_dev": 0.022682791106671363,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.418396455,
      "avg_duration_per_task_ms": 91613.17333333334,
      "avg_output_tokens_per_task": 5586.18,
      "last_run_timestamp": "2026-10-31T11:48:47.453Z"
    },
    {
      "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": 0.7668518333333335,
      "accuracy_std_dev": 0.02021403460126433,
      "run_count": 36,
      "total_questions": 1800,
      "avg_cost_per_task_usd": 0.5330153783333336,
      "avg_duration_per_task_ms": 118459.865,
      "avg_output_tokens_per_task": 8062.295,
      "last_run_timestamp": "2026-10-31T12:26:06.842Z"
    },
    {
      "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": "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-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Venice",
      "accuracy": 0.7666667777777777,
      "accuracy_std_dev": 0.026851093732235272,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.14037539777777777,
      "avg_duration_per_task_ms": 284966.99333333335,
      "avg_output_tokens_per_task": 11990.22,
      "last_run_timestamp": "2026-10-31T19:58:31.913Z"
    },
    {
      "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": "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": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Relace",
      "accuracy": 0.7666666666666667,
      "accuracy_std_dev": 0.010886757164351354,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.031067933333333332,
      "avg_duration_per_task_ms": 60278.37,
      "avg_output_tokens_per_task": 4752.3966666666665,
      "last_run_timestamp": "2026-10-31T12:25:28.648Z"
    },
    {
      "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": 0.7666666666666667,
      "accuracy_std_dev": 0.023517736937610882,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.04758358666666667,
      "avg_duration_per_task_ms": 110427.16444444444,
      "avg_output_tokens_per_task": 4684.946666666667,
      "last_run_timestamp": "2026-10-31T12:08:49.485Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.8-flash-20260902",
      "model_slug": "google/gemini-3.8-flash",
      "model_name": "Google: Gemini 3.8 Flash",
      "provider_name": "Google",
      "accuracy": 0.7666665,
      "accuracy_std_dev": 0.013333499999999998,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.27949224,
      "avg_duration_per_task_ms": 363191.33,
      "avg_output_tokens_per_task": 35460.86,
      "last_run_timestamp": "2026-10-15T21:31:19.497Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "DekaLLM",
      "accuracy": 0.7666665,
      "accuracy_std_dev": 0.026666499999999982,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01352679,
      "avg_duration_per_task_ms": 90848.6,
      "avg_output_tokens_per_task": 4507.38,
      "last_run_timestamp": "2026-10-10T11:18:50.119Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.7661113333333333,
      "accuracy_std_dev": 0.02045003245039534,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.4937038083333333,
      "avg_duration_per_task_ms": 124542.97333333333,
      "avg_output_tokens_per_task": 8042.625,
      "last_run_timestamp": "2026-10-31T12:30:31.663Z"
    },
    {
      "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": 0.766068730941704,
      "accuracy_std_dev": 0.02506483163840778,
      "run_count": 9,
      "total_questions": 446,
      "avg_cost_per_task_usd": 0.06749501121076233,
      "avg_duration_per_task_ms": 62014.139013452914,
      "avg_output_tokens_per_task": 4517.719730941704,
      "last_run_timestamp": "2026-10-31T13:21:43.928Z"
    },
    {
      "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": 0.7660001999999999,
      "accuracy_std_dev": 0.0260256591340162,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.12242998000000002,
      "avg_duration_per_task_ms": 384915.524,
      "avg_output_tokens_per_task": 20927.716,
      "last_run_timestamp": "2026-10-31T13:14:15.095Z"
    },
    {
      "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": 0.766,
      "accuracy_std_dev": 0.036203222693566926,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.017495754,
      "avg_duration_per_task_ms": 289899.184,
      "avg_output_tokens_per_task": 11370.218,
      "last_run_timestamp": "2026-10-31T20:10:47.431Z"
    },
    {
      "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": 0.7659258888888889,
      "accuracy_std_dev": 0.02853523512137403,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.02177499333333333,
      "avg_duration_per_task_ms": 226666.30666666667,
      "avg_output_tokens_per_task": 12033.74888888889,
      "last_run_timestamp": "2026-10-31T19:23:14.281Z"
    },
    {
      "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": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "BaseTen",
      "accuracy": 0.765560294605809,
      "accuracy_std_dev": 0.01753797388069671,
      "run_count": 5,
      "total_questions": 241,
      "avg_cost_per_task_usd": 0.26006869709543573,
      "avg_duration_per_task_ms": 145161.65560165976,
      "avg_output_tokens_per_task": 7387.091286307054,
      "last_run_timestamp": "2026-10-31T11:39:17.172Z"
    },
    {
      "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": 0.7654054594594596,
      "accuracy_std_dev": 0.021965741528503096,
      "run_count": 37,
      "total_questions": 1850,
      "avg_cost_per_task_usd": 0.5855930291891892,
      "avg_duration_per_task_ms": 71348.61243243243,
      "avg_output_tokens_per_task": 5712.135675675676,
      "last_run_timestamp": "2026-10-31T12:33:51.242Z"
    },
    {
      "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": 0.7653333999999999,
      "accuracy_std_dev": 0.023999911113168737,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.088749776,
      "avg_duration_per_task_ms": 386636.328,
      "avg_output_tokens_per_task": 14600.352,
      "last_run_timestamp": "2026-10-31T12:27:43.549Z"
    },
    {
      "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": 0.7651853333333333,
      "accuracy_std_dev": 0.06605486475482165,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.2539050577777778,
      "avg_duration_per_task_ms": 487309.1822222222,
      "avg_output_tokens_per_task": 29336.053333333333,
      "last_run_timestamp": "2026-10-31T19:26:18.381Z"
    },
    {
      "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": 0.7650000833333334,
      "accuracy_std_dev": 0.026822549594381503,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.53746671,
      "avg_duration_per_task_ms": 174605.76,
      "avg_output_tokens_per_task": 10683.748333333333,
      "last_run_timestamp": "2026-10-31T11:55:38.336Z"
    },
    {
      "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": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Cloudflare",
      "accuracy": 0.7647618571428572,
      "accuracy_std_dev": 0.016218418466392492,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.1530029685714286,
      "avg_duration_per_task_ms": 398501.6857142857,
      "avg_output_tokens_per_task": 18823.014285714286,
      "last_run_timestamp": "2026-10-31T20:34:04.127Z"
    },
    {
      "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": 0.7647368421052632,
      "accuracy_std_dev": 0.0276849081598111,
      "run_count": 19,
      "total_questions": 950,
      "avg_cost_per_task_usd": 0.04469010842105263,
      "avg_duration_per_task_ms": 88120.05473684211,
      "avg_output_tokens_per_task": 4333.72,
      "last_run_timestamp": "2026-10-31T13:24:41.102Z"
    },
    {
      "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": 0.7644446666666667,
      "accuracy_std_dev": 0.02474571917187565,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.3338584511111111,
      "avg_duration_per_task_ms": 643731.1911111111,
      "avg_output_tokens_per_task": 28077.5,
      "last_run_timestamp": "2026-10-31T19:49:31.051Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Novita",
      "accuracy": 0.7644444444444445,
      "accuracy_std_dev": 0.019116355871412038,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.11003770888888888,
      "avg_duration_per_task_ms": 252792.1288888889,
      "avg_output_tokens_per_task": 19117.18222222222,
      "last_run_timestamp": "2026-10-31T19:46:00.618Z"
    },
    {
      "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": 0.7644443333333334,
      "accuracy_std_dev": 0.024241503540186792,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.021445343333333335,
      "avg_duration_per_task_ms": 267641.68333333335,
      "avg_output_tokens_per_task": 15646.876666666667,
      "last_run_timestamp": "2026-10-31T13:19:25.796Z"
    },
    {
      "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": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "SambaNova",
      "accuracy": 0.76439777486911,
      "accuracy_std_dev": 0.015263579494993957,
      "run_count": 4,
      "total_questions": 191,
      "avg_cost_per_task_usd": 0.06284435078534031,
      "avg_duration_per_task_ms": 156780.0209424084,
      "avg_output_tokens_per_task": 9570.141361256545,
      "last_run_timestamp": "2026-10-31T19:43:06.929Z"
    },
    {
      "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": 0.7641665000000001,
      "accuracy_std_dev": 0.021586383312171588,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0185151775,
      "avg_duration_per_task_ms": 270096.6575,
      "avg_output_tokens_per_task": 14323.5575,
      "last_run_timestamp": "2026-10-31T12:30:14.640Z"
    },
    {
      "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": 0.764,
      "accuracy_std_dev": 0.020483142122242875,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.080228768,
      "avg_duration_per_task_ms": 494377.212,
      "avg_output_tokens_per_task": 14669.972,
      "last_run_timestamp": "2026-10-31T12:27:34.738Z"
    },
    {
      "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": 0.7638889166666667,
      "accuracy_std_dev": 0.035219828608939255,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.140352225,
      "avg_duration_per_task_ms": 834808.895,
      "avg_output_tokens_per_task": 31424.338333333333,
      "last_run_timestamp": "2026-10-31T20:06:11.663Z"
    },
    {
      "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": 0.7638888333333334,
      "accuracy_std_dev": 0.03795297203740732,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.09057705333333334,
      "avg_duration_per_task_ms": 303333.575,
      "avg_output_tokens_per_task": 15558.463333333333,
      "last_run_timestamp": "2026-10-31T19:58:49.610Z"
    },
    {
      "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": 0.7637037777777778,
      "accuracy_std_dev": 0.021798347556015337,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.06269429111111112,
      "avg_duration_per_task_ms": 113855.65555555555,
      "avg_output_tokens_per_task": 4583.311111111111,
      "last_run_timestamp": "2026-10-31T12:40:08.732Z"
    },
    {
      "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": 0.7637036666666667,
      "accuracy_std_dev": 0.01835498758739252,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.029433213333333333,
      "avg_duration_per_task_ms": 122522.84,
      "avg_output_tokens_per_task": 12030.735555555555,
      "last_run_timestamp": "2026-10-31T19:08:17.847Z"
    },
    {
      "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": 0.7635896153846154,
      "accuracy_std_dev": 0.024476071585053976,
      "run_count": 13,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.03524474461538461,
      "avg_duration_per_task_ms": 67488.58923076923,
      "avg_output_tokens_per_task": 4789.504615384615,
      "last_run_timestamp": "2026-10-31T13:07:57.724Z"
    },
    {
      "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": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "Darkbloom",
      "accuracy": 0.763333375,
      "accuracy_std_dev": 0.024267095921934618,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0630558525,
      "avg_duration_per_task_ms": 1180279.39,
      "avg_output_tokens_per_task": 60139.9825,
      "last_run_timestamp": "2026-10-31T19:16:07.249Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "CoreWeave",
      "accuracy": 0.7633332500000001,
      "accuracy_std_dev": 0.03605539719220273,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.157135485,
      "avg_duration_per_task_ms": 411704.8,
      "avg_output_tokens_per_task": 26150.13,
      "last_run_timestamp": "2026-08-29T08:26:04.467Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Xiaomi",
      "accuracy": 0.7633332500000001,
      "accuracy_std_dev": 0.013743584364622665,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.01189864,
      "avg_duration_per_task_ms": 642869.1,
      "avg_output_tokens_per_task": 23140.98,
      "last_run_timestamp": "2026-10-24T13:28:08.470Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "BaseTen",
      "accuracy": 0.763333,
      "accuracy_std_dev": 0.010000000000000009,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.020903190000000002,
      "avg_duration_per_task_ms": 62428.04,
      "avg_output_tokens_per_task": 8093.56,
      "last_run_timestamp": "2026-10-31T12:43:49.743Z"
    },
    {
      "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": 0.7630302727272728,
      "accuracy_std_dev": 0.03170123269496259,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.15556449636363637,
      "avg_duration_per_task_ms": 593095.6363636364,
      "avg_output_tokens_per_task": 12420.629090909091,
      "last_run_timestamp": "2026-10-31T13:02:37.599Z"
    },
    {
      "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": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "GMICloud",
      "accuracy": 0.7626668,
      "accuracy_std_dev": 0.014966487007978884,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.009859876,
      "avg_duration_per_task_ms": 578544.56,
      "avg_output_tokens_per_task": 20173.776,
      "last_run_timestamp": "2026-10-31T14:48:08.689Z"
    },
    {
      "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": 0.7626666,
      "accuracy_std_dev": 0.02954472165785287,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.009234846,
      "avg_duration_per_task_ms": 195074.734,
      "avg_output_tokens_per_task": 4475.264,
      "last_run_timestamp": "2026-10-31T12:25:34.861Z"
    },
    {
      "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": 0.7625418160535117,
      "accuracy_std_dev": 0.02147753020972668,
      "run_count": 12,
      "total_questions": 598,
      "avg_cost_per_task_usd": 0.09006360702341139,
      "avg_duration_per_task_ms": 48936.09866220736,
      "avg_output_tokens_per_task": 4348.224080267559,
      "last_run_timestamp": "2026-10-31T12:39:14.326Z"
    },
    {
      "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": 0.7625417625418061,
      "accuracy_std_dev": 0.01910679234955175,
      "run_count": 6,
      "total_questions": 299,
      "avg_cost_per_task_usd": 0.03341546153846154,
      "avg_duration_per_task_ms": 144903.0602006689,
      "avg_output_tokens_per_task": 21777.678929765887,
      "last_run_timestamp": "2026-10-31T12:18:15.192Z"
    },
    {
      "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": 0.7624352026726057,
      "accuracy_std_dev": 0.020878013624710118,
      "run_count": 9,
      "total_questions": 449,
      "avg_cost_per_task_usd": 0.010567316258351892,
      "avg_duration_per_task_ms": 295155.5011135857,
      "avg_output_tokens_per_task": 12618.436525612473,
      "last_run_timestamp": "2026-10-31T19:48:48.460Z"
    },
    {
      "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": 0.7623223984962407,
      "accuracy_std_dev": 0.0190669296226313,
      "run_count": 8,
      "total_questions": 399,
      "avg_cost_per_task_usd": 0.011861127819548874,
      "avg_duration_per_task_ms": 357840.4411027569,
      "avg_output_tokens_per_task": 17112.51127819549,
      "last_run_timestamp": "2026-10-31T11:47:32.862Z"
    },
    {
      "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": "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": 0.7622222916666667,
      "accuracy_std_dev": 0.03325923894759566,
      "run_count": 24,
      "total_questions": 1200,
      "avg_cost_per_task_usd": 0.5556523033333334,
      "avg_duration_per_task_ms": 182337.15583333332,
      "avg_output_tokens_per_task": 10601.08,
      "last_run_timestamp": "2026-10-31T12:37:18.963Z"
    },
    {
      "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": 0.7622221111111112,
      "accuracy_std_dev": 0.02931314119959945,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.10448193111111109,
      "avg_duration_per_task_ms": 237398.75333333333,
      "avg_output_tokens_per_task": 11214.604444444445,
      "last_run_timestamp": "2026-10-31T20:03:31.357Z"
    },
    {
      "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": 0.7622099340659342,
      "accuracy_std_dev": 0.03802706823103689,
      "run_count": 11,
      "total_questions": 546,
      "avg_cost_per_task_usd": 0.09558790476190476,
      "avg_duration_per_task_ms": 736797.2087912088,
      "avg_output_tokens_per_task": 10344.02564102564,
      "last_run_timestamp": "2026-10-31T20:25:10.002Z"
    },
    {
      "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": 0.762177653295129,
      "accuracy_std_dev": 0.016008036097981727,
      "run_count": 7,
      "total_questions": 349,
      "avg_cost_per_task_usd": 0.027902553008595985,
      "avg_duration_per_task_ms": 104077.65616045846,
      "avg_output_tokens_per_task": 17491.627507163324,
      "last_run_timestamp": "2026-10-31T11:55:25.455Z"
    },
    {
      "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": 0.7621775042979944,
      "accuracy_std_dev": 0.026194934880387766,
      "run_count": 7,
      "total_questions": 349,
      "avg_cost_per_task_usd": 0.06229761318051576,
      "avg_duration_per_task_ms": 372098.9226361032,
      "avg_output_tokens_per_task": 49911.94555873926,
      "last_run_timestamp": "2026-10-31T19:59:18.193Z"
    },
    {
      "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": 0.7620513846153846,
      "accuracy_std_dev": 0.024443852124782315,
      "run_count": 13,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.02194166153846154,
      "avg_duration_per_task_ms": 173997.9723076923,
      "avg_output_tokens_per_task": 16407.72,
      "last_run_timestamp": "2026-10-31T12:11:52.708Z"
    },
    {
      "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": 0.7619782998454405,
      "accuracy_std_dev": 0.039147489254698645,
      "run_count": 13,
      "total_questions": 647,
      "avg_cost_per_task_usd": 0.06417079598145287,
      "avg_duration_per_task_ms": 203709.1777434312,
      "avg_output_tokens_per_task": 11216.542503863988,
      "last_run_timestamp": "2026-10-31T14:18:08.529Z"
    },
    {
      "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-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "DekaLLM",
      "accuracy": 0.761784404040404,
      "accuracy_std_dev": 0.046884570628299496,
      "run_count": 4,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.03727159090909091,
      "avg_duration_per_task_ms": 138498.65151515152,
      "avg_output_tokens_per_task": 4381.035353535353,
      "last_run_timestamp": "2026-10-31T12:40:46.071Z"
    },
    {
      "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": "Cloudflare",
      "accuracy": 0.7616666666666667,
      "accuracy_std_dev": 0.018282859816110712,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.031528701666666666,
      "avg_duration_per_task_ms": 227802.605,
      "avg_output_tokens_per_task": 15291.826666666666,
      "last_run_timestamp": "2026-10-31T12:46:55.135Z"
    },
    {
      "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": "anthropic/claude-4.6-opus-20260205",
      "model_slug": "anthropic/claude-opus-4.6",
      "model_name": "Anthropic: Claude Opus 4.6",
      "provider_name": "Anthropic",
      "accuracy": 0.7614285714285715,
      "accuracy_std_dev": 0.027276150573386282,
      "run_count": 14,
      "total_questions": 700,
      "avg_cost_per_task_usd": 0.5232190085714287,
      "avg_duration_per_task_ms": 215585.69428571427,
      "avg_output_tokens_per_task": 9953.878571428571,
      "last_run_timestamp": "2026-10-31T12:37:18.962Z"
    },
    {
      "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": 0.7614249072580644,
      "accuracy_std_dev": 0.03191360961423825,
      "run_count": 10,
      "total_questions": 496,
      "avg_cost_per_task_usd": 0.04908138104838709,
      "avg_duration_per_task_ms": 385516.27419354836,
      "avg_output_tokens_per_task": 12910.963709677419,
      "last_run_timestamp": "2026-10-31T13:30:45.600Z"
    },
    {
      "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": 0.761389,
      "accuracy_std_dev": 0.03531505729383243,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.19697044166666666,
      "avg_duration_per_task_ms": 478792.93666666665,
      "avg_output_tokens_per_task": 27912.631666666668,
      "last_run_timestamp": "2026-10-31T19:55:50.954Z"
    },
    {
      "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": 0.7612120909090909,
      "accuracy_std_dev": 0.02288640210595304,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.54388782,
      "avg_duration_per_task_ms": 180374.2618181818,
      "avg_output_tokens_per_task": 10717.156363636364,
      "last_run_timestamp": "2026-10-31T12:58:32.935Z"
    },
    {
      "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": 0.7611111666666667,
      "accuracy_std_dev": 0.02274381750364017,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.03547926999999999,
      "avg_duration_per_task_ms": 167748.38666666666,
      "avg_output_tokens_per_task": 16394.798333333332,
      "last_run_timestamp": "2026-10-31T12:45:56.690Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Sail Research",
      "accuracy": 0.76099995,
      "accuracy_std_dev": 0.018651859903170524,
      "run_count": 20,
      "total_questions": 1000,
      "avg_cost_per_task_usd": 0.010585624,
      "avg_duration_per_task_ms": 322498.887,
      "avg_output_tokens_per_task": 14838.748,
      "last_run_timestamp": "2026-10-31T13:21:16.354Z"
    },
    {
      "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": 0.7609786762860727,
      "accuracy_std_dev": 0.02511202768812936,
      "run_count": 16,
      "total_questions": 797,
      "avg_cost_per_task_usd": 0.054708953575909665,
      "avg_duration_per_task_ms": 274645.58720200753,
      "avg_output_tokens_per_task": 12041.819322459221,
      "last_run_timestamp": "2026-10-31T12:07:50.786Z"
    },
    {
      "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": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Novita",
      "accuracy": 0.7607886785714285,
      "accuracy_std_dev": 0.017367969160654122,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.018826165178571424,
      "avg_duration_per_task_ms": 120325.66071428571,
      "avg_output_tokens_per_task": 14433.707589285714,
      "last_run_timestamp": "2026-10-31T12:07:52.801Z"
    },
    {
      "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": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7606666,
      "accuracy_std_dev": 0.023371542050964467,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.010802634,
      "avg_duration_per_task_ms": 1014797.92,
      "avg_output_tokens_per_task": 13061.106,
      "last_run_timestamp": "2026-10-31T20:29:45.923Z"
    },
    {
      "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": 0.7606666,
      "accuracy_std_dev": 0.03705257046198009,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.00899061,
      "avg_duration_per_task_ms": 185087.438,
      "avg_output_tokens_per_task": 15485.002,
      "last_run_timestamp": "2026-10-31T19:07:06.878Z"
    },
    {
      "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": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "DeepInfra",
      "accuracy": 0.7605362068965518,
      "accuracy_std_dev": 0.008806243921218638,
      "run_count": 7,
      "total_questions": 348,
      "avg_cost_per_task_usd": 0.007250201149425288,
      "avg_duration_per_task_ms": 286627.4166666667,
      "avg_output_tokens_per_task": 12069.367816091954,
      "last_run_timestamp": "2026-10-31T20:10:56.893Z"
    },
    {
      "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": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "AkashML",
      "accuracy": 0.7604166354166666,
      "accuracy_std_dev": 0.010476531320050537,
      "run_count": 4,
      "total_questions": 192,
      "avg_cost_per_task_usd": 0.29691326041666666,
      "avg_duration_per_task_ms": 310718.984375,
      "avg_output_tokens_per_task": 7030.182291666667,
      "last_run_timestamp": "2026-10-31T12:13:14.359Z"
    },
    {
      "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": 0.7603703333333334,
      "accuracy_std_dev": 0.033126907327897526,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.08524436222222224,
      "avg_duration_per_task_ms": 566210.18,
      "avg_output_tokens_per_task": 12520.904444444444,
      "last_run_timestamp": "2026-10-31T12:26:16.049Z"
    },
    {
      "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": 0.7601351621621621,
      "accuracy_std_dev": 0.012280818789432917,
      "run_count": 6,
      "total_questions": 296,
      "avg_cost_per_task_usd": 0.032259290540540546,
      "avg_duration_per_task_ms": 475460.4324324324,
      "avg_output_tokens_per_task": 12381.702702702703,
      "last_run_timestamp": "2026-10-31T12:56:10.594Z"
    },
    {
      "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": 0.7600002499999999,
      "accuracy_std_dev": 0.008164965810553043,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.015787874999999996,
      "avg_duration_per_task_ms": 119015.24,
      "avg_output_tokens_per_task": 12260.915,
      "last_run_timestamp": "2026-10-31T13:06:19.395Z"
    },
    {
      "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": 0.7600002499999999,
      "accuracy_std_dev": 0.027080189561513434,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.02989183,
      "avg_duration_per_task_ms": 125270.795,
      "avg_output_tokens_per_task": 12269.195,
      "last_run_timestamp": "2026-10-31T12:01:23.978Z"
    },
    {
      "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": 0.7600001999999999,
      "accuracy_std_dev": 0.03126929626582598,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.114411384,
      "avg_duration_per_task_ms": 874872.748,
      "avg_output_tokens_per_task": 24203.156,
      "last_run_timestamp": "2026-10-31T12:44:40.958Z"
    },
    {
      "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": 0.7600001,
      "accuracy_std_dev": 0.02905935688018578,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.045049942,
      "avg_duration_per_task_ms": 188166.976,
      "avg_output_tokens_per_task": 9673.868,
      "last_run_timestamp": "2026-10-31T20:17:26.781Z"
    },
    {
      "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": "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": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Novita",
      "accuracy": 0.76,
      "accuracy_std_dev": 0.025991443038046205,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.08891490399999999,
      "avg_duration_per_task_ms": 301406.882,
      "avg_output_tokens_per_task": 12044.934,
      "last_run_timestamp": "2026-10-31T12:46:14.833Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "Cohere",
      "accuracy": 0.76,
      "accuracy_std_dev": 0.0066669999999999785,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.10151575999999998,
      "avg_duration_per_task_ms": 434936.03,
      "avg_output_tokens_per_task": 13962.46,
      "last_run_timestamp": "2026-10-24T12:32:07.608Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Makora",
      "accuracy": 0.76,
      "accuracy_std_dev": 0.0066669999999999785,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.07038507,
      "avg_duration_per_task_ms": 256302.49,
      "avg_output_tokens_per_task": 10045.27,
      "last_run_timestamp": "2026-08-26T00:00:05.638Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Novita",
      "accuracy": 0.7599999000000001,
      "accuracy_std_dev": 0.030840204640858015,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.008153706000000002,
      "avg_duration_per_task_ms": 200971.108,
      "avg_output_tokens_per_task": 4219.412,
      "last_run_timestamp": "2026-10-31T12:49:39.925Z"
    },
    {
      "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": 0.759846288184438,
      "accuracy_std_dev": 0.015452349289717374,
      "run_count": 7,
      "total_questions": 347,
      "avg_cost_per_task_usd": 0.15869831700288184,
      "avg_duration_per_task_ms": 791662.6916426513,
      "avg_output_tokens_per_task": 30669.08069164265,
      "last_run_timestamp": "2026-10-31T19:07:30.175Z"
    },
    {
      "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": 0.7594444166666666,
      "accuracy_std_dev": 0.033715523315080306,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.09267692833333334,
      "avg_duration_per_task_ms": 244983.77333333335,
      "avg_output_tokens_per_task": 12483.865,
      "last_run_timestamp": "2026-10-31T12:53:55.102Z"
    },
    {
      "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": 0.7592155882352941,
      "accuracy_std_dev": 0.027618637596970693,
      "run_count": 17,
      "total_questions": 850,
      "avg_cost_per_task_usd": 0.011689743529411765,
      "avg_duration_per_task_ms": 166650.9011764706,
      "avg_output_tokens_per_task": 5474.670588235294,
      "last_run_timestamp": "2026-10-31T12:46:12.537Z"
    },
    {
      "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": "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": 0.7587878181818182,
      "accuracy_std_dev": 0.01653114932933463,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.17482494545454544,
      "avg_duration_per_task_ms": 315756.4018181818,
      "avg_output_tokens_per_task": 10740.534545454546,
      "last_run_timestamp": "2026-10-31T11:13:27.901Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Cloudflare",
      "accuracy": 0.7585185555555556,
      "accuracy_std_dev": 0.020793455498792097,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.027907260000000003,
      "avg_duration_per_task_ms": 202475.69777777776,
      "avg_output_tokens_per_task": 13617.924444444445,
      "last_run_timestamp": "2026-10-31T19:55:08.023Z"
    },
    {
      "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": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Relace",
      "accuracy": 0.7583333333333333,
      "accuracy_std_dev": 0.03125451671181126,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.06620967333333333,
      "avg_duration_per_task_ms": 161431.49,
      "avg_output_tokens_per_task": 10316.483333333334,
      "last_run_timestamp": "2026-10-31T19:43:05.932Z"
    },
    {
      "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": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Baidu",
      "accuracy": 0.7580951428571429,
      "accuracy_std_dev": 0.016605241912709113,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.048255082857142854,
      "avg_duration_per_task_ms": 69808.27142857143,
      "avg_output_tokens_per_task": 4328.277142857143,
      "last_run_timestamp": "2026-10-31T12:44:24.094Z"
    },
    {
      "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": 0.7579487692307693,
      "accuracy_std_dev": 0.03142044042230091,
      "run_count": 13,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.019486126153846157,
      "avg_duration_per_task_ms": 195283.25692307693,
      "avg_output_tokens_per_task": 14827.041538461539,
      "last_run_timestamp": "2026-10-31T12:48:59.031Z"
    },
    {
      "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": 0.7578806257545272,
      "accuracy_std_dev": 0.03401018014418623,
      "run_count": 10,
      "total_questions": 497,
      "avg_cost_per_task_usd": 0.10939716700201207,
      "avg_duration_per_task_ms": 158414.4527162978,
      "avg_output_tokens_per_task": 9282.295774647888,
      "last_run_timestamp": "2026-10-31T19:46:27.571Z"
    },
    {
      "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": 0.7577777777777778,
      "accuracy_std_dev": 0.026480824650543643,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.025315553333333334,
      "avg_duration_per_task_ms": 238703.38,
      "avg_output_tokens_per_task": 12990.084444444445,
      "last_run_timestamp": "2026-10-31T19:02:15.155Z"
    },
    {
      "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": 0.7577777777777778,
      "accuracy_std_dev": 0.01859267080077988,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.1710706822222222,
      "avg_duration_per_task_ms": 189011.57111111112,
      "avg_output_tokens_per_task": 11157.495555555555,
      "last_run_timestamp": "2026-10-31T19:18:33.522Z"
    },
    {
      "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": 0.75777775,
      "accuracy_std_dev": 0.03177850620866512,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.5383182933333334,
      "avg_duration_per_task_ms": 190201.31833333333,
      "avg_output_tokens_per_task": 10680.48,
      "last_run_timestamp": "2026-10-31T12:48:37.862Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "ParetoAlpha",
      "accuracy": 0.7577776666666667,
      "accuracy_std_dev": 0.013698571612479267,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.28612407999999995,
      "avg_duration_per_task_ms": 2065607.3133333332,
      "avg_output_tokens_per_task": 10347.066666666668,
      "last_run_timestamp": "2026-10-31T11:54:15.773Z"
    },
    {
      "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": 0.7577286370967742,
      "accuracy_std_dev": 0.0230900961972877,
      "run_count": 10,
      "total_questions": 496,
      "avg_cost_per_task_usd": 0.01778327016129032,
      "avg_duration_per_task_ms": 195849.37096774194,
      "avg_output_tokens_per_task": 5723.635080645161,
      "last_run_timestamp": "2026-10-31T12:36:56.960Z"
    },
    {
      "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": 0.7576190714285714,
      "accuracy_std_dev": 0.02056452651139227,
      "run_count": 14,
      "total_questions": 700,
      "avg_cost_per_task_usd": 0.021698402857142855,
      "avg_duration_per_task_ms": 123305.51285714285,
      "avg_output_tokens_per_task": 16890.44,
      "last_run_timestamp": "2026-10-31T12:41:20.328Z"
    },
    {
      "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": 0.757619,
      "accuracy_std_dev": 0.020871170540108334,
      "run_count": 14,
      "total_questions": 700,
      "avg_cost_per_task_usd": 0.0069606071428571425,
      "avg_duration_per_task_ms": 264179.4914285714,
      "avg_output_tokens_per_task": 15717.655714285715,
      "last_run_timestamp": "2026-10-31T13:22:04.008Z"
    },
    {
      "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": 0.7575,
      "accuracy_std_dev": 0.05171382256708548,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.26214604,
      "avg_duration_per_task_ms": 26048.875,
      "avg_output_tokens_per_task": 12109.95,
      "last_run_timestamp": "2026-10-31T11:18:07.571Z"
    },
    {
      "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": 0.7575,
      "accuracy_std_dev": 0.023078766095699303,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.012481125,
      "avg_duration_per_task_ms": 215467.87,
      "avg_output_tokens_per_task": 4890.4975,
      "last_run_timestamp": "2026-10-31T11:56:22.568Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "NextBit",
      "accuracy": 0.7573333999999999,
      "accuracy_std_dev": 0.03991088294989224,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.023026752,
      "avg_duration_per_task_ms": 161149.652,
      "avg_output_tokens_per_task": 18473.992,
      "last_run_timestamp": "2026-10-03T08:55:31.322Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Alibaba",
      "accuracy": 0.7572916875,
      "accuracy_std_dev": 0.027315378536455315,
      "run_count": 16,
      "total_questions": 800,
      "avg_cost_per_task_usd": 0.090318465,
      "avg_duration_per_task_ms": 189685.91875,
      "avg_output_tokens_per_task": 11015.33625,
      "last_run_timestamp": "2026-10-31T19:54:53.711Z"
    },
    {
      "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": 0.757037,
      "accuracy_std_dev": 0.024966444422686845,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.022134202222222222,
      "avg_duration_per_task_ms": 462417.70666666667,
      "avg_output_tokens_per_task": 13881.615555555556,
      "last_run_timestamp": "2026-10-31T19:04:16.045Z"
    },
    {
      "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": 0.756708080495356,
      "accuracy_std_dev": 0.027120750325406145,
      "run_count": 13,
      "total_questions": 646,
      "avg_cost_per_task_usd": 0.19143365634674925,
      "avg_duration_per_task_ms": 498299.07585139317,
      "avg_output_tokens_per_task": 20833.128482972137,
      "last_run_timestamp": "2026-10-31T14:12:32.839Z"
    },
    {
      "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": 0.7566669999999999,
      "accuracy_std_dev": 0.010000000000000009,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.08839812,
      "avg_duration_per_task_ms": 73380.11,
      "avg_output_tokens_per_task": 8719.83,
      "last_run_timestamp": "2026-10-31T12:44:24.100Z"
    },
    {
      "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": 0.7566669999999999,
      "accuracy_std_dev": 0.010000000000000009,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.06446149,
      "avg_duration_per_task_ms": 318929.97,
      "avg_output_tokens_per_task": 13284.66,
      "last_run_timestamp": "2026-10-31T12:25:36.318Z"
    },
    {
      "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": 0.7566668,
      "accuracy_std_dev": 0.01819650760338369,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.151380274,
      "avg_duration_per_task_ms": 165330.14,
      "avg_output_tokens_per_task": 19976.098,
      "last_run_timestamp": "2026-10-31T12:20:13.275Z"
    },
    {
      "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": 0.7566666666666667,
      "accuracy_std_dev": 0.022026835486641215,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.06927284666666667,
      "avg_duration_per_task_ms": 313911.26,
      "avg_output_tokens_per_task": 14319.526666666667,
      "last_run_timestamp": "2026-10-31T12:43:48.271Z"
    },
    {
      "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": 0.7564444666666666,
      "accuracy_std_dev": 0.023958871344776582,
      "run_count": 15,
      "total_questions": 750,
      "avg_cost_per_task_usd": 0.04415017200000001,
      "avg_duration_per_task_ms": 138405.18933333334,
      "avg_output_tokens_per_task": 4510.322666666667,
      "last_run_timestamp": "2026-10-31T12:43:47.323Z"
    },
    {
      "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": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7562962222222223,
      "accuracy_std_dev": 0.028650564911156574,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.18542253777777779,
      "avg_duration_per_task_ms": 171878.6222222222,
      "avg_output_tokens_per_task": 9625.26888888889,
      "last_run_timestamp": "2026-10-31T20:09:19.251Z"
    },
    {
      "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": 0.7562961111111112,
      "accuracy_std_dev": 0.02966657881426859,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.03134207111111111,
      "avg_duration_per_task_ms": 175492.92222222223,
      "avg_output_tokens_per_task": 14182.893333333333,
      "last_run_timestamp": "2026-10-31T11:54:37.503Z"
    },
    {
      "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": 0.7558333333333334,
      "accuracy_std_dev": 0.021131098726495254,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.01632131833333333,
      "avg_duration_per_task_ms": 115260.22166666666,
      "avg_output_tokens_per_task": 4430.61,
      "last_run_timestamp": "2026-10-31T12:09:39.076Z"
    },
    {
      "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": "GMICloud",
      "accuracy": 0.7557575454545454,
      "accuracy_std_dev": 0.021134116315326572,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.0626866909090909,
      "avg_duration_per_task_ms": 166564.66363636364,
      "avg_output_tokens_per_task": 4635.490909090909,
      "last_run_timestamp": "2026-10-31T13:36:23.199Z"
    },
    {
      "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": 0.7555554444444444,
      "accuracy_std_dev": 0.018856167739030906,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.0086613,
      "avg_duration_per_task_ms": 117282.64,
      "avg_output_tokens_per_task": 5380.108888888889,
      "last_run_timestamp": "2026-10-31T13:20:27.978Z"
    },
    {
      "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": 0.7554545454545455,
      "accuracy_std_dev": 0.03830433230509108,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.16801238363636364,
      "avg_duration_per_task_ms": 309650.71636363637,
      "avg_output_tokens_per_task": 24145.685454545455,
      "last_run_timestamp": "2026-10-31T13:11:05.620Z"
    },
    {
      "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": 0.7552378571428572,
      "accuracy_std_dev": 0.014126057041900291,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.009880265714285713,
      "avg_duration_per_task_ms": 245506.26571428572,
      "avg_output_tokens_per_task": 16074.99142857143,
      "last_run_timestamp": "2026-10-31T12:27:44.073Z"
    },
    {
      "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": 0.7552085625,
      "accuracy_std_dev": 0.029782500000000045,
      "run_count": 2,
      "total_questions": 96,
      "avg_cost_per_task_usd": 0.02997340625,
      "avg_duration_per_task_ms": 1679708.6145833333,
      "avg_output_tokens_per_task": 72130.22916666667,
      "last_run_timestamp": "2026-10-31T20:01:41.151Z"
    },
    {
      "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": 0.7551024285714285,
      "accuracy_std_dev": 0.011805499999999969,
      "run_count": 2,
      "total_questions": 98,
      "avg_cost_per_task_usd": 0.27501580612244897,
      "avg_duration_per_task_ms": 65997.72448979592,
      "avg_output_tokens_per_task": 8842.591836734693,
      "last_run_timestamp": "2026-10-31T12:36:21.331Z"
    },
    {
      "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": 0.7550476285714286,
      "accuracy_std_dev": 0.02909609450973075,
      "run_count": 35,
      "total_questions": 1750,
      "avg_cost_per_task_usd": 0.20630642971428564,
      "avg_duration_per_task_ms": 116033.43085714286,
      "avg_output_tokens_per_task": 7739.464571428572,
      "last_run_timestamp": "2026-10-31T12:48:39.187Z"
    },
    {
      "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": 0.7550336026845637,
      "accuracy_std_dev": 0.024862145639416477,
      "run_count": 15,
      "total_questions": 745,
      "avg_cost_per_task_usd": 0.08558708322147651,
      "avg_duration_per_task_ms": 635566.877852349,
      "avg_output_tokens_per_task": 21211.187919463086,
      "last_run_timestamp": "2026-10-31T13:12:07.908Z"
    },
    {
      "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": 0.7550335167785235,
      "accuracy_std_dev": 0.023364430687122125,
      "run_count": 6,
      "total_questions": 298,
      "avg_cost_per_task_usd": 0.012232758389261746,
      "avg_duration_per_task_ms": 267603.8926174497,
      "avg_output_tokens_per_task": 4166.208053691275,
      "last_run_timestamp": "2026-10-31T12:14:04.084Z"
    },
    {
      "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": "Modal",
      "accuracy": 0.7549998,
      "accuracy_std_dev": 0.029636688974310207,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.014858368,
      "avg_duration_per_task_ms": 71120.474,
      "avg_output_tokens_per_task": 4319.272,
      "last_run_timestamp": "2026-10-31T11:35:00.891Z"
    },
    {
      "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": 0.7547041935483871,
      "accuracy_std_dev": 0.023888446777469652,
      "run_count": 10,
      "total_questions": 496,
      "avg_cost_per_task_usd": 0.01950188306451613,
      "avg_duration_per_task_ms": 216554.71370967742,
      "avg_output_tokens_per_task": 4241.487903225807,
      "last_run_timestamp": "2026-10-31T11:51:07.775Z"
    },
    {
      "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": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "ModelRun",
      "accuracy": 0.7541725722460657,
      "accuracy_std_dev": 0.027627708682145743,
      "run_count": 14,
      "total_questions": 699,
      "avg_cost_per_task_usd": 0.0887023748211731,
      "avg_duration_per_task_ms": 129758.05722460659,
      "avg_output_tokens_per_task": 12053.809728183118,
      "last_run_timestamp": "2026-10-31T18:56:42.115Z"
    },
    {
      "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": 0.75416675,
      "accuracy_std_dev": 0.019096569464219906,
      "run_count": 13,
      "total_questions": 640,
      "avg_cost_per_task_usd": 0.0324113640625,
      "avg_duration_per_task_ms": 163294.796875,
      "avg_output_tokens_per_task": 15840.8265625,
      "last_run_timestamp": "2026-10-31T11:41:35.064Z"
    },
    {
      "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": 0.7538461538461538,
      "accuracy_std_dev": 0.029405579849477954,
      "run_count": 13,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.008398384615384618,
      "avg_duration_per_task_ms": 128001.65076923076,
      "avg_output_tokens_per_task": 4472.704615384615,
      "last_run_timestamp": "2026-10-31T13:18:32.761Z"
    },
    {
      "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": 0.75375,
      "accuracy_std_dev": 0.02086800226622888,
      "run_count": 24,
      "total_questions": 1200,
      "avg_cost_per_task_usd": 0.2756704208333333,
      "avg_duration_per_task_ms": 243435.00416666668,
      "avg_output_tokens_per_task": 11821.2175,
      "last_run_timestamp": "2026-10-31T12:36:58.963Z"
    },
    {
      "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": 0.75375,
      "accuracy_std_dev": 0.02114362919711279,
      "run_count": 16,
      "total_questions": 800,
      "avg_cost_per_task_usd": 0.010135285,
      "avg_duration_per_task_ms": 220308.23,
      "avg_output_tokens_per_task": 4260.24875,
      "last_run_timestamp": "2026-10-31T13:20:23.742Z"
    },
    {
      "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": 0.753611125,
      "accuracy_std_dev": 0.03106020227224396,
      "run_count": 24,
      "total_questions": 1200,
      "avg_cost_per_task_usd": 0.20096273666666667,
      "avg_duration_per_task_ms": 97219.05666666667,
      "avg_output_tokens_per_task": 7598.600833333333,
      "last_run_timestamp": "2026-10-31T12:38:13.068Z"
    },
    {
      "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": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "GMICloud",
      "accuracy": 0.7533335,
      "accuracy_std_dev": 0.00666650000000002,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.07113415,
      "avg_duration_per_task_ms": 677434.95,
      "avg_output_tokens_per_task": 13100.77,
      "last_run_timestamp": "2026-10-31T13:06:57.362Z"
    },
    {
      "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": 0.75333348,
      "accuracy_std_dev": 0.022784037298284095,
      "run_count": 25,
      "total_questions": 1250,
      "avg_cost_per_task_usd": 0.24681676960000004,
      "avg_duration_per_task_ms": 219719.8136,
      "avg_output_tokens_per_task": 10167.9728,
      "last_run_timestamp": "2026-10-31T12:43:33.299Z"
    },
    {
      "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": 0.7533334166666666,
      "accuracy_std_dev": 0.10114155829781209,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.15308477166666667,
      "avg_duration_per_task_ms": 578170.1783333333,
      "avg_output_tokens_per_task": 28402.441666666666,
      "last_run_timestamp": "2026-10-31T19:37:30.834Z"
    },
    {
      "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": 0.7533333333333333,
      "accuracy_std_dev": 0.01721340993398913,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.33945626,
      "avg_duration_per_task_ms": 116389.55333333333,
      "avg_output_tokens_per_task": 7044.336666666667,
      "last_run_timestamp": "2026-10-31T12:36:01.147Z"
    },
    {
      "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": 0.7533333333333333,
      "accuracy_std_dev": 0.03925236487426234,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.03720142666666666,
      "avg_duration_per_task_ms": 244561.18666666668,
      "avg_output_tokens_per_task": 13729.346666666666,
      "last_run_timestamp": "2026-10-31T19:28:27.377Z"
    },
    {
      "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": 0.7533333333333333,
      "accuracy_std_dev": 0.02277092037860767,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.020851383333333334,
      "avg_duration_per_task_ms": 133350.03333333333,
      "avg_output_tokens_per_task": 13555.86,
      "last_run_timestamp": "2026-10-31T19:21:19.082Z"
    },
    {
      "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": 0.7533333333333333,
      "accuracy_std_dev": 0.005443174461857933,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.018967786666666667,
      "avg_duration_per_task_ms": 85971.64666666667,
      "avg_output_tokens_per_task": 11730.173333333334,
      "last_run_timestamp": "2026-10-31T12:40:46.055Z"
    },
    {
      "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": 0.7533333333333333,
      "accuracy_std_dev": 0.029186609056590034,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.09943221,
      "avg_duration_per_task_ms": 284246.225,
      "avg_output_tokens_per_task": 12177.723333333333,
      "last_run_timestamp": "2026-10-31T12:41:58.031Z"
    },
    {
      "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": 0.7533332857142857,
      "accuracy_std_dev": 0.015118704911601446,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.3460215285714286,
      "avg_duration_per_task_ms": 119668.60857142857,
      "avg_output_tokens_per_task": 7072.9971428571425,
      "last_run_timestamp": "2026-10-31T12:24:55.233Z"
    },
    {
      "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": 0.7533332222222222,
      "accuracy_std_dev": 0.028109090837812517,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.01274101777777778,
      "avg_duration_per_task_ms": 229472.85111111111,
      "avg_output_tokens_per_task": 13907.048888888889,
      "last_run_timestamp": "2026-10-31T20:16:35.156Z"
    },
    {
      "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": "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": "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": "anthropic/claude-sonnet-5-20260630",
      "model_slug": "anthropic/claude-sonnet-5",
      "model_name": "Anthropic: Claude Sonnet 5",
      "provider_name": "Google",
      "accuracy": 0.7531531351351352,
      "accuracy_std_dev": 0.03358419418792134,
      "run_count": 37,
      "total_questions": 1850,
      "avg_cost_per_task_usd": 0.21483056864864872,
      "avg_duration_per_task_ms": 93416.17405405405,
      "avg_output_tokens_per_task": 7484.544324324324,
      "last_run_timestamp": "2026-10-31T13:20:20.506Z"
    },
    {
      "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": "anthropic/claude-4.5-sonnet-20250929",
      "model_slug": "anthropic/claude-sonnet-4.5",
      "model_name": "Anthropic: Claude Sonnet 4.5",
      "provider_name": "Anthropic",
      "accuracy": 0.7528572857142857,
      "accuracy_std_dev": 0.02855957373238457,
      "run_count": 14,
      "total_questions": 700,
      "avg_cost_per_task_usd": 0.3433513242857144,
      "avg_duration_per_task_ms": 226520.09285714285,
      "avg_output_tokens_per_task": 10840.655714285715,
      "last_run_timestamp": "2026-10-31T12:24:26.627Z"
    },
    {
      "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": 0.7528204615384616,
      "accuracy_std_dev": 0.022412150364050934,
      "run_count": 13,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.00771266,
      "avg_duration_per_task_ms": 231992.75692307693,
      "avg_output_tokens_per_task": 14548.033846153847,
      "last_run_timestamp": "2026-10-31T13:30:46.598Z"
    },
    {
      "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": 0.7527778333333334,
      "accuracy_std_dev": 0.030118007616798902,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.20071909666666668,
      "avg_duration_per_task_ms": 119608.17166666666,
      "avg_output_tokens_per_task": 7765.845,
      "last_run_timestamp": "2026-10-31T11:54:55.856Z"
    },
    {
      "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": 0.7527778333333334,
      "accuracy_std_dev": 0.03696434635301368,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.19816059333333333,
      "avg_duration_per_task_ms": 99704.65333333334,
      "avg_output_tokens_per_task": 7614.635,
      "last_run_timestamp": "2026-10-31T11:28:40.153Z"
    },
    {
      "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": 0.7527776666666667,
      "accuracy_std_dev": 0.016206541129707153,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.03677672666666666,
      "avg_duration_per_task_ms": 268844.9166666667,
      "avg_output_tokens_per_task": 21482.343333333334,
      "last_run_timestamp": "2026-10-31T20:23:01.980Z"
    },
    {
      "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": "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": 0.752715030075188,
      "accuracy_std_dev": 0.009940333684036977,
      "run_count": 8,
      "total_questions": 399,
      "avg_cost_per_task_usd": 0.02512772681704261,
      "avg_duration_per_task_ms": 75665.01253132832,
      "avg_output_tokens_per_task": 12251.558897243107,
      "last_run_timestamp": "2026-10-31T12:26:39.122Z"
    },
    {
      "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": 0.7526667,
      "accuracy_std_dev": 0.02374861961062159,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.007248376000000002,
      "avg_duration_per_task_ms": 194889.812,
      "avg_output_tokens_per_task": 4384.802,
      "last_run_timestamp": "2026-10-31T13:46:40.932Z"
    },
    {
      "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": 0.7525925555555556,
      "accuracy_std_dev": 0.029049942333509857,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.02899676666666667,
      "avg_duration_per_task_ms": 458008.2733333333,
      "avg_output_tokens_per_task": 11690.06,
      "last_run_timestamp": "2026-10-31T19:35:25.708Z"
    },
    {
      "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": 0.7523809047619048,
      "accuracy_std_dev": 0.017612644913472624,
      "run_count": 21,
      "total_questions": 1050,
      "avg_cost_per_task_usd": 0.36889707904761904,
      "avg_duration_per_task_ms": 141180.9876190476,
      "avg_output_tokens_per_task": 7127.063809523809,
      "last_run_timestamp": "2026-10-31T12:26:17.125Z"
    },
    {
      "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": 0.7523076923076923,
      "accuracy_std_dev": 0.0332661532107905,
      "run_count": 13,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.10102499846153845,
      "avg_duration_per_task_ms": 302464.04,
      "avg_output_tokens_per_task": 18714.755384615386,
      "last_run_timestamp": "2026-10-31T13:11:23.394Z"
    },
    {
      "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": 0.7522223333333333,
      "accuracy_std_dev": 0.021914618269294344,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.07768509666666666,
      "avg_duration_per_task_ms": 545757.0816666667,
      "avg_output_tokens_per_task": 35147.68,
      "last_run_timestamp": "2026-10-31T20:08:33.748Z"
    },
    {
      "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": 0.75222225,
      "accuracy_std_dev": 0.028393307985406832,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.5304961100000001,
      "avg_duration_per_task_ms": 187932.66333333333,
      "avg_output_tokens_per_task": 10407.706666666667,
      "last_run_timestamp": "2026-10-31T11:53:29.846Z"
    },
    {
      "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": 0.7519260539291217,
      "accuracy_std_dev": 0.019651146282719818,
      "run_count": 13,
      "total_questions": 649,
      "avg_cost_per_task_usd": 0.027055047765793528,
      "avg_duration_per_task_ms": 539313.875192604,
      "avg_output_tokens_per_task": 22031.878274268103,
      "last_run_timestamp": "2026-10-31T19:23:15.539Z"
    },
    {
      "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": 0.751852,
      "accuracy_std_dev": 0.019316204987292706,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.014403708888888889,
      "avg_duration_per_task_ms": 237123.64,
      "avg_output_tokens_per_task": 14731.191111111111,
      "last_run_timestamp": "2026-10-31T19:19:01.715Z"
    },
    {
      "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": 0.7516778724832214,
      "accuracy_std_dev": 0.030780262062995275,
      "run_count": 3,
      "total_questions": 149,
      "avg_cost_per_task_usd": 0.08526638255033557,
      "avg_duration_per_task_ms": 172059.79865771811,
      "avg_output_tokens_per_task": 8282.208053691274,
      "last_run_timestamp": "2026-10-31T11:50:00.858Z"
    },
    {
      "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": "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": 0.7514287142857142,
      "accuracy_std_dev": 0.02844421520758676,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.032982819999999996,
      "avg_duration_per_task_ms": 221535.9714285714,
      "avg_output_tokens_per_task": 35206.688571428575,
      "last_run_timestamp": "2026-10-31T11:46:18.454Z"
    },
    {
      "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": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "auto-routing",
      "accuracy": 0.7511111111111111,
      "accuracy_std_dev": 0.03744961388747538,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.04471753777777778,
      "avg_duration_per_task_ms": 182484.97555555555,
      "avg_output_tokens_per_task": 9416.271111111111,
      "last_run_timestamp": "2026-10-31T20:15:42.614Z"
    },
    {
      "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": "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": 0.7506666,
      "accuracy_std_dev": 0.03785357285435553,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.007070539999999999,
      "avg_duration_per_task_ms": 292989.264,
      "avg_output_tokens_per_task": 3954.392,
      "last_run_timestamp": "2026-10-31T13:04:37.434Z"
    },
    {
      "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": 0.7505555833333333,
      "accuracy_std_dev": 0.02754919561577778,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.01935584833333333,
      "avg_duration_per_task_ms": 217986.275,
      "avg_output_tokens_per_task": 15601.298333333334,
      "last_run_timestamp": "2026-10-31T12:32:12.444Z"
    },
    {
      "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": 0.7504761428571428,
      "accuracy_std_dev": 0.04779025223509368,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.020752462857142853,
      "avg_duration_per_task_ms": 333009.8742857143,
      "avg_output_tokens_per_task": 11171.528571428571,
      "last_run_timestamp": "2026-10-31T13:00:18.168Z"
    },
    {
      "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": 0.7503704444444445,
      "accuracy_std_dev": 0.0275968517733902,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.025161266666666664,
      "avg_duration_per_task_ms": 319969.4244444444,
      "avg_output_tokens_per_task": 10853.25111111111,
      "last_run_timestamp": "2026-10-31T19:15:14.763Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "BaseTen",
      "accuracy": 0.75,
      "accuracy_std_dev": 0.0033330000000000304,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.28014248,
      "avg_duration_per_task_ms": 223020.56,
      "avg_output_tokens_per_task": 18715.38,
      "last_run_timestamp": "2026-09-12T08:12:13.731Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.4-20260305",
      "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": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7499999999999999,
      "accuracy_std_dev": 0.03349968490896593,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.028200379999999997,
      "avg_duration_per_task_ms": 163955.21,
      "avg_output_tokens_per_task": 15323.823333333334,
      "last_run_timestamp": "2026-10-31T12:28:07.595Z"
    },
    {
      "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": 0.749524,
      "accuracy_std_dev": 0.021633867377662402,
      "run_count": 14,
      "total_questions": 700,
      "avg_cost_per_task_usd": 0.35241304,
      "avg_duration_per_task_ms": 87048.52428571429,
      "avg_output_tokens_per_task": 6993.714285714285,
      "last_run_timestamp": "2026-10-31T11:29:30.760Z"
    },
    {
      "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": 0.7495238571428572,
      "accuracy_std_dev": 0.021338220105372136,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.016605834285714285,
      "avg_duration_per_task_ms": 72972.84285714285,
      "avg_output_tokens_per_task": 4253.202857142857,
      "last_run_timestamp": "2026-10-31T12:03:51.953Z"
    },
    {
      "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": 0.7494445555555556,
      "accuracy_std_dev": 0.024107483087258824,
      "run_count": 36,
      "total_questions": 1800,
      "avg_cost_per_task_usd": 0.37011108833333334,
      "avg_duration_per_task_ms": 153729.88,
      "avg_output_tokens_per_task": 11750.439444444444,
      "last_run_timestamp": "2026-10-31T12:39:28.528Z"
    },
    {
      "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-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "DeepInfra",
      "accuracy": 0.7493333,
      "accuracy_std_dev": 0.015549210771289982,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.012228522,
      "avg_duration_per_task_ms": 315949.068,
      "avg_output_tokens_per_task": 16300.002,
      "last_run_timestamp": "2026-10-31T11:27:38.689Z"
    },
    {
      "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": 0.7492774046242774,
      "accuracy_std_dev": 0.025230016427736986,
      "run_count": 14,
      "total_questions": 692,
      "avg_cost_per_task_usd": 0.013838482658959535,
      "avg_duration_per_task_ms": 56300.32225433526,
      "avg_output_tokens_per_task": 4085.725433526012,
      "last_run_timestamp": "2026-10-31T12:14:05.375Z"
    },
    {
      "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": 0.7492306923076923,
      "accuracy_std_dev": 0.034644866837859514,
      "run_count": 13,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.031206456923076928,
      "avg_duration_per_task_ms": 254745.31692307693,
      "avg_output_tokens_per_task": 16975.855384615385,
      "last_run_timestamp": "2026-10-31T12:25:12.423Z"
    },
    {
      "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": 0.749091,
      "accuracy_std_dev": 0.0224325111713902,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.1957560618181818,
      "avg_duration_per_task_ms": 114755.80545454545,
      "avg_output_tokens_per_task": 7618.518181818182,
      "last_run_timestamp": "2026-10-31T11:08:14.469Z"
    },
    {
      "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": 0.7490679239373601,
      "accuracy_std_dev": 0.028614120182218948,
      "run_count": 9,
      "total_questions": 447,
      "avg_cost_per_task_usd": 0.09606371140939597,
      "avg_duration_per_task_ms": 393542.0715883669,
      "avg_output_tokens_per_task": 24831.977628635348,
      "last_run_timestamp": "2026-10-31T12:39:48.592Z"
    },
    {
      "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": 0.748889,
      "accuracy_std_dev": 0.015713326891527487,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.03620722,
      "avg_duration_per_task_ms": 114735.25333333333,
      "avg_output_tokens_per_task": 14087.753333333334,
      "last_run_timestamp": "2026-10-31T11:35:01.591Z"
    },
    {
      "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": 0.7488888333333333,
      "accuracy_std_dev": 0.03852200108992549,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.04173693,
      "avg_duration_per_task_ms": 185285.73166666666,
      "avg_output_tokens_per_task": 18834.241666666665,
      "last_run_timestamp": "2026-10-31T12:26:48.848Z"
    },
    {
      "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": 0.7487437738693468,
      "accuracy_std_dev": 0.05136696424259468,
      "run_count": 8,
      "total_questions": 398,
      "avg_cost_per_task_usd": 0.08314239447236181,
      "avg_duration_per_task_ms": 784494.41959799,
      "avg_output_tokens_per_task": 50526.82914572865,
      "last_run_timestamp": "2026-10-31T18:44:20.050Z"
    },
    {
      "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": 0.7485714285714286,
      "accuracy_std_dev": 0.023560522731632144,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.34892749714285715,
      "avg_duration_per_task_ms": 98456.37428571428,
      "avg_output_tokens_per_task": 7086.754285714285,
      "last_run_timestamp": "2026-10-31T11:27:20.548Z"
    },
    {
      "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": 0.7483334166666666,
      "accuracy_std_dev": 0.027772250006251255,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.02606768166666667,
      "avg_duration_per_task_ms": 149836.03333333333,
      "avg_output_tokens_per_task": 9361.645,
      "last_run_timestamp": "2026-10-31T19:14:12.566Z"
    },
    {
      "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": 0.748333375,
      "accuracy_std_dev": 0.024664456546503833,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0092696725,
      "avg_duration_per_task_ms": 91481.9675,
      "avg_output_tokens_per_task": 13410.6275,
      "last_run_timestamp": "2026-10-31T13:03:50.748Z"
    },
    {
      "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": "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": 0.7483333333333333,
      "accuracy_std_dev": 0.028037772142633272,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.50234278,
      "avg_duration_per_task_ms": 193180.515,
      "avg_output_tokens_per_task": 10059.131666666666,
      "last_run_timestamp": "2026-10-31T12:10:56.773Z"
    },
    {
      "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": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Modal",
      "accuracy": 0.748322067114094,
      "accuracy_std_dev": 0.02148614665313445,
      "run_count": 3,
      "total_questions": 149,
      "avg_cost_per_task_usd": 0.029814261744966444,
      "avg_duration_per_task_ms": 257681.3489932886,
      "avg_output_tokens_per_task": 10900.818791946309,
      "last_run_timestamp": "2026-10-31T12:26:05.030Z"
    },
    {
      "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": 0.7481482222222222,
      "accuracy_std_dev": 0.020068464436510978,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.19089935333333333,
      "avg_duration_per_task_ms": 286228.22222222225,
      "avg_output_tokens_per_task": 11572.273333333333,
      "last_run_timestamp": "2026-10-31T12:37:16.632Z"
    },
    {
      "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": "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": 0.74791675,
      "accuracy_std_dev": 0.031222055284486006,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.01487202,
      "avg_duration_per_task_ms": 51526.7675,
      "avg_output_tokens_per_task": 4120.2525,
      "last_run_timestamp": "2026-10-31T13:13:13.022Z"
    },
    {
      "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": 0.7478510515759312,
      "accuracy_std_dev": 0.031000808296901307,
      "run_count": 7,
      "total_questions": 349,
      "avg_cost_per_task_usd": 0.018540833810888252,
      "avg_duration_per_task_ms": 126193.2005730659,
      "avg_output_tokens_per_task": 4035.12893982808,
      "last_run_timestamp": "2026-10-31T11:27:58.748Z"
    },
    {
      "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": 0.7477776666666667,
      "accuracy_std_dev": 0.028393017748891882,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.020487799999999997,
      "avg_duration_per_task_ms": 112394.16,
      "avg_output_tokens_per_task": 6879.653333333334,
      "last_run_timestamp": "2026-10-31T12:42:03.915Z"
    },
    {
      "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": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "Baidu",
      "accuracy": 0.7474074444444444,
      "accuracy_std_dev": 0.021417470403134346,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.06300137777777777,
      "avg_duration_per_task_ms": 204272.4,
      "avg_output_tokens_per_task": 11398.04,
      "last_run_timestamp": "2026-10-31T19:58:43.811Z"
    },
    {
      "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": 0.7473684210526316,
      "accuracy_std_dev": 0.02465141176979054,
      "run_count": 19,
      "total_questions": 950,
      "avg_cost_per_task_usd": 0.02123823368421052,
      "avg_duration_per_task_ms": 106810.42210526316,
      "avg_output_tokens_per_task": 14307.62947368421,
      "last_run_timestamp": "2026-10-31T13:23:12.329Z"
    },
    {
      "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": "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": "qwen/qwen3.5-122b-a10b-20260224",
      "model_slug": "qwen/qwen3.5-122b-a10b",
      "model_name": "Qwen: Qwen3.5-122B-A10B",
      "provider_name": "Alibaba",
      "accuracy": 0.7467730182232346,
      "accuracy_std_dev": 0.0416315871638084,
      "run_count": 9,
      "total_questions": 439,
      "avg_cost_per_task_usd": 0.11620976082004557,
      "avg_duration_per_task_ms": 452232.8268792711,
      "avg_output_tokens_per_task": 21963.851936218678,
      "last_run_timestamp": "2026-10-31T19:29:13.230Z"
    },
    {
      "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": "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": "InferenceNet",
      "accuracy": 0.746667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.07074242,
      "avg_duration_per_task_ms": 800280.44,
      "avg_output_tokens_per_task": 15183.42,
      "last_run_timestamp": "2026-10-02T06:00:07.943Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "CoreWeave",
      "accuracy": 0.7466668888888889,
      "accuracy_std_dev": 0.0339933762614766,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.08365215111111111,
      "avg_duration_per_task_ms": 72579.65555555555,
      "avg_output_tokens_per_task": 11821.215555555556,
      "last_run_timestamp": "2026-10-31T19:56:16.658Z"
    },
    {
      "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": 0.7466666666666667,
      "accuracy_std_dev": 0.023726903000602494,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.007031926666666666,
      "avg_duration_per_task_ms": 152232.88666666666,
      "avg_output_tokens_per_task": 12687.024444444445,
      "last_run_timestamp": "2026-10-31T19:25:36.518Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.6-opus-20260205",
      "model_slug": "anthropic/claude-opus-4.6",
      "model_name": "Anthropic: Claude Opus 4.6",
      "provider_name": "Azure",
      "accuracy": 0.7466665833333334,
      "accuracy_std_dev": 0.0259628652195732,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.5233425716666666,
      "avg_duration_per_task_ms": 222591.87333333332,
      "avg_output_tokens_per_task": 9975.42,
      "last_run_timestamp": "2026-10-31T13:21:45.027Z"
    },
    {
      "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": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Reka",
      "accuracy": 0.7466665625000001,
      "accuracy_std_dev": 0.022360772946078904,
      "run_count": 16,
      "total_questions": 800,
      "avg_cost_per_task_usd": 0.013514558749999999,
      "avg_duration_per_task_ms": 139486.7775,
      "avg_output_tokens_per_task": 15486.175,
      "last_run_timestamp": "2026-10-31T13:03:01.037Z"
    },
    {
      "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": 0.7466665,
      "accuracy_std_dev": 0.026666500000000037,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.00520861,
      "avg_duration_per_task_ms": 148091.91,
      "avg_output_tokens_per_task": 12193.49,
      "last_run_timestamp": "2026-10-31T11:47:33.345Z"
    },
    {
      "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": "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": 0.7466663333333334,
      "accuracy_std_dev": 0.033993463423951875,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.01425808,
      "avg_duration_per_task_ms": 139346.08,
      "avg_output_tokens_per_task": 12850.746666666666,
      "last_run_timestamp": "2026-10-31T12:18:46.817Z"
    },
    {
      "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": 0.7463888333333334,
      "accuracy_std_dev": 0.027854064690912814,
      "run_count": 24,
      "total_questions": 1200,
      "avg_cost_per_task_usd": 0.5516252541666667,
      "avg_duration_per_task_ms": 200356.9575,
      "avg_output_tokens_per_task": 9752.154166666667,
      "last_run_timestamp": "2026-10-31T13:01:06.466Z"
    },
    {
      "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-sonnet-20260217",
      "model_slug": "anthropic/claude-sonnet-4.6",
      "model_name": "Anthropic: Claude Sonnet 4.6",
      "provider_name": "Azure",
      "accuracy": 0.7461111666666667,
      "accuracy_std_dev": 0.018946052393719277,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.3487555299999999,
      "avg_duration_per_task_ms": 215726.44333333333,
      "avg_output_tokens_per_task": 11930.17,
      "last_run_timestamp": "2026-10-31T13:31:12.841Z"
    },
    {
      "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": 0.7460141846496107,
      "accuracy_std_dev": 0.015367727629578602,
      "run_count": 18,
      "total_questions": 899,
      "avg_cost_per_task_usd": 0.04452976084538376,
      "avg_duration_per_task_ms": 330627.27141268074,
      "avg_output_tokens_per_task": 25778.021134593993,
      "last_run_timestamp": "2026-10-31T19:20:57.312Z"
    },
    {
      "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": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "AtlasCloud",
      "accuracy": 0.745926,
      "accuracy_std_dev": 0.0403005321856231,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.11525609333333334,
      "avg_duration_per_task_ms": 227448.44222222222,
      "avg_output_tokens_per_task": 9598.077777777778,
      "last_run_timestamp": "2026-10-31T20:27:27.677Z"
    },
    {
      "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": 0.74583325,
      "accuracy_std_dev": 0.025913946803555434,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0561430525,
      "avg_duration_per_task_ms": 815372.045,
      "avg_output_tokens_per_task": 15741.0025,
      "last_run_timestamp": "2026-10-31T20:02:53.411Z"
    },
    {
      "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": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Morph",
      "accuracy": 0.7454545909090909,
      "accuracy_std_dev": 0.02161524816516654,
      "run_count": 22,
      "total_questions": 1100,
      "avg_cost_per_task_usd": 0.4409956472727272,
      "avg_duration_per_task_ms": 587766.49,
      "avg_output_tokens_per_task": 10456.687272727273,
      "last_run_timestamp": "2026-10-31T11:24:57.351Z"
    },
    {
      "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": 0.7453333,
      "accuracy_std_dev": 0.02454934079379731,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.17119798400000003,
      "avg_duration_per_task_ms": 320387.726,
      "avg_output_tokens_per_task": 11023.856,
      "last_run_timestamp": "2026-10-31T12:25:36.323Z"
    },
    {
      "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": "anthropic/claude-4.6-sonnet-20260217",
      "model_slug": "anthropic/claude-sonnet-4.6",
      "model_name": "Anthropic: Claude Sonnet 4.6",
      "provider_name": "Anthropic",
      "accuracy": 0.7452380714285715,
      "accuracy_std_dev": 0.025093932830934848,
      "run_count": 14,
      "total_questions": 700,
      "avg_cost_per_task_usd": 0.3444721642857143,
      "avg_duration_per_task_ms": 207641.8,
      "avg_output_tokens_per_task": 11831.092857142858,
      "last_run_timestamp": "2026-10-31T13:36:32.496Z"
    },
    {
      "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": 0.7451853333333333,
      "accuracy_std_dev": 0.031234315132203188,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.01557802666666667,
      "avg_duration_per_task_ms": 147555.71333333335,
      "avg_output_tokens_per_task": 15042.41111111111,
      "last_run_timestamp": "2026-10-31T12:25:02.563Z"
    },
    {
      "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": 0.7451852222222222,
      "accuracy_std_dev": 0.017716056438394953,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.21677278888888887,
      "avg_duration_per_task_ms": 284791.18666666665,
      "avg_output_tokens_per_task": 11557.691111111111,
      "last_run_timestamp": "2026-10-31T11:21:59.486Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "SambaNova",
      "accuracy": 0.745,
      "accuracy_std_dev": 0.026822730826670128,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.15284815,
      "avg_duration_per_task_ms": 494527.49,
      "avg_output_tokens_per_task": 28246.935,
      "last_run_timestamp": "2026-10-17T19:47:21.934Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Phala",
      "accuracy": 0.744999875,
      "accuracy_std_dev": 0.030867001888900313,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0114373625,
      "avg_duration_per_task_ms": 404697.66,
      "avg_output_tokens_per_task": 5754.185,
      "last_run_timestamp": "2026-10-31T12:24:48.023Z"
    },
    {
      "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": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Wafer",
      "accuracy": 0.7446667,
      "accuracy_std_dev": 0.01956180291818729,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.055664519999999995,
      "avg_duration_per_task_ms": 332682.316,
      "avg_output_tokens_per_task": 14458.648,
      "last_run_timestamp": "2026-10-31T11:42:33.248Z"
    },
    {
      "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": 0.7444445833333333,
      "accuracy_std_dev": 0.031661685769128825,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.5453893666666667,
      "avg_duration_per_task_ms": 180530.81666666668,
      "avg_output_tokens_per_task": 10281.893333333333,
      "last_run_timestamp": "2026-10-31T12:28:38.525Z"
    },
    {
      "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": 0.7444444444444445,
      "accuracy_std_dev": 0.031426873772868374,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.09879648,
      "avg_duration_per_task_ms": 287953.65555555554,
      "avg_output_tokens_per_task": 7148.386666666666,
      "last_run_timestamp": "2026-10-31T19:33:49.408Z"
    },
    {
      "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": 0.7444443333333334,
      "accuracy_std_dev": 0.02199873172606905,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.02140772,
      "avg_duration_per_task_ms": 53184.22,
      "avg_output_tokens_per_task": 14364.18,
      "last_run_timestamp": "2026-10-31T11:41:11.192Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "Reka",
      "accuracy": 0.7444443333333334,
      "accuracy_std_dev": 0.006285236475706844,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.04041968666666667,
      "avg_duration_per_task_ms": 291588.68666666665,
      "avg_output_tokens_per_task": 24330.48,
      "last_run_timestamp": "2026-09-30T11:25:05.946Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "DigitalOcean",
      "accuracy": 0.74416675,
      "accuracy_std_dev": 0.03012731088543251,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.064155875,
      "avg_duration_per_task_ms": 1722921.32,
      "avg_output_tokens_per_task": 8915.945,
      "last_run_timestamp": "2026-08-29T16:20:58.762Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Phala",
      "accuracy": 0.744,
      "accuracy_std_dev": 0.04013309523821955,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.12688716400000002,
      "avg_duration_per_task_ms": 700323.218,
      "avg_output_tokens_per_task": 36239.088,
      "last_run_timestamp": "2026-10-31T12:47:12.433Z"
    },
    {
      "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": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Parasail",
      "accuracy": 0.7438325352839931,
      "accuracy_std_dev": 0.030628998095826893,
      "run_count": 12,
      "total_questions": 581,
      "avg_cost_per_task_usd": 0.023620459552495695,
      "avg_duration_per_task_ms": 98292.74010327022,
      "avg_output_tokens_per_task": 8358.294320137693,
      "last_run_timestamp": "2026-10-31T19:46:13.143Z"
    },
    {
      "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-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "auto-routing",
      "accuracy": 0.7437037777777777,
      "accuracy_std_dev": 0.025936447601258752,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.03902438222222222,
      "avg_duration_per_task_ms": 589876.5155555556,
      "avg_output_tokens_per_task": 15857.186666666666,
      "last_run_timestamp": "2026-10-31T19:47:52.982Z"
    },
    {
      "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": 0.7436363636363637,
      "accuracy_std_dev": 0.0487718196806176,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.0472110309090909,
      "avg_duration_per_task_ms": 288762.1018181818,
      "avg_output_tokens_per_task": 15032.698181818181,
      "last_run_timestamp": "2026-10-31T12:45:59.472Z"
    },
    {
      "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": "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": 0.7433333333333333,
      "accuracy_std_dev": 0.026597020163085114,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.32731308,
      "avg_duration_per_task_ms": 196690.64,
      "avg_output_tokens_per_task": 11548.81,
      "last_run_timestamp": "2026-10-31T12:07:21.658Z"
    },
    {
      "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": "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": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "DeepInfra",
      "accuracy": 0.7433034866071428,
      "accuracy_std_dev": 0.01807640693429757,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.047182370535714294,
      "avg_duration_per_task_ms": 707194.1785714285,
      "avg_output_tokens_per_task": 14275.674107142857,
      "last_run_timestamp": "2026-10-31T19:48:13.441Z"
    },
    {
      "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": 0.7431746190476191,
      "accuracy_std_dev": 0.018008818983714862,
      "run_count": 21,
      "total_questions": 1050,
      "avg_cost_per_task_usd": 0.36043625523809525,
      "avg_duration_per_task_ms": 86437.01714285715,
      "avg_output_tokens_per_task": 7106.278095238095,
      "last_run_timestamp": "2026-10-31T12:25:51.224Z"
    },
    {
      "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": 0.7430302727272727,
      "accuracy_std_dev": 0.027760033635986,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.5088923236363636,
      "avg_duration_per_task_ms": 192892.78363636363,
      "avg_output_tokens_per_task": 9984.365454545454,
      "last_run_timestamp": "2026-10-31T12:42:53.834Z"
    },
    {
      "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": "StreamLake",
      "accuracy": 0.7429720200803213,
      "accuracy_std_dev": 0.03338175982419142,
      "run_count": 10,
      "total_questions": 498,
      "avg_cost_per_task_usd": 0.004595510040160643,
      "avg_duration_per_task_ms": 259445.41365461846,
      "avg_output_tokens_per_task": 12133.893574297188,
      "last_run_timestamp": "2026-10-31T19:49:31.140Z"
    },
    {
      "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": 0.742963,
      "accuracy_std_dev": 0.03221999055796815,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.13625242222222222,
      "avg_duration_per_task_ms": 360519.17555555556,
      "avg_output_tokens_per_task": 11834.471111111112,
      "last_run_timestamp": "2026-10-31T21:07:15.251Z"
    },
    {
      "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": 0.7429168125,
      "accuracy_std_dev": 0.02570869361037903,
      "run_count": 16,
      "total_questions": 800,
      "avg_cost_per_task_usd": 0.05779907625,
      "avg_duration_per_task_ms": 136331.88625,
      "avg_output_tokens_per_task": 9998.7625,
      "last_run_timestamp": "2026-10-31T19:13:33.926Z"
    },
    {
      "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": 0.74277775,
      "accuracy_std_dev": 0.0244128236149399,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.14337107166666668,
      "avg_duration_per_task_ms": 621697.9133333333,
      "avg_output_tokens_per_task": 26711.696666666667,
      "last_run_timestamp": "2026-10-31T19:26:38.406Z"
    },
    {
      "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": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Cloudflare",
      "accuracy": 0.7426668,
      "accuracy_std_dev": 0.0246215659770048,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.079647184,
      "avg_duration_per_task_ms": 303286.308,
      "avg_output_tokens_per_task": 11264.834,
      "last_run_timestamp": "2026-10-31T21:26:40.342Z"
    },
    {
      "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": 0.7426667,
      "accuracy_std_dev": 0.028472143112347543,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.020416034,
      "avg_duration_per_task_ms": 153083.812,
      "avg_output_tokens_per_task": 16185.95,
      "last_run_timestamp": "2026-10-31T12:48:17.235Z"
    },
    {
      "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": 0.7426666000000001,
      "accuracy_std_dev": 0.04434197309367279,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.16675142799999998,
      "avg_duration_per_task_ms": 302047.356,
      "avg_output_tokens_per_task": 9682.606,
      "last_run_timestamp": "2026-10-31T11:21:43.861Z"
    },
    {
      "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": "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": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Mara",
      "accuracy": 0.7425405959079282,
      "accuracy_std_dev": 0.026283120985422872,
      "run_count": 8,
      "total_questions": 391,
      "avg_cost_per_task_usd": 0.12785004603580563,
      "avg_duration_per_task_ms": 203562.63938618926,
      "avg_output_tokens_per_task": 10880.268542199488,
      "last_run_timestamp": "2026-10-31T19:27:07.786Z"
    },
    {
      "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": 0.7424241818181818,
      "accuracy_std_dev": 0.03530779679235989,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.36784247818181814,
      "avg_duration_per_task_ms": 381853.54363636364,
      "avg_output_tokens_per_task": 9773.532727272728,
      "last_run_timestamp": "2026-10-31T12:43:33.294Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7422838796296296,
      "accuracy_std_dev": 0.02605883730643627,
      "run_count": 13,
      "total_questions": 648,
      "avg_cost_per_task_usd": 0.04679922222222222,
      "avg_duration_per_task_ms": 176761.8811728395,
      "avg_output_tokens_per_task": 8966.202160493827,
      "last_run_timestamp": "2026-10-31T19:46:31.278Z"
    },
    {
      "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": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Cohere",
      "accuracy": 0.7422223333333333,
      "accuracy_std_dev": 0.011330936570096763,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.01793052666666667,
      "avg_duration_per_task_ms": 86484.18,
      "avg_output_tokens_per_task": 20778.286666666667,
      "last_run_timestamp": "2026-10-31T11:37:00.194Z"
    },
    {
      "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": 0.7422221111111111,
      "accuracy_std_dev": 0.03442657601151391,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.07352381777777778,
      "avg_duration_per_task_ms": 217584.44444444444,
      "avg_output_tokens_per_task": 9519.468888888889,
      "last_run_timestamp": "2026-10-31T19:53:34.188Z"
    },
    {
      "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": 0.742222,
      "accuracy_std_dev": 0.03861798340151903,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.010298153333333334,
      "avg_duration_per_task_ms": 95025.84666666666,
      "avg_output_tokens_per_task": 13300.326666666666,
      "last_run_timestamp": "2026-10-31T12:36:20.326Z"
    },
    {
      "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": 0.7420636446886446,
      "accuracy_std_dev": 0.016469967104193704,
      "run_count": 11,
      "total_questions": 546,
      "avg_cost_per_task_usd": 0.01720449084249084,
      "avg_duration_per_task_ms": 82040.02197802198,
      "avg_output_tokens_per_task": 4114.483516483517,
      "last_run_timestamp": "2026-10-31T12:00:42.348Z"
    },
    {
      "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": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Phala",
      "accuracy": 0.741666625,
      "accuracy_std_dev": 0.025331164930464133,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.021842495,
      "avg_duration_per_task_ms": 141327.99,
      "avg_output_tokens_per_task": 16250.775,
      "last_run_timestamp": "2026-10-31T12:24:54.205Z"
    },
    {
      "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": 0.741666625,
      "accuracy_std_dev": 0.030866893899198455,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.021115824999999998,
      "avg_duration_per_task_ms": 150544.0025,
      "avg_output_tokens_per_task": 8219.325,
      "last_run_timestamp": "2026-10-31T18:57:59.813Z"
    },
    {
      "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": 0.7416481202672606,
      "accuracy_std_dev": 0.022563192218990436,
      "run_count": 9,
      "total_questions": 449,
      "avg_cost_per_task_usd": 0.21138012694877506,
      "avg_duration_per_task_ms": 402109.59020044544,
      "avg_output_tokens_per_task": 9801.538975501113,
      "last_run_timestamp": "2026-10-31T11:26:03.627Z"
    },
    {
      "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": 0.7414965918367348,
      "accuracy_std_dev": 0.01208300000000001,
      "run_count": 2,
      "total_questions": 98,
      "avg_cost_per_task_usd": 0.014484438775510205,
      "avg_duration_per_task_ms": 90891.24489795919,
      "avg_output_tokens_per_task": 9684.755102040815,
      "last_run_timestamp": "2026-10-31T13:18:30.658Z"
    },
    {
      "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": 0.74136442746114,
      "accuracy_std_dev": 0.023598813561278888,
      "run_count": 8,
      "total_questions": 386,
      "avg_cost_per_task_usd": 0.1499042176165803,
      "avg_duration_per_task_ms": 675715.4041450777,
      "avg_output_tokens_per_task": 29694.012953367877,
      "last_run_timestamp": "2026-10-31T19:55:58.516Z"
    },
    {
      "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": 0.7413334,
      "accuracy_std_dev": 0.02436751571744648,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.097602524,
      "avg_duration_per_task_ms": 217804.16,
      "avg_output_tokens_per_task": 9920.02,
      "last_run_timestamp": "2026-10-31T14:40:25.043Z"
    },
    {
      "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": 0.7413333,
      "accuracy_std_dev": 0.03095523075685272,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.40647999399999996,
      "avg_duration_per_task_ms": 588538.78,
      "avg_output_tokens_per_task": 9896.416,
      "last_run_timestamp": "2026-10-31T12:27:47.241Z"
    },
    {
      "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": 0.7412121818181818,
      "accuracy_std_dev": 0.0275340250329006,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.3449351727272727,
      "avg_duration_per_task_ms": 200135.41454545455,
      "avg_output_tokens_per_task": 11800.205454545454,
      "last_run_timestamp": "2026-10-31T13:12:27.229Z"
    },
    {
      "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": 0.7410254102564102,
      "accuracy_std_dev": 0.015687179899124656,
      "run_count": 4,
      "total_questions": 195,
      "avg_cost_per_task_usd": 0.010222671794871796,
      "avg_duration_per_task_ms": 114819.93846153846,
      "avg_output_tokens_per_task": 3734.005128205128,
      "last_run_timestamp": "2026-10-31T12:06:01.335Z"
    },
    {
      "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": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "StreamLake",
      "accuracy": 0.7406667,
      "accuracy_std_dev": 0.03147484286553311,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.067724174,
      "avg_duration_per_task_ms": 257262.924,
      "avg_output_tokens_per_task": 9632.852,
      "last_run_timestamp": "2026-10-31T20:31:43.469Z"
    },
    {
      "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": 0.7406666000000001,
      "accuracy_std_dev": 0.018961987708043703,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.014243233999999999,
      "avg_duration_per_task_ms": 72756.944,
      "avg_output_tokens_per_task": 4429.298,
      "last_run_timestamp": "2026-10-31T12:42:18.186Z"
    },
    {
      "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": 0.7403698151001541,
      "accuracy_std_dev": 0.028547774310492945,
      "run_count": 13,
      "total_questions": 649,
      "avg_cost_per_task_usd": 0.012115494607087829,
      "avg_duration_per_task_ms": 425113.8289676425,
      "avg_output_tokens_per_task": 14181.97688751926,
      "last_run_timestamp": "2026-10-31T11:58:45.245Z"
    },
    {
      "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": 0.7403334,
      "accuracy_std_dev": 0.020245596504919312,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.379164166,
      "avg_duration_per_task_ms": 189558.308,
      "avg_output_tokens_per_task": 11056.844,
      "last_run_timestamp": "2026-10-31T11:51:20.518Z"
    },
    {
      "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": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "AkashML",
      "accuracy": 0.74014702004008,
      "accuracy_std_dev": 0.022155950173486136,
      "run_count": 10,
      "total_questions": 499,
      "avg_cost_per_task_usd": 0.06005350100200401,
      "avg_duration_per_task_ms": 140448.37274549098,
      "avg_output_tokens_per_task": 4119.851703406814,
      "last_run_timestamp": "2026-10-31T12:27:45.919Z"
    },
    {
      "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": "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-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "auto-routing",
      "accuracy": 0.74,
      "accuracy_std_dev": 0.04773656244431517,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.00978464181818182,
      "avg_duration_per_task_ms": 283078.21636363637,
      "avg_output_tokens_per_task": 15624.66909090909,
      "last_run_timestamp": "2026-10-31T11:40:30.784Z"
    },
    {
      "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": 0.74,
      "accuracy_std_dev": 0.026666999999999996,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.14729807,
      "avg_duration_per_task_ms": 386775.27,
      "avg_output_tokens_per_task": 24280.02,
      "last_run_timestamp": "2026-10-31T19:58:44.255Z"
    },
    {
      "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": 0.74,
      "accuracy_std_dev": 0.03511903567511501,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.023083885,
      "avg_duration_per_task_ms": 150345.7,
      "avg_output_tokens_per_task": 16221.6925,
      "last_run_timestamp": "2026-10-31T11:30:43.640Z"
    },
    {
      "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": "Near AI",
      "accuracy": 0.74,
      "accuracy_std_dev": 0.0066669999999999785,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01862194,
      "avg_duration_per_task_ms": 339225.77,
      "avg_output_tokens_per_task": 12354.38,
      "last_run_timestamp": "2026-10-31T13:22:50.108Z"
    },
    {
      "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": 1.1102230246251565e-16,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.03857922666666667,
      "avg_duration_per_task_ms": 77482.69333333333,
      "avg_output_tokens_per_task": 13512.933333333332,
      "last_run_timestamp": "2026-10-31T11:52:45.166Z"
    },
    {
      "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": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Novita",
      "accuracy": 0.74,
      "accuracy_std_dev": 0.03370161735650745,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.0978457,
      "avg_duration_per_task_ms": 271590.2111111111,
      "avg_output_tokens_per_task": 12524.18888888889,
      "last_run_timestamp": "2026-10-31T20:07:04.555Z"
    },
    {
      "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": 0.74,
      "accuracy_std_dev": 0.026666999999999996,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.06914866,
      "avg_duration_per_task_ms": 152605.49,
      "avg_output_tokens_per_task": 14774.68,
      "last_run_timestamp": "2026-10-31T12:25:52.830Z"
    },
    {
      "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": 0.74,
      "accuracy_std_dev": 0.019860653454128848,
      "run_count": 16,
      "total_questions": 800,
      "avg_cost_per_task_usd": 0.01624762875,
      "avg_duration_per_task_ms": 161764.35125,
      "avg_output_tokens_per_task": 4237.58,
      "last_run_timestamp": "2026-10-31T11:33:42.887Z"
    },
    {
      "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": 0.7399329261744967,
      "accuracy_std_dev": 0.017150616176167584,
      "run_count": 6,
      "total_questions": 298,
      "avg_cost_per_task_usd": 0.24555212751677855,
      "avg_duration_per_task_ms": 357614.23825503356,
      "avg_output_tokens_per_task": 12408.355704697986,
      "last_run_timestamp": "2026-10-31T12:24:39.056Z"
    },
    {
      "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": 0.7395555333333333,
      "accuracy_std_dev": 0.016947170784004716,
      "run_count": 15,
      "total_questions": 750,
      "avg_cost_per_task_usd": 0.013297554666666666,
      "avg_duration_per_task_ms": 100947.12933333333,
      "avg_output_tokens_per_task": 10746.964,
      "last_run_timestamp": "2026-10-31T12:45:41.031Z"
    },
    {
      "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": 0.7394444166666667,
      "accuracy_std_dev": 0.04237910291928153,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.01083521,
      "avg_duration_per_task_ms": 286660.46166666667,
      "avg_output_tokens_per_task": 15769.041666666666,
      "last_run_timestamp": "2026-10-31T11:54:55.854Z"
    },
    {
      "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": 0.7393938181818183,
      "accuracy_std_dev": 0.027627475312607935,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.34960052363636374,
      "avg_duration_per_task_ms": 219975.07454545455,
      "avg_output_tokens_per_task": 10861.59090909091,
      "last_run_timestamp": "2026-10-31T12:46:18.490Z"
    },
    {
      "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": 0.7392380571428574,
      "accuracy_std_dev": 0.02911476386493665,
      "run_count": 35,
      "total_questions": 1750,
      "avg_cost_per_task_usd": 0.369135528,
      "avg_duration_per_task_ms": 199052.34971428572,
      "avg_output_tokens_per_task": 11910.057714285715,
      "last_run_timestamp": "2026-10-31T13:17:20.726Z"
    },
    {
      "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": 0.739166625,
      "accuracy_std_dev": 0.02933284919223455,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0108853675,
      "avg_duration_per_task_ms": 219104.3875,
      "avg_output_tokens_per_task": 14027.29,
      "last_run_timestamp": "2026-10-31T19:46:05.353Z"
    },
    {
      "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": 0.738889,
      "accuracy_std_dev": 0.024695368688615833,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.01095846,
      "avg_duration_per_task_ms": 321436.35333333333,
      "avg_output_tokens_per_task": 22570.04,
      "last_run_timestamp": "2026-10-31T12:52:06.048Z"
    },
    {
      "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": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "InferenceNet",
      "accuracy": 0.738484909090909,
      "accuracy_std_dev": 0.024261459951778173,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.017931623636363636,
      "avg_duration_per_task_ms": 332207.99818181817,
      "avg_output_tokens_per_task": 23710.247272727273,
      "last_run_timestamp": "2026-10-31T13:20:49.854Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "DeepInfra",
      "accuracy": 0.7383335,
      "accuracy_std_dev": 0.0016665000000000152,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.08170183999999998,
      "avg_duration_per_task_ms": 327224.56,
      "avg_output_tokens_per_task": 8775.96,
      "last_run_timestamp": "2026-08-22T16:12:20.461Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "SiliconFlow",
      "accuracy": 0.738333375,
      "accuracy_std_dev": 0.03244649655948039,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.049700385,
      "avg_duration_per_task_ms": 377861.96,
      "avg_output_tokens_per_task": 11236.51,
      "last_run_timestamp": "2026-10-31T21:59:10.427Z"
    },
    {
      "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": "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": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Morph",
      "accuracy": 0.7380000999999999,
      "accuracy_std_dev": 0.02844855708274146,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.016056571999999998,
      "avg_duration_per_task_ms": 534586.188,
      "avg_output_tokens_per_task": 17634.696,
      "last_run_timestamp": "2026-10-31T12:03:50.388Z"
    },
    {
      "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": 0.738,
      "accuracy_std_dev": 0.030954985239860806,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.019570816,
      "avg_duration_per_task_ms": 88243.608,
      "avg_output_tokens_per_task": 5445.588,
      "last_run_timestamp": "2026-10-31T11:13:51.572Z"
    },
    {
      "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": 0.7378222664756448,
      "accuracy_std_dev": 0.026214713234349613,
      "run_count": 7,
      "total_questions": 349,
      "avg_cost_per_task_usd": 0.13115203724928365,
      "avg_duration_per_task_ms": 65255.4899713467,
      "avg_output_tokens_per_task": 9539.232091690545,
      "last_run_timestamp": "2026-10-31T19:12:18.660Z"
    },
    {
      "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": 0.7377779999999999,
      "accuracy_std_dev": 0.027933050471917086,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.02205162,
      "avg_duration_per_task_ms": 230910.26,
      "avg_output_tokens_per_task": 11297.92,
      "last_run_timestamp": "2026-10-31T12:27:40.071Z"
    },
    {
      "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": 0.7377779999999999,
      "accuracy_std_dev": 0.020608136952831695,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.009694766666666667,
      "avg_duration_per_task_ms": 179217.14666666667,
      "avg_output_tokens_per_task": 12873.993333333334,
      "last_run_timestamp": "2026-10-31T11:06:39.132Z"
    },
    {
      "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": 0.7377776666666667,
      "accuracy_std_dev": 0.020366893981273733,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.3511924911111111,
      "avg_duration_per_task_ms": 195777.59333333332,
      "avg_output_tokens_per_task": 9122.282222222222,
      "last_run_timestamp": "2026-10-31T11:28:52.482Z"
    },
    {
      "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": 0.7377776666666667,
      "accuracy_std_dev": 0.030952013836833894,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.18000499999999997,
      "avg_duration_per_task_ms": 513788.9733333333,
      "avg_output_tokens_per_task": 9651.956666666667,
      "last_run_timestamp": "2026-10-31T22:21:14.715Z"
    },
    {
      "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": 0.7376126486486487,
      "accuracy_std_dev": 0.027927098584230564,
      "run_count": 6,
      "total_questions": 296,
      "avg_cost_per_task_usd": 0.011551577702702704,
      "avg_duration_per_task_ms": 61368.75337837838,
      "avg_output_tokens_per_task": 4262.564189189189,
      "last_run_timestamp": "2026-10-31T12:25:51.222Z"
    },
    {
      "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": 0.737499875,
      "accuracy_std_dev": 0.022592425906913488,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.01732587,
      "avg_duration_per_task_ms": 193821.065,
      "avg_output_tokens_per_task": 4493.1475,
      "last_run_timestamp": "2026-10-31T12:42:00.537Z"
    },
    {
      "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": "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": 0.7372221666666667,
      "accuracy_std_dev": 0.025452663641988886,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.07899650500000001,
      "avg_duration_per_task_ms": 156175.89,
      "avg_output_tokens_per_task": 8784.876666666667,
      "last_run_timestamp": "2026-10-31T20:18:12.373Z"
    },
    {
      "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": "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": "auto-routing",
      "accuracy": 0.7363635454545455,
      "accuracy_std_dev": 0.027022348334677756,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.2857510236363636,
      "avg_duration_per_task_ms": 338121.0927272727,
      "avg_output_tokens_per_task": 10844.636363636364,
      "last_run_timestamp": "2026-10-31T11:38:53.393Z"
    },
    {
      "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": 0.7362961111111112,
      "accuracy_std_dev": 0.02157064127952337,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.06493803333333333,
      "avg_duration_per_task_ms": 247546.98,
      "avg_output_tokens_per_task": 12690.464444444444,
      "last_run_timestamp": "2026-10-31T20:18:31.003Z"
    },
    {
      "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": "PrimeIntellect",
      "accuracy": 0.736,
      "accuracy_std_dev": 0.011623577573191504,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.02217724,
      "avg_duration_per_task_ms": 74896.184,
      "avg_output_tokens_per_task": 4355.024,
      "last_run_timestamp": "2026-10-31T12:05:37.260Z"
    },
    {
      "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": 0.7359999333333334,
      "accuracy_std_dev": 0.0231964569463145,
      "run_count": 15,
      "total_questions": 750,
      "avg_cost_per_task_usd": 0.013030309333333335,
      "avg_duration_per_task_ms": 73246.75466666666,
      "avg_output_tokens_per_task": 10779.536,
      "last_run_timestamp": "2026-10-31T11:42:11.484Z"
    },
    {
      "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": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Makora",
      "accuracy": 0.7359261111111111,
      "accuracy_std_dev": 0.02318882707227026,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.04132428666666666,
      "avg_duration_per_task_ms": 135748.28,
      "avg_output_tokens_per_task": 25779.64,
      "last_run_timestamp": "2026-10-31T12:16:10.183Z"
    },
    {
      "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": "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": 0.7355555555555555,
      "accuracy_std_dev": 0.016629543844289166,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.11909873777777778,
      "avg_duration_per_task_ms": 359004.24666666664,
      "avg_output_tokens_per_task": 9944.664444444445,
      "last_run_timestamp": "2026-10-31T20:34:53.249Z"
    },
    {
      "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": "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": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "GMICloud",
      "accuracy": 0.735000125,
      "accuracy_std_dev": 0.04186617734650937,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0273002175,
      "avg_duration_per_task_ms": 144987.4475,
      "avg_output_tokens_per_task": 12792.885,
      "last_run_timestamp": "2026-10-31T18:53:34.804Z"
    },
    {
      "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": 0.7349398099062917,
      "accuracy_std_dev": 0.02455646627618342,
      "run_count": 15,
      "total_questions": 747,
      "avg_cost_per_task_usd": 0.05549164792503348,
      "avg_duration_per_task_ms": 157974.1485943775,
      "avg_output_tokens_per_task": 30263.555555555555,
      "last_run_timestamp": "2026-10-31T12:41:59.422Z"
    },
    {
      "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": 0.7349397911646587,
      "accuracy_std_dev": 0.047428285455095254,
      "run_count": 10,
      "total_questions": 498,
      "avg_cost_per_task_usd": 0.2969604819277108,
      "avg_duration_per_task_ms": 386565.9397590361,
      "avg_output_tokens_per_task": 11351.530120481928,
      "last_run_timestamp": "2026-10-31T12:42:56.001Z"
    },
    {
      "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": 0.7348147777777778,
      "accuracy_std_dev": 0.03043361346631854,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.09666358000000001,
      "avg_duration_per_task_ms": 320248.93111111113,
      "avg_output_tokens_per_task": 11007.382222222222,
      "last_run_timestamp": "2026-10-31T20:47:51.733Z"
    },
    {
      "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": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Minimax",
      "accuracy": 0.7344445,
      "accuracy_std_dev": 0.029165456061066488,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.028961041666666656,
      "avg_duration_per_task_ms": 102116.81166666666,
      "avg_output_tokens_per_task": 10295.041666666666,
      "last_run_timestamp": "2026-10-31T19:09:49.337Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Together",
      "accuracy": 0.7344443333333334,
      "accuracy_std_dev": 0.03241329935981356,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.30671481166666664,
      "avg_duration_per_task_ms": 338981.285,
      "avg_output_tokens_per_task": 9637.978333333333,
      "last_run_timestamp": "2026-10-31T13:04:39.070Z"
    },
    {
      "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": 0.734,
      "accuracy_std_dev": 0.030905008584370283,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.049636292000000005,
      "avg_duration_per_task_ms": 244020.536,
      "avg_output_tokens_per_task": 10552.6,
      "last_run_timestamp": "2026-10-31T19:47:34.571Z"
    },
    {
      "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": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7337622316176471,
      "accuracy_std_dev": 0.042021571236176135,
      "run_count": 11,
      "total_questions": 544,
      "avg_cost_per_task_usd": 0.02766205882352941,
      "avg_duration_per_task_ms": 113870.94669117648,
      "avg_output_tokens_per_task": 10219.617647058823,
      "last_run_timestamp": "2026-10-31T19:48:53.852Z"
    },
    {
      "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": "meta/muse-spark-open-20261015",
      "model_slug": "meta/muse-spark-open",
      "model_name": "Meta: Muse Spark Open",
      "provider_name": "Together",
      "accuracy": 0.7333335,
      "accuracy_std_dev": 0.00666650000000002,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.08901925000000001,
      "avg_duration_per_task_ms": 93612.32,
      "avg_output_tokens_per_task": 8634.62,
      "last_run_timestamp": "2026-10-31T11:29:29.503Z"
    },
    {
      "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": "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": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7333334,
      "accuracy_std_dev": 0.03451234883690184,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.30572972400000004,
      "avg_duration_per_task_ms": 282521.796,
      "avg_output_tokens_per_task": 10813.856,
      "last_run_timestamp": "2026-10-31T12:51:11.835Z"
    },
    {
      "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": "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": 0.7333333333333333,
      "accuracy_std_dev": 0.014401903168987377,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.06451945333333334,
      "avg_duration_per_task_ms": 876085.0533333333,
      "avg_output_tokens_per_task": 15144.64,
      "last_run_timestamp": "2026-10-31T12:31:12.958Z"
    },
    {
      "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": 0.7333333333333333,
      "accuracy_std_dev": 0.03569390445004426,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.021479246666666667,
      "avg_duration_per_task_ms": 120781.28,
      "avg_output_tokens_per_task": 11581.666666666666,
      "last_run_timestamp": "2026-10-31T11:22:01.481Z"
    },
    {
      "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": "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": 0.7333333333333333,
      "accuracy_std_dev": 0.024944382578492966,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.09260214,
      "avg_duration_per_task_ms": 326558.9733333333,
      "avg_output_tokens_per_task": 4729.153333333334,
      "last_run_timestamp": "2026-10-31T12:28:59.426Z"
    },
    {
      "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": 0.7333333333333333,
      "accuracy_std_dev": 0.009428090415820642,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.024898453333333334,
      "avg_duration_per_task_ms": 37126.9,
      "avg_output_tokens_per_task": 4153.473333333333,
      "last_run_timestamp": "2026-10-31T12:07:50.460Z"
    },
    {
      "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": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "Anthropic",
      "accuracy": 0.7333332,
      "accuracy_std_dev": 0.032659904062320846,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.012294704000000002,
      "avg_duration_per_task_ms": 85675.876,
      "avg_output_tokens_per_task": 10698.828,
      "last_run_timestamp": "2026-10-31T11:35:21.238Z"
    },
    {
      "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": "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": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Modal",
      "accuracy": 0.732916625,
      "accuracy_std_dev": 0.018888392049996622,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0482453475,
      "avg_duration_per_task_ms": 162563.185,
      "avg_output_tokens_per_task": 29218.5525,
      "last_run_timestamp": "2026-10-31T12:25:59.335Z"
    },
    {
      "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": 0.7327394365256125,
      "accuracy_std_dev": 0.030733543577653163,
      "run_count": 9,
      "total_questions": 449,
      "avg_cost_per_task_usd": 0.06566840534521158,
      "avg_duration_per_task_ms": 152213.53006681515,
      "avg_output_tokens_per_task": 11216.12026726058,
      "last_run_timestamp": "2026-10-31T19:24:16.029Z"
    },
    {
      "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": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "Venice",
      "accuracy": 0.7325926666666667,
      "accuracy_std_dev": 0.03841877184687483,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.042171924444444446,
      "avg_duration_per_task_ms": 580922.6555555556,
      "avg_output_tokens_per_task": 9134.031111111111,
      "last_run_timestamp": "2026-10-31T19:41:08.087Z"
    },
    {
      "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": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Inceptron",
      "accuracy": 0.7323944162754304,
      "accuracy_std_dev": 0.034116380301140836,
      "run_count": 13,
      "total_questions": 639,
      "avg_cost_per_task_usd": 0.06478910172143976,
      "avg_duration_per_task_ms": 133446.72456964006,
      "avg_output_tokens_per_task": 8336.646322378716,
      "last_run_timestamp": "2026-10-31T19:10:53.479Z"
    },
    {
      "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": 0.732,
      "accuracy_std_dev": 0.02124973202654565,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.036015575999999994,
      "avg_duration_per_task_ms": 148623.836,
      "avg_output_tokens_per_task": 32577.608,
      "last_run_timestamp": "2026-10-31T12:40:24.644Z"
    },
    {
      "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": 0.7319445,
      "accuracy_std_dev": 0.011663532651674044,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.22145578333333332,
      "avg_duration_per_task_ms": 246837.50166666668,
      "avg_output_tokens_per_task": 10031.816666666668,
      "last_run_timestamp": "2026-10-31T11:24:41.391Z"
    },
    {
      "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": 0.7318517777777778,
      "accuracy_std_dev": 0.03154899079836093,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.16188553555555557,
      "avg_duration_per_task_ms": 219860.70222222223,
      "avg_output_tokens_per_task": 9557.291111111112,
      "last_run_timestamp": "2026-10-31T19:57:59.015Z"
    },
    {
      "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": 0.7316666666666667,
      "accuracy_std_dev": 0.02436222789939834,
      "run_count": 24,
      "total_questions": 1200,
      "avg_cost_per_task_usd": 0.3637274400000001,
      "avg_duration_per_task_ms": 227467.92083333334,
      "avg_output_tokens_per_task": 10893.904166666667,
      "last_run_timestamp": "2026-10-31T12:43:31.737Z"
    },
    {
      "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": 0.7316666666666667,
      "accuracy_std_dev": 0.04261151817942603,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.028175713333333335,
      "avg_duration_per_task_ms": 114994.48666666666,
      "avg_output_tokens_per_task": 9908.751666666667,
      "last_run_timestamp": "2026-10-31T19:22:00.212Z"
    },
    {
      "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": 0.7316665,
      "accuracy_std_dev": 0.018333500000000003,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01274185,
      "avg_duration_per_task_ms": 154997.2,
      "avg_output_tokens_per_task": 13696.96,
      "last_run_timestamp": "2026-10-31T19:54:14.421Z"
    },
    {
      "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": 0.7313333,
      "accuracy_std_dev": 0.04022720859331405,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.069409406,
      "avg_duration_per_task_ms": 292502.6,
      "avg_output_tokens_per_task": 12732.744,
      "last_run_timestamp": "2026-10-31T19:48:30.443Z"
    },
    {
      "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": 0.7313332,
      "accuracy_std_dev": 0.020231941744676914,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.012560144,
      "avg_duration_per_task_ms": 74716.332,
      "avg_output_tokens_per_task": 10525.228,
      "last_run_timestamp": "2026-10-31T11:19:23.393Z"
    },
    {
      "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": 0.7311827782258065,
      "accuracy_std_dev": 0.07837008853695393,
      "run_count": 10,
      "total_questions": 496,
      "avg_cost_per_task_usd": 0.009883709677419354,
      "avg_duration_per_task_ms": 170348.1431451613,
      "avg_output_tokens_per_task": 4275.5161290322585,
      "last_run_timestamp": "2026-10-31T12:31:51.998Z"
    },
    {
      "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": 0.7311637573964497,
      "accuracy_std_dev": 0.02813377440238986,
      "run_count": 17,
      "total_questions": 845,
      "avg_cost_per_task_usd": 0.024811487573964494,
      "avg_duration_per_task_ms": 209572.125443787,
      "avg_output_tokens_per_task": 21468.289940828403,
      "last_run_timestamp": "2026-10-31T11:27:16.882Z"
    },
    {
      "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": "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": 0.7311110000000001,
      "accuracy_std_dev": 0.020608136952831695,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.028709900000000003,
      "avg_duration_per_task_ms": 638312.6933333334,
      "avg_output_tokens_per_task": 61738.18,
      "last_run_timestamp": "2026-10-31T12:28:49.803Z"
    },
    {
      "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": "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": 0.7310256538461538,
      "accuracy_std_dev": 0.03399061082116005,
      "run_count": 26,
      "total_questions": 1300,
      "avg_cost_per_task_usd": 0.3526058969230769,
      "avg_duration_per_task_ms": 206233.06846153847,
      "avg_output_tokens_per_task": 10305.186923076923,
      "last_run_timestamp": "2026-10-31T12:42:19.688Z"
    },
    {
      "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-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "Novita",
      "accuracy": 0.7303704444444444,
      "accuracy_std_dev": 0.027417306521859138,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.05101439333333333,
      "avg_duration_per_task_ms": 356298.62666666665,
      "avg_output_tokens_per_task": 10059.124444444444,
      "last_run_timestamp": "2026-10-31T21:01:26.628Z"
    },
    {
      "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": 0.7300002499999999,
      "accuracy_std_dev": 0.028867484592314224,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.03407485,
      "avg_duration_per_task_ms": 264292.92,
      "avg_output_tokens_per_task": 34254.08,
      "last_run_timestamp": "2026-10-31T13:21:05.335Z"
    },
    {
      "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": 0.73,
      "accuracy_std_dev": 0.0268328157299975,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.074570136,
      "avg_duration_per_task_ms": 246781.984,
      "avg_output_tokens_per_task": 30037.412,
      "last_run_timestamp": "2026-10-31T11:52:51.154Z"
    },
    {
      "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": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Parasail",
      "accuracy": 0.73,
      "accuracy_std_dev": 0.0336238300093847,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.07512531000000001,
      "avg_duration_per_task_ms": 210778.66,
      "avg_output_tokens_per_task": 11117.87,
      "last_run_timestamp": "2026-10-31T20:16:51.242Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "AtlasCloud",
      "accuracy": 0.73,
      "accuracy_std_dev": 0.04167786071965147,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.07906948333333333,
      "avg_duration_per_task_ms": 158873.66833333333,
      "avg_output_tokens_per_task": 8236.776666666667,
      "last_run_timestamp": "2026-10-31T19:53:09.865Z"
    },
    {
      "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": 0.7299999166666666,
      "accuracy_std_dev": 0.026034048220418122,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.08839280833333334,
      "avg_duration_per_task_ms": 664802.0966666667,
      "avg_output_tokens_per_task": 9393.481666666667,
      "last_run_timestamp": "2026-10-31T19:40:42.793Z"
    },
    {
      "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": 0.7299998888888889,
      "accuracy_std_dev": 0.02922883744388389,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.014746135555555555,
      "avg_duration_per_task_ms": 251713.60666666666,
      "avg_output_tokens_per_task": 21386.666666666668,
      "last_run_timestamp": "2026-10-31T12:23:57.099Z"
    },
    {
      "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": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Nebius",
      "accuracy": 0.729654525083612,
      "accuracy_std_dev": 0.01804542249436126,
      "run_count": 6,
      "total_questions": 299,
      "avg_cost_per_task_usd": 0.18312612040133777,
      "avg_duration_per_task_ms": 543481.8829431438,
      "avg_output_tokens_per_task": 8776.65551839465,
      "last_run_timestamp": "2026-10-31T21:08:40.848Z"
    },
    {
      "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": 0.7296295555555555,
      "accuracy_std_dev": 0.02759685177339019,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.09735883111111111,
      "avg_duration_per_task_ms": 218490.78666666665,
      "avg_output_tokens_per_task": 12521.58,
      "last_run_timestamp": "2026-10-31T20:36:53.013Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.5-sonnet-20250929",
      "model_slug": "anthropic/claude-sonnet-4.5",
      "model_name": "Anthropic: Claude Sonnet 4.5",
      "provider_name": "auto-routing",
      "accuracy": 0.7294444166666667,
      "accuracy_std_dev": 0.03271169279849418,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.32659671333333334,
      "avg_duration_per_task_ms": 207735.03833333333,
      "avg_output_tokens_per_task": 10355.186666666666,
      "last_run_timestamp": "2026-10-31T12:49:56.822Z"
    },
    {
      "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": 0.7294443333333334,
      "accuracy_std_dev": 0.03346740998178512,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.033491596666666665,
      "avg_duration_per_task_ms": 413814.1533333333,
      "avg_output_tokens_per_task": 8947.89,
      "last_run_timestamp": "2026-10-31T21:01:10.349Z"
    },
    {
      "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": 0.7293334,
      "accuracy_std_dev": 0.026533159130416407,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.012403696,
      "avg_duration_per_task_ms": 92314.648,
      "avg_output_tokens_per_task": 10764.472,
      "last_run_timestamp": "2026-10-31T12:41:01.831Z"
    },
    {
      "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": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Together",
      "accuracy": 0.7291665,
      "accuracy_std_dev": 0.023318429765959808,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0316769325,
      "avg_duration_per_task_ms": 89151.6375,
      "avg_output_tokens_per_task": 22436.51,
      "last_run_timestamp": "2026-10-31T11:37:01.556Z"
    },
    {
      "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-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Novita",
      "accuracy": 0.728889,
      "accuracy_std_dev": 0.01911599421078241,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.017266966666666664,
      "avg_duration_per_task_ms": 122400.14,
      "avg_output_tokens_per_task": 11879.513333333334,
      "last_run_timestamp": "2026-10-31T11:33:22.952Z"
    },
    {
      "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": 0.728889,
      "accuracy_std_dev": 0.01257094435593444,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.04855964666666667,
      "avg_duration_per_task_ms": 115027.81333333334,
      "avg_output_tokens_per_task": 15070.966666666667,
      "last_run_timestamp": "2026-10-31T12:30:34.970Z"
    },
    {
      "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": 0.7288886666666667,
      "accuracy_std_dev": 0.017497738831695454,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.033088113333333335,
      "avg_duration_per_task_ms": 134354.10666666666,
      "avg_output_tokens_per_task": 14586.926666666666,
      "last_run_timestamp": "2026-10-31T12:32:08.127Z"
    },
    {
      "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": 0.7283335,
      "accuracy_std_dev": 0.024626886091898288,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.04604961333333333,
      "avg_duration_per_task_ms": 118630.39,
      "avg_output_tokens_per_task": 26239.556666666667,
      "last_run_timestamp": "2026-10-31T12:19:28.999Z"
    },
    {
      "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": 0.72833325,
      "accuracy_std_dev": 0.041197373838480274,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.011627464999999998,
      "avg_duration_per_task_ms": 86463.365,
      "avg_output_tokens_per_task": 10669.235,
      "last_run_timestamp": "2026-10-31T12:31:32.731Z"
    },
    {
      "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": "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": 0.728,
      "accuracy_std_dev": 0.02362657637492154,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.027497304,
      "avg_duration_per_task_ms": 93734.552,
      "avg_output_tokens_per_task": 12939.792,
      "last_run_timestamp": "2026-10-31T12:24:43.560Z"
    },
    {
      "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": 0.7277778333333333,
      "accuracy_std_dev": 0.029165443361489008,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.00697867,
      "avg_duration_per_task_ms": 90387.44333333333,
      "avg_output_tokens_per_task": 9165.73,
      "last_run_timestamp": "2026-10-31T11:54:38.723Z"
    },
    {
      "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": 0.72777775,
      "accuracy_std_dev": 0.030164136379495988,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.34935106333333343,
      "avg_duration_per_task_ms": 233197.23,
      "avg_output_tokens_per_task": 10873.441666666668,
      "last_run_timestamp": "2026-10-31T12:28:19.030Z"
    },
    {
      "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": 0.7274047715780296,
      "accuracy_std_dev": 0.02610995768010187,
      "run_count": 23,
      "total_questions": 1147,
      "avg_cost_per_task_usd": 0.44006520662598075,
      "avg_duration_per_task_ms": 249359.09677419355,
      "avg_output_tokens_per_task": 10802.551002615519,
      "last_run_timestamp": "2026-10-31T12:24:25.742Z"
    },
    {
      "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": "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": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "BaseTen",
      "accuracy": 0.7270409540816326,
      "accuracy_std_dev": 0.02754153279934688,
      "run_count": 8,
      "total_questions": 392,
      "avg_cost_per_task_usd": 0.04606436989795917,
      "avg_duration_per_task_ms": 81576.88520408163,
      "avg_output_tokens_per_task": 19891.04081632653,
      "last_run_timestamp": "2026-10-31T12:26:53.018Z"
    },
    {
      "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": 0.726667,
      "accuracy_std_dev": 0.019999999999999962,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.24019618,
      "avg_duration_per_task_ms": 932058.35,
      "avg_output_tokens_per_task": 24816.04,
      "last_run_timestamp": "2026-10-31T11:58:06.550Z"
    },
    {
      "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": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Baidu",
      "accuracy": 0.7266667777777778,
      "accuracy_std_dev": 0.02591526553459167,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.06854443777777779,
      "avg_duration_per_task_ms": 220265.38444444444,
      "avg_output_tokens_per_task": 13043.293333333333,
      "last_run_timestamp": "2026-10-31T19:39:42.479Z"
    },
    {
      "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": "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": "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": 0.7266663333333333,
      "accuracy_std_dev": 0.018856180831641284,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.024500659999999997,
      "avg_duration_per_task_ms": 72543.00666666667,
      "avg_output_tokens_per_task": 15175.193333333333,
      "last_run_timestamp": "2026-10-31T11:34:42.984Z"
    },
    {
      "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": 0.7264369310344827,
      "accuracy_std_dev": 0.006057869592521776,
      "run_count": 3,
      "total_questions": 145,
      "avg_cost_per_task_usd": 0.012712737931034484,
      "avg_duration_per_task_ms": 160320.86896551723,
      "avg_output_tokens_per_task": 8003.303448275862,
      "last_run_timestamp": "2026-10-31T12:11:53.396Z"
    },
    {
      "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": 0.726330531092437,
      "accuracy_std_dev": 0.031188754641318872,
      "run_count": 12,
      "total_questions": 595,
      "avg_cost_per_task_usd": 0.07654581848739497,
      "avg_duration_per_task_ms": 189475.03361344538,
      "avg_output_tokens_per_task": 8239.426890756302,
      "last_run_timestamp": "2026-10-31T19:01:31.914Z"
    },
    {
      "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": 0.7262963333333333,
      "accuracy_std_dev": 0.03679916650928086,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.08815544444444444,
      "avg_duration_per_task_ms": 169827.67555555556,
      "avg_output_tokens_per_task": 10443.75111111111,
      "last_run_timestamp": "2026-10-31T19:56:54.162Z"
    },
    {
      "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": 0.7261110833333333,
      "accuracy_std_dev": 0.02924471198484249,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.08110476333333332,
      "avg_duration_per_task_ms": 226307.265,
      "avg_output_tokens_per_task": 8776.583333333334,
      "last_run_timestamp": "2026-10-31T20:37:11.918Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Cloudflare",
      "accuracy": 0.7256409230769231,
      "accuracy_std_dev": 0.024647466274717044,
      "run_count": 13,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.07311417384615386,
      "avg_duration_per_task_ms": 221051.06153846154,
      "avg_output_tokens_per_task": 8508.512307692308,
      "last_run_timestamp": "2026-10-31T20:10:58.659Z"
    },
    {
      "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": "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": "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": 0.7251952608695652,
      "accuracy_std_dev": 0.07868938349065195,
      "run_count": 12,
      "total_questions": 598,
      "avg_cost_per_task_usd": 0.04124413545150502,
      "avg_duration_per_task_ms": 274554.9113712375,
      "avg_output_tokens_per_task": 23809.344481605352,
      "last_run_timestamp": "2026-10-31T19:30:45.204Z"
    },
    {
      "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": 0.7244446666666666,
      "accuracy_std_dev": 0.01662961808207136,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.00495835,
      "avg_duration_per_task_ms": 143781.20333333334,
      "avg_output_tokens_per_task": 12059.35,
      "last_run_timestamp": "2026-10-31T12:51:11.039Z"
    },
    {
      "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": 0.7244444444444444,
      "accuracy_std_dev": 0.022662439694648512,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.06561171777777777,
      "avg_duration_per_task_ms": 200024.29111111112,
      "avg_output_tokens_per_task": 9471.135555555555,
      "last_run_timestamp": "2026-10-31T20:17:42.375Z"
    },
    {
      "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": 0.7244443333333334,
      "accuracy_std_dev": 0.006285236475706843,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.018751046666666667,
      "avg_duration_per_task_ms": 183683.96666666667,
      "avg_output_tokens_per_task": 11588.8,
      "last_run_timestamp": "2026-10-31T11:59:31.865Z"
    },
    {
      "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": "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": 0.724,
      "accuracy_std_dev": 0.02351363670723862,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.26574487199999997,
      "avg_duration_per_task_ms": 177434.862,
      "avg_output_tokens_per_task": 11402.85,
      "last_run_timestamp": "2026-10-31T12:56:09.161Z"
    },
    {
      "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": 0.7238889166666667,
      "accuracy_std_dev": 0.03479662118285417,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.08805141666666667,
      "avg_duration_per_task_ms": 322075.905,
      "avg_output_tokens_per_task": 9307.315,
      "last_run_timestamp": "2026-10-31T19:57:59.098Z"
    },
    {
      "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": 0.7237037777777777,
      "accuracy_std_dev": 0.0396418852639976,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.10080122000000001,
      "avg_duration_per_task_ms": 210300.32,
      "avg_output_tokens_per_task": 11702.788888888888,
      "last_run_timestamp": "2026-10-31T20:14:27.576Z"
    },
    {
      "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": 0.723360643442623,
      "accuracy_std_dev": 0.019132270021092652,
      "run_count": 5,
      "total_questions": 244,
      "avg_cost_per_task_usd": 0.0756424631147541,
      "avg_duration_per_task_ms": 76631.30327868853,
      "avg_output_tokens_per_task": 8337.151639344262,
      "last_run_timestamp": "2026-10-31T11:46:18.075Z"
    },
    {
      "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": "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": 0.7233330000000001,
      "accuracy_std_dev": 0.010000000000000009,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.02918011,
      "avg_duration_per_task_ms": 177614.61,
      "avg_output_tokens_per_task": 35023.7,
      "last_run_timestamp": "2026-10-31T11:38:56.927Z"
    },
    {
      "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": "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": 0.7229627777777778,
      "accuracy_std_dev": 0.03557083313858196,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.09454113555555556,
      "avg_duration_per_task_ms": 220849.31555555554,
      "avg_output_tokens_per_task": 10002.34,
      "last_run_timestamp": "2026-10-31T20:09:58.111Z"
    },
    {
      "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": "meta/muse-spark-open-20261015",
      "model_slug": "meta/muse-spark-open",
      "model_name": "Meta: Muse Spark Open",
      "provider_name": "BaseTen",
      "accuracy": 0.7228067157894738,
      "accuracy_std_dev": 0.002917999999999976,
      "run_count": 2,
      "total_questions": 95,
      "avg_cost_per_task_usd": 0.06760965263157895,
      "avg_duration_per_task_ms": 49831.26315789474,
      "avg_output_tokens_per_task": 5886.4315789473685,
      "last_run_timestamp": "2026-10-31T12:12:21.006Z"
    },
    {
      "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": 0.722650186440678,
      "accuracy_std_dev": 0.03806732610044826,
      "run_count": 13,
      "total_questions": 649,
      "avg_cost_per_task_usd": 0.024282243451463792,
      "avg_duration_per_task_ms": 277851.95069337444,
      "avg_output_tokens_per_task": 9180.11093990755,
      "last_run_timestamp": "2026-10-31T13:24:52.766Z"
    },
    {
      "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": 0.7225,
      "accuracy_std_dev": 0.029615564602924586,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.05079202000000001,
      "avg_duration_per_task_ms": 151914.885,
      "avg_output_tokens_per_task": 28118.625,
      "last_run_timestamp": "2026-10-31T12:23:57.101Z"
    },
    {
      "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": 0.7223807857142858,
      "accuracy_std_dev": 0.05901878731650815,
      "run_count": 14,
      "total_questions": 700,
      "avg_cost_per_task_usd": 0.2836348157142858,
      "avg_duration_per_task_ms": 433854.6257142857,
      "avg_output_tokens_per_task": 10133.661428571428,
      "last_run_timestamp": "2026-10-31T12:33:51.242Z"
    },
    {
      "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": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "SambaNova",
      "accuracy": 0.7222223333333333,
      "accuracy_std_dev": 0.011330936570096713,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.03683,
      "avg_duration_per_task_ms": 122927.25333333333,
      "avg_output_tokens_per_task": 9303.18,
      "last_run_timestamp": "2026-10-31T12:49:19.749Z"
    },
    {
      "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": 0.7222223333333333,
      "accuracy_std_dev": 0.01911642045177101,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.010745106666666667,
      "avg_duration_per_task_ms": 61185.806666666664,
      "avg_output_tokens_per_task": 4195.073333333334,
      "last_run_timestamp": "2026-10-31T11:48:12.980Z"
    },
    {
      "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": 0.72166675,
      "accuracy_std_dev": 0.012801898733683979,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.024317345,
      "avg_duration_per_task_ms": 116922.04,
      "avg_output_tokens_per_task": 11395.83,
      "last_run_timestamp": "2026-10-31T12:46:14.030Z"
    },
    {
      "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": 0.7214610958904109,
      "accuracy_std_dev": 0.02944405326868039,
      "run_count": 3,
      "total_questions": 146,
      "avg_cost_per_task_usd": 0.01699455479452055,
      "avg_duration_per_task_ms": 840567.9794520548,
      "avg_output_tokens_per_task": 10799.780821917808,
      "last_run_timestamp": "2026-10-31T13:36:58.095Z"
    },
    {
      "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": 0.7213334,
      "accuracy_std_dev": 0.01146975927559073,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.012899843999999999,
      "avg_duration_per_task_ms": 152142.784,
      "avg_output_tokens_per_task": 13063.108,
      "last_run_timestamp": "2026-10-31T13:10:04.012Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Parasail",
      "accuracy": 0.721249875,
      "accuracy_std_dev": 0.02449118393135325,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.045012550000000005,
      "avg_duration_per_task_ms": 161807.095,
      "avg_output_tokens_per_task": 30651.7875,
      "last_run_timestamp": "2026-10-31T11:23:43.290Z"
    },
    {
      "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": 0.7211110833333333,
      "accuracy_std_dev": 0.02579616561771126,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.04659319666666666,
      "avg_duration_per_task_ms": 177612.27333333335,
      "avg_output_tokens_per_task": 9673.603333333333,
      "last_run_timestamp": "2026-10-31T20:01:58.593Z"
    },
    {
      "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": 0.7211021935483871,
      "accuracy_std_dev": 0.04800607831972945,
      "run_count": 10,
      "total_questions": 496,
      "avg_cost_per_task_usd": 0.03239710483870968,
      "avg_duration_per_task_ms": 1093021.4314516129,
      "avg_output_tokens_per_task": 51597.49596774193,
      "last_run_timestamp": "2026-10-31T20:03:31.183Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Makora",
      "accuracy": 0.7210882448979592,
      "accuracy_std_dev": 0.0076780598851532815,
      "run_count": 5,
      "total_questions": 245,
      "avg_cost_per_task_usd": 0.009489710204081632,
      "avg_duration_per_task_ms": 145499.9755102041,
      "avg_output_tokens_per_task": 16260.81224489796,
      "last_run_timestamp": "2026-09-26T08:06:43.933Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "auto-routing",
      "accuracy": 0.7207407777777778,
      "accuracy_std_dev": 0.01349684513570796,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.052085299999999994,
      "avg_duration_per_task_ms": 388042.9088888889,
      "avg_output_tokens_per_task": 9288.517777777777,
      "last_run_timestamp": "2026-10-31T20:46:45.843Z"
    },
    {
      "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": "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": "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": "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": 0.72,
      "accuracy_std_dev": 0.01247214674785381,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.08792842000000001,
      "avg_duration_per_task_ms": 120135.21,
      "avg_output_tokens_per_task": 16462.21,
      "last_run_timestamp": "2026-10-31T11:24:22.174Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.7-max-20260520",
      "model_slug": "qwen/qwen3.7-max",
      "model_name": "Qwen: Qwen3.7 Max",
      "provider_name": "AtlasCloud",
      "accuracy": 0.72,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.402901,
      "avg_duration_per_task_ms": 196016.4,
      "avg_output_tokens_per_task": 16256.52,
      "last_run_timestamp": "2026-09-27T00:00:03.103Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "Parasail",
      "accuracy": 0.72,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.028463379999999996,
      "avg_duration_per_task_ms": 583244.64,
      "avg_output_tokens_per_task": 43866.54,
      "last_run_timestamp": "2026-08-08T08:27:52.152Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7199999166666667,
      "accuracy_std_dev": 0.032546251989689814,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.08362974499999999,
      "avg_duration_per_task_ms": 133531.81833333333,
      "avg_output_tokens_per_task": 9309.471666666666,
      "last_run_timestamp": "2026-10-31T19:18:12.479Z"
    },
    {
      "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": 0.7194444166666667,
      "accuracy_std_dev": 0.02063055611812381,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.08198992666666667,
      "avg_duration_per_task_ms": 234565.96333333335,
      "avg_output_tokens_per_task": 9245.488333333333,
      "last_run_timestamp": "2026-10-31T20:11:34.826Z"
    },
    {
      "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": 0.71933335,
      "accuracy_std_dev": 0.026574804694061246,
      "run_count": 20,
      "total_questions": 1000,
      "avg_cost_per_task_usd": 0.065825297,
      "avg_duration_per_task_ms": 115333.499,
      "avg_output_tokens_per_task": 8746.383,
      "last_run_timestamp": "2026-10-31T19:54:14.505Z"
    },
    {
      "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": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "Venice",
      "accuracy": 0.7192592222222223,
      "accuracy_std_dev": 0.03684397384701473,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.07047637111111112,
      "avg_duration_per_task_ms": 259236.1,
      "avg_output_tokens_per_task": 27382.248888888887,
      "last_run_timestamp": "2026-10-31T19:22:37.818Z"
    },
    {
      "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": 0.7186667999999999,
      "accuracy_std_dev": 0.028095019725211078,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.11915163599999999,
      "avg_duration_per_task_ms": 86040.636,
      "avg_output_tokens_per_task": 9027.512,
      "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": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Crusoe",
      "accuracy": 0.7178963360655738,
      "accuracy_std_dev": 0.15543665632919412,
      "run_count": 5,
      "total_questions": 244,
      "avg_cost_per_task_usd": 0.02473220901639344,
      "avg_duration_per_task_ms": 313234.66803278687,
      "avg_output_tokens_per_task": 9615.508196721312,
      "last_run_timestamp": "2026-10-31T19:23:32.973Z"
    },
    {
      "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": 0.717778,
      "accuracy_std_dev": 0.046930323977857506,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.045013793333333336,
      "avg_duration_per_task_ms": 196458.52,
      "avg_output_tokens_per_task": 18662.633333333335,
      "last_run_timestamp": "2026-10-31T19:03:35.054Z"
    },
    {
      "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": 0.7177779999999999,
      "accuracy_std_dev": 0.03499564699597175,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.039637780000000004,
      "avg_duration_per_task_ms": 223451.97333333333,
      "avg_output_tokens_per_task": 45456.886666666665,
      "last_run_timestamp": "2026-10-31T11:34:19.759Z"
    },
    {
      "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": 0.7177777777777777,
      "accuracy_std_dev": 0.02060811299128885,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.05747164444444444,
      "avg_duration_per_task_ms": 334450.40222222224,
      "avg_output_tokens_per_task": 9986.046666666667,
      "last_run_timestamp": "2026-10-31T20:51:04.120Z"
    },
    {
      "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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Alibaba",
      "accuracy": 0.7175000624999999,
      "accuracy_std_dev": 0.03212396291121463,
      "run_count": 32,
      "total_questions": 1600,
      "avg_cost_per_task_usd": 0.11736274687500002,
      "avg_duration_per_task_ms": 128613.640625,
      "avg_output_tokens_per_task": 8558.630625,
      "last_run_timestamp": "2026-10-31T19:56:11.088Z"
    },
    {
      "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": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Novita",
      "accuracy": 0.717037,
      "accuracy_std_dev": 0.02668713823382175,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.1317152511111111,
      "avg_duration_per_task_ms": 254884.2288888889,
      "avg_output_tokens_per_task": 10329.728888888889,
      "last_run_timestamp": "2026-10-31T20:31:35.900Z"
    },
    {
      "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": "DekaLLM",
      "accuracy": 0.716667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.01519146,
      "avg_duration_per_task_ms": 819823.24,
      "avg_output_tokens_per_task": 11470.58,
      "last_run_timestamp": "2026-10-31T20:19:18.995Z"
    },
    {
      "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": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "Phala",
      "accuracy": 0.7166667777777778,
      "accuracy_std_dev": 0.026943095041124358,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.2115978688888889,
      "avg_duration_per_task_ms": 1320901.8111111112,
      "avg_output_tokens_per_task": 13916.793333333333,
      "last_run_timestamp": "2026-10-31T19:47:23.162Z"
    },
    {
      "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": 0.7166665000000001,
      "accuracy_std_dev": 0.0033334999999999892,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.022047789999999998,
      "avg_duration_per_task_ms": 63248.07,
      "avg_output_tokens_per_task": 8992.32,
      "last_run_timestamp": "2026-10-31T12:43:48.267Z"
    },
    {
      "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": 0.7166665000000001,
      "accuracy_std_dev": 0.0033334999999999892,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.015074639999999999,
      "avg_duration_per_task_ms": 124785.23,
      "avg_output_tokens_per_task": 11839.88,
      "last_run_timestamp": "2026-10-31T12:28:39.529Z"
    },
    {
      "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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Relace",
      "accuracy": 0.7166665000000001,
      "accuracy_std_dev": 0.02603405888927041,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.13610847,
      "avg_duration_per_task_ms": 230555.42,
      "avg_output_tokens_per_task": 11694.86,
      "last_run_timestamp": "2026-10-31T19:46:31.005Z"
    },
    {
      "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": 0.7159998000000001,
      "accuracy_std_dev": 0.03295793548085197,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.24450749400000002,
      "avg_duration_per_task_ms": 314057.314,
      "avg_output_tokens_per_task": 21575.564,
      "last_run_timestamp": "2026-10-31T12:46:35.433Z"
    },
    {
      "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": 0.7155556666666666,
      "accuracy_std_dev": 0.03189478845969807,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.01325666,
      "avg_duration_per_task_ms": 86099.67333333334,
      "avg_output_tokens_per_task": 8966.793333333333,
      "last_run_timestamp": "2026-10-31T13:08:15.957Z"
    },
    {
      "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": 0.7155554444444445,
      "accuracy_std_dev": 0.032203028767531616,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.05172563555555556,
      "avg_duration_per_task_ms": 392299.5088888889,
      "avg_output_tokens_per_task": 11718.533333333333,
      "last_run_timestamp": "2026-10-31T20:58:51.733Z"
    },
    {
      "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": 0.7155554444444445,
      "accuracy_std_dev": 0.023094000077514658,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.05959621777777777,
      "avg_duration_per_task_ms": 317385.8333333333,
      "avg_output_tokens_per_task": 9770.851111111111,
      "last_run_timestamp": "2026-10-31T20:35:24.659Z"
    },
    {
      "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": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "Minimax",
      "accuracy": 0.7133333888888889,
      "accuracy_std_dev": 0.022553037627726643,
      "run_count": 18,
      "total_questions": 900,
      "avg_cost_per_task_usd": 0.03885568666666667,
      "avg_duration_per_task_ms": 225043.39444444445,
      "avg_output_tokens_per_task": 9769.26,
      "last_run_timestamp": "2026-10-31T20:38:36.972Z"
    },
    {
      "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": 0.7133333333333334,
      "accuracy_std_dev": 0.03126952723321681,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.04438405333333333,
      "avg_duration_per_task_ms": 115670.52333333333,
      "avg_output_tokens_per_task": 8401.7,
      "last_run_timestamp": "2026-10-31T19:54:42.760Z"
    },
    {
      "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": 0.7133330000000001,
      "accuracy_std_dev": 0.03265986323710907,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.02237405333333333,
      "avg_duration_per_task_ms": 91226.86,
      "avg_output_tokens_per_task": 9082.033333333333,
      "last_run_timestamp": "2026-10-31T12:52:04.992Z"
    },
    {
      "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-v4.1-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": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Fireworks",
      "accuracy": 0.713333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.01617958,
      "avg_duration_per_task_ms": 141600.68,
      "avg_output_tokens_per_task": 14742.64,
      "last_run_timestamp": "2026-10-31T12:42:39.534Z"
    },
    {
      "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": "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": 0.7126666,
      "accuracy_std_dev": 0.019425055985504893,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.02303438,
      "avg_duration_per_task_ms": 273736.758,
      "avg_output_tokens_per_task": 10030.944,
      "last_run_timestamp": "2026-10-31T19:56:09.033Z"
    },
    {
      "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": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "CoreWeave",
      "accuracy": 0.7123077692307692,
      "accuracy_std_dev": 0.03859468338864592,
      "run_count": 13,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.08011615384615385,
      "avg_duration_per_task_ms": 106325.87538461538,
      "avg_output_tokens_per_task": 8598.36,
      "last_run_timestamp": "2026-10-31T19:45:55.386Z"
    },
    {
      "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": 0.71222225,
      "accuracy_std_dev": 0.14864397040699018,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.06236268333333334,
      "avg_duration_per_task_ms": 148178.685,
      "avg_output_tokens_per_task": 17104.778333333332,
      "last_run_timestamp": "2026-10-31T19:40:48.790Z"
    },
    {
      "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": 0.7117822371364653,
      "accuracy_std_dev": 0.056118924341169225,
      "run_count": 9,
      "total_questions": 447,
      "avg_cost_per_task_usd": 0.0462677941834452,
      "avg_duration_per_task_ms": 1033622.2237136465,
      "avg_output_tokens_per_task": 60076.49888143177,
      "last_run_timestamp": "2026-10-31T13:23:14.602Z"
    },
    {
      "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": 0.7113334,
      "accuracy_std_dev": 0.036884212823916955,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.28188106999999996,
      "avg_duration_per_task_ms": 497814.602,
      "avg_output_tokens_per_task": 9499.378,
      "last_run_timestamp": "2026-10-31T22:14:51.015Z"
    },
    {
      "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": 0.7111113333333333,
      "accuracy_std_dev": 0.006285236475706792,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.06754876666666666,
      "avg_duration_per_task_ms": 139695.35333333333,
      "avg_output_tokens_per_task": 33581.94,
      "last_run_timestamp": "2026-10-31T12:01:01.015Z"
    },
    {
      "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": 0.7111111111111111,
      "accuracy_std_dev": 0.012570963999492764,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.027520999999999997,
      "avg_duration_per_task_ms": 233912.26444444444,
      "avg_output_tokens_per_task": 10378.048888888889,
      "last_run_timestamp": "2026-10-31T19:47:12.389Z"
    },
    {
      "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": 0.7111111111111111,
      "accuracy_std_dev": 0.043659620580232664,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.06395731555555556,
      "avg_duration_per_task_ms": 239858.9088888889,
      "avg_output_tokens_per_task": 8206.18888888889,
      "last_run_timestamp": "2026-10-31T20:09:59.492Z"
    },
    {
      "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": 0.7110144347826087,
      "accuracy_std_dev": 0.024991420835482343,
      "run_count": 23,
      "total_questions": 1150,
      "avg_cost_per_task_usd": 0.10159863652173912,
      "avg_duration_per_task_ms": 108613.41913043478,
      "avg_output_tokens_per_task": 9699.877391304348,
      "last_run_timestamp": "2026-10-31T19:37:51.988Z"
    },
    {
      "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": 0.7100000833333333,
      "accuracy_std_dev": 0.030367993020224245,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.12156609333333333,
      "avg_duration_per_task_ms": 163832.635,
      "avg_output_tokens_per_task": 11504.31,
      "last_run_timestamp": "2026-10-31T19:55:04.254Z"
    },
    {
      "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": 0.71,
      "accuracy_std_dev": 0.016666999999999987,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.015407820000000001,
      "avg_duration_per_task_ms": 62181.78,
      "avg_output_tokens_per_task": 11850.44,
      "last_run_timestamp": "2026-10-31T11:31:06.991Z"
    },
    {
      "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": 0.71,
      "accuracy_std_dev": 0.01795033271279393,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.0134254,
      "avg_duration_per_task_ms": 120168.52,
      "avg_output_tokens_per_task": 11670.58,
      "last_run_timestamp": "2026-10-31T12:25:39.138Z"
    },
    {
      "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-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "GMICloud",
      "accuracy": 0.7093333,
      "accuracy_std_dev": 0.07889230465267193,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.010360992,
      "avg_duration_per_task_ms": 240638.89,
      "avg_output_tokens_per_task": 9031.174,
      "last_run_timestamp": "2026-10-31T12:26:55.059Z"
    },
    {
      "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": 0.709166625,
      "accuracy_std_dev": 0.031786845419675976,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.09062656499999999,
      "avg_duration_per_task_ms": 834592.7325,
      "avg_output_tokens_per_task": 9529.9875,
      "last_run_timestamp": "2026-10-31T21:09:07.436Z"
    },
    {
      "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": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "Novita",
      "accuracy": 0.708889,
      "accuracy_std_dev": 0.03311020878822723,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.025858879999999997,
      "avg_duration_per_task_ms": 230061.79333333333,
      "avg_output_tokens_per_task": 9903.024444444445,
      "last_run_timestamp": "2026-10-31T20:20:16.475Z"
    },
    {
      "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": 0.708889,
      "accuracy_std_dev": 0.0031423825355929936,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.017471633333333333,
      "avg_duration_per_task_ms": 219758.70666666667,
      "avg_output_tokens_per_task": 16040.473333333333,
      "last_run_timestamp": "2026-10-31T12:30:16.930Z"
    },
    {
      "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": 0.708889,
      "accuracy_std_dev": 0.033554753486735006,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.07036646666666667,
      "avg_duration_per_task_ms": 317115.9511111111,
      "avg_output_tokens_per_task": 10689.18,
      "last_run_timestamp": "2026-10-31T21:34:09.712Z"
    },
    {
      "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": 0.7088888333333333,
      "accuracy_std_dev": 0.04374444021212551,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.11762815500000003,
      "avg_duration_per_task_ms": 129475.63166666667,
      "avg_output_tokens_per_task": 9546.961666666666,
      "last_run_timestamp": "2026-10-31T21:47:15.954Z"
    },
    {
      "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": 0.7088556290726817,
      "accuracy_std_dev": 0.03423477982234403,
      "run_count": 8,
      "total_questions": 399,
      "avg_cost_per_task_usd": 0.021466571428571425,
      "avg_duration_per_task_ms": 619938.7243107769,
      "avg_output_tokens_per_task": 9440.741854636592,
      "last_run_timestamp": "2026-10-31T19:57:15.103Z"
    },
    {
      "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": "Together",
      "accuracy": 0.7083333333333334,
      "accuracy_std_dev": 0.036119498172347615,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.10219764333333331,
      "avg_duration_per_task_ms": 96190.13,
      "avg_output_tokens_per_task": 9006.453333333333,
      "last_run_timestamp": "2026-10-31T19:33:25.981Z"
    },
    {
      "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": 0.7083331666666667,
      "accuracy_std_dev": 0.017400462957602263,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.25373078000000004,
      "avg_duration_per_task_ms": 176420.9,
      "avg_output_tokens_per_task": 10605.743333333334,
      "last_run_timestamp": "2026-10-31T19:48:27.256Z"
    },
    {
      "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": 0.7082050769230769,
      "accuracy_std_dev": 0.030060181146036768,
      "run_count": 13,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.034733135384615386,
      "avg_duration_per_task_ms": 181977.0169230769,
      "avg_output_tokens_per_task": 9062.112307692307,
      "last_run_timestamp": "2026-10-31T20:27:27.569Z"
    },
    {
      "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": 0.7081482222222222,
      "accuracy_std_dev": 0.10187287917659142,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.09577372888888888,
      "avg_duration_per_task_ms": 240048.98666666666,
      "avg_output_tokens_per_task": 22174.402222222223,
      "last_run_timestamp": "2026-10-31T19:46:05.991Z"
    },
    {
      "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": "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/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": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "GMICloud",
      "accuracy": 0.70666675,
      "accuracy_std_dev": 0.027080004924067123,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.017216295,
      "avg_duration_per_task_ms": 228198.02,
      "avg_output_tokens_per_task": 15657.435,
      "last_run_timestamp": "2026-10-31T13:09:38.196Z"
    },
    {
      "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": "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": 0.7066666666666667,
      "accuracy_std_dev": 0.014401440259764123,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.015417799999999999,
      "avg_duration_per_task_ms": 75266.33333333333,
      "avg_output_tokens_per_task": 12018.02,
      "last_run_timestamp": "2026-10-31T11:20:39.476Z"
    },
    {
      "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": "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": 0.7061111666666666,
      "accuracy_std_dev": 0.023602995992152234,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.023716863333333334,
      "avg_duration_per_task_ms": 111467.81333333334,
      "avg_output_tokens_per_task": 29595.083333333332,
      "last_run_timestamp": "2026-10-31T19:52:48.234Z"
    },
    {
      "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": 0.7059259999999999,
      "accuracy_std_dev": 0.05168229207593467,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.03188982666666667,
      "avg_duration_per_task_ms": 346128.37333333335,
      "avg_output_tokens_per_task": 10303.717777777778,
      "last_run_timestamp": "2026-10-31T20:10:49.937Z"
    },
    {
      "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": 0.7058029271255061,
      "accuracy_std_dev": 0.021947489401751626,
      "run_count": 5,
      "total_questions": 247,
      "avg_cost_per_task_usd": 0.012539862348178138,
      "avg_duration_per_task_ms": 294724.995951417,
      "avg_output_tokens_per_task": 13911.971659919029,
      "last_run_timestamp": "2026-10-31T12:54:14.719Z"
    },
    {
      "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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "DeepInfra",
      "accuracy": 0.7051282307692308,
      "accuracy_std_dev": 0.03309201444725772,
      "run_count": 13,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.04285510153846153,
      "avg_duration_per_task_ms": 142444.70769230768,
      "avg_output_tokens_per_task": 8502.066153846154,
      "last_run_timestamp": "2026-10-31T20:18:13.909Z"
    },
    {
      "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": "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": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "DeepInfra",
      "accuracy": 0.7045706153846154,
      "accuracy_std_dev": 0.020768805526558354,
      "run_count": 6,
      "total_questions": 299,
      "avg_cost_per_task_usd": 0.049576892976588634,
      "avg_duration_per_task_ms": 412685.872909699,
      "avg_output_tokens_per_task": 8662.418060200669,
      "last_run_timestamp": "2026-10-31T21:01:44.534Z"
    },
    {
      "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": "Makora",
      "accuracy": 0.7039998000000001,
      "accuracy_std_dev": 0.025508095447524113,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.021704752,
      "avg_duration_per_task_ms": 119683.016,
      "avg_output_tokens_per_task": 11904.78,
      "last_run_timestamp": "2026-10-31T12:25:42.323Z"
    },
    {
      "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": "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": "Inceptron",
      "accuracy": 0.7031330647149461,
      "accuracy_std_dev": 0.03140917629230264,
      "run_count": 13,
      "total_questions": 649,
      "avg_cost_per_task_usd": 0.07489512018489985,
      "avg_duration_per_task_ms": 165770.67642526966,
      "avg_output_tokens_per_task": 9459.0030816641,
      "last_run_timestamp": "2026-10-31T19:41:05.183Z"
    },
    {
      "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": 0.7025,
      "accuracy_std_dev": 0.025353913425741587,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.11362007500000001,
      "avg_duration_per_task_ms": 91835.93333333333,
      "avg_output_tokens_per_task": 9769.808333333332,
      "last_run_timestamp": "2026-10-31T19:57:32.812Z"
    },
    {
      "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": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "Venice",
      "accuracy": 0.7022223333333333,
      "accuracy_std_dev": 0.027755546660882827,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.04592378666666666,
      "avg_duration_per_task_ms": 539584.4533333334,
      "avg_output_tokens_per_task": 9026.931111111111,
      "last_run_timestamp": "2026-10-31T20:03:54.586Z"
    },
    {
      "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": 0.7019519504504504,
      "accuracy_std_dev": 0.09294428248162816,
      "run_count": 9,
      "total_questions": 444,
      "avg_cost_per_task_usd": 0.1323645135135135,
      "avg_duration_per_task_ms": 1692899.1238738738,
      "avg_output_tokens_per_task": 97433.60585585586,
      "last_run_timestamp": "2026-10-31T19:11:32.112Z"
    },
    {
      "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": "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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "CoreWeave",
      "accuracy": 0.7007407777777778,
      "accuracy_std_dev": 0.04138870710116677,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.06795873333333333,
      "avg_duration_per_task_ms": 72240.74888888889,
      "avg_output_tokens_per_task": 9788.00888888889,
      "last_run_timestamp": "2026-10-31T19:23:52.545Z"
    },
    {
      "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": 0.7004591575757575,
      "accuracy_std_dev": 0.0304287066932544,
      "run_count": 37,
      "total_questions": 1815,
      "avg_cost_per_task_usd": 0.10116235151515152,
      "avg_duration_per_task_ms": 72387.2787878788,
      "avg_output_tokens_per_task": 7992.845730027549,
      "last_run_timestamp": "2026-10-31T20:07:10.133Z"
    },
    {
      "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": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Google",
      "accuracy": 0.7,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.0355865,
      "avg_duration_per_task_ms": 154491.8,
      "avg_output_tokens_per_task": 12574.4,
      "last_run_timestamp": "2026-10-31T11:15:31.400Z"
    },
    {
      "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": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "StreamLake",
      "accuracy": 0.7,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.0161093,
      "avg_duration_per_task_ms": 215051.72,
      "avg_output_tokens_per_task": 15584.08,
      "last_run_timestamp": "2026-10-31T12:24:55.227Z"
    },
    {
      "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": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Mancer 2",
      "accuracy": 0.6995834999999999,
      "accuracy_std_dev": 0.07598327392755594,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0480097725,
      "avg_duration_per_task_ms": 463483.8175,
      "avg_output_tokens_per_task": 15486.0875,
      "last_run_timestamp": "2026-10-31T19:37:26.520Z"
    },
    {
      "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": 0.6992592222222223,
      "accuracy_std_dev": 0.03968313867732682,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.06891943333333334,
      "avg_duration_per_task_ms": 354863.96,
      "avg_output_tokens_per_task": 10019.886666666667,
      "last_run_timestamp": "2026-10-31T20:44:35.166Z"
    },
    {
      "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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Cloudflare",
      "accuracy": 0.6983333125,
      "accuracy_std_dev": 0.03425807589408435,
      "run_count": 16,
      "total_questions": 800,
      "avg_cost_per_task_usd": 0.127335795,
      "avg_duration_per_task_ms": 128403.7125,
      "avg_output_tokens_per_task": 8998.39,
      "last_run_timestamp": "2026-10-31T20:01:39.270Z"
    },
    {
      "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": 0.6977776666666667,
      "accuracy_std_dev": 0.011330936570096727,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.013071946666666667,
      "avg_duration_per_task_ms": 88036.78,
      "avg_output_tokens_per_task": 11978.486666666666,
      "last_run_timestamp": "2026-10-31T11:55:17.583Z"
    },
    {
      "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": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "auto-routing",
      "accuracy": 0.6970371111111111,
      "accuracy_std_dev": 0.02933193693056708,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.05237315555555556,
      "avg_duration_per_task_ms": 363352.25111111114,
      "avg_output_tokens_per_task": 9093.155555555555,
      "last_run_timestamp": "2026-10-31T22:10:13.617Z"
    },
    {
      "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": "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": 0.6963472739726028,
      "accuracy_std_dev": 0.05315590618172168,
      "run_count": 3,
      "total_questions": 146,
      "avg_cost_per_task_usd": 0.01628368493150685,
      "avg_duration_per_task_ms": 275676.23287671234,
      "avg_output_tokens_per_task": 12531.828767123288,
      "last_run_timestamp": "2026-10-31T13:29:36.200Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "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": "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": 0.6946665000000001,
      "accuracy_std_dev": 0.013597418058219743,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.02175149,
      "avg_duration_per_task_ms": 498490.026,
      "avg_output_tokens_per_task": 18620.424,
      "last_run_timestamp": "2026-10-31T19:50:01.310Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-luna-pro-20260709",
      "model_slug": "openai/gpt-5.6-luna-pro",
      "model_name": "OpenAI: GPT-5.6 Luna Pro",
      "provider_name": "Azure",
      "accuracy": 0.6933334999999999,
      "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": "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": "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": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "DigitalOcean",
      "accuracy": 0.6918933673469387,
      "accuracy_std_dev": 0.11325810487735402,
      "run_count": 12,
      "total_questions": 588,
      "avg_cost_per_task_usd": 0.3106834897959183,
      "avg_duration_per_task_ms": 1590211.7602040817,
      "avg_output_tokens_per_task": 51719.284013605444,
      "last_run_timestamp": "2026-10-31T12:54:39.225Z"
    },
    {
      "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": 0.6914284285714286,
      "accuracy_std_dev": 0.02217231732110459,
      "run_count": 28,
      "total_questions": 1400,
      "avg_cost_per_task_usd": 0.10640983571428575,
      "avg_duration_per_task_ms": 75797.385,
      "avg_output_tokens_per_task": 9708.589285714286,
      "last_run_timestamp": "2026-10-31T19:55:55.295Z"
    },
    {
      "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": 0.6909721458333333,
      "accuracy_std_dev": 0.0024635000000000074,
      "run_count": 2,
      "total_questions": 96,
      "avg_cost_per_task_usd": 0.0246803125,
      "avg_duration_per_task_ms": 48638.177083333336,
      "avg_output_tokens_per_task": 11672.458333333334,
      "last_run_timestamp": "2026-10-31T11:12:51.723Z"
    },
    {
      "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": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Minimax",
      "accuracy": 0.69,
      "accuracy_std_dev": 0.009999999999999953,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01750064,
      "avg_duration_per_task_ms": 190406.46,
      "avg_output_tokens_per_task": 15714.76,
      "last_run_timestamp": "2026-10-31T12:31:30.661Z"
    },
    {
      "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": "PrimeIntellect",
      "accuracy": 0.69,
      "accuracy_std_dev": 0.016667000000000043,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01877389,
      "avg_duration_per_task_ms": 125629.58,
      "avg_output_tokens_per_task": 15732.46,
      "last_run_timestamp": "2026-10-31T11:20:36.173Z"
    },
    {
      "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": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Tenstorrent",
      "accuracy": 0.6877778333333333,
      "accuracy_std_dev": 0.1755169025663119,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.0163179,
      "avg_duration_per_task_ms": 360518.31666666665,
      "avg_output_tokens_per_task": 11107.656666666666,
      "last_run_timestamp": "2026-10-31T11:27:39.395Z"
    },
    {
      "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": 0.6874072222222223,
      "accuracy_std_dev": 0.042449001759700566,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.016434591111111112,
      "avg_duration_per_task_ms": 373578.1911111111,
      "avg_output_tokens_per_task": 19878.99111111111,
      "last_run_timestamp": "2026-10-31T20:18:22.353Z"
    },
    {
      "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": 0.6873796242774566,
      "accuracy_std_dev": 0.023430359614685144,
      "run_count": 7,
      "total_questions": 346,
      "avg_cost_per_task_usd": 0.021091329479768786,
      "avg_duration_per_task_ms": 336802.3034682081,
      "avg_output_tokens_per_task": 17067.51156069364,
      "last_run_timestamp": "2026-10-31T21:08:27.571Z"
    },
    {
      "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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Friendli",
      "accuracy": 0.6866666923076923,
      "accuracy_std_dev": 0.06431750303513466,
      "run_count": 13,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.11028715384615387,
      "avg_duration_per_task_ms": 78688.8476923077,
      "avg_output_tokens_per_task": 9043.872307692307,
      "last_run_timestamp": "2026-10-31T19:47:27.755Z"
    },
    {
      "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": "GMICloud",
      "accuracy": 0.6861537692307692,
      "accuracy_std_dev": 0.04555144565493437,
      "run_count": 13,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.09695205384615385,
      "avg_duration_per_task_ms": 189917.99538461538,
      "avg_output_tokens_per_task": 7426.601538461538,
      "last_run_timestamp": "2026-10-31T19:10:43.801Z"
    },
    {
      "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": "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": "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": 0.6833334999999999,
      "accuracy_std_dev": 0.023333500000000007,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.018626550000000002,
      "avg_duration_per_task_ms": 239747.89,
      "avg_output_tokens_per_task": 15949.04,
      "last_run_timestamp": "2026-10-31T19:36:36.917Z"
    },
    {
      "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": "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": "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": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "CoreWeave",
      "accuracy": 0.68125,
      "accuracy_std_dev": 0.0202715728731147,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0174918175,
      "avg_duration_per_task_ms": 123091.5525,
      "avg_output_tokens_per_task": 21340.095,
      "last_run_timestamp": "2026-10-31T11:46:38.147Z"
    },
    {
      "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": "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": "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": 0.6799998571428572,
      "accuracy_std_dev": 0.0198405229080613,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.015568199999999999,
      "avg_duration_per_task_ms": 259045.84285714285,
      "avg_output_tokens_per_task": 18504.7,
      "last_run_timestamp": "2026-10-31T20:09:24.664Z"
    },
    {
      "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": 0.6793939999999999,
      "accuracy_std_dev": 0.03428396622064286,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.017193465454545455,
      "avg_duration_per_task_ms": 303188.84545454546,
      "avg_output_tokens_per_task": 19789.663636363635,
      "last_run_timestamp": "2026-10-31T21:04:41.015Z"
    },
    {
      "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": 0.678333,
      "accuracy_std_dev": 0.0050000000000000044,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.0069528,
      "avg_duration_per_task_ms": 471037.61,
      "avg_output_tokens_per_task": 19230.47,
      "last_run_timestamp": "2026-10-31T13:27:50.961Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "Fireworks",
      "accuracy": 0.6777776666666667,
      "accuracy_std_dev": 0.013698571612479227,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.0340762,
      "avg_duration_per_task_ms": 244093.44,
      "avg_output_tokens_per_task": 27086.806666666667,
      "last_run_timestamp": "2026-09-19T08:01:53.393Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.7-flash-20260119",
      "model_slug": "z-ai/glm-4.7-flash",
      "model_name": "Z.ai: GLM 4.7 Flash",
      "provider_name": "Novita",
      "accuracy": 0.67666675,
      "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": "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": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "DigitalOcean",
      "accuracy": 0.6755556666666667,
      "accuracy_std_dev": 0.00831485358994507,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.014157986666666667,
      "avg_duration_per_task_ms": 117427.1,
      "avg_output_tokens_per_task": 13587.66,
      "last_run_timestamp": "2026-10-31T11:44:53.544Z"
    },
    {
      "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": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "GMICloud",
      "accuracy": 0.6755554444444445,
      "accuracy_std_dev": 0.10270425478377884,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.017111031111111112,
      "avg_duration_per_task_ms": 257153.8,
      "avg_output_tokens_per_task": 10465.493333333334,
      "last_run_timestamp": "2026-10-31T19:58:02.198Z"
    },
    {
      "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": "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": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Tenstorrent",
      "accuracy": 0.6733333333333333,
      "accuracy_std_dev": 0.01962587106064057,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.012727326666666665,
      "avg_duration_per_task_ms": 155983.29333333333,
      "avg_output_tokens_per_task": 15660.926666666666,
      "last_run_timestamp": "2026-10-31T12:26:47.334Z"
    },
    {
      "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": "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": 0.6713710161290322,
      "accuracy_std_dev": 0.029936692690409203,
      "run_count": 10,
      "total_questions": 496,
      "avg_cost_per_task_usd": 0.02159695766129032,
      "avg_duration_per_task_ms": 287822.4233870968,
      "avg_output_tokens_per_task": 16757.65927419355,
      "last_run_timestamp": "2026-10-31T20:14:20.109Z"
    },
    {
      "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": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Friendli",
      "accuracy": 0.6693334,
      "accuracy_std_dev": 0.07285906090281427,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.09112218200000001,
      "avg_duration_per_task_ms": 124279.178,
      "avg_output_tokens_per_task": 8535.108,
      "last_run_timestamp": "2026-10-31T19:19:57.503Z"
    },
    {
      "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": 0.6690909090909091,
      "accuracy_std_dev": 0.14703195632270777,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.12102286727272728,
      "avg_duration_per_task_ms": 359186.4090909091,
      "avg_output_tokens_per_task": 8810.841818181818,
      "last_run_timestamp": "2026-10-31T19:48:58.813Z"
    },
    {
      "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": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "DigitalOcean",
      "accuracy": 0.6684915255474452,
      "accuracy_std_dev": 0.1213625557738751,
      "run_count": 11,
      "total_questions": 548,
      "avg_cost_per_task_usd": 0.040750651459854015,
      "avg_duration_per_task_ms": 1768695.9981751824,
      "avg_output_tokens_per_task": 66240.04562043796,
      "last_run_timestamp": "2026-10-31T11:55:35.800Z"
    },
    {
      "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": "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Sail Research",
      "accuracy": 0.666667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.013241780000000002,
      "avg_duration_per_task_ms": 134332.46,
      "avg_output_tokens_per_task": 19207.38,
      "last_run_timestamp": "2026-10-31T19:46:36.834Z"
    },
    {
      "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": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.6666666666666666,
      "accuracy_std_dev": 0.0524933858267454,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.03491572666666667,
      "avg_duration_per_task_ms": 274075.64666666667,
      "avg_output_tokens_per_task": 23860.97333333333,
      "last_run_timestamp": "2026-10-31T19:43:40.322Z"
    },
    {
      "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": 0.6666665,
      "accuracy_std_dev": 0.006666499999999964,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.013174700000000001,
      "avg_duration_per_task_ms": 181950.26,
      "avg_output_tokens_per_task": 13269.31,
      "last_run_timestamp": "2026-10-31T12:42:39.997Z"
    },
    {
      "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": "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": 0.6661110555555556,
      "accuracy_std_dev": 0.029580425001590006,
      "run_count": 36,
      "total_questions": 1800,
      "avg_cost_per_task_usd": 0.12436681222222222,
      "avg_duration_per_task_ms": 119159.24222222222,
      "avg_output_tokens_per_task": 12659.702222222222,
      "last_run_timestamp": "2026-10-31T12:26:00.225Z"
    },
    {
      "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": 0.66583325,
      "accuracy_std_dev": 0.03639120582211448,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.006095729999999999,
      "avg_duration_per_task_ms": 325369.79,
      "avg_output_tokens_per_task": 18643.27,
      "last_run_timestamp": "2026-10-31T19:46:03.854Z"
    },
    {
      "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": 0.6616665833333334,
      "accuracy_std_dev": 0.0293288743461068,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.11567571166666667,
      "avg_duration_per_task_ms": 123049.12333333334,
      "avg_output_tokens_per_task": 12632.496666666666,
      "last_run_timestamp": "2026-10-31T12:42:18.189Z"
    },
    {
      "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "DekaLLM",
      "accuracy": 0.6604760000000001,
      "accuracy_std_dev": 0.025350253033394797,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.008717225714285714,
      "avg_duration_per_task_ms": 574345.4257142857,
      "avg_output_tokens_per_task": 17634.762857142858,
      "last_run_timestamp": "2026-10-31T20:07:05.983Z"
    },
    {
      "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": 0.6600000833333333,
      "accuracy_std_dev": 0.02652737928838283,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.11395726666666665,
      "avg_duration_per_task_ms": 127166.50666666667,
      "avg_output_tokens_per_task": 12326.945,
      "last_run_timestamp": "2026-10-31T11:44:30.576Z"
    },
    {
      "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": "anthropic/claude-4.5-haiku-20251001",
      "model_slug": "anthropic/claude-haiku-4.5",
      "model_name": "Anthropic: Claude Haiku 4.5",
      "provider_name": "Google",
      "accuracy": 0.6580179999999999,
      "accuracy_std_dev": 0.027893658429732626,
      "run_count": 37,
      "total_questions": 1850,
      "avg_cost_per_task_usd": 0.12081404162162164,
      "avg_duration_per_task_ms": 121162.63837837838,
      "avg_output_tokens_per_task": 12580.847027027026,
      "last_run_timestamp": "2026-10-31T12:17:41.489Z"
    },
    {
      "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": "Parasail",
      "accuracy": 0.6579165,
      "accuracy_std_dev": 0.013326680635101876,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.023098354999999997,
      "avg_duration_per_task_ms": 154244.335,
      "avg_output_tokens_per_task": 17007.6275,
      "last_run_timestamp": "2026-10-31T18:35:38.514Z"
    },
    {
      "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "AkashML",
      "accuracy": 0.6564246703910614,
      "accuracy_std_dev": 0.022405218838489775,
      "run_count": 18,
      "total_questions": 895,
      "avg_cost_per_task_usd": 0.01014628938547486,
      "avg_duration_per_task_ms": 302924.7284916201,
      "avg_output_tokens_per_task": 17044.198882681565,
      "last_run_timestamp": "2026-10-31T19:05:31.409Z"
    },
    {
      "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": 0.6562239298245615,
      "accuracy_std_dev": 0.018946289616967236,
      "run_count": 8,
      "total_questions": 399,
      "avg_cost_per_task_usd": 0.01692722807017544,
      "avg_duration_per_task_ms": 312804.7944862155,
      "avg_output_tokens_per_task": 22811.754385964912,
      "last_run_timestamp": "2026-10-31T12:28:49.050Z"
    },
    {
      "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": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "DeepInfra",
      "accuracy": 0.6555556666666666,
      "accuracy_std_dev": 0.03807065880409451,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.021039175555555557,
      "avg_duration_per_task_ms": 248791.5977777778,
      "avg_output_tokens_per_task": 26269.908888888887,
      "last_run_timestamp": "2026-10-31T12:25:21.635Z"
    },
    {
      "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Crusoe",
      "accuracy": 0.6546664,
      "accuracy_std_dev": 0.03512514515614133,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.01491196,
      "avg_duration_per_task_ms": 86567.944,
      "avg_output_tokens_per_task": 19737.508,
      "last_run_timestamp": "2026-10-31T19:08:06.670Z"
    },
    {
      "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": "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "auto-routing",
      "accuracy": 0.6518518888888889,
      "accuracy_std_dev": 0.03247467220412599,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.012659444444444444,
      "avg_duration_per_task_ms": 517342.6377777778,
      "avg_output_tokens_per_task": 18101.524444444443,
      "last_run_timestamp": "2026-10-31T19:41:05.491Z"
    },
    {
      "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": 0.6518188195991091,
      "accuracy_std_dev": 0.02538535107955173,
      "run_count": 9,
      "total_questions": 449,
      "avg_cost_per_task_usd": 0.012973,
      "avg_duration_per_task_ms": 774691.5612472161,
      "avg_output_tokens_per_task": 26354.719376391982,
      "last_run_timestamp": "2026-10-31T19:22:53.448Z"
    },
    {
      "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": 0.65166675,
      "accuracy_std_dev": 0.028819327042585528,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.01676374,
      "avg_duration_per_task_ms": 406825.29,
      "avg_output_tokens_per_task": 27732.185,
      "last_run_timestamp": "2026-10-31T19:20:36.815Z"
    },
    {
      "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": 0.6505556666666666,
      "accuracy_std_dev": 0.024107646962230795,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.11465736166666668,
      "avg_duration_per_task_ms": 137405.46666666667,
      "avg_output_tokens_per_task": 12547.148333333333,
      "last_run_timestamp": "2026-10-31T11:42:13.882Z"
    },
    {
      "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": 0.6502015887096775,
      "accuracy_std_dev": 0.03709283547317459,
      "run_count": 10,
      "total_questions": 496,
      "avg_cost_per_task_usd": 0.01248410685483871,
      "avg_duration_per_task_ms": 1080016.7157258065,
      "avg_output_tokens_per_task": 22928.112903225807,
      "last_run_timestamp": "2026-10-31T14:30:58.017Z"
    },
    {
      "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": 0.6500001111111111,
      "accuracy_std_dev": 0.032356116448892955,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.03375578888888889,
      "avg_duration_per_task_ms": 60946.455555555556,
      "avg_output_tokens_per_task": 14806.375555555556,
      "last_run_timestamp": "2026-10-31T18:54:40.763Z"
    },
    {
      "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": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "AtlasCloud",
      "accuracy": 0.6451851111111111,
      "accuracy_std_dev": 0.12616958039827408,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.05096927111111111,
      "avg_duration_per_task_ms": 818258.1133333333,
      "avg_output_tokens_per_task": 25672.8,
      "last_run_timestamp": "2026-10-31T19:22:14.979Z"
    },
    {
      "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": "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": 0.6413691205357143,
      "accuracy_std_dev": 0.06151143594868467,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.024266383928571424,
      "avg_duration_per_task_ms": 398610.2589285714,
      "avg_output_tokens_per_task": 15573.261160714286,
      "last_run_timestamp": "2026-10-31T19:07:04.114Z"
    },
    {
      "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": "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": "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": "DeepInfra",
      "accuracy": 0.636978650507328,
      "accuracy_std_dev": 0.041368248430863674,
      "run_count": 18,
      "total_questions": 887,
      "avg_cost_per_task_usd": 0.020942316798196168,
      "avg_duration_per_task_ms": 316972.6538895152,
      "avg_output_tokens_per_task": 16888.583990980835,
      "last_run_timestamp": "2026-10-31T20:19:17.569Z"
    },
    {
      "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": 0.6355553333333334,
      "accuracy_std_dev": 0.003142853940113798,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.008414553333333331,
      "avg_duration_per_task_ms": 1275373.5733333332,
      "avg_output_tokens_per_task": 17754.266666666666,
      "last_run_timestamp": "2026-10-31T21:55:54.854Z"
    },
    {
      "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": "Phala",
      "accuracy": 0.6331658693467337,
      "accuracy_std_dev": 0.03188449882289982,
      "run_count": 4,
      "total_questions": 199,
      "avg_cost_per_task_usd": 0.03419022110552764,
      "avg_duration_per_task_ms": 256833.6432160804,
      "avg_output_tokens_per_task": 16666.94974874372,
      "last_run_timestamp": "2026-10-31T19:04:42.700Z"
    },
    {
      "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": "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "SambaNova",
      "accuracy": 0.6273407438202249,
      "accuracy_std_dev": 0.028511029028200886,
      "run_count": 9,
      "total_questions": 445,
      "avg_cost_per_task_usd": 0.04094152584269662,
      "avg_duration_per_task_ms": 84933.8,
      "avg_output_tokens_per_task": 20903.849438202247,
      "last_run_timestamp": "2026-10-31T18:59:19.181Z"
    },
    {
      "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": 0.6266666666666667,
      "accuracy_std_dev": 0.14676766724392065,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.03268331333333334,
      "avg_duration_per_task_ms": 135271.32666666666,
      "avg_output_tokens_per_task": 12787.166666666666,
      "last_run_timestamp": "2026-10-31T11:51:01.680Z"
    },
    {
      "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": 0.62,
      "accuracy_std_dev": 0.024343133492264774,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.06804065333333334,
      "avg_duration_per_task_ms": 32072.04888888889,
      "avg_output_tokens_per_task": 21507.428888888888,
      "last_run_timestamp": "2026-10-31T19:45:42.530Z"
    },
    {
      "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": 0.6161905714285714,
      "accuracy_std_dev": 0.1331358561415016,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.049797854285714274,
      "avg_duration_per_task_ms": 468411.09142857144,
      "avg_output_tokens_per_task": 9093.771428571428,
      "last_run_timestamp": "2026-10-31T19:32:40.676Z"
    },
    {
      "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": "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": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.6066666666666667,
      "accuracy_std_dev": 0.04343290554406878,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.09464461999999998,
      "avg_duration_per_task_ms": 58154.904444444444,
      "avg_output_tokens_per_task": 3592.282222222222,
      "last_run_timestamp": "2026-10-31T19:06:25.403Z"
    },
    {
      "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": "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Mancer 2",
      "accuracy": 0.59966491959799,
      "accuracy_std_dev": 0.08352485507909829,
      "run_count": 8,
      "total_questions": 398,
      "avg_cost_per_task_usd": 0.028319288944723613,
      "avg_duration_per_task_ms": 1740003.0175879397,
      "avg_output_tokens_per_task": 55612.85427135678,
      "last_run_timestamp": "2026-10-31T19:40:44.417Z"
    },
    {
      "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": "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": 0.5409356315789473,
      "accuracy_std_dev": 0.048637113976719215,
      "run_count": 8,
      "total_questions": 399,
      "avg_cost_per_task_usd": 0.00998410275689223,
      "avg_duration_per_task_ms": 184598.81453634085,
      "avg_output_tokens_per_task": 15438.807017543859,
      "last_run_timestamp": "2026-10-31T19:08:29.434Z"
    },
    {
      "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": "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": "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": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "auto-routing",
      "accuracy": 0.5330330315315315,
      "accuracy_std_dev": 0.017762103976002412,
      "run_count": 9,
      "total_questions": 444,
      "avg_cost_per_task_usd": 0.016922781531531532,
      "avg_duration_per_task_ms": 1163584.25,
      "avg_output_tokens_per_task": 75958.21846846846,
      "last_run_timestamp": "2026-10-31T19:53:39.071Z"
    },
    {
      "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": 0.5296297777777778,
      "accuracy_std_dev": 0.01914500153377888,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.06659953333333334,
      "avg_duration_per_task_ms": 283737.4666666667,
      "avg_output_tokens_per_task": 4050.251111111111,
      "last_run_timestamp": "2026-10-31T19:28:47.220Z"
    },
    {
      "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": 0.5285472267002518,
      "accuracy_std_dev": 0.03054949850959095,
      "run_count": 8,
      "total_questions": 397,
      "avg_cost_per_task_usd": 0.0026572493702770776,
      "avg_duration_per_task_ms": 371704.879093199,
      "avg_output_tokens_per_task": 6936.692695214106,
      "last_run_timestamp": "2026-10-31T19:13:20.193Z"
    },
    {
      "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": "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": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "DeepInfra",
      "accuracy": 0.4972901544715447,
      "accuracy_std_dev": 0.018392858501059588,
      "run_count": 5,
      "total_questions": 246,
      "avg_cost_per_task_usd": 0.021434772357723578,
      "avg_duration_per_task_ms": 1275972.138211382,
      "avg_output_tokens_per_task": 109100.26829268293,
      "last_run_timestamp": "2026-10-31T19:48:14.501Z"
    },
    {
      "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": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "auto-routing",
      "accuracy": 0.4715242060301507,
      "accuracy_std_dev": 0.16796757633971357,
      "run_count": 8,
      "total_questions": 398,
      "avg_cost_per_task_usd": 0.06913073618090453,
      "avg_duration_per_task_ms": 563511.0577889447,
      "avg_output_tokens_per_task": 9865.51256281407,
      "last_run_timestamp": "2026-10-31T19:59:00.038Z"
    },
    {
      "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-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Together",
      "accuracy": 0.46555555555555556,
      "accuracy_std_dev": 0.1375265612956365,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.027833837777777776,
      "avg_duration_per_task_ms": 236061.30666666667,
      "avg_output_tokens_per_task": 18727.45111111111,
      "last_run_timestamp": "2026-10-31T19:46:58.356Z"
    },
    {
      "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": "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": "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": "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": 0.4497487512562814,
      "accuracy_std_dev": 0.043480269304018805,
      "run_count": 8,
      "total_questions": 398,
      "avg_cost_per_task_usd": 0.015482552763819097,
      "avg_duration_per_task_ms": 129064.97989949748,
      "avg_output_tokens_per_task": 3318.743718592965,
      "last_run_timestamp": "2026-10-31T18:58:15.016Z"
    },
    {
      "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": 0.4490197823529411,
      "accuracy_std_dev": 0.021932270978759,
      "run_count": 7,
      "total_questions": 340,
      "avg_cost_per_task_usd": 0.05236282352941176,
      "avg_duration_per_task_ms": 148084.36176470589,
      "avg_output_tokens_per_task": 72432.0411764706,
      "last_run_timestamp": "2026-10-31T19:03:34.962Z"
    },
    {
      "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": "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": 0.44458325000000004,
      "accuracy_std_dev": 0.02852565974587617,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.039974685,
      "avg_duration_per_task_ms": 271844.9625,
      "avg_output_tokens_per_task": 3550.7,
      "last_run_timestamp": "2026-10-31T19:56:10.591Z"
    },
    {
      "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": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Parasail",
      "accuracy": 0.428333375,
      "accuracy_std_dev": 0.02990715029862215,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.007658255,
      "avg_duration_per_task_ms": 180128.6975,
      "avg_output_tokens_per_task": 4519.6725,
      "last_run_timestamp": "2026-10-31T19:46:05.351Z"
    },
    {
      "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": "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": 0.4273334,
      "accuracy_std_dev": 0.03115546180110319,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.024095914,
      "avg_duration_per_task_ms": 401744.36,
      "avg_output_tokens_per_task": 3568.138,
      "last_run_timestamp": "2026-10-31T19:22:37.207Z"
    },
    {
      "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": "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": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Chutes",
      "accuracy": 0.42154647297297293,
      "accuracy_std_dev": 0.02867673607820049,
      "run_count": 9,
      "total_questions": 444,
      "avg_cost_per_task_usd": 0.0049315878378378385,
      "avg_duration_per_task_ms": 343548.7364864865,
      "avg_output_tokens_per_task": 4828.531531531532,
      "last_run_timestamp": "2026-10-31T19:55:29.787Z"
    },
    {
      "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": "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": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "AkashML",
      "accuracy": 0.401667,
      "accuracy_std_dev": 0.025000000000000022,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01796061,
      "avg_duration_per_task_ms": 1329879.6,
      "avg_output_tokens_per_task": 125682.59,
      "last_run_timestamp": "2026-10-31T20:05:39.534Z"
    },
    {
      "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/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "auto-routing",
      "accuracy": 0.3948148888888889,
      "accuracy_std_dev": 0.12040172902084868,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.021518880000000004,
      "avg_duration_per_task_ms": 219251.11555555556,
      "avg_output_tokens_per_task": 3342.9933333333333,
      "last_run_timestamp": "2026-10-31T19:18:14.606Z"
    },
    {
      "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/qwen3-30b-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": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "DeepInfra",
      "accuracy": 0.37745103619909504,
      "accuracy_std_dev": 0.12198880856754195,
      "run_count": 9,
      "total_questions": 442,
      "avg_cost_per_task_usd": 0.05260786199095024,
      "avg_duration_per_task_ms": 782967.8257918552,
      "avg_output_tokens_per_task": 9060.183257918552,
      "last_run_timestamp": "2026-10-31T19:55:08.491Z"
    },
    {
      "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": 0.37416568080094226,
      "accuracy_std_dev": 0.022198267854897723,
      "run_count": 17,
      "total_questions": 849,
      "avg_cost_per_task_usd": 0.013129518256772676,
      "avg_duration_per_task_ms": 40276.608951707894,
      "avg_output_tokens_per_task": 755.5041224970554,
      "last_run_timestamp": "2026-10-31T19:10:53.562Z"
    },
    {
      "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": "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": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.3542132751677853,
      "accuracy_std_dev": 0.13859537037210243,
      "run_count": 18,
      "total_questions": 894,
      "avg_cost_per_task_usd": 0.036926778523489925,
      "avg_duration_per_task_ms": 548291.3724832215,
      "avg_output_tokens_per_task": 133944.0715883669,
      "last_run_timestamp": "2026-10-31T20:07:06.512Z"
    },
    {
      "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": 0.3518517777777778,
      "accuracy_std_dev": 0.03277707804337855,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.09174267777777778,
      "avg_duration_per_task_ms": 26796.88,
      "avg_output_tokens_per_task": 1065.5333333333333,
      "last_run_timestamp": "2026-10-31T18:48:48.352Z"
    },
    {
      "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": "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": 0.3281016860986547,
      "accuracy_std_dev": 0.044158645918249494,
      "run_count": 9,
      "total_questions": 446,
      "avg_cost_per_task_usd": 0.047258513452914805,
      "avg_duration_per_task_ms": 86946.02466367713,
      "avg_output_tokens_per_task": 1321.1367713004483,
      "last_run_timestamp": "2026-10-31T19:07:35.559Z"
    },
    {
      "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": "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "BaseTen",
      "accuracy": 0.27745670520231214,
      "accuracy_std_dev": 0.31828543520589203,
      "run_count": 7,
      "total_questions": 346,
      "avg_cost_per_task_usd": 0.009253416184971098,
      "avg_duration_per_task_ms": 52282.53468208093,
      "avg_output_tokens_per_task": 9707.71098265896,
      "last_run_timestamp": "2026-10-31T19:36:09.009Z"
    },
    {
      "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": "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": "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Nebius",
      "accuracy": 0.20799989999999996,
      "accuracy_std_dev": 0.23746439227827398,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.06683555599999999,
      "avg_duration_per_task_ms": 145217.8,
      "avg_output_tokens_per_task": 16818.696,
      "last_run_timestamp": "2026-10-31T19:46:05.356Z"
    },
    {
      "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": 0.19360285858585857,
      "accuracy_std_dev": 0.23547614599731753,
      "run_count": 8,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.008070734848484848,
      "avg_duration_per_task_ms": 691070.9520202021,
      "avg_output_tokens_per_task": 53901.46464646464,
      "last_run_timestamp": "2026-10-31T19:13:59.240Z"
    },
    {
      "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"
    }
  ]
}
