{
  "dataset": "benchmark-leaderboard",
  "title": "OpenRouter benchmark leaderboard",
  "description": "One row per benchmark, model version, and provider: the aggregate values displayed on the OpenRouter leaderboard pages (GPQA Diamond, tau2-bench airline). Derived from the benchmark-runs dataset.",
  "schema_version": "v1",
  "snapshot_date": "2026-10-04",
  "retrieved_at": "2026-10-04T01:45:32.205Z",
  "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": 2110,
  "rows": [
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.8-flash-20260902",
      "model_slug": "google/gemini-3.8-flash",
      "model_name": "Google: Gemini 3.8 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.95637,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 191,
      "avg_cost_per_task_usd": 0.06822796858638744,
      "avg_duration_per_task_ms": 132197.8062827225,
      "avg_output_tokens_per_task": 18049.068062827224,
      "last_run_timestamp": "2026-10-01T08:31:33.481Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-astra-pro-20260903",
      "model_slug": "openai/gpt-6-astra-pro",
      "model_name": "OpenAI: GPT-6 Astra Pro",
      "provider_name": "auto-routing",
      "accuracy": 0.954545,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.332555398989899,
      "avg_duration_per_task_ms": 75446.5505050505,
      "avg_output_tokens_per_task": 4530.388888888889,
      "last_run_timestamp": "2026-10-01T08:27:35.414Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.8-flash-20260902",
      "model_slug": "google/gemini-3.8-flash",
      "model_name": "Google: Gemini 3.8 Flash",
      "provider_name": "Google AI Studio",
      "accuracy": 0.952862,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.07423223737373738,
      "avg_duration_per_task_ms": 86748.28787878787,
      "avg_output_tokens_per_task": 19646.409090909092,
      "last_run_timestamp": "2026-10-01T08:29:11.347Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.1-pro-preview-20260219",
      "model_slug": "google/gemini-3.1-pro-preview",
      "model_name": "Google: Gemini 3.1 Pro Preview",
      "provider_name": "Google",
      "accuracy": 0.95202,
      "accuracy_std_dev": 0.0008420000000000094,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.20364760606060606,
      "avg_duration_per_task_ms": 135706.18686868687,
      "avg_output_tokens_per_task": 16847.588383838385,
      "last_run_timestamp": "2026-10-01T19:14:26.157Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-astra-pro-20260903",
      "model_slug": "openai/gpt-6-astra-pro",
      "model_name": "OpenAI: GPT-6 Astra Pro",
      "provider_name": "OpenAI",
      "accuracy": 0.9511785,
      "accuracy_std_dev": 0.0016835000000000044,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.33620659090909094,
      "avg_duration_per_task_ms": 70544.52272727272,
      "avg_output_tokens_per_task": 4588.830808080808,
      "last_run_timestamp": "2026-10-02T23:23:01.262Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-astra-pro-20260903",
      "model_slug": "openai/gpt-6-astra-pro",
      "model_name": "OpenAI: GPT-6 Astra Pro",
      "provider_name": "Azure",
      "accuracy": 0.949495,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.35495546717171716,
      "avg_duration_per_task_ms": 72102.17676767676,
      "avg_output_tokens_per_task": 4622.60101010101,
      "last_run_timestamp": "2026-10-01T08:29:03.480Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.1-pro-preview-20260219",
      "model_slug": "google/gemini-3.1-pro-preview",
      "model_name": "Google: Gemini 3.1 Pro Preview",
      "provider_name": "Google AI Studio",
      "accuracy": 0.9483723333333332,
      "accuracy_std_dev": 0.004827144451490511,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.20306945454545455,
      "avg_duration_per_task_ms": 134835.03535353535,
      "avg_output_tokens_per_task": 16798.409090909092,
      "last_run_timestamp": "2026-10-01T19:12:42.155Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6.1-sol-20260929",
      "model_slug": "openai/gpt-6.1-sol",
      "model_name": "OpenAI: GPT-6.1 Sol",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.947811,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.021931015151515153,
      "avg_duration_per_task_ms": 68600.9191919192,
      "avg_output_tokens_per_task": 1835.4242424242425,
      "last_run_timestamp": "2026-10-03T01:13:05.002Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-astra-20260903",
      "model_slug": "openai/gpt-6-astra",
      "model_name": "OpenAI: GPT-6 Astra",
      "provider_name": "OpenAI",
      "accuracy": 0.9472503333333333,
      "accuracy_std_dev": 0.0015872190215034077,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.11162585016835018,
      "avg_duration_per_task_ms": 48234.978114478115,
      "avg_output_tokens_per_task": 2076.191919191919,
      "last_run_timestamp": "2026-10-02T23:16:05.967Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.8-flash-20260902",
      "model_slug": "google/gemini-3.8-flash",
      "model_name": "Google: Gemini 3.8 Flash",
      "provider_name": "Google",
      "accuracy": 0.947009,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 195,
      "avg_cost_per_task_usd": 0.07090916923076923,
      "avg_duration_per_task_ms": 142045.61538461538,
      "avg_output_tokens_per_task": 18762.907692307694,
      "last_run_timestamp": "2026-10-01T08:30:23.202Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "sakana/fugu-ultra",
      "model_slug": "sakana/fugu-ultra",
      "model_name": "Fugu Ultra",
      "provider_name": "auto-routing",
      "accuracy": 0.9461279999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 1.148739303030303,
      "avg_duration_per_task_ms": 328779.10101010103,
      "avg_output_tokens_per_task": 33271.606060606064,
      "last_run_timestamp": "2026-09-15T01:07:47.987Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.1-pro-preview-20260219",
      "model_slug": "google/gemini-3.1-pro-preview",
      "model_name": "Google: Gemini 3.1 Pro Preview",
      "provider_name": "auto-routing",
      "accuracy": 0.945286,
      "accuracy_std_dev": 0.0008420000000000094,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.20203182828282829,
      "avg_duration_per_task_ms": 133595.55303030304,
      "avg_output_tokens_per_task": 16712.858585858587,
      "last_run_timestamp": "2026-10-01T19:16:14.342Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-astra-20260903",
      "model_slug": "openai/gpt-6-astra",
      "model_name": "OpenAI: GPT-6 Astra",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.945286,
      "accuracy_std_dev": 0.0008420000000000094,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.11987038888888889,
      "avg_duration_per_task_ms": 35572.169191919194,
      "avg_output_tokens_per_task": 2020.9419191919192,
      "last_run_timestamp": "2026-10-01T08:24:02.411Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-astra-20260903",
      "model_slug": "openai/gpt-6-astra",
      "model_name": "OpenAI: GPT-6 Astra",
      "provider_name": "auto-routing",
      "accuracy": 0.944444,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.11333201010101009,
      "avg_duration_per_task_ms": 57056.792929292926,
      "avg_output_tokens_per_task": 2108.121212121212,
      "last_run_timestamp": "2026-10-01T08:25:16.514Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6.1-sol-20260929",
      "model_slug": "openai/gpt-6.1-sol",
      "model_name": "OpenAI: GPT-6.1 Sol",
      "provider_name": "OpenAI",
      "accuracy": 0.944444,
      "accuracy_std_dev": 0.0027491439879836154,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.020607563973063972,
      "avg_duration_per_task_ms": 47519.37542087542,
      "avg_output_tokens_per_task": 1902.2373737373737,
      "last_run_timestamp": "2026-10-02T23:20:22.253Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6.1-sol-20260929",
      "model_slug": "openai/gpt-6.1-sol",
      "model_name": "OpenAI: GPT-6.1 Sol",
      "provider_name": "auto-routing",
      "accuracy": 0.944444,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.02136650505050505,
      "avg_duration_per_task_ms": 52411.32323232323,
      "avg_output_tokens_per_task": 1978.1313131313132,
      "last_run_timestamp": "2026-10-01T08:24:51.870Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-astra-20260903",
      "model_slug": "openai/gpt-6-astra",
      "model_name": "OpenAI: GPT-6 Astra",
      "provider_name": "Azure",
      "accuracy": 0.9431713315363881,
      "accuracy_std_dev": 0.006147978549897494,
      "run_count": 4,
      "total_questions": 742,
      "avg_cost_per_task_usd": 0.10982969407008086,
      "avg_duration_per_task_ms": 117613.10242587602,
      "avg_output_tokens_per_task": 1949.2371967654988,
      "last_run_timestamp": "2026-10-01T08:26:36.876Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.7-flash-20260813",
      "model_slug": "google/gemini-3.7-flash",
      "model_name": "Google: Gemini 3.7 Flash",
      "provider_name": "Google AI Studio",
      "accuracy": 0.9427606666666666,
      "accuracy_std_dev": 0.0036369299629708685,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.029872336700336698,
      "avg_duration_per_task_ms": 28492.575757575756,
      "avg_output_tokens_per_task": 7819.474747474747,
      "last_run_timestamp": "2026-10-01T16:16:03.631Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "typesafe/jev-router",
      "model_slug": "typesafe/jev-router",
      "model_name": "TypeSafe: Jev Router",
      "provider_name": "auto-routing",
      "accuracy": 0.939394,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.015926459595959598,
      "avg_duration_per_task_ms": 24665.858585858587,
      "avg_output_tokens_per_task": 2219.030303030303,
      "last_run_timestamp": "2026-09-28T18:11:34.031Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-sol-pro-20260709",
      "model_slug": "openai/gpt-5.6-sol-pro",
      "model_name": "OpenAI: GPT-5.6 Sol Pro",
      "provider_name": "auto-routing",
      "accuracy": 0.9393935,
      "accuracy_std_dev": 0.0016835000000000044,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.29793838636363634,
      "avg_duration_per_task_ms": 106861.7196969697,
      "avg_output_tokens_per_task": 11692.656565656565,
      "last_run_timestamp": "2026-10-01T16:10:33.208Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.7-flash-20260813",
      "model_slug": "google/gemini-3.7-flash",
      "model_name": "Google: Gemini 3.7 Flash",
      "provider_name": "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": "google/gemini-3.7-flash-20260813",
      "model_slug": "google/gemini-3.7-flash",
      "model_name": "Google: Gemini 3.7 Flash",
      "provider_name": "Google",
      "accuracy": 0.937943329787234,
      "accuracy_std_dev": 0.011200313606720516,
      "run_count": 3,
      "total_questions": 564,
      "avg_cost_per_task_usd": 0.023761370567375886,
      "avg_duration_per_task_ms": 52418.78723404255,
      "avg_output_tokens_per_task": 7636.52304964539,
      "last_run_timestamp": "2026-10-01T16:16:37.307Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.5-20260423",
      "model_slug": "openai/gpt-5.5",
      "model_name": "OpenAI: GPT-5.5",
      "provider_name": "auto-routing",
      "accuracy": 0.9371493333333333,
      "accuracy_std_dev": 0.004199386807089248,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.30925494444444446,
      "avg_duration_per_task_ms": 152328.37037037036,
      "avg_output_tokens_per_task": 10179.092592592593,
      "last_run_timestamp": "2026-10-01T19:12:11.505Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.5-20260423",
      "model_slug": "openai/gpt-5.5",
      "model_name": "OpenAI: GPT-5.5",
      "provider_name": "OpenAI",
      "accuracy": 0.9360266666666668,
      "accuracy_std_dev": 0.0036369299629708854,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.3141816919191919,
      "avg_duration_per_task_ms": 142474.1195286195,
      "avg_output_tokens_per_task": 10342.412457912458,
      "last_run_timestamp": "2026-10-02T23:25:41.614Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Parasail",
      "accuracy": 0.9339223333333333,
      "accuracy_std_dev": 0.0062945671724820625,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.13663711026936026,
      "avg_duration_per_task_ms": 176815.70622895623,
      "avg_output_tokens_per_task": 8900.839225589225,
      "last_run_timestamp": "2026-10-03T10:36:57.552Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-sol-pro-20260709",
      "model_slug": "openai/gpt-5.6-sol-pro",
      "model_name": "OpenAI: GPT-5.6 Sol Pro",
      "provider_name": "Azure",
      "accuracy": 0.9337823333333333,
      "accuracy_std_dev": 0.008398773614178557,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.3842955404040404,
      "avg_duration_per_task_ms": 82918.1936026936,
      "avg_output_tokens_per_task": 11996.59090909091,
      "last_run_timestamp": "2026-10-01T16:36:19.657Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-sol-pro-20260709",
      "model_slug": "openai/gpt-5.6-sol-pro",
      "model_name": "OpenAI: GPT-5.6 Sol Pro",
      "provider_name": "OpenAI",
      "accuracy": 0.9332206666666667,
      "accuracy_std_dev": 0.00965467639137749,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.24476150168350166,
      "avg_duration_per_task_ms": 101605.01851851853,
      "avg_output_tokens_per_task": 11508.424242424242,
      "last_run_timestamp": "2026-10-02T23:21:48.863Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "sakana/fugu-ultra-v2",
      "model_slug": "sakana/fugu-ultra-v2",
      "model_name": "Fugu Ultra V2",
      "provider_name": "auto-routing",
      "accuracy": 0.93266,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.5025571414141414,
      "avg_duration_per_task_ms": 194915.41919191918,
      "avg_output_tokens_per_task": 13806.227272727272,
      "last_run_timestamp": "2026-09-15T01:07:48.052Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.5-20260423",
      "model_slug": "openai/gpt-5.5",
      "model_name": "OpenAI: GPT-5.5",
      "provider_name": "Azure",
      "accuracy": 0.9326596250000001,
      "accuracy_std_dev": 0.00862529761714779,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.32610833964646463,
      "avg_duration_per_task_ms": 161836.85416666666,
      "avg_output_tokens_per_task": 10112.390782828283,
      "last_run_timestamp": "2026-10-01T19:14:07.472Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-sonnet-5.5-20260928",
      "model_slug": "anthropic/claude-sonnet-5.5",
      "model_name": "Anthropic: Claude Sonnet 5.5",
      "provider_name": "Anthropic",
      "accuracy": 0.9326595000000001,
      "accuracy_std_dev": 0.0016835000000000044,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.02474653787878788,
      "avg_duration_per_task_ms": 20002.444444444445,
      "avg_output_tokens_per_task": 2234.5,
      "last_run_timestamp": "2026-10-03T09:05:12.762Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.6-flash-20260721",
      "model_slug": "google/gemini-3.6-flash",
      "model_name": "Google: Gemini 3.6 Flash",
      "provider_name": "Google AI Studio",
      "accuracy": 0.9325395714285714,
      "accuracy_std_dev": 0.004401689845956867,
      "run_count": 4,
      "total_questions": 756,
      "avg_cost_per_task_usd": 0.06429072222222224,
      "avg_duration_per_task_ms": 63071.10317460317,
      "avg_output_tokens_per_task": 11529.871693121693,
      "last_run_timestamp": "2026-10-01T19:12:42.162Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-sonnet-5.5-20260928",
      "model_slug": "anthropic/claude-sonnet-5.5",
      "model_name": "Anthropic: Claude Sonnet 5.5",
      "provider_name": "Google",
      "accuracy": 0.931257,
      "accuracy_std_dev": 0.00924235327175929,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.026367194444444447,
      "avg_duration_per_task_ms": 19064.280303030304,
      "avg_output_tokens_per_task": 2249.5185185185187,
      "last_run_timestamp": "2026-10-03T09:39:59.864Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.6-flash-20260721",
      "model_slug": "google/gemini-3.6-flash",
      "model_name": "Google: Gemini 3.6 Flash",
      "provider_name": "Google",
      "accuracy": 0.930415,
      "accuracy_std_dev": 0.004827415699246697,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.06216489141414142,
      "avg_duration_per_task_ms": 64761.851851851854,
      "avg_output_tokens_per_task": 11730.429292929293,
      "last_run_timestamp": "2026-10-01T19:12:42.153Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "x-ai/grok-4.6-20260810",
      "model_slug": "x-ai/grok-4.6",
      "model_name": "SpaceXAI: Grok 4.6",
      "provider_name": "auto-routing",
      "accuracy": 0.928451,
      "accuracy_std_dev": 0.004209000000000018,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.12326839393939394,
      "avg_duration_per_task_ms": 361758.7878787879,
      "avg_output_tokens_per_task": 20322.222222222223,
      "last_run_timestamp": "2026-10-01T16:09:08.142Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Fireworks",
      "accuracy": 0.9278394565075213,
      "accuracy_std_dev": 0.008041238261262993,
      "run_count": 25,
      "total_questions": 4587,
      "avg_cost_per_task_usd": 0.1608165642031829,
      "avg_duration_per_task_ms": 149318.05646391978,
      "avg_output_tokens_per_task": 8864.42119032047,
      "last_run_timestamp": "2026-10-03T10:48:13.331Z"
    },
    {
      "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": 0.9277417729032257,
      "accuracy_std_dev": 0.015966818205265555,
      "run_count": 5,
      "total_questions": 775,
      "avg_cost_per_task_usd": 0.09847772,
      "avg_duration_per_task_ms": 244962.84516129032,
      "avg_output_tokens_per_task": 7582.7083870967745,
      "last_run_timestamp": "2026-10-03T10:05:38.767Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.6-flash-20260721",
      "model_slug": "google/gemini-3.6-flash",
      "model_name": "Google: Gemini 3.6 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.9264870000000001,
      "accuracy_std_dev": 0.004418715424192887,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.05958464141414141,
      "avg_duration_per_task_ms": 58983.43771043771,
      "avg_output_tokens_per_task": 11740.584175084176,
      "last_run_timestamp": "2026-10-01T19:11:04.522Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Relace",
      "accuracy": 0.926400137295082,
      "accuracy_std_dev": 0.007132075389394045,
      "run_count": 5,
      "total_questions": 976,
      "avg_cost_per_task_usd": 0.09721093032786886,
      "avg_duration_per_task_ms": 197311.04713114753,
      "avg_output_tokens_per_task": 8911.119877049181,
      "last_run_timestamp": "2026-10-03T10:09:17.601Z"
    },
    {
      "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": 0.925926,
      "accuracy_std_dev": 0.003367000000000009,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.025347974747474744,
      "avg_duration_per_task_ms": 21521.87373737374,
      "avg_output_tokens_per_task": 2298.931818181818,
      "last_run_timestamp": "2026-10-03T09:35:22.953Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.5-flash-20260519",
      "model_slug": "google/gemini-3.5-flash",
      "model_name": "Google: Gemini 3.5 Flash",
      "provider_name": "Google AI Studio",
      "accuracy": 0.925926,
      "accuracy_std_dev": 0.005498287975967322,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.1441699393939394,
      "avg_duration_per_task_ms": 61658.986531986535,
      "avg_output_tokens_per_task": 15894.8367003367,
      "last_run_timestamp": "2026-10-01T19:12:44.398Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.925926,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.14454752525252526,
      "avg_duration_per_task_ms": 66158.0808080808,
      "avg_output_tokens_per_task": 9383.212121212122,
      "last_run_timestamp": "2026-10-03T19:50:41.468Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.5-flash-20260519",
      "model_slug": "google/gemini-3.5-flash",
      "model_name": "Google: Gemini 3.5 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.9259256666666668,
      "accuracy_std_dev": 0.0036366213562713575,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.1413088872053872,
      "avg_duration_per_task_ms": 82468.60606060606,
      "avg_output_tokens_per_task": 15584.752525252525,
      "last_run_timestamp": "2026-10-01T19:18:40.706Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.9259255000000001,
      "accuracy_std_dev": 0.0016835000000000044,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.025620611111111115,
      "avg_duration_per_task_ms": 19140.497474747473,
      "avg_output_tokens_per_task": 2323.35101010101,
      "last_run_timestamp": "2026-10-03T09:34:06.834Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.5-20260423",
      "model_slug": "openai/gpt-5.5",
      "model_name": "OpenAI: GPT-5.5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.9255440309278352,
      "accuracy_std_dev": 0.0047577793839843644,
      "run_count": 3,
      "total_questions": 582,
      "avg_cost_per_task_usd": 0.3407643092783505,
      "avg_duration_per_task_ms": 132309.46391752578,
      "avg_output_tokens_per_task": 10197.671821305841,
      "last_run_timestamp": "2026-10-01T19:13:06.559Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-sol-20260709",
      "model_slug": "openai/gpt-5.6-sol",
      "model_name": "OpenAI: GPT-5.6 Sol",
      "provider_name": "Azure",
      "accuracy": 0.9248735,
      "accuracy_std_dev": 0.00862277742377712,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.10364464204545455,
      "avg_duration_per_task_ms": 55123.33648989899,
      "avg_output_tokens_per_task": 3517.8642676767677,
      "last_run_timestamp": "2026-10-01T19:42:48.407Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "x-ai/grok-4.6-20260810",
      "model_slug": "x-ai/grok-4.6",
      "model_name": "SpaceXAI: Grok 4.6",
      "provider_name": "xAI",
      "accuracy": 0.9246375695652174,
      "accuracy_std_dev": 0.006863146440316182,
      "run_count": 7,
      "total_questions": 1380,
      "avg_cost_per_task_usd": 0.12292555942028985,
      "avg_duration_per_task_ms": 340480.1579710145,
      "avg_output_tokens_per_task": 19662.684057971015,
      "last_run_timestamp": "2026-10-01T16:09:31.184Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-sol-20260922",
      "model_slug": "openai/gpt-6-sol",
      "model_name": "OpenAI: GPT-6 Sol",
      "provider_name": "Azure",
      "accuracy": 0.924599700084246,
      "accuracy_std_dev": 0.006819818440072705,
      "run_count": 6,
      "total_questions": 1187,
      "avg_cost_per_task_usd": 0.02751449789385004,
      "avg_duration_per_task_ms": 24957.139848357205,
      "avg_output_tokens_per_task": 2420.7127211457455,
      "last_run_timestamp": "2026-10-01T08:24:52.927Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.5-flash-20260519",
      "model_slug": "google/gemini-3.5-flash",
      "model_name": "Google: Gemini 3.5 Flash",
      "provider_name": "Google",
      "accuracy": 0.9242421666666667,
      "accuracy_std_dev": 0.0048599613653014136,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.14878865656565657,
      "avg_duration_per_task_ms": 85040.50420875421,
      "avg_output_tokens_per_task": 15641.649831649831,
      "last_run_timestamp": "2026-10-01T19:13:26.156Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "unbiased/pareto",
      "model_slug": "unbiased/pareto",
      "model_name": "Pareto",
      "provider_name": "auto-routing",
      "accuracy": 0.9242420000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.03377419696969697,
      "avg_duration_per_task_ms": 114763.54545454546,
      "avg_output_tokens_per_task": 4358.363636363636,
      "last_run_timestamp": "2026-09-29T03:11:33.733Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-sol-20260709",
      "model_slug": "openai/gpt-5.6-sol",
      "model_name": "OpenAI: GPT-5.6 Sol",
      "provider_name": "OpenAI",
      "accuracy": 0.9234005000000001,
      "accuracy_std_dev": 0.008625346413333228,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.05533751388888889,
      "avg_duration_per_task_ms": 61705.84722222222,
      "avg_output_tokens_per_task": 3564.286616161616,
      "last_run_timestamp": "2026-10-02T23:20:36.509Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-sol-20260709",
      "model_slug": "openai/gpt-5.6-sol",
      "model_name": "OpenAI: GPT-5.6 Sol",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.9227054347826086,
      "accuracy_std_dev": 0.0037342422507786015,
      "run_count": 4,
      "total_questions": 759,
      "avg_cost_per_task_usd": 0.09927257048748353,
      "avg_duration_per_task_ms": 52555.73781291173,
      "avg_output_tokens_per_task": 3495.6416337285905,
      "last_run_timestamp": "2026-10-01T19:12:57.119Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Modal",
      "accuracy": 0.9219427970033833,
      "accuracy_std_dev": 0.006990580536078893,
      "run_count": 11,
      "total_questions": 2069,
      "avg_cost_per_task_usd": 0.15209735766070562,
      "avg_duration_per_task_ms": 132397.76655389077,
      "avg_output_tokens_per_task": 9993.401159980667,
      "last_run_timestamp": "2026-10-03T10:21:23.522Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Decart",
      "accuracy": 0.9212965,
      "accuracy_std_dev": 0.0012624999999999997,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.13693505808080808,
      "avg_duration_per_task_ms": 189958.2702020202,
      "avg_output_tokens_per_task": 8925.60101010101,
      "last_run_timestamp": "2026-10-03T09:18:30.358Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-sol-20260709",
      "model_slug": "openai/gpt-5.6-sol",
      "model_name": "OpenAI: GPT-5.6 Sol",
      "provider_name": "auto-routing",
      "accuracy": 0.9208750000000001,
      "accuracy_std_dev": 0.004761657064510223,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.06043574579124578,
      "avg_duration_per_task_ms": 62971.38047138047,
      "avg_output_tokens_per_task": 3586.8703703703704,
      "last_run_timestamp": "2026-10-01T19:12:28.573Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.920875,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.025261131313131315,
      "avg_duration_per_task_ms": 21541.484848484848,
      "avg_output_tokens_per_task": 2294.9949494949497,
      "last_run_timestamp": "2026-10-03T09:35:43.283Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-sol-20260922",
      "model_slug": "openai/gpt-6-sol",
      "model_name": "OpenAI: GPT-6 Sol",
      "provider_name": "OpenAI",
      "accuracy": 0.9208749999999999,
      "accuracy_std_dev": 0.002749143987983661,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.025590255892255893,
      "avg_duration_per_task_ms": 33502.22222222222,
      "avg_output_tokens_per_task": 2400.506734006734,
      "last_run_timestamp": "2026-10-02T23:17:42.236Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Sail Research",
      "accuracy": 0.9205012222222222,
      "accuracy_std_dev": 0.00850227984821038,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.1576766610549944,
      "avg_duration_per_task_ms": 236410.12177328844,
      "avg_output_tokens_per_task": 12595.254208754208,
      "last_run_timestamp": "2026-09-26T08:55:21.775Z"
    },
    {
      "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": 0.920314,
      "accuracy_std_dev": 0.007316769300722847,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.1692038207070707,
      "avg_duration_per_task_ms": 320867.6077441077,
      "avg_output_tokens_per_task": 9628.20622895623,
      "last_run_timestamp": "2026-10-03T09:31:32.854Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-sol-20260922",
      "model_slug": "openai/gpt-6-sol",
      "model_name": "OpenAI: GPT-6 Sol",
      "provider_name": "auto-routing",
      "accuracy": 0.9191919999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.02528019191919192,
      "avg_duration_per_task_ms": 32477.28787878788,
      "avg_output_tokens_per_task": 2369.5,
      "last_run_timestamp": "2026-10-01T08:24:24.034Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-sol-20260922",
      "model_slug": "openai/gpt-6-sol",
      "model_name": "OpenAI: GPT-6 Sol",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.9191915,
      "accuracy_std_dev": 0.005050500000000013,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.026391550505050505,
      "avg_duration_per_task_ms": 26392.646464646463,
      "avg_output_tokens_per_task": 2240.712121212121,
      "last_run_timestamp": "2026-10-01T08:37:17.903Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "NextBit",
      "accuracy": 0.918981375,
      "accuracy_std_dev": 0.00679620392089402,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.12805398863636364,
      "avg_duration_per_task_ms": 541652.8712121212,
      "avg_output_tokens_per_task": 37116.541666666664,
      "last_run_timestamp": "2026-10-03T09:05:48.324Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "auto-routing",
      "accuracy": 0.9185433141891891,
      "accuracy_std_dev": 0.011935651348422994,
      "run_count": 9,
      "total_questions": 1776,
      "avg_cost_per_task_usd": 0.13230360641891895,
      "avg_duration_per_task_ms": 318469.4380630631,
      "avg_output_tokens_per_task": 10386.162725225226,
      "last_run_timestamp": "2026-10-03T09:44:31.992Z"
    },
    {
      "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": 0.9181816,
      "accuracy_std_dev": 0.007721961838807562,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.0816559797979798,
      "avg_duration_per_task_ms": 327979.47777777776,
      "avg_output_tokens_per_task": 24710.48888888889,
      "last_run_timestamp": "2026-10-03T11:44:39.356Z"
    },
    {
      "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": 0.9173829326647566,
      "accuracy_std_dev": 0.013022336741090185,
      "run_count": 11,
      "total_questions": 2094,
      "avg_cost_per_task_usd": 0.12426453438395414,
      "avg_duration_per_task_ms": 122014.65329512894,
      "avg_output_tokens_per_task": 8324.876313276027,
      "last_run_timestamp": "2026-10-03T09:17:37.556Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "AkashML",
      "accuracy": 0.9171931536842105,
      "accuracy_std_dev": 0.011022305393496383,
      "run_count": 4,
      "total_questions": 475,
      "avg_cost_per_task_usd": 0.10596987368421053,
      "avg_duration_per_task_ms": 176112.00210526315,
      "avg_output_tokens_per_task": 6923.242105263158,
      "last_run_timestamp": "2026-10-03T10:04:15.247Z"
    },
    {
      "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": 0.9161053333333333,
      "accuracy_std_dev": 0.004900048423797018,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.10286850841750843,
      "avg_duration_per_task_ms": 630668.6658249159,
      "avg_output_tokens_per_task": 39966.40488215488,
      "last_run_timestamp": "2026-10-03T09:15:20.023Z"
    },
    {
      "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": 0.9158830997830802,
      "accuracy_std_dev": 0.0087081660666812,
      "run_count": 7,
      "total_questions": 1383,
      "avg_cost_per_task_usd": 0.2615176702819957,
      "avg_duration_per_task_ms": 379032.86912509036,
      "avg_output_tokens_per_task": 37727.92263195951,
      "last_run_timestamp": "2026-10-03T09:17:56.331Z"
    },
    {
      "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": 0.9153725336438923,
      "accuracy_std_dev": 0.007489152329202574,
      "run_count": 10,
      "total_questions": 1932,
      "avg_cost_per_task_usd": 0.14073902795031057,
      "avg_duration_per_task_ms": 293484.7536231884,
      "avg_output_tokens_per_task": 9227.590062111802,
      "last_run_timestamp": "2026-10-03T10:17:13.505Z"
    },
    {
      "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": 0.9148461928609483,
      "accuracy_std_dev": 0.040300990523310966,
      "run_count": 10,
      "total_questions": 1877,
      "avg_cost_per_task_usd": 0.11435862759722962,
      "avg_duration_per_task_ms": 215504.55833777305,
      "avg_output_tokens_per_task": 9109.937133724028,
      "last_run_timestamp": "2026-10-03T10:08:37.048Z"
    },
    {
      "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": 0.9145619999999999,
      "accuracy_std_dev": 0.003005841230005351,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.025954439393939394,
      "avg_duration_per_task_ms": 19982.071969696968,
      "avg_output_tokens_per_task": 2234.438131313131,
      "last_run_timestamp": "2026-10-03T10:12:04.782Z"
    },
    {
      "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": 0.9141412499999999,
      "accuracy_std_dev": 0.012579376886561864,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.03262016877104377,
      "avg_duration_per_task_ms": 228754.26725589225,
      "avg_output_tokens_per_task": 26998.277356902356,
      "last_run_timestamp": "2026-10-03T16:51:25.531Z"
    },
    {
      "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": 0.9136839532631578,
      "accuracy_std_dev": 0.030012698133341075,
      "run_count": 12,
      "total_questions": 2375,
      "avg_cost_per_task_usd": 0.017909636210526318,
      "avg_duration_per_task_ms": 57780.396631578944,
      "avg_output_tokens_per_task": 3003.1595789473686,
      "last_run_timestamp": "2026-09-30T03:11:21.812Z"
    },
    {
      "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": 0.9132995,
      "accuracy_std_dev": 0.010546877796295947,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.15848487878787879,
      "avg_duration_per_task_ms": 520006.99494949495,
      "avg_output_tokens_per_task": 39704.125,
      "last_run_timestamp": "2026-10-03T09:15:53.269Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "GMICloud",
      "accuracy": 0.9117844,
      "accuracy_std_dev": 0.006079420781620585,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.03790721616161616,
      "avg_duration_per_task_ms": 129456.97070707071,
      "avg_output_tokens_per_task": 28484.87575757576,
      "last_run_timestamp": "2026-10-03T09:28:48.247Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Morph",
      "accuracy": 0.9115787584210527,
      "accuracy_std_dev": 0.03150249392869556,
      "run_count": 20,
      "total_questions": 3800,
      "avg_cost_per_task_usd": 0.14629828526315786,
      "avg_duration_per_task_ms": 443992.2734210526,
      "avg_output_tokens_per_task": 8936.269473684211,
      "last_run_timestamp": "2026-10-03T10:26:01.662Z"
    },
    {
      "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": 0.9098249433962265,
      "accuracy_std_dev": 0.01556456113130081,
      "run_count": 10,
      "total_questions": 1961,
      "avg_cost_per_task_usd": 0.18924138296787354,
      "avg_duration_per_task_ms": 516166.16930137685,
      "avg_output_tokens_per_task": 12560.821009688934,
      "last_run_timestamp": "2026-10-03T08:12:32.366Z"
    },
    {
      "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": 0.9095505370786516,
      "accuracy_std_dev": 0.004402180380080482,
      "run_count": 9,
      "total_questions": 1780,
      "avg_cost_per_task_usd": 0.02965778258426966,
      "avg_duration_per_task_ms": 474993.9230337079,
      "avg_output_tokens_per_task": 30670.006179775282,
      "last_run_timestamp": "2026-10-03T17:37:28.270Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-luna-pro-20260709",
      "model_slug": "openai/gpt-5.6-luna-pro",
      "model_name": "OpenAI: GPT-5.6 Luna Pro",
      "provider_name": "Azure",
      "accuracy": 0.9094274,
      "accuracy_std_dev": 0.002474202384608023,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.07117444141414142,
      "avg_duration_per_task_ms": 135218.17474747475,
      "avg_output_tokens_per_task": 27695.128282828282,
      "last_run_timestamp": "2026-10-01T16:11:23.394Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "NextBit",
      "accuracy": 0.9094272,
      "accuracy_std_dev": 0.00547072746168187,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.025930728282828285,
      "avg_duration_per_task_ms": 177886.11212121212,
      "avg_output_tokens_per_task": 28121.050505050505,
      "last_run_timestamp": "2026-10-03T09:34:35.533Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Reka",
      "accuracy": 0.9090909999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.1305062777777778,
      "avg_duration_per_task_ms": 540279.6767676767,
      "avg_output_tokens_per_task": 38448.40404040404,
      "last_run_timestamp": "2026-10-01T09:14:23.001Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "GLM 5.3 Flash",
      "provider_name": "digitalocean",
      "accuracy": 0.9090909999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0054125,
      "avg_duration_per_task_ms": 235839.91414141413,
      "avg_output_tokens_per_task": 10749.717171717171,
      "last_run_timestamp": "2026-10-01T23:50:45.024Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.7-max-20260520",
      "model_slug": "qwen/qwen3.7-max",
      "model_name": "Qwen: Qwen3.7 Max",
      "provider_name": "AtlasCloud",
      "accuracy": 0.9090909999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.2221867676767677,
      "avg_duration_per_task_ms": 298883.5202020202,
      "avg_output_tokens_per_task": 29352.99494949495,
      "last_run_timestamp": "2026-09-27T00:06:19.757Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Relace",
      "accuracy": 0.9075092967032967,
      "accuracy_std_dev": 0.0008110000000000062,
      "run_count": 2,
      "total_questions": 364,
      "avg_cost_per_task_usd": 0.15048144505494504,
      "avg_duration_per_task_ms": 345399.21428571426,
      "avg_output_tokens_per_task": 38131.52197802198,
      "last_run_timestamp": "2026-10-03T10:04:03.257Z"
    },
    {
      "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": 0.9074822295247725,
      "accuracy_std_dev": 0.016471532022249782,
      "run_count": 10,
      "total_questions": 1978,
      "avg_cost_per_task_usd": 0.03557361324570273,
      "avg_duration_per_task_ms": 322803.7876643074,
      "avg_output_tokens_per_task": 29631.699696663298,
      "last_run_timestamp": "2026-10-03T19:03:35.841Z"
    },
    {
      "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": 0.907407,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.11648837373737374,
      "avg_duration_per_task_ms": 581221.1060606061,
      "avg_output_tokens_per_task": 46418.58585858586,
      "last_run_timestamp": "2026-10-03T09:31:25.870Z"
    },
    {
      "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": 0.9072541818181818,
      "accuracy_std_dev": 0.02180228378286919,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.03718532782369146,
      "avg_duration_per_task_ms": 268057.74655647384,
      "avg_output_tokens_per_task": 31182.948576675848,
      "last_run_timestamp": "2026-10-03T16:39:40.370Z"
    },
    {
      "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": 0.906355125,
      "accuracy_std_dev": 0.018459945425417037,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.13796338068181818,
      "avg_duration_per_task_ms": 561092.216540404,
      "avg_output_tokens_per_task": 39407.719696969696,
      "last_run_timestamp": "2026-10-03T09:49:39.245Z"
    },
    {
      "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": 0.9060606,
      "accuracy_std_dev": 0.006060466668500054,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.023695626262626262,
      "avg_duration_per_task_ms": 214413.58686868686,
      "avg_output_tokens_per_task": 29199.484848484848,
      "last_run_timestamp": "2026-10-03T09:22:46.333Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Alibaba",
      "accuracy": 0.90572375,
      "accuracy_std_dev": 0.009744176680843784,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.03296931691919192,
      "avg_duration_per_task_ms": 228601.25883838383,
      "avg_output_tokens_per_task": 27296.915404040403,
      "last_run_timestamp": "2026-10-03T10:17:41.561Z"
    },
    {
      "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": 0.9053872,
      "accuracy_std_dev": 0.006933048201188252,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.035172772727272726,
      "avg_duration_per_task_ms": 190581.2101010101,
      "avg_output_tokens_per_task": 29132.872727272726,
      "last_run_timestamp": "2026-10-03T09:10:43.528Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "InferenceNet",
      "accuracy": 0.905376,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 155,
      "avg_cost_per_task_usd": 0.09219292258064517,
      "avg_duration_per_task_ms": 892432.9419354838,
      "avg_output_tokens_per_task": 36790.02580645162,
      "last_run_timestamp": "2026-10-02T06:44:44.722Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Wafer",
      "accuracy": 0.9053669934810953,
      "accuracy_std_dev": 0.014124276768541987,
      "run_count": 8,
      "total_questions": 1534,
      "avg_cost_per_task_usd": 0.012634106910039114,
      "avg_duration_per_task_ms": 286058.7731421121,
      "avg_output_tokens_per_task": 17810.844198174706,
      "last_run_timestamp": "2026-10-03T10:17:33.683Z"
    },
    {
      "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": 0.9041413987901349,
      "accuracy_std_dev": 0.012307350103169372,
      "run_count": 11,
      "total_questions": 2149,
      "avg_cost_per_task_usd": 0.02970322568636575,
      "avg_duration_per_task_ms": 256023.29827826895,
      "avg_output_tokens_per_task": 24623.136342484875,
      "last_run_timestamp": "2026-10-03T16:51:44.428Z"
    },
    {
      "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": 0.9040978022813688,
      "accuracy_std_dev": 0.025559857939936256,
      "run_count": 12,
      "total_questions": 2367,
      "avg_cost_per_task_usd": 0.040857811998310095,
      "avg_duration_per_task_ms": 257266.3920574567,
      "avg_output_tokens_per_task": 30699.54752851711,
      "last_run_timestamp": "2026-10-03T16:57:18.499Z"
    },
    {
      "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": 0.903872,
      "accuracy_std_dev": 0.028683231540396564,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.04186602929292929,
      "avg_duration_per_task_ms": 291657.99141414143,
      "avg_output_tokens_per_task": 32019.437373737375,
      "last_run_timestamp": "2026-10-03T16:15:35.791Z"
    },
    {
      "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": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "auto-routing",
      "accuracy": 0.9034975865992416,
      "accuracy_std_dev": 0.019591261601512255,
      "run_count": 12,
      "total_questions": 2373,
      "avg_cost_per_task_usd": 0.03484569616519174,
      "avg_duration_per_task_ms": 327613.8377581121,
      "avg_output_tokens_per_task": 28971.93510324484,
      "last_run_timestamp": "2026-10-03T18:02:12.491Z"
    },
    {
      "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": 0.903424887142055,
      "accuracy_std_dev": 0.056166981894488544,
      "run_count": 9,
      "total_questions": 1781,
      "avg_cost_per_task_usd": 0.14229890510948906,
      "avg_duration_per_task_ms": 308136.8270634475,
      "avg_output_tokens_per_task": 10189.49241998877,
      "last_run_timestamp": "2026-10-03T10:42:35.637Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Baidu",
      "accuracy": 0.9031985,
      "accuracy_std_dev": 0.002525500000000014,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.033975815656565655,
      "avg_duration_per_task_ms": 141490.74494949495,
      "avg_output_tokens_per_task": 28136.507575757576,
      "last_run_timestamp": "2026-10-03T09:38:38.471Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.4-20260305",
      "model_slug": "openai/gpt-5.4",
      "model_name": "OpenAI: GPT-5.4",
      "provider_name": "auto-routing",
      "accuracy": 0.902918,
      "accuracy_std_dev": 0.0007933738084913016,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.14672305387205387,
      "avg_duration_per_task_ms": 133620.78956228957,
      "avg_output_tokens_per_task": 9652.131313131313,
      "last_run_timestamp": "2026-10-01T19:14:23.097Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-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": 0.9028655747851002,
      "accuracy_std_dev": 0.042913716403447925,
      "run_count": 9,
      "total_questions": 1745,
      "avg_cost_per_task_usd": 0.010968459025787967,
      "avg_duration_per_task_ms": 489384.2842406877,
      "avg_output_tokens_per_task": 20816.36446991404,
      "last_run_timestamp": "2026-10-03T09:16:27.631Z"
    },
    {
      "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": 0.9024267704752275,
      "accuracy_std_dev": 0.008260579897319528,
      "run_count": 5,
      "total_questions": 989,
      "avg_cost_per_task_usd": 0.03290473609706774,
      "avg_duration_per_task_ms": 147318.88675429727,
      "avg_output_tokens_per_task": 29037.182002022244,
      "last_run_timestamp": "2026-10-03T09:38:12.362Z"
    },
    {
      "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": 0.9023569999999999,
      "accuracy_std_dev": 0.003367000000000009,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.05417848484848485,
      "avg_duration_per_task_ms": 36580.51515151515,
      "avg_output_tokens_per_task": 2499.1969696969695,
      "last_run_timestamp": "2026-10-03T09:27:36.763Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "Novita",
      "accuracy": 0.9023565,
      "accuracy_std_dev": 0.005050499999999958,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.032186345959595956,
      "avg_duration_per_task_ms": 777530.196969697,
      "avg_output_tokens_per_task": 36633.931818181816,
      "last_run_timestamp": "2026-10-03T10:59:16.091Z"
    },
    {
      "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": 0.9021696666666666,
      "accuracy_std_dev": 0.023279077501768247,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.15281446520763184,
      "avg_duration_per_task_ms": 592828.4741863075,
      "avg_output_tokens_per_task": 38634.81593714927,
      "last_run_timestamp": "2026-10-03T10:13:07.289Z"
    },
    {
      "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": "Alibaba",
      "accuracy": 0.9021161428571426,
      "accuracy_std_dev": 0.016357597294997403,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.027857989898989896,
      "avg_duration_per_task_ms": 208577.13636363635,
      "avg_output_tokens_per_task": 26073.319624819625,
      "last_run_timestamp": "2026-10-03T09:35:22.002Z"
    },
    {
      "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": 0.9017359124999998,
      "accuracy_std_dev": 0.023311648922467388,
      "run_count": 8,
      "total_questions": 1440,
      "avg_cost_per_task_usd": 0.15431451527777776,
      "avg_duration_per_task_ms": 567983.5055555556,
      "avg_output_tokens_per_task": 37740.19097222222,
      "last_run_timestamp": "2026-10-03T09:09:53.172Z"
    },
    {
      "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": 0.9016834000000001,
      "accuracy_std_dev": 0.04222344006165297,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.1424910212121212,
      "avg_duration_per_task_ms": 211878.58636363636,
      "avg_output_tokens_per_task": 9531.086363636363,
      "last_run_timestamp": "2026-10-03T10:10:50.167Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "GMICloud",
      "accuracy": 0.9016834,
      "accuracy_std_dev": 0.01216780761846602,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.028648670707070705,
      "avg_duration_per_task_ms": 147445.17373737373,
      "avg_output_tokens_per_task": 28162.69292929293,
      "last_run_timestamp": "2026-10-03T09:28:15.105Z"
    },
    {
      "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": 0.9014598562605278,
      "accuracy_std_dev": 0.03309502797836693,
      "run_count": 9,
      "total_questions": 1781,
      "avg_cost_per_task_usd": 0.10678652554744525,
      "avg_duration_per_task_ms": 569188.9988770353,
      "avg_output_tokens_per_task": 39083.0421111735,
      "last_run_timestamp": "2026-10-03T10:32:36.645Z"
    },
    {
      "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": 0.9010467416413374,
      "accuracy_std_dev": 0.006919339835562336,
      "run_count": 5,
      "total_questions": 987,
      "avg_cost_per_task_usd": 0.017340083080040522,
      "avg_duration_per_task_ms": 440190.9321175279,
      "avg_output_tokens_per_task": 28500.98986828774,
      "last_run_timestamp": "2026-10-03T09:52:27.621Z"
    },
    {
      "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": 0.9007667777777778,
      "accuracy_std_dev": 0.005624838482180366,
      "run_count": 18,
      "total_questions": 3564,
      "avg_cost_per_task_usd": 0.20504940852974188,
      "avg_duration_per_task_ms": 98780.56790123456,
      "avg_output_tokens_per_task": 7889.076038159372,
      "last_run_timestamp": "2026-10-03T09:47:38.062Z"
    },
    {
      "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": 0.9004328214285713,
      "accuracy_std_dev": 0.005913167322978487,
      "run_count": 28,
      "total_questions": 5544,
      "avg_cost_per_task_usd": 0.21622311616161613,
      "avg_duration_per_task_ms": 94665.56475468975,
      "avg_output_tokens_per_task": 7881.3840187590185,
      "last_run_timestamp": "2026-10-03T10:10:31.743Z"
    },
    {
      "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": 0.9001119999999999,
      "accuracy_std_dev": 0.0034590473640392213,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.053618132996633,
      "avg_duration_per_task_ms": 27125.141414141413,
      "avg_output_tokens_per_task": 2464.4410774410776,
      "last_run_timestamp": "2026-10-03T09:16:11.366Z"
    },
    {
      "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": 0.9000926755555556,
      "accuracy_std_dev": 0.009531680514027292,
      "run_count": 11,
      "total_questions": 1800,
      "avg_cost_per_task_usd": 0.022115208888888887,
      "avg_duration_per_task_ms": 208945.33944444446,
      "avg_output_tokens_per_task": 18295.701666666668,
      "last_run_timestamp": "2026-10-03T16:51:49.235Z"
    },
    {
      "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": 0.8999998999999999,
      "accuracy_std_dev": 0.005644066804175849,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.012073083333333333,
      "avg_duration_per_task_ms": 179513.8106060606,
      "avg_output_tokens_per_task": 26194.345454545455,
      "last_run_timestamp": "2026-10-03T09:29:36.040Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.4-20260305",
      "model_slug": "openai/gpt-5.4",
      "model_name": "OpenAI: GPT-5.4",
      "provider_name": "Azure",
      "accuracy": 0.8999248888888889,
      "accuracy_std_dev": 0.006354339085213833,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.15479229124579125,
      "avg_duration_per_task_ms": 130090.1857463524,
      "avg_output_tokens_per_task": 9547.675645342311,
      "last_run_timestamp": "2026-10-01T19:18:24.536Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "Xiaomi",
      "accuracy": 0.8998314999999999,
      "accuracy_std_dev": 0.000841499999999995,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.0343437398989899,
      "avg_duration_per_task_ms": 823628.1161616162,
      "avg_output_tokens_per_task": 39114.656565656565,
      "last_run_timestamp": "2026-10-03T11:02:12.320Z"
    },
    {
      "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": "DeepSeek",
      "accuracy": 0.8996633999999999,
      "accuracy_std_dev": 0.005792805834826492,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.007343192929292929,
      "avg_duration_per_task_ms": 219337.0404040404,
      "avg_output_tokens_per_task": 25809.762626262625,
      "last_run_timestamp": "2026-08-08T09:30:52.456Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.4-20260305",
      "model_slug": "openai/gpt-5.4",
      "model_name": "OpenAI: GPT-5.4",
      "provider_name": "OpenAI",
      "accuracy": 0.899551,
      "accuracy_std_dev": 0.006198077174952426,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.14109027272727273,
      "avg_duration_per_task_ms": 131303,
      "avg_output_tokens_per_task": 9278.422558922559,
      "last_run_timestamp": "2026-10-01T19:14:55.028Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-luna-pro-20260709",
      "model_slug": "openai/gpt-5.6-luna-pro",
      "model_name": "OpenAI: GPT-5.6 Luna Pro",
      "provider_name": "auto-routing",
      "accuracy": 0.899551,
      "accuracy_std_dev": 0.006198077174952426,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0412280202020202,
      "avg_duration_per_task_ms": 141835.40572390571,
      "avg_output_tokens_per_task": 27211.210437710437,
      "last_run_timestamp": "2026-10-01T16:12:34.647Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "Novita",
      "accuracy": 0.8994372877637131,
      "accuracy_std_dev": 0.01977995405637288,
      "run_count": 6,
      "total_questions": 1185,
      "avg_cost_per_task_usd": 0.03544181097046414,
      "avg_duration_per_task_ms": 152645.21687763714,
      "avg_output_tokens_per_task": 26597.288607594935,
      "last_run_timestamp": "2026-10-03T09:28:39.969Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-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": 0.8994104999999998,
      "accuracy_std_dev": 0.004817079950550935,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.22572103156565654,
      "avg_duration_per_task_ms": 108993.70707070707,
      "avg_output_tokens_per_task": 8821.869949494949,
      "last_run_timestamp": "2026-10-03T09:42:54.478Z"
    },
    {
      "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": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "DeepInfra",
      "accuracy": 0.8993387243502051,
      "accuracy_std_dev": 0.026386926357751014,
      "run_count": 8,
      "total_questions": 1462,
      "avg_cost_per_task_usd": 0.027840691518467853,
      "avg_duration_per_task_ms": 740625.9062927497,
      "avg_output_tokens_per_task": 24883.40560875513,
      "last_run_timestamp": "2026-09-12T16:17:57.063Z"
    },
    {
      "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-4.8-opus-20260528",
      "model_slug": "anthropic/claude-opus-4.8",
      "model_name": "Anthropic: Claude Opus 4.8",
      "provider_name": "Google",
      "accuracy": 0.8992219655172418,
      "accuracy_std_dev": 0.0066569301119999745,
      "run_count": 29,
      "total_questions": 5742,
      "avg_cost_per_task_usd": 0.22289186381051904,
      "avg_duration_per_task_ms": 94799.24817136886,
      "avg_output_tokens_per_task": 8115.898293277603,
      "last_run_timestamp": "2026-10-03T10:16:05.445Z"
    },
    {
      "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": 0.8991767777777778,
      "accuracy_std_dev": 0.005290828285875197,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.058499294612794614,
      "avg_duration_per_task_ms": 32080.961840628508,
      "avg_output_tokens_per_task": 2531.0488215488217,
      "last_run_timestamp": "2026-10-03T10:16:57.881Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-luna-pro-20260709",
      "model_slug": "openai/gpt-5.6-luna-pro",
      "model_name": "OpenAI: GPT-5.6 Luna Pro",
      "provider_name": "OpenAI",
      "accuracy": 0.8989897499999999,
      "accuracy_std_dev": 0.005188652155184398,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.03249949747474747,
      "avg_duration_per_task_ms": 137621.66666666666,
      "avg_output_tokens_per_task": 26156.715909090908,
      "last_run_timestamp": "2026-10-02T23:26:10.109Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": "z-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": 0.8980803802559414,
      "accuracy_std_dev": 0.010966029181066412,
      "run_count": 6,
      "total_questions": 1094,
      "avg_cost_per_task_usd": 0.006137671846435101,
      "avg_duration_per_task_ms": 419028.80987202923,
      "avg_output_tokens_per_task": 12654.702925045704,
      "last_run_timestamp": "2026-10-03T10:14:40.227Z"
    },
    {
      "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": 0.8980545555555557,
      "accuracy_std_dev": 0.0033774129188909115,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.057495572951739626,
      "avg_duration_per_task_ms": 25176.90740740741,
      "avg_output_tokens_per_task": 2488.7480359147025,
      "last_run_timestamp": "2026-10-03T09:48:22.257Z"
    },
    {
      "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": "Azure",
      "accuracy": 0.8978673333333335,
      "accuracy_std_dev": 0.005382399950042904,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.057627444444444446,
      "avg_duration_per_task_ms": 26162.272727272728,
      "avg_output_tokens_per_task": 2526.015993265993,
      "last_run_timestamp": "2026-10-03T10:06:47.904Z"
    },
    {
      "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": 0.8977272500000001,
      "accuracy_std_dev": 0.008615029218609743,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.19720289898989898,
      "avg_duration_per_task_ms": 95317.51073232324,
      "avg_output_tokens_per_task": 7684.227904040404,
      "last_run_timestamp": "2026-10-03T09:39:18.026Z"
    },
    {
      "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": 0.8975330208728654,
      "accuracy_std_dev": 0.015763870376592163,
      "run_count": 8,
      "total_questions": 1581,
      "avg_cost_per_task_usd": 0.11425380961416824,
      "avg_duration_per_task_ms": 628521.0765338393,
      "avg_output_tokens_per_task": 36072.062618595824,
      "last_run_timestamp": "2026-10-03T10:20:39.400Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.8-opus-20260528",
      "model_slug": "anthropic/claude-opus-4.8",
      "model_name": "Anthropic: Claude Opus 4.8",
      "provider_name": "auto-routing",
      "accuracy": 0.8974932222222222,
      "accuracy_std_dev": 0.012382476271622079,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.20160183445566776,
      "avg_duration_per_task_ms": 97321.57126823794,
      "avg_output_tokens_per_task": 7848.616722783389,
      "last_run_timestamp": "2026-10-03T09:45:23.245Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.4-20260305",
      "model_slug": "openai/gpt-5.4",
      "model_name": "OpenAI: GPT-5.4",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.8968568964143425,
      "accuracy_std_dev": 0.005278733293840445,
      "run_count": 4,
      "total_questions": 753,
      "avg_cost_per_task_usd": 0.15754132403718457,
      "avg_duration_per_task_ms": 84099.57768924303,
      "avg_output_tokens_per_task": 9420.674634794157,
      "last_run_timestamp": "2026-10-01T19:12:25.243Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-opus-5-20260723",
      "model_slug": "anthropic/claude-opus-5",
      "model_name": "Anthropic: Claude Opus 5",
      "provider_name": "Azure",
      "accuracy": 0.8967218206896553,
      "accuracy_std_dev": 0.007385261912078655,
      "run_count": 20,
      "total_questions": 3915,
      "avg_cost_per_task_usd": 0.12559540842911876,
      "avg_duration_per_task_ms": 57752.34942528736,
      "avg_output_tokens_per_task": 4741.332567049809,
      "last_run_timestamp": "2026-10-03T10:05:55.795Z"
    },
    {
      "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": 0.8964932281368821,
      "accuracy_std_dev": 0.008729450995767121,
      "run_count": 4,
      "total_questions": 789,
      "avg_cost_per_task_usd": 0.01934689100126743,
      "avg_duration_per_task_ms": 616078.9974651458,
      "avg_output_tokens_per_task": 17884.557667934092,
      "last_run_timestamp": "2026-10-03T10:04:37.350Z"
    },
    {
      "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": 0.8960090269814502,
      "accuracy_std_dev": 0.005525506754035197,
      "run_count": 3,
      "total_questions": 593,
      "avg_cost_per_task_usd": 0.01363839797639123,
      "avg_duration_per_task_ms": 405081.17875210795,
      "avg_output_tokens_per_task": 20090.210792580103,
      "last_run_timestamp": "2026-10-03T11:58:57.361Z"
    },
    {
      "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": 0.8956340674536255,
      "accuracy_std_dev": 0.02132427030288766,
      "run_count": 9,
      "total_questions": 1779,
      "avg_cost_per_task_usd": 0.13012930916245083,
      "avg_duration_per_task_ms": 507416.96458684653,
      "avg_output_tokens_per_task": 37964.30185497471,
      "last_run_timestamp": "2026-10-03T10:12:16.841Z"
    },
    {
      "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": 0.8956227499999999,
      "accuracy_std_dev": 0.005791223203046117,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.20478679124579122,
      "avg_duration_per_task_ms": 99940.12121212122,
      "avg_output_tokens_per_task": 7979.755892255892,
      "last_run_timestamp": "2026-10-03T09:40:16.985Z"
    },
    {
      "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": 0.8951176,
      "accuracy_std_dev": 0.00756834082477789,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.12048651767676768,
      "avg_duration_per_task_ms": 57579.538383838386,
      "avg_output_tokens_per_task": 4603.499494949495,
      "last_run_timestamp": "2026-10-03T09:25:15.659Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-terra-20260709",
      "model_slug": "openai/gpt-5.6-terra",
      "model_name": "OpenAI: GPT-5.6 Terra",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.8950130708661417,
      "accuracy_std_dev": 0.00891334037763057,
      "run_count": 4,
      "total_questions": 762,
      "avg_cost_per_task_usd": 0.048151682414698166,
      "avg_duration_per_task_ms": 43476.24934383202,
      "avg_output_tokens_per_task": 3503.036745406824,
      "last_run_timestamp": "2026-10-01T19:11:43.444Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Novita",
      "accuracy": 0.8948596495327101,
      "accuracy_std_dev": 0.0171199232602798,
      "run_count": 13,
      "total_questions": 2568,
      "avg_cost_per_task_usd": 0.022189622274143304,
      "avg_duration_per_task_ms": 231984.90109034267,
      "avg_output_tokens_per_task": 26737.70015576324,
      "last_run_timestamp": "2026-10-03T09:42:10.726Z"
    },
    {
      "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": 0.8945527678958785,
      "accuracy_std_dev": 0.010412875331697246,
      "run_count": 14,
      "total_questions": 2766,
      "avg_cost_per_task_usd": 0.012522617136659437,
      "avg_duration_per_task_ms": 291164.3582791034,
      "avg_output_tokens_per_task": 27224.147143890095,
      "last_run_timestamp": "2026-10-03T09:48:09.405Z"
    },
    {
      "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": 0.8942841547799696,
      "accuracy_std_dev": 0.008875212006481873,
      "run_count": 10,
      "total_questions": 1977,
      "avg_cost_per_task_usd": 0.22388517248356093,
      "avg_duration_per_task_ms": 107507.08952959029,
      "avg_output_tokens_per_task": 8731.16894284269,
      "last_run_timestamp": "2026-10-03T09:18:11.817Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-terra-20260709",
      "model_slug": "openai/gpt-5.6-terra",
      "model_name": "OpenAI: GPT-5.6 Terra",
      "provider_name": "Azure",
      "accuracy": 0.8941264444444444,
      "accuracy_std_dev": 0.009636502052221474,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.05178651795735129,
      "avg_duration_per_task_ms": 45445.343434343435,
      "avg_output_tokens_per_task": 3671.5897867564536,
      "last_run_timestamp": "2026-10-01T19:13:06.514Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "ModelRun",
      "accuracy": 0.8939392,
      "accuracy_std_dev": 0.021664239228738207,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.08778542626262627,
      "avg_duration_per_task_ms": 210435.61212121212,
      "avg_output_tokens_per_task": 29520.061616161616,
      "last_run_timestamp": "2026-10-03T16:44:09.301Z"
    },
    {
      "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": 0.893939,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.021992727272727275,
      "avg_duration_per_task_ms": 693845.0404040404,
      "avg_output_tokens_per_task": 46651.48484848485,
      "last_run_timestamp": "2026-10-01T08:47:48.972Z"
    },
    {
      "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": 0.8938831666666668,
      "accuracy_std_dev": 0.007122347684499023,
      "run_count": 30,
      "total_questions": 5940,
      "avg_cost_per_task_usd": 0.20402719309764314,
      "avg_duration_per_task_ms": 70642.39461279461,
      "avg_output_tokens_per_task": 7430.633333333333,
      "last_run_timestamp": "2026-10-03T10:09:58.833Z"
    },
    {
      "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": 0.89322933125,
      "accuracy_std_dev": 0.0012764999999999582,
      "run_count": 2,
      "total_questions": 320,
      "avg_cost_per_task_usd": 0.033160396875,
      "avg_duration_per_task_ms": 1020849.565625,
      "avg_output_tokens_per_task": 37888.675,
      "last_run_timestamp": "2026-10-03T11:52:13.119Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "DeepInfra",
      "accuracy": 0.8931597755794235,
      "accuracy_std_dev": 0.01575096391955302,
      "run_count": 9,
      "total_questions": 1769,
      "avg_cost_per_task_usd": 0.11365974788015829,
      "avg_duration_per_task_ms": 394066.868852459,
      "avg_output_tokens_per_task": 7770.012436404749,
      "last_run_timestamp": "2026-10-03T10:26:28.328Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-terra-20260709",
      "model_slug": "openai/gpt-5.6-terra",
      "model_name": "OpenAI: GPT-5.6 Terra",
      "provider_name": "OpenAI",
      "accuracy": 0.89267675,
      "accuracy_std_dev": 0.006979218773437292,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.037834488636363635,
      "avg_duration_per_task_ms": 50710.07575757576,
      "avg_output_tokens_per_task": 3480.1502525252527,
      "last_run_timestamp": "2026-10-02T23:21:06.556Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "BaseTen",
      "accuracy": 0.892567903763697,
      "accuracy_std_dev": 0.02022831814050264,
      "run_count": 11,
      "total_questions": 2099,
      "avg_cost_per_task_usd": 0.15797696188661267,
      "avg_duration_per_task_ms": 338559.5883754169,
      "avg_output_tokens_per_task": 39638.616484040016,
      "last_run_timestamp": "2026-09-19T08:51:08.108Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.892548239520958,
      "accuracy_std_dev": 0.006915975724525221,
      "run_count": 3,
      "total_questions": 501,
      "avg_cost_per_task_usd": 0.013722602794411178,
      "avg_duration_per_task_ms": 355406.4910179641,
      "avg_output_tokens_per_task": 23503.43512974052,
      "last_run_timestamp": "2026-10-03T10:04:24.528Z"
    },
    {
      "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": 0.8922556666666667,
      "accuracy_std_dev": 0.0023810642345154617,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.05459258585858586,
      "avg_duration_per_task_ms": 34041.639730639734,
      "avg_output_tokens_per_task": 2518.429292929293,
      "last_run_timestamp": "2026-10-03T09:50:45.552Z"
    },
    {
      "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": "Wafer",
      "accuracy": 0.8919671427077888,
      "accuracy_std_dev": 0.027075382802174704,
      "run_count": 11,
      "total_questions": 1913,
      "avg_cost_per_task_usd": 0.2187960825927862,
      "avg_duration_per_task_ms": 381468.48144276004,
      "avg_output_tokens_per_task": 43318.19968635651,
      "last_run_timestamp": "2026-10-03T17:02:39.089Z"
    },
    {
      "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": 0.8909932500000002,
      "accuracy_std_dev": 0.010197164639619184,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.00789814393939394,
      "avg_duration_per_task_ms": 268244.18434343435,
      "avg_output_tokens_per_task": 27958.415404040403,
      "last_run_timestamp": "2026-08-08T10:06:09.471Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "DeepInfra",
      "accuracy": 0.89087581459854,
      "accuracy_std_dev": 0.012535536331663226,
      "run_count": 14,
      "total_questions": 2740,
      "avg_cost_per_task_usd": 0.0037107835766423355,
      "avg_duration_per_task_ms": 283929.7711678832,
      "avg_output_tokens_per_task": 20325.415693430656,
      "last_run_timestamp": "2026-10-03T09:45:10.741Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3-flash-preview-20251217",
      "model_slug": "google/gemini-3-flash-preview",
      "model_name": "Google: Gemini 3 Flash Preview",
      "provider_name": "Google",
      "accuracy": 0.8905723333333333,
      "accuracy_std_dev": 0.021339306148253502,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.13216584343434343,
      "avg_duration_per_task_ms": 251067.54713804714,
      "avg_output_tokens_per_task": 43934.003367003366,
      "last_run_timestamp": "2026-10-01T19:18:35.588Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "SiliconFlow",
      "accuracy": 0.890572,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.12533664393939395,
      "avg_duration_per_task_ms": 808381.0631313132,
      "avg_output_tokens_per_task": 49868.60353535353,
      "last_run_timestamp": "2026-10-03T09:33:15.909Z"
    },
    {
      "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": 0.8903411328976034,
      "accuracy_std_dev": 0.020961163458907063,
      "run_count": 7,
      "total_questions": 1377,
      "avg_cost_per_task_usd": 0.1928877625272331,
      "avg_duration_per_task_ms": 1363104.034858388,
      "avg_output_tokens_per_task": 56628.1917211329,
      "last_run_timestamp": "2026-10-03T17:43:47.691Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "amazon/nova-micro-v1",
      "model_slug": "amazon/nova-micro-v1",
      "model_name": "Amazon: Nova Micro 1.0",
      "provider_name": "Anthropic",
      "accuracy": 0.8900113333333334,
      "accuracy_std_dev": 0.005722799567422281,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.19800758417508418,
      "avg_duration_per_task_ms": 96141.14141414141,
      "avg_output_tokens_per_task": 7695.558922558923,
      "last_run_timestamp": "2026-10-01T19:11:18.427Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-luna-pro-20260922",
      "model_slug": "openai/gpt-6-luna-pro",
      "model_name": "OpenAI: GPT-6 Luna Pro",
      "provider_name": "OpenAI",
      "accuracy": 0.8900113333333334,
      "accuracy_std_dev": 0.0015872190215034077,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.008230473063973063,
      "avg_duration_per_task_ms": 80128.80303030302,
      "avg_output_tokens_per_task": 12935.998316498317,
      "last_run_timestamp": "2026-10-02T23:23:43.005Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "amazon/nova-micro-v1",
      "model_slug": "amazon/nova-micro-v1",
      "model_name": "Amazon: Nova Micro 1.0",
      "provider_name": "auto-routing",
      "accuracy": 0.8900110000000001,
      "accuracy_std_dev": 0.0055552329084086844,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.2164912121212121,
      "avg_duration_per_task_ms": 104660.01851851853,
      "avg_output_tokens_per_task": 8434.90404040404,
      "last_run_timestamp": "2026-10-01T19:10:42.213Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "GMICloud",
      "accuracy": 0.8898507142857144,
      "accuracy_std_dev": 0.013463734020527524,
      "run_count": 14,
      "total_questions": 2772,
      "avg_cost_per_task_usd": 0.013538019841269842,
      "avg_duration_per_task_ms": 210147.86976911977,
      "avg_output_tokens_per_task": 24623.09126984127,
      "last_run_timestamp": "2026-10-03T09:29:34.537Z"
    },
    {
      "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": 0.8898480650076179,
      "accuracy_std_dev": 0.00936748461292215,
      "run_count": 30,
      "total_questions": 5907,
      "avg_cost_per_task_usd": 0.13325874741831725,
      "avg_duration_per_task_ms": 59111.40138818351,
      "avg_output_tokens_per_task": 4767.174030810902,
      "last_run_timestamp": "2026-10-03T09:22:12.774Z"
    },
    {
      "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": 0.8895201249999999,
      "accuracy_std_dev": 0.018150193638343783,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.1383345025252525,
      "avg_duration_per_task_ms": 580496.523989899,
      "avg_output_tokens_per_task": 34909.79734848485,
      "last_run_timestamp": "2026-10-03T10:33:13.241Z"
    },
    {
      "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": 0.88945,
      "accuracy_std_dev": 0.00975207530050226,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.024394900673400673,
      "avg_duration_per_task_ms": 319196.3552188552,
      "avg_output_tokens_per_task": 27886,
      "last_run_timestamp": "2026-10-03T10:36:49.762Z"
    },
    {
      "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": "anthropic/claude-4.7-opus-20260416",
      "model_slug": "anthropic/claude-opus-4.7",
      "model_name": "Anthropic: Claude Opus 4.7",
      "provider_name": "Anthropic",
      "accuracy": 0.8894080654205608,
      "accuracy_std_dev": 0.00941594536414648,
      "run_count": 13,
      "total_questions": 2568,
      "avg_cost_per_task_usd": 0.21110783605919006,
      "avg_duration_per_task_ms": 101949.81853582554,
      "avg_output_tokens_per_task": 8216.18808411215,
      "last_run_timestamp": "2026-10-03T09:19:25.458Z"
    },
    {
      "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": 0.8892253999999999,
      "accuracy_std_dev": 0.008511315206241623,
      "run_count": 20,
      "total_questions": 3960,
      "avg_cost_per_task_usd": 0.24073908914141412,
      "avg_duration_per_task_ms": 84900.21515151515,
      "avg_output_tokens_per_task": 8956.617424242424,
      "last_run_timestamp": "2026-10-03T09:30:01.468Z"
    },
    {
      "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": 0.8888886666666669,
      "accuracy_std_dev": 0.007653072992087717,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.008966420875420876,
      "avg_duration_per_task_ms": 148767.4797979798,
      "avg_output_tokens_per_task": 31547.74074074074,
      "last_run_timestamp": "2026-10-03T10:11:42.574Z"
    },
    {
      "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": 0.8885382849631968,
      "accuracy_std_dev": 0.024250159666278505,
      "run_count": 5,
      "total_questions": 951,
      "avg_cost_per_task_usd": 0.1142291840168244,
      "avg_duration_per_task_ms": 715234.0883280757,
      "avg_output_tokens_per_task": 30845.972660357518,
      "last_run_timestamp": "2026-10-03T09:27:52.133Z"
    },
    {
      "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": 0.8883276666666667,
      "accuracy_std_dev": 0.0028614651647170097,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.008124902356902356,
      "avg_duration_per_task_ms": 491163.50673400675,
      "avg_output_tokens_per_task": 22629.957912457914,
      "last_run_timestamp": "2026-10-01T11:20:35.509Z"
    },
    {
      "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": 0.8881577165991904,
      "accuracy_std_dev": 0.008283430245979028,
      "run_count": 10,
      "total_questions": 1976,
      "avg_cost_per_task_usd": 0.015942041497975706,
      "avg_duration_per_task_ms": 164141.9762145749,
      "avg_output_tokens_per_task": 26128.10374493927,
      "last_run_timestamp": "2026-10-03T13:59:16.761Z"
    },
    {
      "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": 0.8880469999999999,
      "accuracy_std_dev": 0.004209000000000018,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.11352663888888888,
      "avg_duration_per_task_ms": 879783.9520202021,
      "avg_output_tokens_per_task": 45200.48484848485,
      "last_run_timestamp": "2026-10-03T10:42:46.764Z"
    },
    {
      "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": 0.8880445319148936,
      "accuracy_std_dev": 0.009804096960454846,
      "run_count": 10,
      "total_questions": 1974,
      "avg_cost_per_task_usd": 0.20574746403242147,
      "avg_duration_per_task_ms": 100391.19402228977,
      "avg_output_tokens_per_task": 8003.147416413374,
      "last_run_timestamp": "2026-10-03T09:42:47.873Z"
    },
    {
      "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": 0.887854866328257,
      "accuracy_std_dev": 0.036201773425286826,
      "run_count": 9,
      "total_questions": 1773,
      "avg_cost_per_task_usd": 0.13917155047941343,
      "avg_duration_per_task_ms": 799755.2588832488,
      "avg_output_tokens_per_task": 34841.37958262832,
      "last_run_timestamp": "2026-10-03T10:47:57.305Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-terra-20260709",
      "model_slug": "openai/gpt-5.6-terra",
      "model_name": "OpenAI: GPT-5.6 Terra",
      "provider_name": "auto-routing",
      "accuracy": 0.8877663333333334,
      "accuracy_std_dev": 0.005555502277522277,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.03440870875420875,
      "avg_duration_per_task_ms": 50195.98148148148,
      "avg_output_tokens_per_task": 3309.2491582491584,
      "last_run_timestamp": "2026-10-01T19:13:07.806Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "NextBit",
      "accuracy": 0.887537924012158,
      "accuracy_std_dev": 0.020250680116973837,
      "run_count": 5,
      "total_questions": 987,
      "avg_cost_per_task_usd": 0.12033997061803446,
      "avg_duration_per_task_ms": 580755.4113475177,
      "avg_output_tokens_per_task": 34357.5663627153,
      "last_run_timestamp": "2026-10-03T16:44:56.199Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-fable-5.1-20260831",
      "model_slug": "anthropic/claude-fable-5.1",
      "model_name": "Anthropic: Claude Fable 5.1",
      "provider_name": "Azure",
      "accuracy": 0.8872051666666666,
      "accuracy_std_dev": 0.024761493125527695,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.1434989882154882,
      "avg_duration_per_task_ms": 42121.643939393936,
      "avg_output_tokens_per_task": 2660.465488215488,
      "last_run_timestamp": "2026-10-01T08:24:34.535Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Parasail",
      "accuracy": 0.8867842499999999,
      "accuracy_std_dev": 0.003834422340783562,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.027946079545454545,
      "avg_duration_per_task_ms": 154476.702020202,
      "avg_output_tokens_per_task": 23124.396464646463,
      "last_run_timestamp": "2026-10-03T10:53:39.908Z"
    },
    {
      "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": 0.8867689502093997,
      "accuracy_std_dev": 0.008568113097385388,
      "run_count": 11,
      "total_questions": 2149,
      "avg_cost_per_task_usd": 0.09704398464402049,
      "avg_duration_per_task_ms": 550284.3764541647,
      "avg_output_tokens_per_task": 29436.060027919964,
      "last_run_timestamp": "2026-10-03T17:04:25.569Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Fireworks",
      "accuracy": 0.8867051248554914,
      "accuracy_std_dev": 0.021208571037200977,
      "run_count": 5,
      "total_questions": 865,
      "avg_cost_per_task_usd": 0.16137210289017342,
      "avg_duration_per_task_ms": 430828.78150289017,
      "avg_output_tokens_per_task": 40500.787283237,
      "last_run_timestamp": "2026-09-19T08:53:57.162Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "auto-routing",
      "accuracy": 0.8863635000000001,
      "accuracy_std_dev": 0.004208500000000004,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.031905555555555555,
      "avg_duration_per_task_ms": 767151.1060606061,
      "avg_output_tokens_per_task": 36810.59595959596,
      "last_run_timestamp": "2026-10-03T10:37:19.082Z"
    },
    {
      "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": 0.8860007604485219,
      "accuracy_std_dev": 0.0277911335645022,
      "run_count": 5,
      "total_questions": 981,
      "avg_cost_per_task_usd": 0.04559343119266055,
      "avg_duration_per_task_ms": 698776.9031600407,
      "avg_output_tokens_per_task": 26722.172273190623,
      "last_run_timestamp": "2026-08-22T10:03:34.994Z"
    },
    {
      "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": 0.8855853621621622,
      "accuracy_std_dev": 0.025702889954244423,
      "run_count": 3,
      "total_questions": 555,
      "avg_cost_per_task_usd": 0.012490181981981983,
      "avg_duration_per_task_ms": 346127.65405405406,
      "avg_output_tokens_per_task": 22279.807207207206,
      "last_run_timestamp": "2026-10-03T10:21:30.026Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3-flash-preview-20251217",
      "model_slug": "google/gemini-3-flash-preview",
      "model_name": "Google: Gemini 3 Flash Preview",
      "provider_name": "Google AI Studio",
      "accuracy": 0.8855220000000001,
      "accuracy_std_dev": 0.012598160421267844,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.1415003787878788,
      "avg_duration_per_task_ms": 184472.2676767677,
      "avg_output_tokens_per_task": 47043.42424242424,
      "last_run_timestamp": "2026-10-01T19:12:30.168Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "sakana/fugu-max",
      "model_slug": "sakana/fugu-max",
      "model_name": "Fugu Max",
      "provider_name": "auto-routing",
      "accuracy": 0.885522,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.07555469191919191,
      "avg_duration_per_task_ms": 225091.94949494948,
      "avg_output_tokens_per_task": 12279.550505050505,
      "last_run_timestamp": "2026-09-15T01:07:47.957Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-flash-20260826",
      "model_slug": "qwen/qwen3.8-flash",
      "model_name": "Qwen: Qwen3.8 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.885522,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.02146120202020202,
      "avg_duration_per_task_ms": 674351.9393939395,
      "avg_output_tokens_per_task": 45464.17676767677,
      "last_run_timestamp": "2026-10-01T08:31:57.634Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "Parasail",
      "accuracy": 0.8853753412384717,
      "accuracy_std_dev": 0.014100336545185728,
      "run_count": 8,
      "total_questions": 1518,
      "avg_cost_per_task_usd": 0.06779363899868249,
      "avg_duration_per_task_ms": 300937.79380764166,
      "avg_output_tokens_per_task": 18734.31488801054,
      "last_run_timestamp": "2026-10-03T18:59:42.075Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-luna-20260709",
      "model_slug": "openai/gpt-5.6-luna",
      "model_name": "OpenAI: GPT-5.6 Luna",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.8853190481927712,
      "accuracy_std_dev": 0.005795847495189989,
      "run_count": 4,
      "total_questions": 747,
      "avg_cost_per_task_usd": 0.0097919718875502,
      "avg_duration_per_task_ms": 50542.31994645247,
      "avg_output_tokens_per_task": 7272.5394912985275,
      "last_run_timestamp": "2026-10-01T19:09:12.892Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Relace",
      "accuracy": 0.8846155538461539,
      "accuracy_std_dev": 0.0024989999999999735,
      "run_count": 2,
      "total_questions": 390,
      "avg_cost_per_task_usd": 0.10645926410256411,
      "avg_duration_per_task_ms": 514880.74102564104,
      "avg_output_tokens_per_task": 34685.94102564103,
      "last_run_timestamp": "2026-10-03T17:17:05.949Z"
    },
    {
      "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": 0.8845953462138955,
      "accuracy_std_dev": 0.012778996436692273,
      "run_count": 13,
      "total_questions": 2562,
      "avg_cost_per_task_usd": 0.007367392271662762,
      "avg_duration_per_task_ms": 467182.3048399688,
      "avg_output_tokens_per_task": 26353.96487119438,
      "last_run_timestamp": "2026-10-03T09:22:28.114Z"
    },
    {
      "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": 0.8845707161572053,
      "accuracy_std_dev": 0.02143221713073307,
      "run_count": 6,
      "total_questions": 1145,
      "avg_cost_per_task_usd": 0.16085208820960697,
      "avg_duration_per_task_ms": 756655.7283842795,
      "avg_output_tokens_per_task": 40813.194759825325,
      "last_run_timestamp": "2026-09-16T00:37:42.883Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-luna-20260709",
      "model_slug": "openai/gpt-5.6-luna",
      "model_name": "OpenAI: GPT-5.6 Luna",
      "provider_name": "Azure",
      "accuracy": 0.884269595505618,
      "accuracy_std_dev": 0.010583520080008851,
      "run_count": 9,
      "total_questions": 1780,
      "avg_cost_per_task_usd": 0.018860816853932583,
      "avg_duration_per_task_ms": 74171.62584269662,
      "avg_output_tokens_per_task": 7862.741573033708,
      "last_run_timestamp": "2026-10-01T19:12:42.994Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Together",
      "accuracy": 0.884259,
      "accuracy_std_dev": 0.003834285135458768,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.028357917929292934,
      "avg_duration_per_task_ms": 90698.91287878787,
      "avg_output_tokens_per_task": 23491.551767676767,
      "last_run_timestamp": "2026-10-03T10:07:51.005Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Phala",
      "accuracy": 0.8841843306451613,
      "accuracy_std_dev": 0.016935757620711884,
      "run_count": 4,
      "total_questions": 744,
      "avg_cost_per_task_usd": 0.04011120698924731,
      "avg_duration_per_task_ms": 259031.17204301077,
      "avg_output_tokens_per_task": 34712.397849462366,
      "last_run_timestamp": "2026-10-03T11:48:18.674Z"
    },
    {
      "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": 0.8840485,
      "accuracy_std_dev": 0.014246189262395764,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.18107499936868682,
      "avg_duration_per_task_ms": 298705.5252525252,
      "avg_output_tokens_per_task": 45444.8851010101,
      "last_run_timestamp": "2026-10-03T10:13:34.085Z"
    },
    {
      "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": 0.8839911818181818,
      "accuracy_std_dev": 0.015783702865574983,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.1157818631772268,
      "avg_duration_per_task_ms": 609822.4752066116,
      "avg_output_tokens_per_task": 30247.83930211203,
      "last_run_timestamp": "2026-10-03T16:57:22.957Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3-flash-preview-20251217",
      "model_slug": "google/gemini-3-flash-preview",
      "model_name": "Google: Gemini 3 Flash Preview",
      "provider_name": "auto-routing",
      "accuracy": 0.8838383333333335,
      "accuracy_std_dev": 0.001374572014200141,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.13180945791245793,
      "avg_duration_per_task_ms": 248970.3754208754,
      "avg_output_tokens_per_task": 43817.097643097644,
      "last_run_timestamp": "2026-10-01T19:15:31.426Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-luna-pro-20260922",
      "model_slug": "openai/gpt-6-luna-pro",
      "model_name": "OpenAI: GPT-6 Luna Pro",
      "provider_name": "auto-routing",
      "accuracy": 0.883838,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.008301368686868687,
      "avg_duration_per_task_ms": 72985.27777777778,
      "avg_output_tokens_per_task": 13098.969696969696,
      "last_run_timestamp": "2026-10-01T08:24:38.949Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.8832300830755635,
      "accuracy_std_dev": 0.01311356273493304,
      "run_count": 23,
      "total_questions": 4526,
      "avg_cost_per_task_usd": 0.1913586323464428,
      "avg_duration_per_task_ms": 377955.7154220062,
      "avg_output_tokens_per_task": 31161.343570481662,
      "last_run_timestamp": "2026-10-03T16:59:23.370Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Phala",
      "accuracy": 0.8829965000000001,
      "accuracy_std_dev": 0.009135988652028878,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.02256884242424242,
      "avg_duration_per_task_ms": 361325.05656565656,
      "avg_output_tokens_per_task": 25803.312121212122,
      "last_run_timestamp": "2026-10-03T11:18:47.195Z"
    },
    {
      "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": 0.882652041571654,
      "accuracy_std_dev": 0.010061769676464596,
      "run_count": 30,
      "total_questions": 5701,
      "avg_cost_per_task_usd": 0.13625698438870373,
      "avg_duration_per_task_ms": 62308.83844939484,
      "avg_output_tokens_per_task": 4947.5786704087,
      "last_run_timestamp": "2026-10-03T09:33:58.131Z"
    },
    {
      "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": 0.882236963045085,
      "accuracy_std_dev": 0.03276554586072114,
      "run_count": 7,
      "total_questions": 1353,
      "avg_cost_per_task_usd": 0.006029221729490022,
      "avg_duration_per_task_ms": 323803.38359201775,
      "avg_output_tokens_per_task": 25311.461197339246,
      "last_run_timestamp": "2026-10-03T09:49:39.330Z"
    },
    {
      "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": 0.8816051722365038,
      "accuracy_std_dev": 0.07456878886603124,
      "run_count": 13,
      "total_questions": 2334,
      "avg_cost_per_task_usd": 0.010089964010282775,
      "avg_duration_per_task_ms": 149337.25192802056,
      "avg_output_tokens_per_task": 23612.06083976007,
      "last_run_timestamp": "2026-10-03T10:12:41.974Z"
    },
    {
      "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-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Relace",
      "accuracy": 0.8815301729019211,
      "accuracy_std_dev": 0.0059611119566738745,
      "run_count": 5,
      "total_questions": 989,
      "avg_cost_per_task_usd": 0.014240532861476238,
      "avg_duration_per_task_ms": 332365.4651162791,
      "avg_output_tokens_per_task": 26073.906976744187,
      "last_run_timestamp": "2026-10-03T10:46:54.027Z"
    },
    {
      "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": 0.8815094743833018,
      "accuracy_std_dev": 0.009777163469023105,
      "run_count": 8,
      "total_questions": 1581,
      "avg_cost_per_task_usd": 0.008091048070841239,
      "avg_duration_per_task_ms": 269562.41682479443,
      "avg_output_tokens_per_task": 28593.919038583175,
      "last_run_timestamp": "2026-10-03T16:57:09.267Z"
    },
    {
      "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": 0.8813523333333333,
      "accuracy_std_dev": 0.012418655370047113,
      "run_count": 5,
      "total_questions": 774,
      "avg_cost_per_task_usd": 0.007396838501291989,
      "avg_duration_per_task_ms": 996564.0413436693,
      "avg_output_tokens_per_task": 29356.904392764856,
      "last_run_timestamp": "2026-10-03T09:53:30.020Z"
    },
    {
      "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": 0.8813128999999998,
      "accuracy_std_dev": 0.018445584861695216,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.024727196464646464,
      "avg_duration_per_task_ms": 216333.7505050505,
      "avg_output_tokens_per_task": 25254.168181818182,
      "last_run_timestamp": "2026-10-03T09:13:55.562Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Inceptron",
      "accuracy": 0.880903752293578,
      "accuracy_std_dev": 0.010625848964200456,
      "run_count": 10,
      "total_questions": 1962,
      "avg_cost_per_task_usd": 0.005325050458715596,
      "avg_duration_per_task_ms": 400241.4791029562,
      "avg_output_tokens_per_task": 18619.075433231395,
      "last_run_timestamp": "2026-10-03T10:33:52.640Z"
    },
    {
      "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": 0.8807518333333332,
      "accuracy_std_dev": 0.008440691666300569,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.015967906565656567,
      "avg_duration_per_task_ms": 289503.96885521885,
      "avg_output_tokens_per_task": 24394.69781144781,
      "last_run_timestamp": "2026-10-03T09:36:17.297Z"
    },
    {
      "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": 0.8804712499999999,
      "accuracy_std_dev": 0.006838655510222777,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.028478232323232322,
      "avg_duration_per_task_ms": 124855.71338383839,
      "avg_output_tokens_per_task": 23539.155303030304,
      "last_run_timestamp": "2026-10-03T10:18:21.349Z"
    },
    {
      "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": 0.880471,
      "accuracy_std_dev": 0.006734000000000018,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.006361709595959596,
      "avg_duration_per_task_ms": 516019.3207070707,
      "avg_output_tokens_per_task": 34951.01515151515,
      "last_run_timestamp": "2026-10-03T11:41:19.088Z"
    },
    {
      "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": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "GMICloud",
      "accuracy": 0.880471,
      "accuracy_std_dev": 0.003367000000000009,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.07564930808080808,
      "avg_duration_per_task_ms": 291613.64898989897,
      "avg_output_tokens_per_task": 20887.656565656565,
      "last_run_timestamp": "2026-08-29T19:12:43.653Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "auto-routing",
      "accuracy": 0.8801906666666667,
      "accuracy_std_dev": 0.011226983784119793,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.0227881835016835,
      "avg_duration_per_task_ms": 234700.68434343435,
      "avg_output_tokens_per_task": 25264.190235690236,
      "last_run_timestamp": "2026-10-03T09:30:18.263Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Venice",
      "accuracy": 0.8798590909090911,
      "accuracy_std_dev": 0.03263267695105222,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.14463881542699725,
      "avg_duration_per_task_ms": 346447.52938475664,
      "avg_output_tokens_per_task": 33431.59090909091,
      "last_run_timestamp": "2026-10-03T17:43:07.447Z"
    },
    {
      "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": 0.8796295,
      "accuracy_std_dev": 0.015267782951365261,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.028045234848484847,
      "avg_duration_per_task_ms": 222260.06818181818,
      "avg_output_tokens_per_task": 27117.741161616163,
      "last_run_timestamp": "2026-10-03T09:35:14.478Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openrouter/auto",
      "model_slug": "openrouter/auto",
      "model_name": "Auto Router",
      "provider_name": "auto-routing",
      "accuracy": 0.8795957999999999,
      "accuracy_std_dev": 0.026351833738091142,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.08917041111111111,
      "avg_duration_per_task_ms": 196788.44343434344,
      "avg_output_tokens_per_task": 37445.42121212121,
      "last_run_timestamp": "2026-09-29T14:28:17.204Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-opus-5-20260723",
      "model_slug": "anthropic/claude-opus-5",
      "model_name": "Anthropic: Claude Opus 5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.8794991295774648,
      "accuracy_std_dev": 0.01942029328961192,
      "run_count": 11,
      "total_questions": 2130,
      "avg_cost_per_task_usd": 0.12990812910798122,
      "avg_duration_per_task_ms": 63213.040375586854,
      "avg_output_tokens_per_task": 4970.991549295774,
      "last_run_timestamp": "2026-10-03T10:06:49.792Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Reka",
      "accuracy": 0.879189,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 189,
      "avg_cost_per_task_usd": 0.020723571428571428,
      "avg_duration_per_task_ms": 141095.54497354498,
      "avg_output_tokens_per_task": 17734.21164021164,
      "last_run_timestamp": "2026-09-15T00:33:26.568Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Fireworks",
      "accuracy": 0.8790546552363299,
      "accuracy_std_dev": 0.030844451113161732,
      "run_count": 6,
      "total_questions": 1079,
      "avg_cost_per_task_usd": 0.1346450278035218,
      "avg_duration_per_task_ms": 284974.0537534754,
      "avg_output_tokens_per_task": 31126.302131603337,
      "last_run_timestamp": "2026-09-19T09:14:09.704Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "GMICloud",
      "accuracy": 0.8790282857142856,
      "accuracy_std_dev": 0.014040008533923795,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.005288715728715729,
      "avg_duration_per_task_ms": 276008.51948051946,
      "avg_output_tokens_per_task": 28469.592352092353,
      "last_run_timestamp": "2026-10-03T17:07:12.901Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.2-20251211",
      "model_slug": "openai/gpt-5.2",
      "model_name": "OpenAI: GPT-5.2",
      "provider_name": "OpenAI",
      "accuracy": 0.878788,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.13332395959595958,
      "avg_duration_per_task_ms": 138478.24242424243,
      "avg_output_tokens_per_task": 9426.934343434343,
      "last_run_timestamp": "2026-08-01T08:50:28.136Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.7-plus-20260602",
      "model_slug": "qwen/qwen3.7-plus",
      "model_name": "Qwen: Qwen3.7 Plus",
      "provider_name": "auto-routing",
      "accuracy": 0.878788,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.043995565656565656,
      "avg_duration_per_task_ms": 588519.7272727273,
      "avg_output_tokens_per_task": 34166.82828282828,
      "last_run_timestamp": "2026-10-01T08:47:14.818Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-luna-20260709",
      "model_slug": "openai/gpt-5.6-luna",
      "model_name": "OpenAI: GPT-5.6 Luna",
      "provider_name": "auto-routing",
      "accuracy": 0.8787876666666666,
      "accuracy_std_dev": 0.005991714074916775,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.007907055555555554,
      "avg_duration_per_task_ms": 79561.398989899,
      "avg_output_tokens_per_task": 7834.653198653199,
      "last_run_timestamp": "2026-10-01T19:14:22.294Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "StreamLake",
      "accuracy": 0.8787876363636363,
      "accuracy_std_dev": 0.006091232690182722,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.005402353535353536,
      "avg_duration_per_task_ms": 380915.31404958677,
      "avg_output_tokens_per_task": 29413.866391184572,
      "last_run_timestamp": "2026-10-03T09:51:11.430Z"
    },
    {
      "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": 0.8787874999999999,
      "accuracy_std_dev": 0.005050500000000013,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.04341706060606061,
      "avg_duration_per_task_ms": 244250.4595959596,
      "avg_output_tokens_per_task": 37643.704545454544,
      "last_run_timestamp": "2026-10-03T10:51:03.639Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Fireworks",
      "accuracy": 0.8787801424227066,
      "accuracy_std_dev": 0.016401176693323544,
      "run_count": 11,
      "total_questions": 1973,
      "avg_cost_per_task_usd": 0.011341015205271159,
      "avg_duration_per_task_ms": 261740.6791687785,
      "avg_output_tokens_per_task": 24571.48302078054,
      "last_run_timestamp": "2026-09-19T08:54:58.652Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Together",
      "accuracy": 0.876843640328698,
      "accuracy_std_dev": 0.014698550138959288,
      "run_count": 8,
      "total_questions": 1582,
      "avg_cost_per_task_usd": 0.08638480594184576,
      "avg_duration_per_task_ms": 222833.85398230088,
      "avg_output_tokens_per_task": 21940.641592920354,
      "last_run_timestamp": "2026-10-03T09:40:59.529Z"
    },
    {
      "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": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "WandB",
      "accuracy": 0.8762625,
      "accuracy_std_dev": 0.002525500000000014,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.008190393939393938,
      "avg_duration_per_task_ms": 699109.8106060605,
      "avg_output_tokens_per_task": 28884.636363636364,
      "last_run_timestamp": "2026-07-18T08:31:23.912Z"
    },
    {
      "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": "WandB",
      "accuracy": 0.8762625,
      "accuracy_std_dev": 0.000841499999999995,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.15284478282828282,
      "avg_duration_per_task_ms": 339007.8914141414,
      "avg_output_tokens_per_task": 34511.28535353536,
      "last_run_timestamp": "2026-07-18T08:26:40.271Z"
    },
    {
      "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": 0.8761422857142855,
      "accuracy_std_dev": 0.023342020751758195,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.08516392496392496,
      "avg_duration_per_task_ms": 500065.10461760464,
      "avg_output_tokens_per_task": 35027.99062049062,
      "last_run_timestamp": "2026-10-03T17:33:23.858Z"
    },
    {
      "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": 0.8757659457720588,
      "accuracy_std_dev": 0.09251418420879137,
      "run_count": 11,
      "total_questions": 2176,
      "avg_cost_per_task_usd": 0.2399917977941176,
      "avg_duration_per_task_ms": 482095.9535845588,
      "avg_output_tokens_per_task": 59848.47564338235,
      "last_run_timestamp": "2026-10-03T10:18:13.963Z"
    },
    {
      "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": 0.8756311250000001,
      "accuracy_std_dev": 0.007549310638023532,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.04973767171717172,
      "avg_duration_per_task_ms": 283165.73863636365,
      "avg_output_tokens_per_task": 21128.222222222223,
      "last_run_timestamp": "2026-10-03T17:26:10.527Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Ionstream",
      "accuracy": 0.8754206666666666,
      "accuracy_std_dev": 0.00901385812081723,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.061844909090909096,
      "avg_duration_per_task_ms": 262173.4595959596,
      "avg_output_tokens_per_task": 22965.14478114478,
      "last_run_timestamp": "2026-09-26T09:19:06.449Z"
    },
    {
      "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": 0.8753905728900255,
      "accuracy_std_dev": 0.012094409094012564,
      "run_count": 6,
      "total_questions": 1173,
      "avg_cost_per_task_usd": 0.035846069053708435,
      "avg_duration_per_task_ms": 251997.9505541347,
      "avg_output_tokens_per_task": 29905.09377664109,
      "last_run_timestamp": "2026-10-03T12:29:29.078Z"
    },
    {
      "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": 0.8748596666666666,
      "accuracy_std_dev": 0.03480145848157005,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.10809361952861953,
      "avg_duration_per_task_ms": 516538.27384960715,
      "avg_output_tokens_per_task": 34956.31930415264,
      "last_run_timestamp": "2026-10-03T09:34:05.323Z"
    },
    {
      "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": 0.8747592697495183,
      "accuracy_std_dev": 0.04128530357388087,
      "run_count": 4,
      "total_questions": 519,
      "avg_cost_per_task_usd": 0.012099612716763006,
      "avg_duration_per_task_ms": 465229.36030828516,
      "avg_output_tokens_per_task": 26222.763005780347,
      "last_run_timestamp": "2026-10-03T09:28:13.430Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-opus-5-20260723",
      "model_slug": "anthropic/claude-opus-5",
      "model_name": "Anthropic: Claude Opus 5",
      "provider_name": "auto-routing",
      "accuracy": 0.8746725555555557,
      "accuracy_std_dev": 0.018426695846040054,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.13061521717171717,
      "avg_duration_per_task_ms": 63177.591470258136,
      "avg_output_tokens_per_task": 5005.783950617284,
      "last_run_timestamp": "2026-10-03T09:54:43.475Z"
    },
    {
      "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": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "CoreWeave",
      "accuracy": 0.8743494545454547,
      "accuracy_std_dev": 0.0150964078771537,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.1106280968778696,
      "avg_duration_per_task_ms": 234771.5486685032,
      "avg_output_tokens_per_task": 31461.72359963269,
      "last_run_timestamp": "2026-10-03T16:40:49.490Z"
    },
    {
      "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": 0.874209816687737,
      "accuracy_std_dev": 0.005925415300582321,
      "run_count": 8,
      "total_questions": 1582,
      "avg_cost_per_task_usd": 0.12289277560050568,
      "avg_duration_per_task_ms": 965329.4336283186,
      "avg_output_tokens_per_task": 43216.86472819216,
      "last_run_timestamp": "2026-10-03T17:24:45.959Z"
    },
    {
      "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": 0.8741581249999999,
      "accuracy_std_dev": 0.004890233185582785,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.07734060606060605,
      "avg_duration_per_task_ms": 324821.60984848486,
      "avg_output_tokens_per_task": 21989.39962121212,
      "last_run_timestamp": "2026-10-03T19:00:32.081Z"
    },
    {
      "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": 0.8740979999999999,
      "accuracy_std_dev": 0.06811118283890162,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.12654391486291486,
      "avg_duration_per_task_ms": 1351779.769119769,
      "avg_output_tokens_per_task": 56150.230158730155,
      "last_run_timestamp": "2026-10-03T17:53:34.568Z"
    },
    {
      "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": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "GMICloud",
      "accuracy": 0.8736602893401014,
      "accuracy_std_dev": 0.018240759785588822,
      "run_count": 3,
      "total_questions": 591,
      "avg_cost_per_task_usd": 0.033190517766497465,
      "avg_duration_per_task_ms": 731117.2707275804,
      "avg_output_tokens_per_task": 38434.73942470389,
      "last_run_timestamp": "2026-10-03T11:12:58.942Z"
    },
    {
      "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": 0.87288105,
      "accuracy_std_dev": 0.008961225878751214,
      "run_count": 6,
      "total_questions": 1180,
      "avg_cost_per_task_usd": 0.11474949322033899,
      "avg_duration_per_task_ms": 657555.8720338983,
      "avg_output_tokens_per_task": 32804.78050847458,
      "last_run_timestamp": "2026-10-03T16:53:50.088Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2-thinking-20251106",
      "model_slug": "moonshotai/kimi-k2-thinking",
      "model_name": "MoonshotAI: Kimi K2 Thinking",
      "provider_name": "auto-routing",
      "accuracy": 0.8727692590673575,
      "accuracy_std_dev": 0.006680537619749563,
      "run_count": 3,
      "total_questions": 579,
      "avg_cost_per_task_usd": 0.10616132815198619,
      "avg_duration_per_task_ms": 1070108.6269430052,
      "avg_output_tokens_per_task": 42555.205526770296,
      "last_run_timestamp": "2026-08-01T17:53:10.275Z"
    },
    {
      "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": 0.8727401757049892,
      "accuracy_std_dev": 0.012263522167761748,
      "run_count": 7,
      "total_questions": 1383,
      "avg_cost_per_task_usd": 0.00852685032537961,
      "avg_duration_per_task_ms": 434756.32754880696,
      "avg_output_tokens_per_task": 21739.99276934201,
      "last_run_timestamp": "2026-10-03T10:06:09.405Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2-thinking-20251106",
      "model_slug": "moonshotai/kimi-k2-thinking",
      "model_name": "MoonshotAI: Kimi K2 Thinking",
      "provider_name": "Novita",
      "accuracy": 0.8726002303664923,
      "accuracy_std_dev": 0.013852891908270337,
      "run_count": 3,
      "total_questions": 573,
      "avg_cost_per_task_usd": 0.10812300000000001,
      "avg_duration_per_task_ms": 1136520.2198952879,
      "avg_output_tokens_per_task": 43220.56020942408,
      "last_run_timestamp": "2026-08-01T17:46:48.312Z"
    },
    {
      "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": 0.8725159619400473,
      "accuracy_std_dev": 0.0310416281682088,
      "run_count": 15,
      "total_questions": 2969,
      "avg_cost_per_task_usd": 0.14988813371505558,
      "avg_duration_per_task_ms": 412802.2492421691,
      "avg_output_tokens_per_task": 11323.438531492086,
      "last_run_timestamp": "2026-10-03T09:38:56.367Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-luna-20260709",
      "model_slug": "openai/gpt-5.6-luna",
      "model_name": "OpenAI: GPT-5.6 Luna",
      "provider_name": "OpenAI",
      "accuracy": 0.8724745,
      "accuracy_std_dev": 0.008858340547190563,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.008325782828282828,
      "avg_duration_per_task_ms": 77307.88005050505,
      "avg_output_tokens_per_task": 7842.704545454545,
      "last_run_timestamp": "2026-10-02T23:19:10.742Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Venice",
      "accuracy": 0.872409485103627,
      "accuracy_std_dev": 0.04752937711760483,
      "run_count": 8,
      "total_questions": 1544,
      "avg_cost_per_task_usd": 0.056010884067357505,
      "avg_duration_per_task_ms": 651555.3283678757,
      "avg_output_tokens_per_task": 19822.61658031088,
      "last_run_timestamp": "2026-10-03T17:12:28.043Z"
    },
    {
      "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": 0.8722849626411816,
      "accuracy_std_dev": 0.03667207040635816,
      "run_count": 6,
      "total_questions": 1151,
      "avg_cost_per_task_usd": 0.01310653171155517,
      "avg_duration_per_task_ms": 596394.4969591659,
      "avg_output_tokens_per_task": 34928.014769765425,
      "last_run_timestamp": "2026-10-03T09:08:44.457Z"
    },
    {
      "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": 0.8722010494296577,
      "accuracy_std_dev": 0.0020781669055925365,
      "run_count": 4,
      "total_questions": 789,
      "avg_cost_per_task_usd": 0.045619747782002534,
      "avg_duration_per_task_ms": 189717.51711026617,
      "avg_output_tokens_per_task": 39672.14828897338,
      "last_run_timestamp": "2026-10-03T10:44:50.348Z"
    },
    {
      "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": 0.8720537142857142,
      "accuracy_std_dev": 0.014113751204453896,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.22260263059163057,
      "avg_duration_per_task_ms": 1170078.2395382395,
      "avg_output_tokens_per_task": 55880.77417027417,
      "last_run_timestamp": "2026-10-03T16:37:45.002Z"
    },
    {
      "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": 0.8718853999999999,
      "accuracy_std_dev": 0.004467018719459339,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.06354945656565657,
      "avg_duration_per_task_ms": 279330.9616161616,
      "avg_output_tokens_per_task": 21096.323232323233,
      "last_run_timestamp": "2026-08-15T15:55:02.752Z"
    },
    {
      "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-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Together",
      "accuracy": 0.8715861506361324,
      "accuracy_std_dev": 0.009632287238761115,
      "run_count": 10,
      "total_questions": 1965,
      "avg_cost_per_task_usd": 0.008127365394402036,
      "avg_duration_per_task_ms": 324970.47837150126,
      "avg_output_tokens_per_task": 28742.415267175573,
      "last_run_timestamp": "2026-10-03T09:52:27.516Z"
    },
    {
      "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": 0.8713843774220725,
      "accuracy_std_dev": 0.007143278612327366,
      "run_count": 6,
      "total_questions": 1187,
      "avg_cost_per_task_usd": 0.07020013732097725,
      "avg_duration_per_task_ms": 276688.6663858467,
      "avg_output_tokens_per_task": 11428.55096882898,
      "last_run_timestamp": "2026-10-03T10:04:35.741Z"
    },
    {
      "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": 0.8710750330630069,
      "accuracy_std_dev": 0.028110456226890915,
      "run_count": 9,
      "total_questions": 1603,
      "avg_cost_per_task_usd": 0.09870775982532752,
      "avg_duration_per_task_ms": 221158.11790393014,
      "avg_output_tokens_per_task": 26829.437305053027,
      "last_run_timestamp": "2026-09-26T09:19:13.012Z"
    },
    {
      "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": 0.8710353744745891,
      "accuracy_std_dev": 0.16260220955169502,
      "run_count": 15,
      "total_questions": 2617,
      "avg_cost_per_task_usd": 0.142784888803974,
      "avg_duration_per_task_ms": 265667.35957202903,
      "avg_output_tokens_per_task": 10833.450133740926,
      "last_run_timestamp": "2026-10-03T10:04:25.516Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "thinkingmachines/inkling-small-20260730",
      "model_slug": "thinkingmachines/inkling-small",
      "model_name": "Thinking Machines: Inkling Small",
      "provider_name": "Together",
      "accuracy": 0.8710116425661915,
      "accuracy_std_dev": 0.01025225770257459,
      "run_count": 5,
      "total_questions": 982,
      "avg_cost_per_task_usd": 0.025059682281059068,
      "avg_duration_per_task_ms": 328330.6405295316,
      "avg_output_tokens_per_task": 20642.19348268839,
      "last_run_timestamp": "2026-09-12T08:54:29.103Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "StreamLake",
      "accuracy": 0.87037025,
      "accuracy_std_dev": 0.013624975814198723,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.028775198232323235,
      "avg_duration_per_task_ms": 719468.7878787878,
      "avg_output_tokens_per_task": 34862.15782828283,
      "last_run_timestamp": "2026-10-03T16:57:03.670Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "thinkingmachines/inkling-small-20260730",
      "model_slug": "thinkingmachines/inkling-small",
      "model_name": "Thinking Machines: Inkling Small",
      "provider_name": "BaseTen",
      "accuracy": 0.8703701666666666,
      "accuracy_std_dev": 0.006941288747215641,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.024583196127946128,
      "avg_duration_per_task_ms": 150981.7457912458,
      "avg_output_tokens_per_task": 20283.477272727272,
      "last_run_timestamp": "2026-09-19T08:49:51.735Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "WandB",
      "accuracy": 0.870085,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 195,
      "avg_cost_per_task_usd": 0.11565565128205128,
      "avg_duration_per_task_ms": 2243159.2153846156,
      "avg_output_tokens_per_task": 32886.94358974359,
      "last_run_timestamp": "2026-07-11T08:42:46.205Z"
    },
    {
      "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": 0.8699495000000002,
      "accuracy_std_dev": 0.018722910197135487,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.030484161616161617,
      "avg_duration_per_task_ms": 134273.03535353535,
      "avg_output_tokens_per_task": 21527.786616161615,
      "last_run_timestamp": "2026-10-03T12:27:06.948Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Ionstream",
      "accuracy": 0.8699495,
      "accuracy_std_dev": 0.002104500000000009,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.00502325,
      "avg_duration_per_task_ms": 389434.8611111111,
      "avg_output_tokens_per_task": 17944.10101010101,
      "last_run_timestamp": "2026-08-08T10:05:07.443Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Parasail",
      "accuracy": 0.8696488571428571,
      "accuracy_std_dev": 0.005236127806706348,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.008200564935064936,
      "avg_duration_per_task_ms": 437965.4329004329,
      "avg_output_tokens_per_task": 29027.78354978355,
      "last_run_timestamp": "2026-10-03T16:58:32.515Z"
    },
    {
      "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": 0.8695323612368023,
      "accuracy_std_dev": 0.009903791381808026,
      "run_count": 8,
      "total_questions": 1326,
      "avg_cost_per_task_usd": 0.0508528371040724,
      "avg_duration_per_task_ms": 417388.269984917,
      "avg_output_tokens_per_task": 16859.24886877828,
      "last_run_timestamp": "2026-10-03T18:37:36.091Z"
    },
    {
      "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": "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": 0.8695285000000001,
      "accuracy_std_dev": 0.010411879221190886,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.07784478367003367,
      "avg_duration_per_task_ms": 107880.84595959596,
      "avg_output_tokens_per_task": 12619.909932659933,
      "last_run_timestamp": "2026-10-03T10:07:47.560Z"
    },
    {
      "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": 0.8692988181818181,
      "accuracy_std_dev": 0.01574807678773497,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.008268114784205692,
      "avg_duration_per_task_ms": 218163.90817263545,
      "avg_output_tokens_per_task": 29305.03948576676,
      "last_run_timestamp": "2026-10-03T10:06:44.341Z"
    },
    {
      "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-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Venice",
      "accuracy": 0.869276,
      "accuracy_std_dev": 0.008908612349855623,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.006512984848484849,
      "avg_duration_per_task_ms": 255242.29242424242,
      "avg_output_tokens_per_task": 18235.289393939394,
      "last_run_timestamp": "2026-10-03T09:02:06.104Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.1-20251113",
      "model_slug": "openai/gpt-5.1",
      "model_name": "OpenAI: GPT-5.1",
      "provider_name": "OpenAI",
      "accuracy": 0.8692479999999999,
      "accuracy_std_dev": 0.004827531943619518,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.22171807575757574,
      "avg_duration_per_task_ms": 260359.50673400675,
      "avg_output_tokens_per_task": 22076.11111111111,
      "last_run_timestamp": "2026-10-01T16:12:49.000Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.8692478333333333,
      "accuracy_std_dev": 0.012955574635611068,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.08235681228956229,
      "avg_duration_per_task_ms": 202641.79545454544,
      "avg_output_tokens_per_task": 13365.021885521886,
      "last_run_timestamp": "2026-10-03T09:43:24.655Z"
    },
    {
      "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": 0.8691446346328195,
      "accuracy_std_dev": 0.025234963139792104,
      "run_count": 6,
      "total_questions": 1103,
      "avg_cost_per_task_usd": 0.01014118223028105,
      "avg_duration_per_task_ms": 374535.42611060746,
      "avg_output_tokens_per_task": 21668.27742520399,
      "last_run_timestamp": "2026-10-03T09:34:29.241Z"
    },
    {
      "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": 0.8691187682758621,
      "accuracy_std_dev": 0.026187627614316582,
      "run_count": 11,
      "total_questions": 2175,
      "avg_cost_per_task_usd": 0.1136188703448276,
      "avg_duration_per_task_ms": 769491.6009195403,
      "avg_output_tokens_per_task": 30218.06206896552,
      "last_run_timestamp": "2026-10-03T16:47:49.662Z"
    },
    {
      "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": 0.8691185958620689,
      "accuracy_std_dev": 0.09871592059712983,
      "run_count": 11,
      "total_questions": 2175,
      "avg_cost_per_task_usd": 0.026795262528735632,
      "avg_duration_per_task_ms": 215775.31816091953,
      "avg_output_tokens_per_task": 23251.332413793105,
      "last_run_timestamp": "2026-10-03T17:06:35.452Z"
    },
    {
      "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": "Ambient",
      "accuracy": 0.868687,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.13888562626262627,
      "avg_duration_per_task_ms": 1028019.2020202021,
      "avg_output_tokens_per_task": 46184.969696969696,
      "last_run_timestamp": "2026-07-18T16:16:27.436Z"
    },
    {
      "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": 0.8686868571428572,
      "accuracy_std_dev": 0.007365644001105447,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.025638126262626258,
      "avg_duration_per_task_ms": 240783.4256854257,
      "avg_output_tokens_per_task": 21122.886002886004,
      "last_run_timestamp": "2026-10-01T16:07:33.196Z"
    },
    {
      "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": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Fireworks",
      "accuracy": 0.8684572837465565,
      "accuracy_std_dev": 0.006350273025626523,
      "run_count": 4,
      "total_questions": 726,
      "avg_cost_per_task_usd": 0.0059327286501377405,
      "avg_duration_per_task_ms": 153900.47520661156,
      "avg_output_tokens_per_task": 21060.526170798898,
      "last_run_timestamp": "2026-08-08T09:25:45.716Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "Novita",
      "accuracy": 0.8679651428571429,
      "accuracy_std_dev": 0.004916957259716384,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.07019697691197692,
      "avg_duration_per_task_ms": 226467.18253968254,
      "avg_output_tokens_per_task": 11432.188311688311,
      "last_run_timestamp": "2026-10-03T09:14:11.063Z"
    },
    {
      "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": 0.8675644999999997,
      "accuracy_std_dev": 0.008112634749779732,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.07749805892255893,
      "avg_duration_per_task_ms": 84869.17171717172,
      "avg_output_tokens_per_task": 12670.825757575758,
      "last_run_timestamp": "2026-10-03T09:34:51.690Z"
    },
    {
      "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": "Alibaba",
      "accuracy": 0.86726020610687,
      "accuracy_std_dev": 0.0104972985066397,
      "run_count": 8,
      "total_questions": 1572,
      "avg_cost_per_task_usd": 0.008138582061068702,
      "avg_duration_per_task_ms": 283436.84605597967,
      "avg_output_tokens_per_task": 30216.644402035625,
      "last_run_timestamp": "2026-10-03T17:03:44.554Z"
    },
    {
      "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": 0.8669514,
      "accuracy_std_dev": 0.020262840823783834,
      "run_count": 6,
      "total_questions": 1170,
      "avg_cost_per_task_usd": 0.07919805042735042,
      "avg_duration_per_task_ms": 415100.8598290598,
      "avg_output_tokens_per_task": 23580.724786324787,
      "last_run_timestamp": "2026-10-03T09:45:55.357Z"
    },
    {
      "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": 0.866915421641791,
      "accuracy_std_dev": 0.02922832992755772,
      "run_count": 12,
      "total_questions": 2144,
      "avg_cost_per_task_usd": 0.010834175839552238,
      "avg_duration_per_task_ms": 328640.2140858209,
      "avg_output_tokens_per_task": 19735.994869402984,
      "last_run_timestamp": "2026-10-03T09:27:10.840Z"
    },
    {
      "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": 0.8667625714285713,
      "accuracy_std_dev": 0.021717036478022374,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.11692872871572872,
      "avg_duration_per_task_ms": 576009.277056277,
      "avg_output_tokens_per_task": 34814.431457431456,
      "last_run_timestamp": "2026-10-03T17:19:05.730Z"
    },
    {
      "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": 0.8663944783673468,
      "accuracy_std_dev": 0.025544015942873897,
      "run_count": 7,
      "total_questions": 1225,
      "avg_cost_per_task_usd": 0.14541799102040817,
      "avg_duration_per_task_ms": 562916.0302040817,
      "avg_output_tokens_per_task": 51076.03020408163,
      "last_run_timestamp": "2026-10-03T17:01:24.374Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "StreamLake",
      "accuracy": 0.8663719999999998,
      "accuracy_std_dev": 0.0061821103395199835,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.0037658851010101005,
      "avg_duration_per_task_ms": 453040.5378787879,
      "avg_output_tokens_per_task": 29702.50568181818,
      "last_run_timestamp": "2026-10-03T17:32:07.433Z"
    },
    {
      "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": 0.8660411428571428,
      "accuracy_std_dev": 0.02151917954774678,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.1960267453102453,
      "avg_duration_per_task_ms": 1120544.6767676768,
      "avg_output_tokens_per_task": 58159.38816738817,
      "last_run_timestamp": "2026-10-03T17:03:53.018Z"
    },
    {
      "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": 0.8658246999999999,
      "accuracy_std_dev": 0.016649641828279688,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.08718337929292928,
      "avg_duration_per_task_ms": 202685.16717171718,
      "avg_output_tokens_per_task": 24084.222222222223,
      "last_run_timestamp": "2026-10-03T16:06:47.442Z"
    },
    {
      "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": 0.8658007142857141,
      "accuracy_std_dev": 0.009679853093842238,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.06976018037518038,
      "avg_duration_per_task_ms": 223743.95815295816,
      "avg_output_tokens_per_task": 11357.300144300145,
      "last_run_timestamp": "2026-10-03T09:30:30.956Z"
    },
    {
      "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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Inceptron",
      "accuracy": 0.865536908952959,
      "accuracy_std_dev": 0.032953765579824115,
      "run_count": 10,
      "total_questions": 1977,
      "avg_cost_per_task_usd": 0.10668413707637835,
      "avg_duration_per_task_ms": 700483.68892261,
      "avg_output_tokens_per_task": 37983.61153262519,
      "last_run_timestamp": "2026-10-03T16:56:36.095Z"
    },
    {
      "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": 0.8654539693682195,
      "accuracy_std_dev": 0.014877891890633394,
      "run_count": 8,
      "total_questions": 1567,
      "avg_cost_per_task_usd": 0.07507311359285258,
      "avg_duration_per_task_ms": 281173.1014677728,
      "avg_output_tokens_per_task": 20728.63433312061,
      "last_run_timestamp": "2026-10-03T17:17:14.917Z"
    },
    {
      "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": 0.8653197142857143,
      "accuracy_std_dev": 0.01749528025100555,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.16135840259740258,
      "avg_duration_per_task_ms": 1133723.20995671,
      "avg_output_tokens_per_task": 54192.0101010101,
      "last_run_timestamp": "2026-10-03T17:43:48.193Z"
    },
    {
      "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": 0.8653196666666666,
      "accuracy_std_dev": 0.010378007269649058,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.029422861952861954,
      "avg_duration_per_task_ms": 503984.98484848486,
      "avg_output_tokens_per_task": 34386.4898989899,
      "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": "BaseTen",
      "accuracy": 0.8651969388477632,
      "accuracy_std_dev": 0.0392226856144323,
      "run_count": 16,
      "total_questions": 3107,
      "avg_cost_per_task_usd": 0.004922328290955906,
      "avg_duration_per_task_ms": 103773.42162858062,
      "avg_output_tokens_per_task": 18562.228194399744,
      "last_run_timestamp": "2026-10-03T10:10:22.431Z"
    },
    {
      "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": 0.8649764115566037,
      "accuracy_std_dev": 0.05470934774460394,
      "run_count": 5,
      "total_questions": 848,
      "avg_cost_per_task_usd": 0.08730664976415095,
      "avg_duration_per_task_ms": 645412.391509434,
      "avg_output_tokens_per_task": 26303.754716981133,
      "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": "GMICloud",
      "accuracy": 0.8649528761766837,
      "accuracy_std_dev": 0.01859749373173817,
      "run_count": 7,
      "total_questions": 1381,
      "avg_cost_per_task_usd": 0.0065336495293265755,
      "avg_duration_per_task_ms": 348157.4265025344,
      "avg_output_tokens_per_task": 21785.366401158582,
      "last_run_timestamp": "2026-10-03T10:20:35.518Z"
    },
    {
      "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": 0.864636,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 197,
      "avg_cost_per_task_usd": 0.07161002030456853,
      "avg_duration_per_task_ms": 868248.1269035533,
      "avg_output_tokens_per_task": 28363.54314720812,
      "last_run_timestamp": "2026-10-03T14:58:52.397Z"
    },
    {
      "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": 0.8642566289198608,
      "accuracy_std_dev": 0.016122613680445513,
      "run_count": 6,
      "total_questions": 1148,
      "avg_cost_per_task_usd": 0.08601719425087108,
      "avg_duration_per_task_ms": 391195.4764808362,
      "avg_output_tokens_per_task": 24802.604529616725,
      "last_run_timestamp": "2026-10-03T09:33:56.518Z"
    },
    {
      "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": 0.864098116850127,
      "accuracy_std_dev": 0.012040853347389158,
      "run_count": 6,
      "total_questions": 1181,
      "avg_cost_per_task_usd": 0.06984978577476715,
      "avg_duration_per_task_ms": 203374.89754445385,
      "avg_output_tokens_per_task": 24608.592718035565,
      "last_run_timestamp": "2026-10-03T09:19:06.149Z"
    },
    {
      "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": 0.8639189175738725,
      "accuracy_std_dev": 0.02340391523335358,
      "run_count": 4,
      "total_questions": 643,
      "avg_cost_per_task_usd": 0.06298315552099534,
      "avg_duration_per_task_ms": 414941.54743390356,
      "avg_output_tokens_per_task": 21263.947122861588,
      "last_run_timestamp": "2026-08-15T08:47:51.369Z"
    },
    {
      "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": 0.8638768571428571,
      "accuracy_std_dev": 0.019098135020606234,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.1702009898989899,
      "avg_duration_per_task_ms": 1092527.1017316016,
      "avg_output_tokens_per_task": 59548.17532467532,
      "last_run_timestamp": "2026-10-03T17:43:01.166Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Fireworks",
      "accuracy": 0.8636704174363595,
      "accuracy_std_dev": 0.014066506514437398,
      "run_count": 26,
      "total_questions": 4439,
      "avg_cost_per_task_usd": 0.1642218321694075,
      "avg_duration_per_task_ms": 249845.5370578959,
      "avg_output_tokens_per_task": 30059.050236539762,
      "last_run_timestamp": "2026-09-26T08:54:22.466Z"
    },
    {
      "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": 0.8636361249999999,
      "accuracy_std_dev": 0.010205573274411139,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.008622361111111112,
      "avg_duration_per_task_ms": 290441.0890151515,
      "avg_output_tokens_per_task": 30478.675505050505,
      "last_run_timestamp": "2026-10-03T17:35:32.100Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Nebius",
      "accuracy": 0.863636,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0938164393939394,
      "avg_duration_per_task_ms": 667365.3939393939,
      "avg_output_tokens_per_task": 23446.50505050505,
      "last_run_timestamp": "2026-09-10T01:44:54.935Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "OpenInference",
      "accuracy": 0.863636,
      "accuracy_std_dev": 0.007273551310513057,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.020974939393939393,
      "avg_duration_per_task_ms": 363755.6632996633,
      "avg_output_tokens_per_task": 34831.013468013465,
      "last_run_timestamp": "2026-10-03T11:53:46.126Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.1-20251113",
      "model_slug": "openai/gpt-5.1",
      "model_name": "OpenAI: GPT-5.1",
      "provider_name": "auto-routing",
      "accuracy": 0.863636,
      "accuracy_std_dev": 0.004761657064510171,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.22125699158249157,
      "avg_duration_per_task_ms": 259075.24915824915,
      "avg_output_tokens_per_task": 22030.511784511786,
      "last_run_timestamp": "2026-10-01T16:12:54.241Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "AtlasCloud",
      "accuracy": 0.863005,
      "accuracy_std_dev": 0.016749196577746643,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.07378729608585859,
      "avg_duration_per_task_ms": 283540.4873737374,
      "avg_output_tokens_per_task": 20998.69318181818,
      "last_run_timestamp": "2026-10-03T17:13:56.806Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "GMICloud",
      "accuracy": 0.8628582711496745,
      "accuracy_std_dev": 0.021376331990005663,
      "run_count": 7,
      "total_questions": 1383,
      "avg_cost_per_task_usd": 0.058844402024584234,
      "avg_duration_per_task_ms": 505421.9110629067,
      "avg_output_tokens_per_task": 33506.749096167754,
      "last_run_timestamp": "2026-10-03T17:23:31.069Z"
    },
    {
      "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": 0.8625138888888889,
      "accuracy_std_dev": 0.02716878807440808,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.00797921548821549,
      "avg_duration_per_task_ms": 504816.8608305275,
      "avg_output_tokens_per_task": 25080.10830527497,
      "last_run_timestamp": "2026-10-03T10:11:44.532Z"
    },
    {
      "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": 0.8625137777777779,
      "accuracy_std_dev": 0.00880146315838422,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.8542647558922559,
      "avg_duration_per_task_ms": 483293.1627384961,
      "avg_output_tokens_per_task": 33979.86139169472,
      "last_run_timestamp": "2026-10-03T09:32:40.286Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.1",
      "model_slug": "minimax/minimax-m2.1",
      "model_name": "MiniMax: MiniMax M2.1",
      "provider_name": "auto-routing",
      "accuracy": 0.8623800507614213,
      "accuracy_std_dev": 0.004695390245289045,
      "run_count": 3,
      "total_questions": 591,
      "avg_cost_per_task_usd": 0.04025210152284264,
      "avg_duration_per_task_ms": 352805.59390862944,
      "avg_output_tokens_per_task": 24224.402707275804,
      "last_run_timestamp": "2026-08-01T17:57:47.536Z"
    },
    {
      "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": 0.8620937891696752,
      "accuracy_std_dev": 0.031861011712545156,
      "run_count": 7,
      "total_questions": 1385,
      "avg_cost_per_task_usd": 0.12809203682310472,
      "avg_duration_per_task_ms": 701690.7270758123,
      "avg_output_tokens_per_task": 42897.71407942238,
      "last_run_timestamp": "2026-08-29T18:25:54.650Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Ionstream",
      "accuracy": 0.8619530000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.008309641414141413,
      "avg_duration_per_task_ms": 730613.3737373737,
      "avg_output_tokens_per_task": 29358.666666666668,
      "last_run_timestamp": "2026-07-31T04:09:58.076Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.5-opus-20251124",
      "model_slug": "anthropic/claude-opus-4.5",
      "model_name": "Anthropic: Claude Opus 4.5",
      "provider_name": "auto-routing",
      "accuracy": 0.8617657777777777,
      "accuracy_std_dev": 0.009167555942050057,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.8453705723905725,
      "avg_duration_per_task_ms": 415699.1722783389,
      "avg_output_tokens_per_task": 33623.734006734005,
      "last_run_timestamp": "2026-10-03T09:39:37.575Z"
    },
    {
      "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": 0.861742375,
      "accuracy_std_dev": 0.01475920397529538,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.054362910353535354,
      "avg_duration_per_task_ms": 712745.5782828283,
      "avg_output_tokens_per_task": 33336.231691919194,
      "last_run_timestamp": "2026-10-03T17:03:53.250Z"
    },
    {
      "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": 0.8616955969987624,
      "accuracy_std_dev": 0.021294415333698184,
      "run_count": 33,
      "total_questions": 6464,
      "avg_cost_per_task_usd": 0.2610163531868812,
      "avg_duration_per_task_ms": 338765.33369430696,
      "avg_output_tokens_per_task": 47192.89201732673,
      "last_run_timestamp": "2026-10-03T17:31:59.643Z"
    },
    {
      "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": 0.8615786666666667,
      "accuracy_std_dev": 0.009382650075775203,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.8094348821548821,
      "avg_duration_per_task_ms": 461988.8821548821,
      "avg_output_tokens_per_task": 32205.768237934906,
      "last_run_timestamp": "2026-10-03T10:13:16.654Z"
    },
    {
      "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": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "CoreWeave",
      "accuracy": 0.8613915000000001,
      "accuracy_std_dev": 0.015601910243620808,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.19432850925925926,
      "avg_duration_per_task_ms": 386703.5462962963,
      "avg_output_tokens_per_task": 56841.80892255892,
      "last_run_timestamp": "2026-10-03T16:58:25.704Z"
    },
    {
      "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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "DeepInfra",
      "accuracy": 0.8611875454545455,
      "accuracy_std_dev": 0.05143184709419506,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.07813096464646466,
      "avg_duration_per_task_ms": 519431.6776859504,
      "avg_output_tokens_per_task": 37464.50964187328,
      "last_run_timestamp": "2026-10-03T17:22:47.587Z"
    },
    {
      "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": 0.8611535819295559,
      "accuracy_std_dev": 0.018922304290795815,
      "run_count": 7,
      "total_questions": 1306,
      "avg_cost_per_task_usd": 0.005349909647779478,
      "avg_duration_per_task_ms": 768463.6661562021,
      "avg_output_tokens_per_task": 29577.52986217458,
      "last_run_timestamp": "2026-10-03T17:40:37.174Z"
    },
    {
      "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": 0.8611110624999999,
      "accuracy_std_dev": 0.01746489113360268,
      "run_count": 16,
      "total_questions": 3168,
      "avg_cost_per_task_usd": 0.11371201957070708,
      "avg_duration_per_task_ms": 408202.1710858586,
      "avg_output_tokens_per_task": 40614.27525252525,
      "last_run_timestamp": "2026-10-03T17:10:11.381Z"
    },
    {
      "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": "ModelRun",
      "accuracy": 0.861111,
      "accuracy_std_dev": 0.005892000000000008,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.1977418813131313,
      "avg_duration_per_task_ms": 587463.0883838384,
      "avg_output_tokens_per_task": 58044.29797979798,
      "last_run_timestamp": "2026-07-18T08:26:48.377Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.1",
      "model_slug": "minimax/minimax-m2.1",
      "model_name": "MiniMax: MiniMax M2.1",
      "provider_name": "Minimax",
      "accuracy": 0.8610391240310077,
      "accuracy_std_dev": 0.010899356367337553,
      "run_count": 6,
      "total_questions": 1161,
      "avg_cost_per_task_usd": 0.05256056416881998,
      "avg_duration_per_task_ms": 426986.81912144704,
      "avg_output_tokens_per_task": 27212.199827734712,
      "last_run_timestamp": "2026-08-01T17:44:01.464Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "DekaLLM",
      "accuracy": 0.8608923543307087,
      "accuracy_std_dev": 0.017439499999999997,
      "run_count": 2,
      "total_questions": 254,
      "avg_cost_per_task_usd": 0.003837814960629921,
      "avg_duration_per_task_ms": 298144.1771653543,
      "avg_output_tokens_per_task": 11523.807086614173,
      "last_run_timestamp": "2026-10-03T17:04:09.464Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.8607369444444446,
      "accuracy_std_dev": 0.008186229884332315,
      "run_count": 18,
      "total_questions": 3564,
      "avg_cost_per_task_usd": 0.8829101122334456,
      "avg_duration_per_task_ms": 416465.39506172837,
      "avg_output_tokens_per_task": 33444.42985409652,
      "last_run_timestamp": "2026-10-03T09:27:25.920Z"
    },
    {
      "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": 0.8602695000000001,
      "accuracy_std_dev": 0.011784499999999976,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.07283564141414141,
      "avg_duration_per_task_ms": 493206.3787878788,
      "avg_output_tokens_per_task": 56881.042929292926,
      "last_run_timestamp": "2026-10-03T17:41:27.941Z"
    },
    {
      "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": 0.8602692222222225,
      "accuracy_std_dev": 0.01392791126381983,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.7468858333333334,
      "avg_duration_per_task_ms": 493290.462962963,
      "avg_output_tokens_per_task": 29714.63973063973,
      "last_run_timestamp": "2026-10-03T09:34:14.060Z"
    },
    {
      "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": "ModelRun",
      "accuracy": 0.860269,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.07998095959595959,
      "avg_duration_per_task_ms": 537979.2878787878,
      "avg_output_tokens_per_task": 41976.78787878788,
      "last_run_timestamp": "2026-07-11T08:27:26.027Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "CoreWeave",
      "accuracy": 0.8601822310030396,
      "accuracy_std_dev": 0.013547575585321513,
      "run_count": 5,
      "total_questions": 987,
      "avg_cost_per_task_usd": 0.007783161094224923,
      "avg_duration_per_task_ms": 526386.7112462006,
      "avg_output_tokens_per_task": 27793.246200607904,
      "last_run_timestamp": "2026-08-29T18:49:16.416Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Together",
      "accuracy": 0.8599929535603715,
      "accuracy_std_dev": 0.009491817962856216,
      "run_count": 5,
      "total_questions": 969,
      "avg_cost_per_task_usd": 0.1207670278637771,
      "avg_duration_per_task_ms": 352384.2745098039,
      "avg_output_tokens_per_task": 30417.953560371516,
      "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": "Ambient",
      "accuracy": 0.8599289999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 188,
      "avg_cost_per_task_usd": 0.023036143617021276,
      "avg_duration_per_task_ms": 1599986.2287234042,
      "avg_output_tokens_per_task": 29232.622340425532,
      "last_run_timestamp": "2026-07-18T08:57:54.026Z"
    },
    {
      "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": 0.8598355714285714,
      "accuracy_std_dev": 0.011856733262862344,
      "run_count": 6,
      "total_questions": 1176,
      "avg_cost_per_task_usd": 0.14287056972789117,
      "avg_duration_per_task_ms": 268723.07568027213,
      "avg_output_tokens_per_task": 28901.64880952381,
      "last_run_timestamp": "2026-10-03T09:42:23.512Z"
    },
    {
      "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": 0.8596616913043478,
      "accuracy_std_dev": 0.02724773313095075,
      "run_count": 7,
      "total_questions": 1380,
      "avg_cost_per_task_usd": 0.10258134202898551,
      "avg_duration_per_task_ms": 343569.69492753624,
      "avg_output_tokens_per_task": 23077.884057971016,
      "last_run_timestamp": "2026-10-03T09:34:01.485Z"
    },
    {
      "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": "WandB",
      "accuracy": 0.8594275000000001,
      "accuracy_std_dev": 0.0025254999999999583,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.005964128787878787,
      "avg_duration_per_task_ms": 438651.28282828286,
      "avg_output_tokens_per_task": 16769.555555555555,
      "last_run_timestamp": "2026-07-18T08:04:02.357Z"
    },
    {
      "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": 0.8592168750000002,
      "accuracy_std_dev": 0.04194354929377548,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.17683030808080807,
      "avg_duration_per_task_ms": 45565.645833333336,
      "avg_output_tokens_per_task": 3337.2266414141413,
      "last_run_timestamp": "2026-10-03T09:18:12.856Z"
    },
    {
      "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": 0.8591573843017328,
      "accuracy_std_dev": 0.02115249701193692,
      "run_count": 5,
      "total_questions": 981,
      "avg_cost_per_task_usd": 0.10401832415902142,
      "avg_duration_per_task_ms": 479500.45667686034,
      "avg_output_tokens_per_task": 30593.96636085627,
      "last_run_timestamp": "2026-08-22T16:49:15.579Z"
    },
    {
      "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": 0.8590881190053286,
      "accuracy_std_dev": 0.02052541308201858,
      "run_count": 4,
      "total_questions": 563,
      "avg_cost_per_task_usd": 0.08836215808170515,
      "avg_duration_per_task_ms": 517852.0088809947,
      "avg_output_tokens_per_task": 22138.96802841918,
      "last_run_timestamp": "2026-08-15T08:47:41.523Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "Alibaba",
      "accuracy": 0.8590668571428572,
      "accuracy_std_dev": 0.0075671013591461535,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.06653240404040404,
      "avg_duration_per_task_ms": 215829.40548340548,
      "avg_output_tokens_per_task": 10820.84126984127,
      "last_run_timestamp": "2026-10-03T10:11:33.574Z"
    },
    {
      "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": 0.8583545954198474,
      "accuracy_std_dev": 0.04093739371296128,
      "run_count": 10,
      "total_questions": 1965,
      "avg_cost_per_task_usd": 0.08452414249363867,
      "avg_duration_per_task_ms": 513092.07786259544,
      "avg_output_tokens_per_task": 22657.731297709925,
      "last_run_timestamp": "2026-10-03T17:07:22.491Z"
    },
    {
      "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": 0.8583051666666667,
      "accuracy_std_dev": 0.011142254012192586,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.006246021043771043,
      "avg_duration_per_task_ms": 610030.5193602693,
      "avg_output_tokens_per_task": 29029.87626262626,
      "last_run_timestamp": "2026-10-03T17:45:03.327Z"
    },
    {
      "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": 0.8578642857142857,
      "accuracy_std_dev": 0.00981070615944329,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.005024529581529581,
      "avg_duration_per_task_ms": 238361.04761904763,
      "avg_output_tokens_per_task": 28338.91414141414,
      "last_run_timestamp": "2026-10-03T16:57:52.806Z"
    },
    {
      "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": 0.8578204545454546,
      "accuracy_std_dev": 0.010161020030789983,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.8425653397612488,
      "avg_duration_per_task_ms": 451495.49678604223,
      "avg_output_tokens_per_task": 33526.43617998163,
      "last_run_timestamp": "2026-10-03T10:12:25.141Z"
    },
    {
      "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": 0.8577779119999999,
      "accuracy_std_dev": 0.006748318809155334,
      "run_count": 4,
      "total_questions": 750,
      "avg_cost_per_task_usd": 0.23260101733333335,
      "avg_duration_per_task_ms": 433991.3,
      "avg_output_tokens_per_task": 51935.09333333333,
      "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": "Morph",
      "accuracy": 0.8577439999999998,
      "accuracy_std_dev": 0.005891999999999953,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.008277098484848486,
      "avg_duration_per_task_ms": 710306.3914141414,
      "avg_output_tokens_per_task": 29437.88888888889,
      "last_run_timestamp": "2026-07-18T08:30:06.336Z"
    },
    {
      "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": 0.8576669736716133,
      "accuracy_std_dev": 0.03054982749271608,
      "run_count": 11,
      "total_questions": 2089,
      "avg_cost_per_task_usd": 0.025787480612733366,
      "avg_duration_per_task_ms": 476213.8056486357,
      "avg_output_tokens_per_task": 24583.015797032072,
      "last_run_timestamp": "2026-10-03T09:49:53.481Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3-ultra-550b-a55b-20260604",
      "model_slug": "nvidia/nemotron-3-ultra-550b-a55b",
      "model_name": "NVIDIA: Nemotron 3 Ultra",
      "provider_name": "BaseTen",
      "accuracy": 0.857418,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 173,
      "avg_cost_per_task_usd": 0.08778269942196533,
      "avg_duration_per_task_ms": 317630.5202312139,
      "avg_output_tokens_per_task": 36575.57225433526,
      "last_run_timestamp": "2026-08-04T03:12:38.954Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Ambient",
      "accuracy": 0.8569513442340791,
      "accuracy_std_dev": 0.013334663992859697,
      "run_count": 9,
      "total_questions": 1743,
      "avg_cost_per_task_usd": 0.10127325989672976,
      "avg_duration_per_task_ms": 639996.5054503729,
      "avg_output_tokens_per_task": 35441.2759609868,
      "last_run_timestamp": "2026-09-19T08:55:10.654Z"
    },
    {
      "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": 0.8569023333333334,
      "accuracy_std_dev": 0.016379967131292483,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.01228687542087542,
      "avg_duration_per_task_ms": 639472.5816498316,
      "avg_output_tokens_per_task": 33004.78787878788,
      "last_run_timestamp": "2026-10-03T09:19:24.461Z"
    },
    {
      "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": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Alibaba",
      "accuracy": 0.8568595349182764,
      "accuracy_std_dev": 0.021080058706304028,
      "run_count": 7,
      "total_questions": 1346,
      "avg_cost_per_task_usd": 0.0749669205052006,
      "avg_duration_per_task_ms": 467671.9881129272,
      "avg_output_tokens_per_task": 31991.333580980685,
      "last_run_timestamp": "2026-10-03T16:56:42.848Z"
    },
    {
      "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": 0.8561403397129187,
      "accuracy_std_dev": 0.00800597163426111,
      "run_count": 16,
      "total_questions": 3135,
      "avg_cost_per_task_usd": 0.062250919936204144,
      "avg_duration_per_task_ms": 609156.9333333333,
      "avg_output_tokens_per_task": 34308.0692185008,
      "last_run_timestamp": "2026-10-03T17:02:38.592Z"
    },
    {
      "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": 0.8550316666666666,
      "accuracy_std_dev": 0.01908702254878369,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.8377942901234569,
      "avg_duration_per_task_ms": 451775.2300785634,
      "avg_output_tokens_per_task": 33341.289001122335,
      "last_run_timestamp": "2026-10-03T09:45:17.956Z"
    },
    {
      "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": 0.8549221865284975,
      "accuracy_std_dev": 0.006241046279804878,
      "run_count": 3,
      "total_questions": 579,
      "avg_cost_per_task_usd": 0.02481670293609672,
      "avg_duration_per_task_ms": 142245.9913644214,
      "avg_output_tokens_per_task": 37982.34369602763,
      "last_run_timestamp": "2026-10-03T10:08:27.319Z"
    },
    {
      "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": 0.8548752091836734,
      "accuracy_std_dev": 0.01338517551144632,
      "run_count": 6,
      "total_questions": 1176,
      "avg_cost_per_task_usd": 0.1090017406462585,
      "avg_duration_per_task_ms": 191170.07142857142,
      "avg_output_tokens_per_task": 28383.128401360544,
      "last_run_timestamp": "2026-10-03T09:07:14.738Z"
    },
    {
      "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": 0.8548644736842105,
      "accuracy_std_dev": 0.024533883813741147,
      "run_count": 19,
      "total_questions": 3762,
      "avg_cost_per_task_usd": 0.15887166959064328,
      "avg_duration_per_task_ms": 152796.97049441785,
      "avg_output_tokens_per_task": 15519.810738968634,
      "last_run_timestamp": "2026-10-03T09:27:53.801Z"
    },
    {
      "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": 0.8547135999999999,
      "accuracy_std_dev": 0.03570492302526363,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.1789077186868687,
      "avg_duration_per_task_ms": 44801.785858585856,
      "avg_output_tokens_per_task": 3372.5843434343433,
      "last_run_timestamp": "2026-10-03T09:14:11.064Z"
    },
    {
      "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": 0.8546573333333333,
      "accuracy_std_dev": 0.015141104458467416,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.007946281144781145,
      "avg_duration_per_task_ms": 637289.4747474748,
      "avg_output_tokens_per_task": 29143.16835016835,
      "last_run_timestamp": "2026-08-29T18:38:15.907Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "BaseTen",
      "accuracy": 0.8545491290566037,
      "accuracy_std_dev": 0.016485433040158266,
      "run_count": 21,
      "total_questions": 3975,
      "avg_cost_per_task_usd": 0.15811767823899373,
      "avg_duration_per_task_ms": 199794.28352201259,
      "avg_output_tokens_per_task": 28179.00402515723,
      "last_run_timestamp": "2026-10-03T09:59:27.517Z"
    },
    {
      "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": 0.8545453999999999,
      "accuracy_std_dev": 0.04176428250119952,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.17536612575757574,
      "avg_duration_per_task_ms": 43901.6994949495,
      "avg_output_tokens_per_task": 3302.6388888888887,
      "last_run_timestamp": "2026-10-03T09:19:04.660Z"
    },
    {
      "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": 0.8544685739130434,
      "accuracy_std_dev": 0.006277655854338451,
      "run_count": 14,
      "total_questions": 2760,
      "avg_cost_per_task_usd": 0.054625573913043464,
      "avg_duration_per_task_ms": 429842.79927536234,
      "avg_output_tokens_per_task": 27957.209420289855,
      "last_run_timestamp": "2026-08-29T19:17:08.288Z"
    },
    {
      "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": "WandB",
      "accuracy": 0.854377,
      "accuracy_std_dev": 0.004208999999999963,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.252522148989899,
      "avg_duration_per_task_ms": 295859.83838383836,
      "avg_output_tokens_per_task": 62942.86616161616,
      "last_run_timestamp": "2026-07-18T08:26:40.483Z"
    },
    {
      "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": 0.8540964999999999,
      "accuracy_std_dev": 0.014793979437032262,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.09252428872053871,
      "avg_duration_per_task_ms": 604262.7205387205,
      "avg_output_tokens_per_task": 27745.595117845118,
      "last_run_timestamp": "2026-10-03T09:59:46.432Z"
    },
    {
      "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": 0.8537222222222223,
      "accuracy_std_dev": 0.017912548902175803,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.7545602300785634,
      "avg_duration_per_task_ms": 501662.9904601571,
      "avg_output_tokens_per_task": 30067.62962962963,
      "last_run_timestamp": "2026-10-03T09:46:07.607Z"
    },
    {
      "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": 0.8535444426523295,
      "accuracy_std_dev": 0.03729714924788522,
      "run_count": 29,
      "total_questions": 5022,
      "avg_cost_per_task_usd": 0.012806513739546001,
      "avg_duration_per_task_ms": 408091.7891278375,
      "avg_output_tokens_per_task": 16584.42453205894,
      "last_run_timestamp": "2026-10-03T10:45:26.547Z"
    },
    {
      "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": 0.8535352222222222,
      "accuracy_std_dev": 0.01802737382598276,
      "run_count": 18,
      "total_questions": 3564,
      "avg_cost_per_task_usd": 0.8012126506734006,
      "avg_duration_per_task_ms": 478739.47053872055,
      "avg_output_tokens_per_task": 30372.589225589225,
      "last_run_timestamp": "2026-10-03T09:41:04.765Z"
    },
    {
      "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": 0.8534097609942638,
      "accuracy_std_dev": 0.02277588037678235,
      "run_count": 8,
      "total_questions": 1569,
      "avg_cost_per_task_usd": 0.04192918291905673,
      "avg_duration_per_task_ms": 666623.9623964309,
      "avg_output_tokens_per_task": 35056.57106437221,
      "last_run_timestamp": "2026-10-03T16:51:59.968Z"
    },
    {
      "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": 0.8533247499999999,
      "accuracy_std_dev": 0.03034305490861954,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.8320753535353536,
      "avg_duration_per_task_ms": 427272.283459596,
      "avg_output_tokens_per_task": 33208.628787878784,
      "last_run_timestamp": "2026-10-03T09:44:19.593Z"
    },
    {
      "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": 0.8531986,
      "accuracy_std_dev": 0.030700851614246794,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.12264740303030303,
      "avg_duration_per_task_ms": 564764.8696969697,
      "avg_output_tokens_per_task": 32450.073737373736,
      "last_run_timestamp": "2026-10-03T16:54:37.572Z"
    },
    {
      "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": 0.8531398217391304,
      "accuracy_std_dev": 0.028934812364545037,
      "run_count": 7,
      "total_questions": 1380,
      "avg_cost_per_task_usd": 0.012572355072463768,
      "avg_duration_per_task_ms": 541534.9311594203,
      "avg_output_tokens_per_task": 33634.00579710145,
      "last_run_timestamp": "2026-10-03T09:08:08.594Z"
    },
    {
      "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": 0.853024801369863,
      "accuracy_std_dev": 0.02604945551361019,
      "run_count": 6,
      "total_questions": 1168,
      "avg_cost_per_task_usd": 0.13878068065068494,
      "avg_duration_per_task_ms": 378053.25171232875,
      "avg_output_tokens_per_task": 31592.83133561644,
      "last_run_timestamp": "2026-10-03T09:40:06.763Z"
    },
    {
      "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": 0.8528984782608696,
      "accuracy_std_dev": 0.014334577107719047,
      "run_count": 7,
      "total_questions": 1380,
      "avg_cost_per_task_usd": 0.018874540579710147,
      "avg_duration_per_task_ms": 358751.04855072463,
      "avg_output_tokens_per_task": 37538.637681159424,
      "last_run_timestamp": "2026-10-03T10:13:52.682Z"
    },
    {
      "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": 0.8527699090909091,
      "accuracy_std_dev": 0.015435615118376213,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.7526788682277317,
      "avg_duration_per_task_ms": 500624.6831955923,
      "avg_output_tokens_per_task": 30015.131313131315,
      "last_run_timestamp": "2026-10-03T09:32:02.382Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Sail Research",
      "accuracy": 0.8522881577868852,
      "accuracy_std_dev": 0.009664339658766143,
      "run_count": 5,
      "total_questions": 976,
      "avg_cost_per_task_usd": 0.09954490881147542,
      "avg_duration_per_task_ms": 487265.8411885246,
      "avg_output_tokens_per_task": 30959.861680327867,
      "last_run_timestamp": "2026-08-29T11:19:17.684Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-sonnet-5-20260630",
      "model_slug": "anthropic/claude-sonnet-5",
      "model_name": "Anthropic: Claude Sonnet 5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.8522125357142856,
      "accuracy_std_dev": 0.025081070583327737,
      "run_count": 28,
      "total_questions": 5544,
      "avg_cost_per_task_usd": 0.1703613856421356,
      "avg_duration_per_task_ms": 157589.71464646465,
      "avg_output_tokens_per_task": 15942.862554112555,
      "last_run_timestamp": "2026-10-03T09:43:58.916Z"
    },
    {
      "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-m2.1",
      "model_slug": "minimax/minimax-m2.1",
      "model_name": "MiniMax: MiniMax M2.1",
      "provider_name": "Novita",
      "accuracy": 0.8518516666666667,
      "accuracy_std_dev": 0.011904378363535908,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.025817747474747476,
      "avg_duration_per_task_ms": 277441.0841750842,
      "avg_output_tokens_per_task": 21342.33164983165,
      "last_run_timestamp": "2026-08-01T17:57:21.740Z"
    },
    {
      "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": 0.8518308614084509,
      "accuracy_std_dev": 0.02345464140331345,
      "run_count": 9,
      "total_questions": 1775,
      "avg_cost_per_task_usd": 0.006422030422535212,
      "avg_duration_per_task_ms": 229362.44,
      "avg_output_tokens_per_task": 12608.14704225352,
      "last_run_timestamp": "2026-10-03T13:30:57.186Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "auto-routing",
      "accuracy": 0.8515010833333332,
      "accuracy_std_dev": 0.020334357614057767,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.09762173148148147,
      "avg_duration_per_task_ms": 616187.7521043771,
      "avg_output_tokens_per_task": 42780.512205387204,
      "last_run_timestamp": "2026-10-03T16:57:15.084Z"
    },
    {
      "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": 0.85145952259887,
      "accuracy_std_dev": 0.01855480961260448,
      "run_count": 5,
      "total_questions": 708,
      "avg_cost_per_task_usd": 0.003704364406779661,
      "avg_duration_per_task_ms": 411396.5324858757,
      "avg_output_tokens_per_task": 7770.766949152542,
      "last_run_timestamp": "2026-10-03T09:41:45.583Z"
    },
    {
      "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": "Fireworks",
      "accuracy": 0.8513253096590909,
      "accuracy_std_dev": 0.016587499999999977,
      "run_count": 2,
      "total_questions": 352,
      "avg_cost_per_task_usd": 0.105763125,
      "avg_duration_per_task_ms": 418327.64204545453,
      "avg_output_tokens_per_task": 31556.65625,
      "last_run_timestamp": "2026-07-18T08:30:34.727Z"
    },
    {
      "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": 0.8508898571428573,
      "accuracy_std_dev": 0.059458570338462385,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.1082869906204906,
      "avg_duration_per_task_ms": 468673.2698412698,
      "avg_output_tokens_per_task": 27557.489898989897,
      "last_run_timestamp": "2026-10-03T17:44:09.402Z"
    },
    {
      "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": 0.8507314900265956,
      "accuracy_std_dev": 0.014339170843684643,
      "run_count": 8,
      "total_questions": 1504,
      "avg_cost_per_task_usd": 0.018270000664893617,
      "avg_duration_per_task_ms": 282286.72473404254,
      "avg_output_tokens_per_task": 34448.984707446805,
      "last_run_timestamp": "2026-10-03T09:44:00.880Z"
    },
    {
      "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": 0.8506987191867852,
      "accuracy_std_dev": 0.008019330452707517,
      "run_count": 12,
      "total_questions": 2361,
      "avg_cost_per_task_usd": 0.07234735196950444,
      "avg_duration_per_task_ms": 223593.9834815756,
      "avg_output_tokens_per_task": 22507.26048284625,
      "last_run_timestamp": "2026-10-03T09:46:11.970Z"
    },
    {
      "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": 0.8506616837545127,
      "accuracy_std_dev": 0.011680143920346185,
      "run_count": 7,
      "total_questions": 1385,
      "avg_cost_per_task_usd": 0.01837306498194946,
      "avg_duration_per_task_ms": 455469.79205776175,
      "avg_output_tokens_per_task": 36513.78483754513,
      "last_run_timestamp": "2026-10-03T09:41:16.696Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Mara",
      "accuracy": 0.850575,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 145,
      "avg_cost_per_task_usd": 0.015425489655172412,
      "avg_duration_per_task_ms": 62324.744827586204,
      "avg_output_tokens_per_task": 15913.096551724138,
      "last_run_timestamp": "2026-08-01T18:26:19.887Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "Xiaomi",
      "accuracy": 0.8504743636363636,
      "accuracy_std_dev": 0.010422752256244077,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.02813888934802571,
      "avg_duration_per_task_ms": 684580.927456382,
      "avg_output_tokens_per_task": 32088.465564738293,
      "last_run_timestamp": "2026-10-03T09:34:29.314Z"
    },
    {
      "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": 0.8504488333333333,
      "accuracy_std_dev": 0.010441946137520948,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.10296402020202021,
      "avg_duration_per_task_ms": 627829.3468013467,
      "avg_output_tokens_per_task": 32815.65572390572,
      "last_run_timestamp": "2026-10-03T17:01:05.431Z"
    },
    {
      "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": 0.8503154279475983,
      "accuracy_std_dev": 0.03140494390885563,
      "run_count": 7,
      "total_questions": 1374,
      "avg_cost_per_task_usd": 0.09674545997088793,
      "avg_duration_per_task_ms": 1349515.0895196507,
      "avg_output_tokens_per_task": 43276.89374090247,
      "last_run_timestamp": "2026-08-29T18:36:10.818Z"
    },
    {
      "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": 0.8502597003792668,
      "accuracy_std_dev": 0.008291599964555819,
      "run_count": 12,
      "total_questions": 2373,
      "avg_cost_per_task_usd": 0.05148469405815423,
      "avg_duration_per_task_ms": 290540.5630004214,
      "avg_output_tokens_per_task": 46134.28908554572,
      "last_run_timestamp": "2026-10-03T17:05:07.642Z"
    },
    {
      "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": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Cloudflare",
      "accuracy": 0.8500861979274612,
      "accuracy_std_dev": 0.01857196094762209,
      "run_count": 5,
      "total_questions": 965,
      "avg_cost_per_task_usd": 0.016829795854922277,
      "avg_duration_per_task_ms": 389325.4870466321,
      "avg_output_tokens_per_task": 22513.535751295338,
      "last_run_timestamp": "2026-08-29T18:07:13.805Z"
    },
    {
      "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": 0.8499464590032153,
      "accuracy_std_dev": 0.022950616654380467,
      "run_count": 7,
      "total_questions": 1244,
      "avg_cost_per_task_usd": 0.010361853697749194,
      "avg_duration_per_task_ms": 360749.26848874596,
      "avg_output_tokens_per_task": 20512.316720257233,
      "last_run_timestamp": "2026-10-03T09:28:36.120Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "Parasail",
      "accuracy": 0.8499387272727273,
      "accuracy_std_dev": 0.010489521646870508,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.040005734159779614,
      "avg_duration_per_task_ms": 395404.1997245179,
      "avg_output_tokens_per_task": 40289.957300275484,
      "last_run_timestamp": "2026-10-03T19:06:53.371Z"
    },
    {
      "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": 0.8498315000000001,
      "accuracy_std_dev": 0.03952326621940551,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.17816506616161618,
      "avg_duration_per_task_ms": 45707.15909090909,
      "avg_output_tokens_per_task": 3356.5939393939393,
      "last_run_timestamp": "2026-10-03T09:19:41.728Z"
    },
    {
      "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": 0.8498314,
      "accuracy_std_dev": 0.02688129444874262,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.15721394040404038,
      "avg_duration_per_task_ms": 152201.82676767677,
      "avg_output_tokens_per_task": 15506.534343434343,
      "last_run_timestamp": "2026-10-03T09:23:24.359Z"
    },
    {
      "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": 0.8492828305531167,
      "accuracy_std_dev": 0.030679553487986455,
      "run_count": 6,
      "total_questions": 1139,
      "avg_cost_per_task_usd": 0.1125236496927129,
      "avg_duration_per_task_ms": 549482.1018437225,
      "avg_output_tokens_per_task": 33342.05618964003,
      "last_run_timestamp": "2026-10-03T10:16:13.056Z"
    },
    {
      "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": 0.8492264101265824,
      "accuracy_std_dev": 0.04905497536664576,
      "run_count": 6,
      "total_questions": 1185,
      "avg_cost_per_task_usd": 0.1081006405063291,
      "avg_duration_per_task_ms": 612102.1265822785,
      "avg_output_tokens_per_task": 27354.124050632912,
      "last_run_timestamp": "2026-08-29T11:31:24.094Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8487653333333334,
      "accuracy_std_dev": 0.011791380929400174,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.01937495622895623,
      "avg_duration_per_task_ms": 442506.297979798,
      "avg_output_tokens_per_task": 38526.9595959596,
      "last_run_timestamp": "2026-10-03T09:49:43.878Z"
    },
    {
      "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": 0.8486995349544073,
      "accuracy_std_dev": 0.009718197845279743,
      "run_count": 5,
      "total_questions": 987,
      "avg_cost_per_task_usd": 0.10196987639311045,
      "avg_duration_per_task_ms": 302885.430597771,
      "avg_output_tokens_per_task": 30304.676798378925,
      "last_run_timestamp": "2026-10-03T09:33:45.532Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Crusoe",
      "accuracy": 0.848597158186864,
      "accuracy_std_dev": 0.013754474089263532,
      "run_count": 6,
      "total_questions": 1081,
      "avg_cost_per_task_usd": 0.006980295097132284,
      "avg_duration_per_task_ms": 490407.8658649399,
      "avg_output_tokens_per_task": 17200.208140610546,
      "last_run_timestamp": "2026-10-03T16:56:50.425Z"
    },
    {
      "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": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "Novita",
      "accuracy": 0.8484290068965519,
      "accuracy_std_dev": 0.010032386386676414,
      "run_count": 11,
      "total_questions": 2175,
      "avg_cost_per_task_usd": 0.03036523402298851,
      "avg_duration_per_task_ms": 666184.1756321839,
      "avg_output_tokens_per_task": 31675.185287356322,
      "last_run_timestamp": "2026-10-03T10:03:21.013Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Parasail",
      "accuracy": 0.8482143392857143,
      "accuracy_std_dev": 0.04522948547653644,
      "run_count": 6,
      "total_questions": 1176,
      "avg_cost_per_task_usd": 0.17130176700680272,
      "avg_duration_per_task_ms": 985601.3954081633,
      "avg_output_tokens_per_task": 51250.37244897959,
      "last_run_timestamp": "2026-10-03T17:02:58.648Z"
    },
    {
      "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": 0.8480139539435808,
      "accuracy_std_dev": 0.013068837655060239,
      "run_count": 9,
      "total_questions": 1737,
      "avg_cost_per_task_usd": 0.02671003742084053,
      "avg_duration_per_task_ms": 67728.42774899252,
      "avg_output_tokens_per_task": 19651.77605066206,
      "last_run_timestamp": "2026-10-03T09:55:20.843Z"
    },
    {
      "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": 0.8480035714285714,
      "accuracy_std_dev": 0.037501057988340775,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.045334733044733046,
      "avg_duration_per_task_ms": 477671.12987012987,
      "avg_output_tokens_per_task": 32379.50144300144,
      "last_run_timestamp": "2026-10-03T17:33:25.840Z"
    },
    {
      "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": "Chutes",
      "accuracy": 0.8479612021276596,
      "accuracy_std_dev": 0.006925500000000029,
      "run_count": 2,
      "total_questions": 376,
      "avg_cost_per_task_usd": 0.07202128191489361,
      "avg_duration_per_task_ms": 1392719.3138297873,
      "avg_output_tokens_per_task": 36985.77127659575,
      "last_run_timestamp": "2026-07-18T16:22:39.277Z"
    },
    {
      "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": 0.8478198075949368,
      "accuracy_std_dev": 0.015598435238474664,
      "run_count": 6,
      "total_questions": 1185,
      "avg_cost_per_task_usd": 0.07492492320675105,
      "avg_duration_per_task_ms": 285938.56540084386,
      "avg_output_tokens_per_task": 19472.943459915612,
      "last_run_timestamp": "2026-10-03T10:12:41.882Z"
    },
    {
      "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": 0.8478035581395349,
      "accuracy_std_dev": 0.061331422730297465,
      "run_count": 7,
      "total_questions": 1290,
      "avg_cost_per_task_usd": 0.010003304651162791,
      "avg_duration_per_task_ms": 534481.4193798449,
      "avg_output_tokens_per_task": 19807.151937984498,
      "last_run_timestamp": "2026-10-03T10:05:20.841Z"
    },
    {
      "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": "Chutes",
      "accuracy": 0.8476773606557377,
      "accuracy_std_dev": 0.0038120000000000376,
      "run_count": 2,
      "total_questions": 244,
      "avg_cost_per_task_usd": 0.014137975409836067,
      "avg_duration_per_task_ms": 314874.34016393445,
      "avg_output_tokens_per_task": 11748.155737704918,
      "last_run_timestamp": "2026-07-18T16:53:38.018Z"
    },
    {
      "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": 0.8475308251851853,
      "accuracy_std_dev": 0.02209777163862587,
      "run_count": 14,
      "total_questions": 2700,
      "avg_cost_per_task_usd": 0.16495124851851853,
      "avg_duration_per_task_ms": 300213.63333333336,
      "avg_output_tokens_per_task": 41197.25444444444,
      "last_run_timestamp": "2026-10-03T17:08:51.266Z"
    },
    {
      "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": 0.8474135454545455,
      "accuracy_std_dev": 0.02039593415527014,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.12871944444444444,
      "avg_duration_per_task_ms": 433056.7759412305,
      "avg_output_tokens_per_task": 35508.42056932966,
      "last_run_timestamp": "2026-10-03T16:40:33.972Z"
    },
    {
      "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": 0.8468523098859316,
      "accuracy_std_dev": 0.015292048274757053,
      "run_count": 8,
      "total_questions": 1578,
      "avg_cost_per_task_usd": 0.03700323827629911,
      "avg_duration_per_task_ms": 627930.3586818757,
      "avg_output_tokens_per_task": 34553.87896070976,
      "last_run_timestamp": "2026-10-03T17:17:28.936Z"
    },
    {
      "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": 0.8467494829721361,
      "accuracy_std_dev": 0.006308499999999995,
      "run_count": 2,
      "total_questions": 323,
      "avg_cost_per_task_usd": 0.009286786377708979,
      "avg_duration_per_task_ms": 438695.826625387,
      "avg_output_tokens_per_task": 23117.473684210527,
      "last_run_timestamp": "2026-10-03T09:32:36.521Z"
    },
    {
      "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": 0.8466377271407838,
      "accuracy_std_dev": 0.016590356861005438,
      "run_count": 7,
      "total_questions": 1378,
      "avg_cost_per_task_usd": 0.009304855587808419,
      "avg_duration_per_task_ms": 444689.5050798258,
      "avg_output_tokens_per_task": 28194.9412191582,
      "last_run_timestamp": "2026-10-03T10:04:48.522Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Sail Research",
      "accuracy": 0.8466045281660314,
      "accuracy_std_dev": 0.011422067985981737,
      "run_count": 12,
      "total_questions": 2361,
      "avg_cost_per_task_usd": 0.1267483756882677,
      "avg_duration_per_task_ms": 355086.75984752225,
      "avg_output_tokens_per_task": 33407.23930537908,
      "last_run_timestamp": "2026-10-03T09:40:39.339Z"
    },
    {
      "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": 0.8462925557680051,
      "accuracy_std_dev": 0.011842625025068353,
      "run_count": 8,
      "total_questions": 1569,
      "avg_cost_per_task_usd": 0.006960335245379223,
      "avg_duration_per_task_ms": 537888.3499043977,
      "avg_output_tokens_per_task": 17232.114722753347,
      "last_run_timestamp": "2026-10-03T17:01:29.670Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Relace",
      "accuracy": 0.8461756369863013,
      "accuracy_std_dev": 0.003462455005724478,
      "run_count": 3,
      "total_questions": 584,
      "avg_cost_per_task_usd": 0.07262447602739726,
      "avg_duration_per_task_ms": 181056.27226027398,
      "avg_output_tokens_per_task": 31634.909246575342,
      "last_run_timestamp": "2026-10-03T09:37:27.212Z"
    },
    {
      "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": 0.8461699999999999,
      "accuracy_std_dev": 0.03781276874945815,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.6916325820707071,
      "avg_duration_per_task_ms": 471363.1818181818,
      "avg_output_tokens_per_task": 27729.57702020202,
      "last_run_timestamp": "2026-10-03T09:31:25.951Z"
    },
    {
      "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": 0.8456601854103344,
      "accuracy_std_dev": 0.00860173579226889,
      "run_count": 5,
      "total_questions": 987,
      "avg_cost_per_task_usd": 0.02046987537993921,
      "avg_duration_per_task_ms": 602884.1205673759,
      "avg_output_tokens_per_task": 23203.004052684904,
      "last_run_timestamp": "2026-08-15T14:57:08.773Z"
    },
    {
      "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": 0.8454851888607595,
      "accuracy_std_dev": 0.015145407351405253,
      "run_count": 10,
      "total_questions": 1975,
      "avg_cost_per_task_usd": 0.08874519088607594,
      "avg_duration_per_task_ms": 411899.82784810127,
      "avg_output_tokens_per_task": 22073.469873417722,
      "last_run_timestamp": "2026-10-03T17:44:12.665Z"
    },
    {
      "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": 0.8453581428571427,
      "accuracy_std_dev": 0.013659783740481936,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.10694367099567102,
      "avg_duration_per_task_ms": 329051.74314574315,
      "avg_output_tokens_per_task": 29197.39393939394,
      "last_run_timestamp": "2026-10-03T09:28:28.144Z"
    },
    {
      "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": 0.8453473636363636,
      "accuracy_std_dev": 0.010732273090354644,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.10142870936639119,
      "avg_duration_per_task_ms": 595868.2428833792,
      "avg_output_tokens_per_task": 32109.968778696053,
      "last_run_timestamp": "2026-10-03T18:05:42.597Z"
    },
    {
      "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": 0.8453441605206073,
      "accuracy_std_dev": 0.024165885254643357,
      "run_count": 21,
      "total_questions": 4149,
      "avg_cost_per_task_usd": 0.23573006531694388,
      "avg_duration_per_task_ms": 512868.35454326344,
      "avg_output_tokens_per_task": 44701.53217642805,
      "last_run_timestamp": "2026-10-03T16:58:25.654Z"
    },
    {
      "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": 0.8450359687930242,
      "accuracy_std_dev": 0.011573586447164944,
      "run_count": 12,
      "total_questions": 2179,
      "avg_cost_per_task_usd": 0.015310958237723726,
      "avg_duration_per_task_ms": 581526.1060119321,
      "avg_output_tokens_per_task": 16391.44194584672,
      "last_run_timestamp": "2026-08-29T18:29:51.235Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "auto-routing",
      "accuracy": 0.8450058498194947,
      "accuracy_std_dev": 0.034613668932657105,
      "run_count": 7,
      "total_questions": 1385,
      "avg_cost_per_task_usd": 0.17534774657039712,
      "avg_duration_per_task_ms": 1010638.444765343,
      "avg_output_tokens_per_task": 53792.70180505415,
      "last_run_timestamp": "2026-10-03T16:47:29.532Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8446367142857142,
      "accuracy_std_dev": 0.0059496991692085995,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.009103200577200578,
      "avg_duration_per_task_ms": 400631.7698412698,
      "avg_output_tokens_per_task": 18223.642135642134,
      "last_run_timestamp": "2026-10-03T17:22:47.678Z"
    },
    {
      "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": 0.8444975263157894,
      "accuracy_std_dev": 0.05215908171451967,
      "run_count": 19,
      "total_questions": 3762,
      "avg_cost_per_task_usd": 0.18362076209463052,
      "avg_duration_per_task_ms": 40742.116427432215,
      "avg_output_tokens_per_task": 3305.715842636895,
      "last_run_timestamp": "2026-10-03T09:56:46.597Z"
    },
    {
      "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": 0.8444934569536424,
      "accuracy_std_dev": 0.014326304586390738,
      "run_count": 7,
      "total_questions": 1359,
      "avg_cost_per_task_usd": 0.07969532008830021,
      "avg_duration_per_task_ms": 899418.5172921265,
      "avg_output_tokens_per_task": 40442.24208977189,
      "last_run_timestamp": "2026-08-30T00:19:02.095Z"
    },
    {
      "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": 0.8443601352718078,
      "accuracy_std_dev": 0.03374720321086705,
      "run_count": 12,
      "total_questions": 2373,
      "avg_cost_per_task_usd": 0.054606860092709646,
      "avg_duration_per_task_ms": 224549.4858828487,
      "avg_output_tokens_per_task": 22651.43194268858,
      "last_run_timestamp": "2026-10-03T10:19:52.322Z"
    },
    {
      "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": 0.8442346282818979,
      "accuracy_std_dev": 0.01874547476103309,
      "run_count": 11,
      "total_questions": 2171,
      "avg_cost_per_task_usd": 0.2081649216950714,
      "avg_duration_per_task_ms": 311887.5845232612,
      "avg_output_tokens_per_task": 47232.94426531552,
      "last_run_timestamp": "2026-10-03T16:37:10.188Z"
    },
    {
      "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": 0.8441076000000001,
      "accuracy_std_dev": 0.00842413538827578,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.1428492090909091,
      "avg_duration_per_task_ms": 160872.74545454545,
      "avg_output_tokens_per_task": 32292.333333333332,
      "last_run_timestamp": "2026-10-03T09:56:48.670Z"
    },
    {
      "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": "Azure",
      "accuracy": 0.8441076,
      "accuracy_std_dev": 0.036707935445077826,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.015406010101010103,
      "avg_duration_per_task_ms": 387191.47878787876,
      "avg_output_tokens_per_task": 27797.58787878788,
      "last_run_timestamp": "2026-10-03T16:57:48.932Z"
    },
    {
      "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": 0.8437161084142395,
      "accuracy_std_dev": 0.013475684773786068,
      "run_count": 7,
      "total_questions": 1236,
      "avg_cost_per_task_usd": 0.006951031553398057,
      "avg_duration_per_task_ms": 652065.4733009709,
      "avg_output_tokens_per_task": 17156.162621359224,
      "last_run_timestamp": "2026-10-03T17:20:08.224Z"
    },
    {
      "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": 0.8435618989169676,
      "accuracy_std_dev": 0.03832760617920939,
      "run_count": 7,
      "total_questions": 1385,
      "avg_cost_per_task_usd": 0.16082656317689534,
      "avg_duration_per_task_ms": 1412866.806498195,
      "avg_output_tokens_per_task": 51968.49458483754,
      "last_run_timestamp": "2026-10-03T17:37:27.208Z"
    },
    {
      "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": 0.8434342499999999,
      "accuracy_std_dev": 0.008881818827667008,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.14890805050505052,
      "avg_duration_per_task_ms": 162687.35606060605,
      "avg_output_tokens_per_task": 32857.1712962963,
      "last_run_timestamp": "2026-10-03T10:18:11.626Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "thinkingmachines/inkling-20260715",
      "model_slug": "thinkingmachines/inkling",
      "model_name": "Thinking Machines: Inkling",
      "provider_name": "Together",
      "accuracy": 0.8425437868421053,
      "accuracy_std_dev": 0.00573439306339873,
      "run_count": 6,
      "total_questions": 1140,
      "avg_cost_per_task_usd": 0.08756267807017544,
      "avg_duration_per_task_ms": 282086.36140350875,
      "avg_output_tokens_per_task": 21480.616666666665,
      "last_run_timestamp": "2026-10-01T16:06:12.801Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Novita",
      "accuracy": 0.8423878991712708,
      "accuracy_std_dev": 0.05766154228422131,
      "run_count": 11,
      "total_questions": 2172,
      "avg_cost_per_task_usd": 0.07142284576427256,
      "avg_duration_per_task_ms": 676306.011510129,
      "avg_output_tokens_per_task": 44005.55662983425,
      "last_run_timestamp": "2026-10-03T17:06:14.822Z"
    },
    {
      "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": 0.842382125,
      "accuracy_std_dev": 0.024293570854433386,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.06867556881313132,
      "avg_duration_per_task_ms": 555486.8421717172,
      "avg_output_tokens_per_task": 21945.48106060606,
      "last_run_timestamp": "2026-10-03T18:21:53.448Z"
    },
    {
      "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": 0.8423120833333333,
      "accuracy_std_dev": 0.02273594124016837,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.06854938888888887,
      "avg_duration_per_task_ms": 673113.2765151515,
      "avg_output_tokens_per_task": 51666.62584175084,
      "last_run_timestamp": "2026-10-03T16:57:33.328Z"
    },
    {
      "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": 0.8422381263537906,
      "accuracy_std_dev": 0.013606259020414556,
      "run_count": 7,
      "total_questions": 1385,
      "avg_cost_per_task_usd": 0.1362164772563177,
      "avg_duration_per_task_ms": 189944.0108303249,
      "avg_output_tokens_per_task": 30733.362454873648,
      "last_run_timestamp": "2026-10-03T09:32:18.488Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Parasail",
      "accuracy": 0.842151959435626,
      "accuracy_std_dev": 0.020139741016761404,
      "run_count": 3,
      "total_questions": 567,
      "avg_cost_per_task_usd": 0.02213832098765432,
      "avg_duration_per_task_ms": 283612.92063492065,
      "avg_output_tokens_per_task": 18387.772486772486,
      "last_run_timestamp": "2026-08-01T18:21:26.176Z"
    },
    {
      "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": 0.8419798924812032,
      "accuracy_std_dev": 0.0161432408803948,
      "run_count": 7,
      "total_questions": 1330,
      "avg_cost_per_task_usd": 0.01795621127819549,
      "avg_duration_per_task_ms": 242875.14962406014,
      "avg_output_tokens_per_task": 35697.367669172934,
      "last_run_timestamp": "2026-10-03T09:24:57.334Z"
    },
    {
      "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": 0.8417268575037897,
      "accuracy_std_dev": 0.03274733184905639,
      "run_count": 30,
      "total_questions": 5937,
      "avg_cost_per_task_usd": 0.16544357267980464,
      "avg_duration_per_task_ms": 149160.101061142,
      "avg_output_tokens_per_task": 15291.950816910898,
      "last_run_timestamp": "2026-10-03T10:45:13.551Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.5-flash-lite-20260721",
      "model_slug": "google/gemini-3.5-flash-lite",
      "model_name": "Google: Gemini 3.5 Flash Lite",
      "provider_name": "Google AI Studio",
      "accuracy": 0.8413977096774194,
      "accuracy_std_dev": 0.007842951293996439,
      "run_count": 4,
      "total_questions": 744,
      "avg_cost_per_task_usd": 0.03688066532258065,
      "avg_duration_per_task_ms": 47685.283602150535,
      "avg_output_tokens_per_task": 14664.099462365592,
      "last_run_timestamp": "2026-10-01T19:13:29.958Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.5-flash-lite-20260721",
      "model_slug": "google/gemini-3.5-flash-lite",
      "model_name": "Google: Gemini 3.5 Flash Lite",
      "provider_name": "auto-routing",
      "accuracy": 0.8411896666666667,
      "accuracy_std_dev": 0.008729468916008341,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.036658867003367,
      "avg_duration_per_task_ms": 51092.77104377104,
      "avg_output_tokens_per_task": 14574.808080808081,
      "last_run_timestamp": "2026-10-01T19:12:26.032Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "auto-routing",
      "accuracy": 0.8410581923809524,
      "accuracy_std_dev": 0.018771452424076705,
      "run_count": 8,
      "total_questions": 1575,
      "avg_cost_per_task_usd": 0.04652745714285714,
      "avg_duration_per_task_ms": 641076.9504761904,
      "avg_output_tokens_per_task": 37228.5180952381,
      "last_run_timestamp": "2026-10-03T16:58:17.410Z"
    },
    {
      "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": 0.8409091945454545,
      "accuracy_std_dev": 0.023300437613580863,
      "run_count": 8,
      "total_questions": 1100,
      "avg_cost_per_task_usd": 0.01265548181818182,
      "avg_duration_per_task_ms": 72073.20181818181,
      "avg_output_tokens_per_task": 11616.710909090909,
      "last_run_timestamp": "2026-10-03T17:24:07.121Z"
    },
    {
      "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": "Fireworks",
      "accuracy": 0.8409091666666667,
      "accuracy_std_dev": 0.012885450586050746,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.12154702777777776,
      "avg_duration_per_task_ms": 341680.0117845118,
      "avg_output_tokens_per_task": 27429.664983164985,
      "last_run_timestamp": "2026-10-03T09:27:41.412Z"
    },
    {
      "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": 0.8408324545454547,
      "accuracy_std_dev": 0.07268828033879719,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.18564792516069786,
      "avg_duration_per_task_ms": 528994.2309458219,
      "avg_output_tokens_per_task": 42221.45684113866,
      "last_run_timestamp": "2026-10-03T18:53:59.567Z"
    },
    {
      "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": 0.840817973713034,
      "accuracy_std_dev": 0.020791979459397293,
      "run_count": 5,
      "total_questions": 913,
      "avg_cost_per_task_usd": 0.1426375947426068,
      "avg_duration_per_task_ms": 229737.46549835705,
      "avg_output_tokens_per_task": 33714.36035049288,
      "last_run_timestamp": "2026-10-03T09:33:59.008Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.4-mini-20260317",
      "model_slug": "openai/gpt-5.4-mini",
      "model_name": "OpenAI: GPT-5.4 Mini",
      "provider_name": "OpenAI",
      "accuracy": 0.84048825,
      "accuracy_std_dev": 0.004667830993887862,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.06197197222222223,
      "avg_duration_per_task_ms": 84839.70707070707,
      "avg_output_tokens_per_task": 13643.420454545454,
      "last_run_timestamp": "2026-10-02T23:24:30.140Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Reka",
      "accuracy": 0.84048825,
      "accuracy_std_dev": 0.007080042491927587,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.0629129558080808,
      "avg_duration_per_task_ms": 337433.2537878788,
      "avg_output_tokens_per_task": 20687.39898989899,
      "last_run_timestamp": "2026-10-03T14:07:19.340Z"
    },
    {
      "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": 0.8404742875816994,
      "accuracy_std_dev": 0.009920155987499742,
      "run_count": 7,
      "total_questions": 1377,
      "avg_cost_per_task_usd": 0.027662923747276687,
      "avg_duration_per_task_ms": 532773.5497458243,
      "avg_output_tokens_per_task": 22865.482207697893,
      "last_run_timestamp": "2026-08-29T18:16:39.840Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.5-flash-lite-20260721",
      "model_slug": "google/gemini-3.5-flash-lite",
      "model_name": "Google: Gemini 3.5 Flash Lite",
      "provider_name": "Google",
      "accuracy": 0.8402775531250001,
      "accuracy_std_dev": 0.007466846475587956,
      "run_count": 10,
      "total_questions": 1920,
      "avg_cost_per_task_usd": 0.039408469791666664,
      "avg_duration_per_task_ms": 51791.722395833334,
      "avg_output_tokens_per_task": 14757.431770833333,
      "last_run_timestamp": "2026-10-01T19:13:57.987Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Phala",
      "accuracy": 0.8400541048387097,
      "accuracy_std_dev": 0.0023684999999999956,
      "run_count": 2,
      "total_questions": 248,
      "avg_cost_per_task_usd": 0.016170879032258067,
      "avg_duration_per_task_ms": 368527.8064516129,
      "avg_output_tokens_per_task": 11745.584677419354,
      "last_run_timestamp": "2026-07-18T16:47:35.300Z"
    },
    {
      "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": 0.8398692413632118,
      "accuracy_std_dev": 0.024400746937490225,
      "run_count": 11,
      "total_questions": 2142,
      "avg_cost_per_task_usd": 0.20761488702147526,
      "avg_duration_per_task_ms": 295169.1633986928,
      "avg_output_tokens_per_task": 47026.597572362276,
      "last_run_timestamp": "2026-10-03T16:57:30.005Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "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-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Relace",
      "accuracy": 0.8391689163108454,
      "accuracy_std_dev": 0.01637841911378101,
      "run_count": 6,
      "total_questions": 1171,
      "avg_cost_per_task_usd": 0.014026462852263022,
      "avg_duration_per_task_ms": 450960.66951323656,
      "avg_output_tokens_per_task": 38615.76686592656,
      "last_run_timestamp": "2026-10-03T09:28:28.466Z"
    },
    {
      "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": 0.8389739548751007,
      "accuracy_std_dev": 0.023840636777552385,
      "run_count": 7,
      "total_questions": 1241,
      "avg_cost_per_task_usd": 0.07291945044319098,
      "avg_duration_per_task_ms": 368316.54069298954,
      "avg_output_tokens_per_task": 24012.563255439163,
      "last_run_timestamp": "2026-10-03T16:56:41.261Z"
    },
    {
      "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": 0.8389449999999999,
      "accuracy_std_dev": 0.010132054118818512,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.027894321548821547,
      "avg_duration_per_task_ms": 402197.4781144781,
      "avg_output_tokens_per_task": 14811.760942760942,
      "last_run_timestamp": "2026-10-03T10:19:34.012Z"
    },
    {
      "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": 0.8388887961538462,
      "accuracy_std_dev": 0.016188823238869932,
      "run_count": 8,
      "total_questions": 1560,
      "avg_cost_per_task_usd": 0.034323907051282056,
      "avg_duration_per_task_ms": 565277.3544871794,
      "avg_output_tokens_per_task": 32958.441666666666,
      "last_run_timestamp": "2026-10-03T16:57:14.717Z"
    },
    {
      "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": 0.8388647142857143,
      "accuracy_std_dev": 0.020435361360112217,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.02562200432900433,
      "avg_duration_per_task_ms": 254478.84992784992,
      "avg_output_tokens_per_task": 21304.302308802307,
      "last_run_timestamp": "2026-08-29T18:12:43.943Z"
    },
    {
      "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": 0.8388431247933885,
      "accuracy_std_dev": 0.0097334078077013,
      "run_count": 7,
      "total_questions": 1210,
      "avg_cost_per_task_usd": 0.08136244214876033,
      "avg_duration_per_task_ms": 548386.6462809917,
      "avg_output_tokens_per_task": 20121.671074380167,
      "last_run_timestamp": "2026-10-01T16:08:32.295Z"
    },
    {
      "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": "Fireworks",
      "accuracy": 0.8386665466666666,
      "accuracy_std_dev": 0.00007150000000000212,
      "run_count": 2,
      "total_questions": 375,
      "avg_cost_per_task_usd": 0.031960357333333335,
      "avg_duration_per_task_ms": 164702.43466666667,
      "avg_output_tokens_per_task": 26929.237333333334,
      "last_run_timestamp": "2026-07-18T08:31:24.271Z"
    },
    {
      "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": 0.8384329244791666,
      "accuracy_std_dev": 0.01468840893357752,
      "run_count": 8,
      "total_questions": 1536,
      "avg_cost_per_task_usd": 0.221448376953125,
      "avg_duration_per_task_ms": 429166.3111979167,
      "avg_output_tokens_per_task": 50182.490234375,
      "last_run_timestamp": "2026-10-03T16:59:26.755Z"
    },
    {
      "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": 0.8383837142857142,
      "accuracy_std_dev": 0.026235324298762266,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.11822585137085138,
      "avg_duration_per_task_ms": 408512.2987012987,
      "avg_output_tokens_per_task": 35583.771284271286,
      "last_run_timestamp": "2026-10-03T10:08:41.581Z"
    },
    {
      "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": 0.8381875026968716,
      "accuracy_std_dev": 0.003092760668974499,
      "run_count": 6,
      "total_questions": 927,
      "avg_cost_per_task_usd": 0.13592745091693637,
      "avg_duration_per_task_ms": 121247.60086299892,
      "avg_output_tokens_per_task": 30661.07551240561,
      "last_run_timestamp": "2026-10-03T22:03:15.220Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8381033333333334,
      "accuracy_std_dev": 0.0831951431348142,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.007828982323232323,
      "avg_duration_per_task_ms": 443392.71717171714,
      "avg_output_tokens_per_task": 28136.66245791246,
      "last_run_timestamp": "2026-10-03T17:03:58.997Z"
    },
    {
      "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": 0.837962875,
      "accuracy_std_dev": 0.08099089361378461,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.007212295454545454,
      "avg_duration_per_task_ms": 692973.9797979798,
      "avg_output_tokens_per_task": 29648.13005050505,
      "last_run_timestamp": "2026-10-03T11:41:47.254Z"
    },
    {
      "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": 0.8376334722439405,
      "accuracy_std_dev": 0.027219006084059767,
      "run_count": 7,
      "total_questions": 1279,
      "avg_cost_per_task_usd": 0.0889748506645817,
      "avg_duration_per_task_ms": 1571235.6520719312,
      "avg_output_tokens_per_task": 41303.13604378421,
      "last_run_timestamp": "2026-08-29T19:12:33.941Z"
    },
    {
      "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": 0.8375589327830189,
      "accuracy_std_dev": 0.020056898198209345,
      "run_count": 9,
      "total_questions": 1696,
      "avg_cost_per_task_usd": 0.054286748820754724,
      "avg_duration_per_task_ms": 282159.9534198113,
      "avg_output_tokens_per_task": 22660.685141509435,
      "last_run_timestamp": "2026-09-26T09:50:39.842Z"
    },
    {
      "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": "DeepInfra",
      "accuracy": 0.837017,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 181,
      "avg_cost_per_task_usd": 0.02255581215469613,
      "avg_duration_per_task_ms": 455797.3591160221,
      "avg_output_tokens_per_task": 19569.629834254145,
      "last_run_timestamp": "2026-07-11T15:51:15.190Z"
    },
    {
      "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": 0.8367149173913044,
      "accuracy_std_dev": 0.019766126930603665,
      "run_count": 7,
      "total_questions": 1380,
      "avg_cost_per_task_usd": 0.0784595304347826,
      "avg_duration_per_task_ms": 736815.4384057971,
      "avg_output_tokens_per_task": 42336.87536231884,
      "last_run_timestamp": "2026-10-04T00:07:08.217Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Morph",
      "accuracy": 0.8366436799091941,
      "accuracy_std_dev": 0.015820054069611535,
      "run_count": 9,
      "total_questions": 1762,
      "avg_cost_per_task_usd": 0.08522557775255392,
      "avg_duration_per_task_ms": 546245.1770715096,
      "avg_output_tokens_per_task": 31173.529511918274,
      "last_run_timestamp": "2026-10-03T11:03:00.010Z"
    },
    {
      "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": 0.8362222851165569,
      "accuracy_std_dev": 0.014310164504994333,
      "run_count": 10,
      "total_questions": 1673,
      "avg_cost_per_task_usd": 0.0965615690376569,
      "avg_duration_per_task_ms": 627840.9587567245,
      "avg_output_tokens_per_task": 30100.354453078304,
      "last_run_timestamp": "2026-10-03T17:45:46.427Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "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": "BaseTen",
      "accuracy": 0.8360065448008849,
      "accuracy_std_dev": 0.008779454087962683,
      "run_count": 11,
      "total_questions": 1808,
      "avg_cost_per_task_usd": 0.018033527101769913,
      "avg_duration_per_task_ms": 187582.5735619469,
      "avg_output_tokens_per_task": 35877.24225663717,
      "last_run_timestamp": "2026-10-03T09:31:41.635Z"
    },
    {
      "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": 0.8359790095238095,
      "accuracy_std_dev": 0.03660393871642171,
      "run_count": 6,
      "total_questions": 630,
      "avg_cost_per_task_usd": 0.10253282222222224,
      "avg_duration_per_task_ms": 301607.7301587302,
      "avg_output_tokens_per_task": 23860.25873015873,
      "last_run_timestamp": "2026-08-22T17:29:49.714Z"
    },
    {
      "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": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "DigitalOcean",
      "accuracy": 0.8357464991482112,
      "accuracy_std_dev": 0.060956979201182,
      "run_count": 6,
      "total_questions": 1174,
      "avg_cost_per_task_usd": 0.08090285604770017,
      "avg_duration_per_task_ms": 586630.9148211244,
      "avg_output_tokens_per_task": 22899.433560477002,
      "last_run_timestamp": "2026-10-03T09:16:10.553Z"
    },
    {
      "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": 0.835648125,
      "accuracy_std_dev": 0.012901937649414363,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.03973862878787879,
      "avg_duration_per_task_ms": 223092.5542929293,
      "avg_output_tokens_per_task": 14589.925505050505,
      "last_run_timestamp": "2026-10-03T09:51:44.286Z"
    },
    {
      "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": 0.8356289090909089,
      "accuracy_std_dev": 0.029442978957653503,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.21974994260789715,
      "avg_duration_per_task_ms": 332523.40220385673,
      "avg_output_tokens_per_task": 49815.44995408632,
      "last_run_timestamp": "2026-10-03T17:09:33.799Z"
    },
    {
      "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": 0.8354408285280729,
      "accuracy_std_dev": 0.043474653742611905,
      "run_count": 10,
      "total_questions": 1977,
      "avg_cost_per_task_usd": 0.1658476181082448,
      "avg_duration_per_task_ms": 161596.665149216,
      "avg_output_tokens_per_task": 16364.909458775923,
      "last_run_timestamp": "2026-10-03T09:33:36.803Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "inclusionai/ling-3.0-flash-vl-20260910",
      "model_slug": "inclusionai/ling-3.0-flash-vl",
      "model_name": "inclusionAI: Ling 3.0 Flash VL",
      "provider_name": "DeepInfra",
      "accuracy": 0.8351849999999998,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 180,
      "avg_cost_per_task_usd": 0.0030483999999999997,
      "avg_duration_per_task_ms": 288802.22222222225,
      "avg_output_tokens_per_task": 16703.5,
      "last_run_timestamp": "2026-10-01T08:43:11.415Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.4-mini-20260317",
      "model_slug": "openai/gpt-5.4-mini",
      "model_name": "OpenAI: GPT-5.4 Mini",
      "provider_name": "auto-routing",
      "accuracy": 0.8350166666666667,
      "accuracy_std_dev": 0.005991526757754575,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.06194655555555556,
      "avg_duration_per_task_ms": 84396.87373737374,
      "avg_output_tokens_per_task": 13638.5,
      "last_run_timestamp": "2026-10-01T19:12:44.401Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Modal",
      "accuracy": 0.8348964285714285,
      "accuracy_std_dev": 0.013870860266422272,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.13617697113997115,
      "avg_duration_per_task_ms": 172785.43217893218,
      "avg_output_tokens_per_task": 30709.33621933622,
      "last_run_timestamp": "2026-10-03T09:19:24.458Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "auto-routing",
      "accuracy": 0.8347762857142856,
      "accuracy_std_dev": 0.039321677971442086,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.10741473737373737,
      "avg_duration_per_task_ms": 1184256.5779220778,
      "avg_output_tokens_per_task": 43225.256132756134,
      "last_run_timestamp": "2026-08-29T18:42:20.541Z"
    },
    {
      "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": 0.834656,
      "accuracy_std_dev": 0.01293003014028089,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.006965619769119769,
      "avg_duration_per_task_ms": 380792.71356421354,
      "avg_output_tokens_per_task": 26227.490620490622,
      "last_run_timestamp": "2026-10-03T17:23:05.193Z"
    },
    {
      "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": 0.8345736842105264,
      "accuracy_std_dev": 0.012051692648327507,
      "run_count": 19,
      "total_questions": 3762,
      "avg_cost_per_task_usd": 0.2574891374269006,
      "avg_duration_per_task_ms": 275645.54146730463,
      "avg_output_tokens_per_task": 56198.80940988836,
      "last_run_timestamp": "2026-10-03T17:33:23.935Z"
    },
    {
      "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": 0.834551244519393,
      "accuracy_std_dev": 0.02463277117085194,
      "run_count": 9,
      "total_questions": 1779,
      "avg_cost_per_task_usd": 0.15877912478920744,
      "avg_duration_per_task_ms": 826328.9359190556,
      "avg_output_tokens_per_task": 52149.6323777403,
      "last_run_timestamp": "2026-10-03T17:11:42.929Z"
    },
    {
      "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": "WandB",
      "accuracy": 0.8344516577181208,
      "accuracy_std_dev": 0.048501499999999975,
      "run_count": 2,
      "total_questions": 298,
      "avg_cost_per_task_usd": 0.0903846644295302,
      "avg_duration_per_task_ms": 304392.4194630872,
      "avg_output_tokens_per_task": 22866.939597315435,
      "last_run_timestamp": "2026-07-18T08:27:05.304Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Phala",
      "accuracy": 0.8344179826464209,
      "accuracy_std_dev": 0.05856560161712257,
      "run_count": 7,
      "total_questions": 1383,
      "avg_cost_per_task_usd": 0.22784925741142445,
      "avg_duration_per_task_ms": 1484141.7903109184,
      "avg_output_tokens_per_task": 53229.01156905278,
      "last_run_timestamp": "2026-10-03T17:35:18.039Z"
    },
    {
      "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/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "WandB",
      "accuracy": 0.834197,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 193,
      "avg_cost_per_task_usd": 0.02599202590673575,
      "avg_duration_per_task_ms": 128468.99481865285,
      "avg_output_tokens_per_task": 20660.492227979274,
      "last_run_timestamp": "2026-07-18T16:00:08.986Z"
    },
    {
      "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": "Together",
      "accuracy": 0.83417930964467,
      "accuracy_std_dev": 0.010994500000000018,
      "run_count": 2,
      "total_questions": 394,
      "avg_cost_per_task_usd": 0.033813327411167514,
      "avg_duration_per_task_ms": 296242.83248730964,
      "avg_output_tokens_per_task": 28735.548223350255,
      "last_run_timestamp": "2026-07-18T08:29:12.575Z"
    },
    {
      "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": 0.8340991776416539,
      "accuracy_std_dev": 0.012440982886713574,
      "run_count": 4,
      "total_questions": 653,
      "avg_cost_per_task_usd": 0.04871676722817765,
      "avg_duration_per_task_ms": 96441.48238897396,
      "avg_output_tokens_per_task": 15147.393568147014,
      "last_run_timestamp": "2026-10-03T09:04:19.662Z"
    },
    {
      "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": 0.8340809047085203,
      "accuracy_std_dev": 0.01732897775403963,
      "run_count": 5,
      "total_questions": 892,
      "avg_cost_per_task_usd": 0.020895510089686098,
      "avg_duration_per_task_ms": 283776.2948430493,
      "avg_output_tokens_per_task": 41571.40470852018,
      "last_run_timestamp": "2026-10-03T09:31:15.820Z"
    },
    {
      "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": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "Phala",
      "accuracy": 0.8339628488664986,
      "accuracy_std_dev": 0.062186325431239316,
      "run_count": 5,
      "total_questions": 794,
      "avg_cost_per_task_usd": 0.1341619395465995,
      "avg_duration_per_task_ms": 1520039.8690176322,
      "avg_output_tokens_per_task": 46924.88035264484,
      "last_run_timestamp": "2026-08-29T18:25:52.753Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-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": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Cloudflare",
      "accuracy": 0.8338943333333333,
      "accuracy_std_dev": 0.05954210180106922,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0609229090909091,
      "avg_duration_per_task_ms": 629153.6734006734,
      "avg_output_tokens_per_task": 22578.223905723906,
      "last_run_timestamp": "2026-10-03T16:58:19.180Z"
    },
    {
      "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": 0.8337931059602649,
      "accuracy_std_dev": 0.013355036895202588,
      "run_count": 12,
      "total_questions": 1812,
      "avg_cost_per_task_usd": 0.03549873675496689,
      "avg_duration_per_task_ms": 241451.8725165563,
      "avg_output_tokens_per_task": 28296.954194260485,
      "last_run_timestamp": "2026-10-03T16:12:07.493Z"
    },
    {
      "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": 0.8337923636363636,
      "accuracy_std_dev": 0.031059653654548318,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.12979151698806246,
      "avg_duration_per_task_ms": 553080.608815427,
      "avg_output_tokens_per_task": 39262.16712580349,
      "last_run_timestamp": "2026-10-03T16:58:32.612Z"
    },
    {
      "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": 0.8334570734966593,
      "accuracy_std_dev": 0.0419490749322732,
      "run_count": 7,
      "total_questions": 1347,
      "avg_cost_per_task_usd": 0.02924405196733482,
      "avg_duration_per_task_ms": 281996.3355605048,
      "avg_output_tokens_per_task": 9032.244988864142,
      "last_run_timestamp": "2026-10-03T10:36:42.102Z"
    },
    {
      "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": "BaseTen",
      "accuracy": 0.8333332105263158,
      "accuracy_std_dev": 0.01599350000000005,
      "run_count": 2,
      "total_questions": 342,
      "avg_cost_per_task_usd": 0.16854262573099413,
      "avg_duration_per_task_ms": 685525.8274853801,
      "avg_output_tokens_per_task": 47357.02046783626,
      "last_run_timestamp": "2026-07-18T08:29:37.313Z"
    },
    {
      "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": 0.8333331818181818,
      "accuracy_std_dev": 0.00830971445541776,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.08045208172635446,
      "avg_duration_per_task_ms": 795789.0610651975,
      "avg_output_tokens_per_task": 47822.912764003675,
      "last_run_timestamp": "2026-10-03T16:25:22.307Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-flash-20260826",
      "model_slug": "qwen/qwen3.8-flash",
      "model_name": "Qwen: Qwen3.8 Flash",
      "provider_name": "Makora",
      "accuracy": 0.8333329999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 156,
      "avg_cost_per_task_usd": 0.003982064102564103,
      "avg_duration_per_task_ms": 34178.44871794872,
      "avg_output_tokens_per_task": 8341.692307692309,
      "last_run_timestamp": "2026-09-09T15:03:02.824Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-122b-a10b-20260224",
      "model_slug": "qwen/qwen3.5-122b-a10b",
      "model_name": "Qwen: Qwen3.5-122B-A10B",
      "provider_name": "auto-routing",
      "accuracy": 0.8331929166666667,
      "accuracy_std_dev": 0.03339137345547584,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.055467366161616155,
      "avg_duration_per_task_ms": 273367.7276936027,
      "avg_output_tokens_per_task": 23750.69654882155,
      "last_run_timestamp": "2026-10-03T09:29:36.121Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Cloudflare",
      "accuracy": 0.8330725023474179,
      "accuracy_std_dev": 0.010168177287347969,
      "run_count": 7,
      "total_questions": 1278,
      "avg_cost_per_task_usd": 0.029710525039123633,
      "avg_duration_per_task_ms": 655555.5469483568,
      "avg_output_tokens_per_task": 42323.18544600939,
      "last_run_timestamp": "2026-10-03T09:40:53.626Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Parasail",
      "accuracy": 0.8329019637618638,
      "accuracy_std_dev": 0.014650670481433776,
      "run_count": 6,
      "total_questions": 1159,
      "avg_cost_per_task_usd": 0.1486290198446937,
      "avg_duration_per_task_ms": 266639.7955133736,
      "avg_output_tokens_per_task": 33590.909404659185,
      "last_run_timestamp": "2026-10-03T10:00:12.438Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8327431107738998,
      "accuracy_std_dev": 0.03352054924908005,
      "run_count": 10,
      "total_questions": 1977,
      "avg_cost_per_task_usd": 0.1104804410723318,
      "avg_duration_per_task_ms": 986804.3333333334,
      "avg_output_tokens_per_task": 34519.57764289327,
      "last_run_timestamp": "2026-09-26T10:24:21.190Z"
    },
    {
      "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": 0.832620296969697,
      "accuracy_std_dev": 0.019550200858985223,
      "run_count": 15,
      "total_questions": 2805,
      "avg_cost_per_task_usd": 0.1869388680926916,
      "avg_duration_per_task_ms": 373798.6777183601,
      "avg_output_tokens_per_task": 38438.90623885918,
      "last_run_timestamp": "2026-10-03T16:59:44.345Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "InferenceNet",
      "accuracy": 0.8324915000000002,
      "accuracy_std_dev": 0.007575500000000013,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.13435382575757576,
      "avg_duration_per_task_ms": 466692.7878787879,
      "avg_output_tokens_per_task": 61036.83838383838,
      "last_run_timestamp": "2026-10-03T17:03:50.322Z"
    },
    {
      "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": 0.8324914999999999,
      "accuracy_std_dev": 0.01902788623310536,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.02279124747474747,
      "avg_duration_per_task_ms": 532516.4444444445,
      "avg_output_tokens_per_task": 41107.4779040404,
      "last_run_timestamp": "2026-10-03T10:07:54.348Z"
    },
    {
      "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": 0.8324914999999999,
      "accuracy_std_dev": 0.000841499999999995,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.023044303030303033,
      "avg_duration_per_task_ms": 640134.2070707071,
      "avg_output_tokens_per_task": 54011.681818181816,
      "last_run_timestamp": "2026-09-26T08:54:28.265Z"
    },
    {
      "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": 0.8323712857142858,
      "accuracy_std_dev": 0.006903653726506745,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.24708544516594516,
      "avg_duration_per_task_ms": 346092.02308802307,
      "avg_output_tokens_per_task": 40400.90404040404,
      "last_run_timestamp": "2026-10-03T10:13:08.858Z"
    },
    {
      "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": 0.8322391838074399,
      "accuracy_std_dev": 0.02245308913522337,
      "run_count": 7,
      "total_questions": 1371,
      "avg_cost_per_task_usd": 0.08690679649890591,
      "avg_duration_per_task_ms": 951013.2662290299,
      "avg_output_tokens_per_task": 39963.18307804522,
      "last_run_timestamp": "2026-08-30T00:28:09.702Z"
    },
    {
      "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": 0.8314979893324156,
      "accuracy_std_dev": 0.10730469177701411,
      "run_count": 15,
      "total_questions": 2906,
      "avg_cost_per_task_usd": 0.007892868203716447,
      "avg_duration_per_task_ms": 643389.9057123193,
      "avg_output_tokens_per_task": 27155.329318651067,
      "last_run_timestamp": "2026-10-03T09:52:45.061Z"
    },
    {
      "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": 0.8313018608082102,
      "accuracy_std_dev": 0.013695449328918986,
      "run_count": 9,
      "total_questions": 1559,
      "avg_cost_per_task_usd": 0.11615080372033355,
      "avg_duration_per_task_ms": 231131.58178319436,
      "avg_output_tokens_per_task": 28108.629249518923,
      "last_run_timestamp": "2026-10-03T09:35:09.089Z"
    },
    {
      "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": 0.8312665185995624,
      "accuracy_std_dev": 0.02000709051964389,
      "run_count": 7,
      "total_questions": 1371,
      "avg_cost_per_task_usd": 0.07119963603209337,
      "avg_duration_per_task_ms": 887802.9110138585,
      "avg_output_tokens_per_task": 40565.85995623632,
      "last_run_timestamp": "2026-08-30T00:37:59.094Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Io Net",
      "accuracy": 0.8312629347826086,
      "accuracy_std_dev": 0.044536486139877485,
      "run_count": 4,
      "total_questions": 644,
      "avg_cost_per_task_usd": 0.005694647515527951,
      "avg_duration_per_task_ms": 270388.4906832298,
      "avg_output_tokens_per_task": 13521.194099378881,
      "last_run_timestamp": "2026-08-22T09:56:05.550Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.4-mini-20260317",
      "model_slug": "openai/gpt-5.4-mini",
      "model_name": "OpenAI: GPT-5.4 Mini",
      "provider_name": "Azure",
      "accuracy": 0.8310885000000001,
      "accuracy_std_dev": 0.012955632388399007,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.06418621717171719,
      "avg_duration_per_task_ms": 117116.24831649831,
      "avg_output_tokens_per_task": 13483.090067340067,
      "last_run_timestamp": "2026-10-01T19:14:05.661Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "BaseTen",
      "accuracy": 0.8309340194750211,
      "accuracy_std_dev": 0.0227049955198312,
      "run_count": 6,
      "total_questions": 1181,
      "avg_cost_per_task_usd": 0.0975369652836579,
      "avg_duration_per_task_ms": 214278.25232853513,
      "avg_output_tokens_per_task": 28093.950889077052,
      "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-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Nebius",
      "accuracy": 0.8308079999999999,
      "accuracy_std_dev": 0.019360000000000044,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.16246437626262625,
      "avg_duration_per_task_ms": 214673.86111111112,
      "avg_output_tokens_per_task": 36669.38383838384,
      "last_run_timestamp": "2026-10-03T09:42:09.121Z"
    },
    {
      "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": 0.8307627712082263,
      "accuracy_std_dev": 0.005791790283452965,
      "run_count": 8,
      "total_questions": 1556,
      "avg_cost_per_task_usd": 0.03552349742930591,
      "avg_duration_per_task_ms": 197284.01863753214,
      "avg_output_tokens_per_task": 13956.56233933162,
      "last_run_timestamp": "2026-10-03T09:16:27.631Z"
    },
    {
      "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": 0.8304254545454545,
      "accuracy_std_dev": 0.08702562758005528,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.09945538659320477,
      "avg_duration_per_task_ms": 844364.1143250689,
      "avg_output_tokens_per_task": 42647.71625344353,
      "last_run_timestamp": "2026-10-03T16:52:01.287Z"
    },
    {
      "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": 0.8302311218274112,
      "accuracy_std_dev": 0.009959264201340341,
      "run_count": 9,
      "total_questions": 1773,
      "avg_cost_per_task_usd": 0.0075188776085730385,
      "avg_duration_per_task_ms": 370069.358714044,
      "avg_output_tokens_per_task": 19121.66271855612,
      "last_run_timestamp": "2026-10-04T00:00:36.971Z"
    },
    {
      "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": 0.829966,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.022576904040404043,
      "avg_duration_per_task_ms": 10045.939393939394,
      "avg_output_tokens_per_task": 14604.111111111111,
      "last_run_timestamp": "2026-10-03T08:20:05.754Z"
    },
    {
      "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": 0.8298018056195302,
      "accuracy_std_dev": 0.03397957835387955,
      "run_count": 11,
      "total_questions": 2171,
      "avg_cost_per_task_usd": 0.06617050253339476,
      "avg_duration_per_task_ms": 1041423.7001381852,
      "avg_output_tokens_per_task": 33323.82865039152,
      "last_run_timestamp": "2026-10-03T17:39:05.807Z"
    },
    {
      "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": 0.82975575,
      "accuracy_std_dev": 0.03242017665262638,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.1467550763888889,
      "avg_duration_per_task_ms": 981685.6369949495,
      "avg_output_tokens_per_task": 52890.121212121216,
      "last_run_timestamp": "2026-10-03T17:37:18.132Z"
    },
    {
      "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": 0.8297257142857143,
      "accuracy_std_dev": 0.01460544534181185,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.02442155772005772,
      "avg_duration_per_task_ms": 463793.1847041847,
      "avg_output_tokens_per_task": 22682.81313131313,
      "last_run_timestamp": "2026-08-29T18:09:03.379Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "Venice",
      "accuracy": 0.8295454999999999,
      "accuracy_std_dev": 0.049917598154358346,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.09026368434343435,
      "avg_duration_per_task_ms": 461794.2853535354,
      "avg_output_tokens_per_task": 20466.646464646463,
      "last_run_timestamp": "2026-10-03T17:40:21.237Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.2-exp",
      "model_slug": "deepseek/deepseek-v3.2-exp",
      "model_name": "DeepSeek: DeepSeek V3.2 Exp",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8294053333333333,
      "accuracy_std_dev": 0.0031744380430068155,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.008978000000000002,
      "avg_duration_per_task_ms": 380339.61616161617,
      "avg_output_tokens_per_task": 21414.491582491584,
      "last_run_timestamp": "2026-08-01T20:03:34.768Z"
    },
    {
      "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": 0.8292847495662232,
      "accuracy_std_dev": 0.024481615398972544,
      "run_count": 9,
      "total_questions": 1729,
      "avg_cost_per_task_usd": 0.021951503759398496,
      "avg_duration_per_task_ms": 881311.056680162,
      "avg_output_tokens_per_task": 45066.87854251012,
      "last_run_timestamp": "2026-10-03T09:15:02.962Z"
    },
    {
      "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": 0.8288054812419146,
      "accuracy_std_dev": 0.014299354744511224,
      "run_count": 8,
      "total_questions": 1546,
      "avg_cost_per_task_usd": 0.03668785575679172,
      "avg_duration_per_task_ms": 429133.35187580856,
      "avg_output_tokens_per_task": 46204.894566623545,
      "last_run_timestamp": "2026-10-03T09:52:10.218Z"
    },
    {
      "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": 0.8286333058568329,
      "accuracy_std_dev": 0.029611903710969927,
      "run_count": 7,
      "total_questions": 1383,
      "avg_cost_per_task_usd": 0.025387288503253794,
      "avg_duration_per_task_ms": 480511.26970354305,
      "avg_output_tokens_per_task": 23613.104121475055,
      "last_run_timestamp": "2026-08-29T18:05:36.579Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "CoreWeave",
      "accuracy": 0.828497570984456,
      "accuracy_std_dev": 0.015230029187102692,
      "run_count": 5,
      "total_questions": 965,
      "avg_cost_per_task_usd": 0.11257541554404144,
      "avg_duration_per_task_ms": 438124.0186528497,
      "avg_output_tokens_per_task": 31261.672538860104,
      "last_run_timestamp": "2026-08-29T10:53:09.444Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "AtlasCloud",
      "accuracy": 0.82849325,
      "accuracy_std_dev": 0.05083663500800876,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.09963171022727271,
      "avg_duration_per_task_ms": 814676.8005050505,
      "avg_output_tokens_per_task": 39824.98926767677,
      "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-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "AkashML",
      "accuracy": 0.828283,
      "accuracy_std_dev": 0.016835000000000044,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.00548029292929293,
      "avg_duration_per_task_ms": 1403265.9545454546,
      "avg_output_tokens_per_task": 25437.79292929293,
      "last_run_timestamp": "2026-07-18T08:42:46.001Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "CoreWeave",
      "accuracy": 0.8279570000000002,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 186,
      "avg_cost_per_task_usd": 0.024709930107526885,
      "avg_duration_per_task_ms": 419969.8602150538,
      "avg_output_tokens_per_task": 20901.392473118278,
      "last_run_timestamp": "2026-08-01T18:03:43.857Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-sonnet-5-20260630",
      "model_slug": "anthropic/claude-sonnet-5",
      "model_name": "Anthropic: Claude Sonnet 5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.827862,
      "accuracy_std_dev": 0.0497186507736282,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.16292082386363638,
      "avg_duration_per_task_ms": 160402.9981060606,
      "avg_output_tokens_per_task": 16138.333333333334,
      "last_run_timestamp": "2026-10-03T09:35:14.442Z"
    },
    {
      "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": 0.8277963918918919,
      "accuracy_std_dev": 0.03021182471021881,
      "run_count": 9,
      "total_questions": 1776,
      "avg_cost_per_task_usd": 0.01481789527027027,
      "avg_duration_per_task_ms": 129146.45101351352,
      "avg_output_tokens_per_task": 17764.291666666668,
      "last_run_timestamp": "2026-10-03T17:12:18.981Z"
    },
    {
      "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": 0.82744125,
      "accuracy_std_dev": 0.11256498383905848,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.10842245328282829,
      "avg_duration_per_task_ms": 1501711.7184343433,
      "avg_output_tokens_per_task": 26982.195707070707,
      "last_run_timestamp": "2026-10-03T16:52:29.704Z"
    },
    {
      "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": 0.8274407499999998,
      "accuracy_std_dev": 0.002791768917281665,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.06294326136363637,
      "avg_duration_per_task_ms": 238417.01893939395,
      "avg_output_tokens_per_task": 15108.617424242424,
      "last_run_timestamp": "2026-10-03T10:04:25.365Z"
    },
    {
      "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": 0.8270997818897639,
      "accuracy_std_dev": 0.023868839957632546,
      "run_count": 13,
      "total_questions": 2540,
      "avg_cost_per_task_usd": 0.04439519330708661,
      "avg_duration_per_task_ms": 367160.94881889765,
      "avg_output_tokens_per_task": 47330.37204724409,
      "last_run_timestamp": "2026-10-03T17:12:02.690Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "BaseTen",
      "accuracy": 0.8269358,
      "accuracy_std_dev": 0.006767606146932625,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.04578402121212121,
      "avg_duration_per_task_ms": 60291.652525252524,
      "avg_output_tokens_per_task": 11285.754545454545,
      "last_run_timestamp": "2026-09-26T09:19:03.901Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "Phala",
      "accuracy": 0.8267676,
      "accuracy_std_dev": 0.014373097433747551,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.06195891818181818,
      "avg_duration_per_task_ms": 501765.85909090907,
      "avg_output_tokens_per_task": 48813.566666666666,
      "last_run_timestamp": "2026-10-03T17:23:58.689Z"
    },
    {
      "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": "WandB",
      "accuracy": 0.8265989999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.02378419696969697,
      "avg_duration_per_task_ms": 421976.08585858584,
      "avg_output_tokens_per_task": 20387.959595959597,
      "last_run_timestamp": "2026-07-18T16:18:27.151Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Crusoe",
      "accuracy": 0.8264543613323679,
      "accuracy_std_dev": 0.018188360323715465,
      "run_count": 7,
      "total_questions": 1381,
      "avg_cost_per_task_usd": 0.20616153874004342,
      "avg_duration_per_task_ms": 315117.1817523534,
      "avg_output_tokens_per_task": 46875.9225199131,
      "last_run_timestamp": "2026-09-12T08:53:53.147Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Friendli",
      "accuracy": 0.8263587142857144,
      "accuracy_std_dev": 0.007278841366086593,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.1363799206349206,
      "avg_duration_per_task_ms": 152940.04473304472,
      "avg_output_tokens_per_task": 33517.63347763348,
      "last_run_timestamp": "2026-10-03T09:28:23.102Z"
    },
    {
      "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": 0.8262166363636364,
      "accuracy_std_dev": 0.05839918106566644,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.05209585123966942,
      "avg_duration_per_task_ms": 600270.5303030303,
      "avg_output_tokens_per_task": 24943.10101010101,
      "last_run_timestamp": "2026-10-03T08:20:09.017Z"
    },
    {
      "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": 0.8261182857142857,
      "accuracy_std_dev": 0.010787762507506711,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.043446235209235205,
      "avg_duration_per_task_ms": 226902.49855699856,
      "avg_output_tokens_per_task": 14365.386002886004,
      "last_run_timestamp": "2026-10-03T09:30:12.452Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Crusoe",
      "accuracy": 0.8261094014447884,
      "accuracy_std_dev": 0.026551028658038853,
      "run_count": 5,
      "total_questions": 969,
      "avg_cost_per_task_usd": 0.2076026150670795,
      "avg_duration_per_task_ms": 857894.439628483,
      "avg_output_tokens_per_task": 47560.28689370485,
      "last_run_timestamp": "2026-08-29T18:53:23.077Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "CoreWeave",
      "accuracy": 0.8258873707224335,
      "accuracy_std_dev": 0.009676826643527784,
      "run_count": 6,
      "total_questions": 1052,
      "avg_cost_per_task_usd": 0.025350776615969583,
      "avg_duration_per_task_ms": 120122.08840304182,
      "avg_output_tokens_per_task": 20151.38783269962,
      "last_run_timestamp": "2026-08-29T18:29:13.946Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.2-exp",
      "model_slug": "deepseek/deepseek-v3.2-exp",
      "model_name": "DeepSeek: DeepSeek V3.2 Exp",
      "provider_name": "Novita",
      "accuracy": 0.8254766666666666,
      "accuracy_std_dev": 0.009654676391377487,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.008892052188552189,
      "avg_duration_per_task_ms": 1093286.5,
      "avg_output_tokens_per_task": 21288.892255892257,
      "last_run_timestamp": "2026-08-01T22:57:01.346Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "CoreWeave",
      "accuracy": 0.8254038475770924,
      "accuracy_std_dev": 0.017871570248271126,
      "run_count": 6,
      "total_questions": 1135,
      "avg_cost_per_task_usd": 0.024938470484581497,
      "avg_duration_per_task_ms": 321173.1885462555,
      "avg_output_tokens_per_task": 49662.51806167401,
      "last_run_timestamp": "2026-10-03T09:30:56.352Z"
    },
    {
      "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": 0.8253399469387755,
      "accuracy_std_dev": 0.06202361079266511,
      "run_count": 5,
      "total_questions": 980,
      "avg_cost_per_task_usd": 0.048117852040816325,
      "avg_duration_per_task_ms": 1077180.0714285714,
      "avg_output_tokens_per_task": 26155.976530612246,
      "last_run_timestamp": "2026-08-29T19:26:41.546Z"
    },
    {
      "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": 0.8249158181818181,
      "accuracy_std_dev": 0.02292623538782263,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.010601494031221303,
      "avg_duration_per_task_ms": 636029.4292929292,
      "avg_output_tokens_per_task": 37631.881542699724,
      "last_run_timestamp": "2026-10-03T09:23:44.336Z"
    },
    {
      "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": 0.8249157142857142,
      "accuracy_std_dev": 0.019837994431424387,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.1225871738816739,
      "avg_duration_per_task_ms": 339042.2626262626,
      "avg_output_tokens_per_task": 50612.39971139971,
      "last_run_timestamp": "2026-09-26T08:56:23.149Z"
    },
    {
      "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": 0.8246,
      "accuracy_std_dev": 0.010795510467319271,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.09903964393939395,
      "avg_duration_per_task_ms": 375586.2720959596,
      "avg_output_tokens_per_task": 45638.01262626263,
      "last_run_timestamp": "2026-10-04T00:05:42.411Z"
    },
    {
      "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": 0.824444492063492,
      "accuracy_std_dev": 0.011481171155826198,
      "run_count": 8,
      "total_questions": 1575,
      "avg_cost_per_task_usd": 0.02515456126984127,
      "avg_duration_per_task_ms": 461294.5841269841,
      "avg_output_tokens_per_task": 25119.664126984128,
      "last_run_timestamp": "2026-10-03T16:15:37.555Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.2-exp",
      "model_slug": "deepseek/deepseek-v3.2-exp",
      "model_name": "DeepSeek: DeepSeek V3.2 Exp",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8243546666666668,
      "accuracy_std_dev": 0.0067804647005611455,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0070816599326599325,
      "avg_duration_per_task_ms": 648845.1127946128,
      "avg_output_tokens_per_task": 16910.563973063974,
      "last_run_timestamp": "2026-08-01T18:25:37.046Z"
    },
    {
      "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": 0.8243546666666666,
      "accuracy_std_dev": 0.012396478769751095,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.038657538720538724,
      "avg_duration_per_task_ms": 268369.8787878788,
      "avg_output_tokens_per_task": 15453.594276094276,
      "last_run_timestamp": "2026-10-03T09:38:33.682Z"
    },
    {
      "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": 0.8238636249999999,
      "accuracy_std_dev": 0.022648157226016746,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.03663502967171717,
      "avg_duration_per_task_ms": 270780.9375,
      "avg_output_tokens_per_task": 14336.900252525253,
      "last_run_timestamp": "2026-10-03T09:40:25.832Z"
    },
    {
      "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": 0.8237297855579868,
      "accuracy_std_dev": 0.03088832039090874,
      "run_count": 7,
      "total_questions": 1371,
      "avg_cost_per_task_usd": 0.025826647702407,
      "avg_duration_per_task_ms": 464740.68708971556,
      "avg_output_tokens_per_task": 21465.053975200582,
      "last_run_timestamp": "2026-08-29T19:19:06.625Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Mancer 2",
      "accuracy": 0.823569,
      "accuracy_std_dev": 0.003264448253533828,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.039843759595959594,
      "avg_duration_per_task_ms": 211450.69292929294,
      "avg_output_tokens_per_task": 14991.045454545454,
      "last_run_timestamp": "2026-10-03T09:53:48.526Z"
    },
    {
      "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": 0.8234428750000001,
      "accuracy_std_dev": 0.0073110015462571785,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.04899790467171717,
      "avg_duration_per_task_ms": 455414.46527777775,
      "avg_output_tokens_per_task": 27141.95202020202,
      "last_run_timestamp": "2026-10-03T16:27:31.571Z"
    },
    {
      "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": 0.8233685148148148,
      "accuracy_std_dev": 0.018414157846885104,
      "run_count": 10,
      "total_questions": 1890,
      "avg_cost_per_task_usd": 0.0329625582010582,
      "avg_duration_per_task_ms": 257300.5761904762,
      "avg_output_tokens_per_task": 14146.621164021164,
      "last_run_timestamp": "2026-10-03T09:29:11.126Z"
    },
    {
      "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": 0.8233255173210161,
      "accuracy_std_dev": 0.0467612856991764,
      "run_count": 5,
      "total_questions": 866,
      "avg_cost_per_task_usd": 0.16841330484988454,
      "avg_duration_per_task_ms": 1763547.682448037,
      "avg_output_tokens_per_task": 49858.13856812933,
      "last_run_timestamp": "2026-10-03T17:19:48.286Z"
    },
    {
      "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": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Cloudflare",
      "accuracy": 0.8231693085966202,
      "accuracy_std_dev": 0.01720174744208347,
      "run_count": 7,
      "total_questions": 1361,
      "avg_cost_per_task_usd": 0.13977533284349744,
      "avg_duration_per_task_ms": 264582.22189566493,
      "avg_output_tokens_per_task": 31542.92799412197,
      "last_run_timestamp": "2026-10-03T09:52:37.682Z"
    },
    {
      "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": 0.8227512857142858,
      "accuracy_std_dev": 0.013638696342332996,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.03510758802308803,
      "avg_duration_per_task_ms": 250326.75108225108,
      "avg_output_tokens_per_task": 13920.564213564214,
      "last_run_timestamp": "2026-10-03T09:31:25.768Z"
    },
    {
      "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": 0.8223905999999999,
      "accuracy_std_dev": 0.023475537008554266,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.03811036666666667,
      "avg_duration_per_task_ms": 320984.0297979798,
      "avg_output_tokens_per_task": 39454.10808080808,
      "last_run_timestamp": "2026-10-03T16:50:17.392Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-2.5-pro",
      "model_slug": "google/gemini-2.5-pro",
      "model_name": "Google: Gemini 2.5 Pro",
      "provider_name": "Google AI Studio",
      "accuracy": 0.8223905000000001,
      "accuracy_std_dev": 0.0244105,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.2415952904040404,
      "avg_duration_per_task_ms": 181953.0101010101,
      "avg_output_tokens_per_task": 24066.49494949495,
      "last_run_timestamp": "2026-10-01T19:12:31.045Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "Novita",
      "accuracy": 0.821789142857143,
      "accuracy_std_dev": 0.03130762753446411,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.09338334415584416,
      "avg_duration_per_task_ms": 1431277.56998557,
      "avg_output_tokens_per_task": 43321.73593073593,
      "last_run_timestamp": "2026-10-03T17:44:28.471Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.2-exp",
      "model_slug": "deepseek/deepseek-v3.2-exp",
      "model_name": "DeepSeek: DeepSeek V3.2 Exp",
      "provider_name": "auto-routing",
      "accuracy": 0.8216522615384615,
      "accuracy_std_dev": 0.004955863351178644,
      "run_count": 3,
      "total_questions": 585,
      "avg_cost_per_task_usd": 0.008033046153846153,
      "avg_duration_per_task_ms": 616226.1504273504,
      "avg_output_tokens_per_task": 19179.20512820513,
      "last_run_timestamp": "2026-08-01T18:07:34.003Z"
    },
    {
      "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": 0.8215487142857143,
      "accuracy_std_dev": 0.020161900680769788,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.10321117243867245,
      "avg_duration_per_task_ms": 633991.6284271284,
      "avg_output_tokens_per_task": 38277.07575757576,
      "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-5.2-20260616",
      "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": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8212558434782609,
      "accuracy_std_dev": 0.006277707202019371,
      "run_count": 7,
      "total_questions": 1380,
      "avg_cost_per_task_usd": 0.007720972463768116,
      "avg_duration_per_task_ms": 593072.697826087,
      "avg_output_tokens_per_task": 17972.839130434782,
      "last_run_timestamp": "2026-08-29T18:40:24.101Z"
    },
    {
      "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": "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": 0.8210436999999998,
      "accuracy_std_dev": 0.014005426348740687,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.3147447828282829,
      "avg_duration_per_task_ms": 289184.82020202023,
      "avg_output_tokens_per_task": 20793.45303030303,
      "last_run_timestamp": "2026-10-03T09:11:51.622Z"
    },
    {
      "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": 0.8209994459203036,
      "accuracy_std_dev": 0.013381158348004657,
      "run_count": 8,
      "total_questions": 1581,
      "avg_cost_per_task_usd": 0.18810200189753318,
      "avg_duration_per_task_ms": 919106.2055660974,
      "avg_output_tokens_per_task": 50318.87096774193,
      "last_run_timestamp": "2026-10-03T16:39:56.568Z"
    },
    {
      "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": 0.820920596529284,
      "accuracy_std_dev": 0.007605756273362683,
      "run_count": 7,
      "total_questions": 1383,
      "avg_cost_per_task_usd": 0.019052912509038324,
      "avg_duration_per_task_ms": 539380.6218365871,
      "avg_output_tokens_per_task": 18914.664497469268,
      "last_run_timestamp": "2026-08-29T18:35:57.732Z"
    },
    {
      "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": 0.8207728695652173,
      "accuracy_std_dev": 0.01735923745738968,
      "run_count": 7,
      "total_questions": 1380,
      "avg_cost_per_task_usd": 0.021539910869565217,
      "avg_duration_per_task_ms": 499775.89130434784,
      "avg_output_tokens_per_task": 22209.624637681158,
      "last_run_timestamp": "2026-08-29T18:16:26.161Z"
    },
    {
      "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": 0.82070725,
      "accuracy_std_dev": 0.0036691031857253717,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.1439596148989899,
      "avg_duration_per_task_ms": 393442.0984848485,
      "avg_output_tokens_per_task": 28794.420454545456,
      "last_run_timestamp": "2026-10-03T09:54:41.255Z"
    },
    {
      "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": 0.820012649122807,
      "accuracy_std_dev": 0.022660477062222247,
      "run_count": 4,
      "total_questions": 513,
      "avg_cost_per_task_usd": 0.08905661598440547,
      "avg_duration_per_task_ms": 310577.5692007797,
      "avg_output_tokens_per_task": 36897.34307992203,
      "last_run_timestamp": "2026-10-03T17:40:40.949Z"
    },
    {
      "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": 0.8198652,
      "accuracy_std_dev": 0.057672987104536215,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.027881877272727274,
      "avg_duration_per_task_ms": 630281.2954545454,
      "avg_output_tokens_per_task": 28888.00202020202,
      "last_run_timestamp": "2026-10-03T10:05:26.294Z"
    },
    {
      "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": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "Novita",
      "accuracy": 0.8191919000000001,
      "accuracy_std_dev": 0.0203168642878275,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.012244186363636363,
      "avg_duration_per_task_ms": 659276.7813131313,
      "avg_output_tokens_per_task": 36210.856060606064,
      "last_run_timestamp": "2026-10-03T10:16:01.896Z"
    },
    {
      "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": 0.818813,
      "accuracy_std_dev": 0.021492541590049312,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.042027822601010095,
      "avg_duration_per_task_ms": 181041.06944444444,
      "avg_output_tokens_per_task": 23246.748106060608,
      "last_run_timestamp": "2026-10-03T17:30:27.921Z"
    },
    {
      "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": 0.8184125485714286,
      "accuracy_std_dev": 0.011787855675970066,
      "run_count": 8,
      "total_questions": 1575,
      "avg_cost_per_task_usd": 0.022701838095238094,
      "avg_duration_per_task_ms": 349381.28571428574,
      "avg_output_tokens_per_task": 20227.981587301587,
      "last_run_timestamp": "2026-10-04T00:17:30.035Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-5-fable-20260609",
      "model_slug": "anthropic/claude-fable-5",
      "model_name": "Anthropic: Claude Fable 5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.8181816666666667,
      "accuracy_std_dev": 0.013255807590972678,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.17135338047138046,
      "avg_duration_per_task_ms": 45814.95117845118,
      "avg_output_tokens_per_task": 3234.355218855219,
      "last_run_timestamp": "2026-08-15T09:15:35.213Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "thinkingmachines/inkling-20260715",
      "model_slug": "thinkingmachines/inkling",
      "model_name": "Thinking Machines: Inkling",
      "provider_name": "BaseTen",
      "accuracy": 0.818035862576065,
      "accuracy_std_dev": 0.01278056208505712,
      "run_count": 10,
      "total_questions": 1972,
      "avg_cost_per_task_usd": 0.0909173985801217,
      "avg_duration_per_task_ms": 165913.36257606492,
      "avg_output_tokens_per_task": 22308.299188640973,
      "last_run_timestamp": "2026-09-19T10:14:07.966Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "auto-routing",
      "accuracy": 0.8179412857142858,
      "accuracy_std_dev": 0.021435433107119505,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.00647507287157287,
      "avg_duration_per_task_ms": 501341.06277056277,
      "avg_output_tokens_per_task": 17357.66810966811,
      "last_run_timestamp": "2026-10-03T17:03:40.743Z"
    },
    {
      "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": 0.8179307705544933,
      "accuracy_std_dev": 0.0314446048740877,
      "run_count": 8,
      "total_questions": 1569,
      "avg_cost_per_task_usd": 0.23604747928616954,
      "avg_duration_per_task_ms": 962378.0981516889,
      "avg_output_tokens_per_task": 53848.4913957935,
      "last_run_timestamp": "2026-10-03T17:05:40.300Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta/muse-glimmer-30b-20260810",
      "model_slug": "meta/muse-glimmer-30b",
      "model_name": "Meta: Muse Glimmer 30B",
      "provider_name": "Together",
      "accuracy": 0.8178077777777777,
      "accuracy_std_dev": 0.009999885885768644,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.024910318742985414,
      "avg_duration_per_task_ms": 190389.1571268238,
      "avg_output_tokens_per_task": 16514.1734006734,
      "last_run_timestamp": "2026-10-03T10:10:35.343Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "CoreWeave",
      "accuracy": 0.8178076666666667,
      "accuracy_std_dev": 0.020010550661532987,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.04244091638608306,
      "avg_duration_per_task_ms": 255376.0824915825,
      "avg_output_tokens_per_task": 14176.158810325476,
      "last_run_timestamp": "2026-10-03T10:33:26.506Z"
    },
    {
      "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": 0.8177005714285713,
      "accuracy_std_dev": 0.009638124000286448,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.01912177994227994,
      "avg_duration_per_task_ms": 549930.608946609,
      "avg_output_tokens_per_task": 18923.724386724385,
      "last_run_timestamp": "2026-08-30T00:16:38.394Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta/muse-glimmer-30b-20260810",
      "model_slug": "meta/muse-glimmer-30b",
      "model_name": "Meta: Muse Glimmer 30B",
      "provider_name": "Phala",
      "accuracy": 0.816967559819413,
      "accuracy_std_dev": 0.022532849378432182,
      "run_count": 9,
      "total_questions": 1772,
      "avg_cost_per_task_usd": 0.021389000564334085,
      "avg_duration_per_task_ms": 297135.54796839727,
      "avg_output_tokens_per_task": 17946.336343115123,
      "last_run_timestamp": "2026-10-03T09:25:02.390Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.5-sonnet-20250929",
      "model_slug": "anthropic/claude-sonnet-4.5",
      "model_name": "Anthropic: Claude Sonnet 4.5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.8165714782608695,
      "accuracy_std_dev": 0.012063277850998734,
      "run_count": 23,
      "total_questions": 4554,
      "avg_cost_per_task_usd": 0.3338848906455862,
      "avg_duration_per_task_ms": 299201.53820816864,
      "avg_output_tokens_per_task": 21149.658322353975,
      "last_run_timestamp": "2026-10-03T09:43:11.426Z"
    },
    {
      "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": 0.8160116235741445,
      "accuracy_std_dev": 0.012319818494376241,
      "run_count": 8,
      "total_questions": 1578,
      "avg_cost_per_task_usd": 0.007714593789607097,
      "avg_duration_per_task_ms": 439013.47782002535,
      "avg_output_tokens_per_task": 19841.0525982256,
      "last_run_timestamp": "2026-10-04T00:12:29.471Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.1-flash-lite-20260507",
      "model_slug": "google/gemini-3.1-flash-lite",
      "model_name": "Google: Gemini 3.1 Flash Lite",
      "provider_name": "auto-routing",
      "accuracy": 0.815937,
      "accuracy_std_dev": 0.021042110493009022,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.04137441245791246,
      "avg_duration_per_task_ms": 82874.20875420875,
      "avg_output_tokens_per_task": 27465.196969696968,
      "last_run_timestamp": "2026-10-01T19:23:01.327Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "DeepInfra",
      "accuracy": 0.8158749778364116,
      "accuracy_std_dev": 0.06466277255569855,
      "run_count": 20,
      "total_questions": 3790,
      "avg_cost_per_task_usd": 0.0072713517150395784,
      "avg_duration_per_task_ms": 485517.56992084434,
      "avg_output_tokens_per_task": 17092.549340369394,
      "last_run_timestamp": "2026-10-03T17:48:04.648Z"
    },
    {
      "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": 0.8156565,
      "accuracy_std_dev": 0.01038452301023016,
      "run_count": 20,
      "total_questions": 3960,
      "avg_cost_per_task_usd": 0.36446038207070713,
      "avg_duration_per_task_ms": 359197.3244949495,
      "avg_output_tokens_per_task": 22956.0851010101,
      "last_run_timestamp": "2026-10-03T09:59:43.843Z"
    },
    {
      "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": 0.8152740000000002,
      "accuracy_std_dev": 0.03956053484563532,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.11835203581267217,
      "avg_duration_per_task_ms": 493386.7966023875,
      "avg_output_tokens_per_task": 33958.68686868687,
      "last_run_timestamp": "2026-10-03T16:58:37.335Z"
    },
    {
      "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": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "StreamLake",
      "accuracy": 0.8143337142857142,
      "accuracy_std_dev": 0.0065951553586008314,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.006175107503607504,
      "avg_duration_per_task_ms": 547758.3585858586,
      "avg_output_tokens_per_task": 18782.603174603173,
      "last_run_timestamp": "2026-08-29T19:09:30.030Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Z.AI",
      "accuracy": 0.8141798076190477,
      "accuracy_std_dev": 0.050441669383897034,
      "run_count": 8,
      "total_questions": 1575,
      "avg_cost_per_task_usd": 0.2300899580952381,
      "avg_duration_per_task_ms": 1003655.4222222222,
      "avg_output_tokens_per_task": 52717.56507936508,
      "last_run_timestamp": "2026-10-03T17:12:36.980Z"
    },
    {
      "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": 0.8140323536776214,
      "accuracy_std_dev": 0.032856006129862406,
      "run_count": 6,
      "total_questions": 639,
      "avg_cost_per_task_usd": 0.06527732394366197,
      "avg_duration_per_task_ms": 147029.73082942097,
      "avg_output_tokens_per_task": 28553.190923317685,
      "last_run_timestamp": "2026-08-15T14:54:40.470Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Relace",
      "accuracy": 0.814024,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 164,
      "avg_cost_per_task_usd": 0.010100225609756097,
      "avg_duration_per_task_ms": 515065.23780487804,
      "avg_output_tokens_per_task": 33245.84146341463,
      "last_run_timestamp": "2026-09-28T09:46:11.365Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "DeepInfra",
      "accuracy": 0.8137084857142858,
      "accuracy_std_dev": 0.0321649704013025,
      "run_count": 6,
      "total_questions": 1155,
      "avg_cost_per_task_usd": 0.07626592467532467,
      "avg_duration_per_task_ms": 1453708.5913419914,
      "avg_output_tokens_per_task": 43644.7670995671,
      "last_run_timestamp": "2026-08-22T17:50:04.003Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Morph",
      "accuracy": 0.8136272825651302,
      "accuracy_std_dev": 0.008773473238499465,
      "run_count": 3,
      "total_questions": 499,
      "avg_cost_per_task_usd": 0.0068851422845691386,
      "avg_duration_per_task_ms": 1090136,
      "avg_output_tokens_per_task": 17404.21643286573,
      "last_run_timestamp": "2026-08-15T10:05:32.709Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Chutes",
      "accuracy": 0.8133964770240699,
      "accuracy_std_dev": 0.02361801216297875,
      "run_count": 7,
      "total_questions": 1371,
      "avg_cost_per_task_usd": 0.006552795769511306,
      "avg_duration_per_task_ms": 681556.7833698031,
      "avg_output_tokens_per_task": 17687.919766593728,
      "last_run_timestamp": "2026-10-03T17:09:07.230Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2-thinking-20251106",
      "model_slug": "moonshotai/kimi-k2-thinking",
      "model_name": "MoonshotAI: Kimi K2 Thinking",
      "provider_name": "Google",
      "accuracy": 0.8133104923857867,
      "accuracy_std_dev": 0.012471288866833284,
      "run_count": 3,
      "total_questions": 591,
      "avg_cost_per_task_usd": 0.07992016920473774,
      "avg_duration_per_task_ms": 392391.41285956005,
      "avg_output_tokens_per_task": 33688.16074450085,
      "last_run_timestamp": "2026-08-01T17:52:55.742Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-2.5-pro",
      "model_slug": "google/gemini-2.5-pro",
      "model_name": "Google: Gemini 2.5 Pro",
      "provider_name": "auto-routing",
      "accuracy": 0.8131313333333333,
      "accuracy_std_dev": 0.02212166805545088,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.2550713585858586,
      "avg_duration_per_task_ms": 227479.37878787878,
      "avg_output_tokens_per_task": 25491.23232323232,
      "last_run_timestamp": "2026-10-01T19:12:42.150Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.2-chat-20251211",
      "model_slug": "openai/gpt-5.2-chat",
      "model_name": "OpenAI: GPT-5.2 Chat",
      "provider_name": "OpenAI",
      "accuracy": 0.813131,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.037260742424242425,
      "avg_duration_per_task_ms": 35156.99494949495,
      "avg_output_tokens_per_task": 2566.8030303030305,
      "last_run_timestamp": "2026-08-01T08:32:23.484Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Io Net",
      "accuracy": 0.8130078536585367,
      "accuracy_std_dev": 0.0073984999999999745,
      "run_count": 2,
      "total_questions": 369,
      "avg_cost_per_task_usd": 0.07202867479674796,
      "avg_duration_per_task_ms": 332408.0243902439,
      "avg_output_tokens_per_task": 16877.368563685635,
      "last_run_timestamp": "2026-07-18T08:30:38.228Z"
    },
    {
      "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": 0.8128504999999999,
      "accuracy_std_dev": 0.008607050515904558,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.03321095707070707,
      "avg_duration_per_task_ms": 706759.8703703703,
      "avg_output_tokens_per_task": 15781.767676767677,
      "last_run_timestamp": "2026-10-03T09:29:30.693Z"
    },
    {
      "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": 0.8127103749999999,
      "accuracy_std_dev": 0.009056161989737969,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.03019831565656566,
      "avg_duration_per_task_ms": 209380.85795454544,
      "avg_output_tokens_per_task": 23957.397095959597,
      "last_run_timestamp": "2026-10-03T17:45:01.711Z"
    },
    {
      "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": 0.8122894999999999,
      "accuracy_std_dev": 0.020078789204448232,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.006274926767676768,
      "avg_duration_per_task_ms": 514195.62205387204,
      "avg_output_tokens_per_task": 18652.397306397306,
      "last_run_timestamp": "2026-10-03T16:58:24.968Z"
    },
    {
      "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": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "StreamLake",
      "accuracy": 0.8112072857142857,
      "accuracy_std_dev": 0.030080678790376323,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.014970598124098124,
      "avg_duration_per_task_ms": 908622.7445887446,
      "avg_output_tokens_per_task": 44238.631313131315,
      "last_run_timestamp": "2026-10-03T08:04:44.082Z"
    },
    {
      "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": 0.8111109391304346,
      "accuracy_std_dev": 0.03310361291213897,
      "run_count": 7,
      "total_questions": 1380,
      "avg_cost_per_task_usd": 0.07639000724637682,
      "avg_duration_per_task_ms": 954401.9471014492,
      "avg_output_tokens_per_task": 39369.77246376812,
      "last_run_timestamp": "2026-08-30T00:30:19.495Z"
    },
    {
      "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": "Chutes",
      "accuracy": 0.8105691219512196,
      "accuracy_std_dev": 0.043372999999999995,
      "run_count": 2,
      "total_questions": 205,
      "avg_cost_per_task_usd": 0.04714223902439025,
      "avg_duration_per_task_ms": 880440.2341463414,
      "avg_output_tokens_per_task": 20870.317073170732,
      "last_run_timestamp": "2026-07-18T16:43:10.588Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.1-flash-lite-20260507",
      "model_slug": "google/gemini-3.1-flash-lite",
      "model_name": "Google: Gemini 3.1 Flash Lite",
      "provider_name": "Google AI Studio",
      "accuracy": 0.8103253333333333,
      "accuracy_std_dev": 0.011527714209195522,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.04107354882154882,
      "avg_duration_per_task_ms": 75169.15151515152,
      "avg_output_tokens_per_task": 27258.996632996634,
      "last_run_timestamp": "2026-10-01T19:16:58.595Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8100831277533042,
      "accuracy_std_dev": 0.021002816511805526,
      "run_count": 7,
      "total_questions": 1362,
      "avg_cost_per_task_usd": 0.16732624743024965,
      "avg_duration_per_task_ms": 899402.9625550661,
      "avg_output_tokens_per_task": 53272.42290748899,
      "last_run_timestamp": "2026-08-29T19:16:33.097Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "StreamLake",
      "accuracy": 0.8099512222222223,
      "accuracy_std_dev": 0.02558631077040038,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.10966612121212123,
      "avg_duration_per_task_ms": 777813.3344556678,
      "avg_output_tokens_per_task": 56577.68855218855,
      "last_run_timestamp": "2026-10-03T17:13:30.899Z"
    },
    {
      "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": 0.8095018965517242,
      "accuracy_std_dev": 0.0486761308529131,
      "run_count": 11,
      "total_questions": 2175,
      "avg_cost_per_task_usd": 0.08429347448275862,
      "avg_duration_per_task_ms": 792182.6556321839,
      "avg_output_tokens_per_task": 31699.048275862067,
      "last_run_timestamp": "2026-10-03T19:00:29.739Z"
    },
    {
      "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": 0.8092201778202676,
      "accuracy_std_dev": 0.025362105349011843,
      "run_count": 8,
      "total_questions": 1569,
      "avg_cost_per_task_usd": 0.19599221542383682,
      "avg_duration_per_task_ms": 884302.3785850861,
      "avg_output_tokens_per_task": 53656.268961121736,
      "last_run_timestamp": "2026-10-03T17:03:42.533Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Relace",
      "accuracy": 0.8091166153846153,
      "accuracy_std_dev": 0.03183613126622014,
      "run_count": 3,
      "total_questions": 585,
      "avg_cost_per_task_usd": 0.19830280854700855,
      "avg_duration_per_task_ms": 1060541.5008547008,
      "avg_output_tokens_per_task": 55592.3264957265,
      "last_run_timestamp": "2026-10-03T16:58:49.489Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-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": 0.8087552731239093,
      "accuracy_std_dev": 0.0546199990705989,
      "run_count": 6,
      "total_questions": 1146,
      "avg_cost_per_task_usd": 0.024335609075043632,
      "avg_duration_per_task_ms": 1150538.0907504363,
      "avg_output_tokens_per_task": 49299.98254799302,
      "last_run_timestamp": "2026-10-03T10:11:21.294Z"
    },
    {
      "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": 0.808728230769231,
      "accuracy_std_dev": 0.016181671721919724,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.34003047086247085,
      "avg_duration_per_task_ms": 354646.60295260296,
      "avg_output_tokens_per_task": 22500.726495726496,
      "last_run_timestamp": "2026-10-03T09:08:27.457Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5-mini-2025-08-07",
      "model_slug": "openai/gpt-5-mini",
      "model_name": "OpenAI: GPT-5 Mini",
      "provider_name": "auto-routing",
      "accuracy": 0.808081,
      "accuracy_std_dev": 0.018027342695657245,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.044340436026936034,
      "avg_duration_per_task_ms": 217888.1734006734,
      "avg_output_tokens_per_task": 22075.030303030304,
      "last_run_timestamp": "2026-10-01T19:19:28.473Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8079455869980879,
      "accuracy_std_dev": 0.03309943783864578,
      "run_count": 8,
      "total_questions": 1569,
      "avg_cost_per_task_usd": 0.2356793072020395,
      "avg_duration_per_task_ms": 903521.8578712556,
      "avg_output_tokens_per_task": 59416.81580624601,
      "last_run_timestamp": "2026-10-03T16:51:04.342Z"
    },
    {
      "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": 0.8073004516129032,
      "accuracy_std_dev": 0.05300431140963913,
      "run_count": 10,
      "total_questions": 1767,
      "avg_cost_per_task_usd": 0.046523333333333326,
      "avg_duration_per_task_ms": 546722.9988681381,
      "avg_output_tokens_per_task": 25674.418222976798,
      "last_run_timestamp": "2026-09-26T09:56:00.145Z"
    },
    {
      "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": 0.8072389999999999,
      "accuracy_std_dev": 0.0870155432092451,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.010984919191919193,
      "avg_duration_per_task_ms": 659527.0782828283,
      "avg_output_tokens_per_task": 25386.00420875421,
      "last_run_timestamp": "2026-10-03T17:34:53.587Z"
    },
    {
      "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": 0.807121686731666,
      "accuracy_std_dev": 0.1063175543594388,
      "run_count": 12,
      "total_questions": 2359,
      "avg_cost_per_task_usd": 0.047212891055532,
      "avg_duration_per_task_ms": 190020.95082662144,
      "avg_output_tokens_per_task": 22597.004662992793,
      "last_run_timestamp": "2026-10-03T10:14:18.163Z"
    },
    {
      "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": "WandB",
      "accuracy": 0.8068935127478752,
      "accuracy_std_dev": 0.009178499999999978,
      "run_count": 2,
      "total_questions": 353,
      "avg_cost_per_task_usd": 0.033640501416430595,
      "avg_duration_per_task_ms": 174824.1133144476,
      "avg_output_tokens_per_task": 26792.110481586402,
      "last_run_timestamp": "2026-07-18T08:04:57.082Z"
    },
    {
      "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": 0.8065657,
      "accuracy_std_dev": 0.022453683925137975,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.3406123287878787,
      "avg_duration_per_task_ms": 355782.31666666665,
      "avg_output_tokens_per_task": 22552.313636363637,
      "last_run_timestamp": "2026-10-03T09:49:05.353Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-2.5-pro",
      "model_slug": "google/gemini-2.5-pro",
      "model_name": "Google: Gemini 2.5 Pro",
      "provider_name": "Google",
      "accuracy": 0.806210222222222,
      "accuracy_std_dev": 0.03697273211969953,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.25474633557800225,
      "avg_duration_per_task_ms": 231596.29349046017,
      "avg_output_tokens_per_task": 25466.63243546577,
      "last_run_timestamp": "2026-10-01T19:17:25.275Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta/muse-glimmer-30b-20260810",
      "model_slug": "meta/muse-glimmer-30b",
      "model_name": "Meta: Muse Glimmer 30B",
      "provider_name": "auto-routing",
      "accuracy": 0.8059763749999999,
      "accuracy_std_dev": 0.01367028747811745,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.022531703914141412,
      "avg_duration_per_task_ms": 251092.985479798,
      "avg_output_tokens_per_task": 17458.415404040403,
      "last_run_timestamp": "2026-10-03T09:13:36.277Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Chutes",
      "accuracy": 0.805556,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 192,
      "avg_cost_per_task_usd": 0.16585933333333333,
      "avg_duration_per_task_ms": 688743.8177083334,
      "avg_output_tokens_per_task": 42426.447916666664,
      "last_run_timestamp": "2026-07-31T00:21:35.641Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "DekaLLM",
      "accuracy": 0.805555546875,
      "accuracy_std_dev": 0.01288499999999998,
      "run_count": 2,
      "total_questions": 384,
      "avg_cost_per_task_usd": 0.0362872109375,
      "avg_duration_per_task_ms": 729621.7864583334,
      "avg_output_tokens_per_task": 36271.864583333336,
      "last_run_timestamp": "2026-09-12T16:45:54.034Z"
    },
    {
      "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": 0.8053298279132792,
      "accuracy_std_dev": 0.03337153373515719,
      "run_count": 7,
      "total_questions": 738,
      "avg_cost_per_task_usd": 0.029580651761517616,
      "avg_duration_per_task_ms": 101041.89430894308,
      "avg_output_tokens_per_task": 30707.855013550135,
      "last_run_timestamp": "2026-08-29T19:44:46.543Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.3-chat-20260303",
      "model_slug": "openai/gpt-5.3-chat",
      "model_name": "OpenAI: GPT-5.3 Chat",
      "provider_name": "auto-routing",
      "accuracy": 0.804714,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.024081237373737373,
      "avg_duration_per_task_ms": 24354.015151515152,
      "avg_output_tokens_per_task": 1625.5555555555557,
      "last_run_timestamp": "2026-08-01T08:38:11.900Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8045310368763557,
      "accuracy_std_dev": 0.02305431337923024,
      "run_count": 7,
      "total_questions": 1383,
      "avg_cost_per_task_usd": 0.027800933477946495,
      "avg_duration_per_task_ms": 458826.73535791755,
      "avg_output_tokens_per_task": 23174.896601590746,
      "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-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "auto-routing",
      "accuracy": 0.8042325714285714,
      "accuracy_std_dev": 0.02234715177030169,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.007311367243867243,
      "avg_duration_per_task_ms": 505999.5836940837,
      "avg_output_tokens_per_task": 18157.529581529583,
      "last_run_timestamp": "2026-08-29T19:12:36.104Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Mara",
      "accuracy": 0.8034189230769232,
      "accuracy_std_dev": 0.024712500000000026,
      "run_count": 2,
      "total_questions": 195,
      "avg_cost_per_task_usd": 0.08983526153846154,
      "avg_duration_per_task_ms": 329694.73846153845,
      "avg_output_tokens_per_task": 37199.13846153846,
      "last_run_timestamp": "2026-10-03T16:53:50.090Z"
    },
    {
      "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": 0.8032406250000002,
      "accuracy_std_dev": 0.023560394356512266,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.3350150732323232,
      "avg_duration_per_task_ms": 352876.84217171714,
      "avg_output_tokens_per_task": 22221.229166666668,
      "last_run_timestamp": "2026-10-03T09:34:21.327Z"
    },
    {
      "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": 0.8026253078034682,
      "accuracy_std_dev": 0.11466736789371888,
      "run_count": 7,
      "total_questions": 1384,
      "avg_cost_per_task_usd": 0.12527993786127167,
      "avg_duration_per_task_ms": 694994.5281791907,
      "avg_output_tokens_per_task": 39948.101156069366,
      "last_run_timestamp": "2026-08-29T18:23:09.250Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.1-flash-lite-20260507",
      "model_slug": "google/gemini-3.1-flash-lite",
      "model_name": "Google: Gemini 3.1 Flash Lite",
      "provider_name": "Google",
      "accuracy": 0.8020951111111111,
      "accuracy_std_dev": 0.027492606765554845,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.04329651571268237,
      "avg_duration_per_task_ms": 84651.98765432098,
      "avg_output_tokens_per_task": 26957.542087542086,
      "last_run_timestamp": "2026-10-01T19:25:58.919Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-thinking-2507",
      "model_slug": "qwen/qwen3-235b-a22b-thinking-2507",
      "model_name": "Qwen: Qwen3 235B A22B Thinking 2507",
      "provider_name": "DeepInfra",
      "accuracy": 0.8019926891304349,
      "accuracy_std_dev": 0.02443714833108715,
      "run_count": 6,
      "total_questions": 920,
      "avg_cost_per_task_usd": 0.05089078369565218,
      "avg_duration_per_task_ms": 796111.2945652173,
      "avg_output_tokens_per_task": 22057.085869565217,
      "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-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Io Net",
      "accuracy": 0.8016385562742562,
      "accuracy_std_dev": 0.007745278218856984,
      "run_count": 4,
      "total_questions": 773,
      "avg_cost_per_task_usd": 0.0594386532988357,
      "avg_duration_per_task_ms": 220979.64165588617,
      "avg_output_tokens_per_task": 17329.77360931436,
      "last_run_timestamp": "2026-09-26T08:53:08.262Z"
    },
    {
      "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": 0.8013468181818181,
      "accuracy_std_dev": 0.10160515424186116,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.09367279752066117,
      "avg_duration_per_task_ms": 528447.3296602387,
      "avg_output_tokens_per_task": 34556.12350780533,
      "last_run_timestamp": "2026-10-03T17:43:17.222Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5-mini-2025-08-07",
      "model_slug": "openai/gpt-5-mini",
      "model_name": "OpenAI: GPT-5 Mini",
      "provider_name": "OpenAI",
      "accuracy": 0.8007853333333332,
      "accuracy_std_dev": 0.008398773614178595,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0439676531986532,
      "avg_duration_per_task_ms": 201662.1632996633,
      "avg_output_tokens_per_task": 21888.12962962963,
      "last_run_timestamp": "2026-10-01T19:26:25.385Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "DigitalOcean",
      "accuracy": 0.8007155,
      "accuracy_std_dev": 0.04390787837905629,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.039541025883838385,
      "avg_duration_per_task_ms": 942411.1104797979,
      "avg_output_tokens_per_task": 24048.43876262626,
      "last_run_timestamp": "2026-09-26T09:34:13.954Z"
    },
    {
      "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": "Phala",
      "accuracy": 0.800505,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.14201096464646465,
      "avg_duration_per_task_ms": 1387672.616161616,
      "avg_output_tokens_per_task": 46225.35353535353,
      "last_run_timestamp": "2026-07-11T16:02:27.787Z"
    },
    {
      "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": 0.7996632000000001,
      "accuracy_std_dev": 0.17472434887834037,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.09227538585858586,
      "avg_duration_per_task_ms": 490452.89090909093,
      "avg_output_tokens_per_task": 24835.06313131313,
      "last_run_timestamp": "2026-10-03T16:43:18.077Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5-mini-2025-08-07",
      "model_slug": "openai/gpt-5-mini",
      "model_name": "OpenAI: GPT-5 Mini",
      "provider_name": "Azure",
      "accuracy": 0.7996631666666666,
      "accuracy_std_dev": 0.024569953550469384,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.045980551346801346,
      "avg_duration_per_task_ms": 194150.27525252526,
      "avg_output_tokens_per_task": 21808.128787878788,
      "last_run_timestamp": "2026-10-01T19:34:33.810Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "Friendli",
      "accuracy": 0.799242375,
      "accuracy_std_dev": 0.022425038149006003,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.02821298611111111,
      "avg_duration_per_task_ms": 287144.1205808081,
      "avg_output_tokens_per_task": 18650.696338383837,
      "last_run_timestamp": "2026-10-04T00:12:29.472Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.1-terminus",
      "model_slug": "deepseek/deepseek-v3.1-terminus",
      "model_name": "DeepSeek: DeepSeek V3.1 Terminus",
      "provider_name": "StreamLake",
      "accuracy": 0.7989418571428571,
      "accuracy_std_dev": 0.010175039872558895,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.01636965007215007,
      "avg_duration_per_task_ms": 355837.20057720057,
      "avg_output_tokens_per_task": 15678.383116883117,
      "last_run_timestamp": "2026-08-29T16:10:33.070Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "auto-routing",
      "accuracy": 0.7984005000000001,
      "accuracy_std_dev": 0.04195903080386868,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.02730416792929293,
      "avg_duration_per_task_ms": 362984.9494949495,
      "avg_output_tokens_per_task": 25613.62626262626,
      "last_run_timestamp": "2026-10-03T19:56:06.052Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.2-chat-20251211",
      "model_slug": "openai/gpt-5.2-chat",
      "model_name": "OpenAI: GPT-5.2 Chat",
      "provider_name": "auto-routing",
      "accuracy": 0.79798,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.03602157070707071,
      "avg_duration_per_task_ms": 32858.843434343435,
      "avg_output_tokens_per_task": 2479.818181818182,
      "last_run_timestamp": "2026-08-01T08:32:36.352Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.7979796666666666,
      "accuracy_std_dev": 0.009816464683830397,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.04952355303030303,
      "avg_duration_per_task_ms": 458017.70707070705,
      "avg_output_tokens_per_task": 24647.512626262625,
      "last_run_timestamp": "2026-08-22T17:18:01.115Z"
    },
    {
      "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": "deepseek/deepseek-v3.1-terminus",
      "model_slug": "deepseek/deepseek-v3.1-terminus",
      "model_name": "DeepSeek: DeepSeek V3.1 Terminus",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7974987142857143,
      "accuracy_std_dev": 0.009164318316387839,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.01350666378066378,
      "avg_duration_per_task_ms": 299956.4531024531,
      "avg_output_tokens_per_task": 14063.395382395382,
      "last_run_timestamp": "2026-08-29T19:12:57.906Z"
    },
    {
      "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": "Fireworks",
      "accuracy": 0.797043,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 124,
      "avg_cost_per_task_usd": 0.11967770967741935,
      "avg_duration_per_task_ms": 375576.0080645161,
      "avg_output_tokens_per_task": 29097.58064516129,
      "last_run_timestamp": "2026-07-18T08:32:17.535Z"
    },
    {
      "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": 0.79671725,
      "accuracy_std_dev": 0.006560703749408305,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.05368239015151515,
      "avg_duration_per_task_ms": 1192594.2481060605,
      "avg_output_tokens_per_task": 24468.87436868687,
      "last_run_timestamp": "2026-10-03T18:46:58.291Z"
    },
    {
      "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": 0.7964375923664123,
      "accuracy_std_dev": 0.11378331267012533,
      "run_count": 4,
      "total_questions": 655,
      "avg_cost_per_task_usd": 0.011028116030534352,
      "avg_duration_per_task_ms": 270893.28854961833,
      "avg_output_tokens_per_task": 23218.567938931297,
      "last_run_timestamp": "2026-09-12T08:53:48.637Z"
    },
    {
      "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": 0.7963679903288201,
      "accuracy_std_dev": 0.04084804193727499,
      "run_count": 8,
      "total_questions": 1551,
      "avg_cost_per_task_usd": 0.1404589497098646,
      "avg_duration_per_task_ms": 1543378.2327530626,
      "avg_output_tokens_per_task": 46550.5467440361,
      "last_run_timestamp": "2026-10-03T16:58:27.160Z"
    },
    {
      "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": 0.7956650000000001,
      "accuracy_std_dev": 0.0069381316108877815,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.014228089646464647,
      "avg_duration_per_task_ms": 327580.9936868687,
      "avg_output_tokens_per_task": 14697.917297979799,
      "last_run_timestamp": "2026-10-04T00:00:36.942Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "Novita",
      "accuracy": 0.7956228,
      "accuracy_std_dev": 0.021522407276138952,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.006844225252525253,
      "avg_duration_per_task_ms": 363476.91515151516,
      "avg_output_tokens_per_task": 16652.98888888889,
      "last_run_timestamp": "2026-08-29T19:20:50.265Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-r1-0528",
      "model_slug": "deepseek/deepseek-r1-0528",
      "model_name": "DeepSeek: R1 0528",
      "provider_name": "DeepInfra",
      "accuracy": 0.795444345060893,
      "accuracy_std_dev": 0.01127012798285806,
      "run_count": 8,
      "total_questions": 1478,
      "avg_cost_per_task_usd": 0.04254500879566983,
      "avg_duration_per_task_ms": 730941.7388362652,
      "avg_output_tokens_per_task": 19660.586603518266,
      "last_run_timestamp": "2026-10-03T16:16:31.572Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Io Net",
      "accuracy": 0.7950336666666666,
      "accuracy_std_dev": 0.03307482054205114,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.03994933501683502,
      "avg_duration_per_task_ms": 251252.56818181818,
      "avg_output_tokens_per_task": 12806.098484848484,
      "last_run_timestamp": "2026-09-19T09:05:41.015Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Phala",
      "accuracy": 0.7949320752351098,
      "accuracy_std_dev": 0.09853873704264732,
      "run_count": 10,
      "total_questions": 1914,
      "avg_cost_per_task_usd": 0.08235360710553814,
      "avg_duration_per_task_ms": 843777.6013584117,
      "avg_output_tokens_per_task": 30560.44775339603,
      "last_run_timestamp": "2026-10-03T19:06:55.619Z"
    },
    {
      "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": 0.7947657272727272,
      "accuracy_std_dev": 0.03887966022132421,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.042441610651974285,
      "avg_duration_per_task_ms": 470794.4618916437,
      "avg_output_tokens_per_task": 42513.95821854913,
      "last_run_timestamp": "2026-10-03T17:31:01.734Z"
    },
    {
      "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": 0.7947205871946083,
      "accuracy_std_dev": 0.04278072577075129,
      "run_count": 6,
      "total_questions": 1187,
      "avg_cost_per_task_usd": 0.1992195315922494,
      "avg_duration_per_task_ms": 633124.2401010952,
      "avg_output_tokens_per_task": 47420.48188711036,
      "last_run_timestamp": "2026-10-03T17:09:08.099Z"
    },
    {
      "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": 0.7929293333333333,
      "accuracy_std_dev": 0.008361060791284552,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.01913193771043771,
      "avg_duration_per_task_ms": 454912.569023569,
      "avg_output_tokens_per_task": 25421.24579124579,
      "last_run_timestamp": "2026-10-03T18:54:19.540Z"
    },
    {
      "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": 0.7928527272727273,
      "accuracy_std_dev": 0.04874799814088574,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.17375368870523417,
      "avg_duration_per_task_ms": 837428.4205693296,
      "avg_output_tokens_per_task": 56964.92470156107,
      "last_run_timestamp": "2026-10-03T17:12:34.213Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2-0905",
      "model_slug": "moonshotai/kimi-k2-0905",
      "model_name": "MoonshotAI: Kimi K2 0905",
      "provider_name": "Novita",
      "accuracy": 0.7924422284263961,
      "accuracy_std_dev": 0.011171723004482763,
      "run_count": 3,
      "total_questions": 591,
      "avg_cost_per_task_usd": 0.016949575296108292,
      "avg_duration_per_task_ms": 173392.4331641286,
      "avg_output_tokens_per_task": 6588.478849407784,
      "last_run_timestamp": "2026-08-01T17:44:11.144Z"
    },
    {
      "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": 0.7918084721906924,
      "accuracy_std_dev": 0.14984325869964388,
      "run_count": 9,
      "total_questions": 1762,
      "avg_cost_per_task_usd": 0.016061496027241773,
      "avg_duration_per_task_ms": 528116.8059023836,
      "avg_output_tokens_per_task": 25582.743473325765,
      "last_run_timestamp": "2026-10-03T08:06:43.537Z"
    },
    {
      "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": 0.7915958188529246,
      "accuracy_std_dev": 0.016282913854712845,
      "run_count": 10,
      "total_questions": 1761,
      "avg_cost_per_task_usd": 0.004947599091425326,
      "avg_duration_per_task_ms": 1178125.1942078364,
      "avg_output_tokens_per_task": 32554.07495741056,
      "last_run_timestamp": "2026-09-26T21:21:20.008Z"
    },
    {
      "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": 0.7914820056925996,
      "accuracy_std_dev": 0.03882771144407526,
      "run_count": 8,
      "total_questions": 1581,
      "avg_cost_per_task_usd": 0.014992018342821002,
      "avg_duration_per_task_ms": 477486.2156862745,
      "avg_output_tokens_per_task": 15890.679316888045,
      "last_run_timestamp": "2026-10-04T00:00:36.969Z"
    },
    {
      "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": 0.7912458333333334,
      "accuracy_std_dev": 0.01576267340287031,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.01776989141414141,
      "avg_duration_per_task_ms": 61590.219696969696,
      "avg_output_tokens_per_task": 5770.594276094276,
      "last_run_timestamp": "2026-10-03T16:56:50.423Z"
    },
    {
      "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": 0.7910545355932204,
      "accuracy_std_dev": 0.011679465336358934,
      "run_count": 9,
      "total_questions": 1770,
      "avg_cost_per_task_usd": 0.018264465536723163,
      "avg_duration_per_task_ms": 145761.14463276835,
      "avg_output_tokens_per_task": 15068.749717514123,
      "last_run_timestamp": "2026-10-03T09:28:19.724Z"
    },
    {
      "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": 0.7910339219088935,
      "accuracy_std_dev": 0.007596404447540465,
      "run_count": 7,
      "total_questions": 1383,
      "avg_cost_per_task_usd": 0.18853212509038325,
      "avg_duration_per_task_ms": 939271.1323210412,
      "avg_output_tokens_per_task": 47171.86912509039,
      "last_run_timestamp": "2026-10-03T17:06:15.877Z"
    },
    {
      "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": 0.7891350392405064,
      "accuracy_std_dev": 0.020945155322890284,
      "run_count": 8,
      "total_questions": 1580,
      "avg_cost_per_task_usd": 0.06166090379746835,
      "avg_duration_per_task_ms": 176165.68037974683,
      "avg_output_tokens_per_task": 19378.685443037975,
      "last_run_timestamp": "2026-10-03T10:40:10.028Z"
    },
    {
      "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": 0.7884000133843213,
      "accuracy_std_dev": 0.06448360442960967,
      "run_count": 8,
      "total_questions": 1569,
      "avg_cost_per_task_usd": 0.21083066475462078,
      "avg_duration_per_task_ms": 943051.9751434034,
      "avg_output_tokens_per_task": 64724.986615678776,
      "last_run_timestamp": "2026-10-03T17:10:01.034Z"
    },
    {
      "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": 0.78761423645894,
      "accuracy_std_dev": 0.05455970909816173,
      "run_count": 11,
      "total_questions": 1717,
      "avg_cost_per_task_usd": 0.05547815375655212,
      "avg_duration_per_task_ms": 804006.846825859,
      "avg_output_tokens_per_task": 46685.46825859057,
      "last_run_timestamp": "2026-08-29T19:07:57.500Z"
    },
    {
      "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": 0.787571717325228,
      "accuracy_std_dev": 0.005353629893819705,
      "run_count": 5,
      "total_questions": 987,
      "avg_cost_per_task_usd": 0.01375579837892604,
      "avg_duration_per_task_ms": 469828.6362715299,
      "avg_output_tokens_per_task": 14334.128672745694,
      "last_run_timestamp": "2026-08-15T14:56:55.269Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Phala",
      "accuracy": 0.7857744999999999,
      "accuracy_std_dev": 0.05357356939993824,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.18102450063131315,
      "avg_duration_per_task_ms": 1433174.245580808,
      "avg_output_tokens_per_task": 45945.68939393939,
      "last_run_timestamp": "2026-10-03T17:00:45.414Z"
    },
    {
      "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": 0.7847189879865448,
      "accuracy_std_dev": 0.015912532710814054,
      "run_count": 11,
      "total_questions": 2081,
      "avg_cost_per_task_usd": 0.008999504565112927,
      "avg_duration_per_task_ms": 423343.8265257088,
      "avg_output_tokens_per_task": 35697.484863046615,
      "last_run_timestamp": "2026-10-03T18:48:40.173Z"
    },
    {
      "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": "WandB",
      "accuracy": 0.784512,
      "accuracy_std_dev": 0.003367000000000009,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.02099649242424242,
      "avg_duration_per_task_ms": 284075.095959596,
      "avg_output_tokens_per_task": 12479.631313131313,
      "last_run_timestamp": "2026-07-18T16:16:42.667Z"
    },
    {
      "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": "Morph",
      "accuracy": 0.784512,
      "accuracy_std_dev": 0.043771000000000004,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.030820969696969697,
      "avg_duration_per_task_ms": 734052.446969697,
      "avg_output_tokens_per_task": 25681.72474747475,
      "last_run_timestamp": "2026-07-18T09:15:42.531Z"
    },
    {
      "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": 0.7842310833333334,
      "accuracy_std_dev": 0.022077051155964113,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.004981357744107744,
      "avg_duration_per_task_ms": 268156.8005050505,
      "avg_output_tokens_per_task": 32664.053030303032,
      "last_run_timestamp": "2026-10-03T17:02:47.316Z"
    },
    {
      "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": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "WandB",
      "accuracy": 0.7830773446153846,
      "accuracy_std_dev": 0.015297999999999978,
      "run_count": 2,
      "total_questions": 325,
      "avg_cost_per_task_usd": 0.07515540923076923,
      "avg_duration_per_task_ms": 234345.21538461538,
      "avg_output_tokens_per_task": 20795.20923076923,
      "last_run_timestamp": "2026-07-18T08:37:00.147Z"
    },
    {
      "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": 0.7825476666666666,
      "accuracy_std_dev": 0.020923567929224913,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.055650603535353546,
      "avg_duration_per_task_ms": 666314.5917508417,
      "avg_output_tokens_per_task": 24452.340909090908,
      "last_run_timestamp": "2026-08-22T17:09:41.940Z"
    },
    {
      "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": 0.782495077073808,
      "accuracy_std_dev": 0.1325801523111939,
      "run_count": 8,
      "total_questions": 1531,
      "avg_cost_per_task_usd": 0.023154647289353365,
      "avg_duration_per_task_ms": 345079.2540822991,
      "avg_output_tokens_per_task": 23624.08883082952,
      "last_run_timestamp": "2026-10-03T19:19:20.712Z"
    },
    {
      "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": 0.7822107419689119,
      "accuracy_std_dev": 0.054227012196505914,
      "run_count": 5,
      "total_questions": 965,
      "avg_cost_per_task_usd": 0.007577522279792746,
      "avg_duration_per_task_ms": 591909.3450777202,
      "avg_output_tokens_per_task": 17266.958549222796,
      "last_run_timestamp": "2026-08-29T19:08:46.030Z"
    },
    {
      "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": 0.7817837115749525,
      "accuracy_std_dev": 0.04901063227759871,
      "run_count": 8,
      "total_questions": 1581,
      "avg_cost_per_task_usd": 0.13180700759013284,
      "avg_duration_per_task_ms": 1013578.3162555344,
      "avg_output_tokens_per_task": 60177.89626818469,
      "last_run_timestamp": "2026-10-03T17:38:07.604Z"
    },
    {
      "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": 0.7815378781725889,
      "accuracy_std_dev": 0.023695801760263955,
      "run_count": 9,
      "total_questions": 1773,
      "avg_cost_per_task_usd": 0.05647687704455725,
      "avg_duration_per_task_ms": 566671.4066553863,
      "avg_output_tokens_per_task": 24552.29046813311,
      "last_run_timestamp": "2026-10-03T16:18:18.642Z"
    },
    {
      "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": 0.7804232857142855,
      "accuracy_std_dev": 0.00964416607835736,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.021076865079365077,
      "avg_duration_per_task_ms": 281668.99494949495,
      "avg_output_tokens_per_task": 12536.775613275613,
      "last_run_timestamp": "2026-10-04T00:50:16.196Z"
    },
    {
      "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": 0.7799144730769231,
      "accuracy_std_dev": 0.07491524224415215,
      "run_count": 8,
      "total_questions": 1560,
      "avg_cost_per_task_usd": 0.2146196794871795,
      "avg_duration_per_task_ms": 1130627.516025641,
      "avg_output_tokens_per_task": 68716.01923076923,
      "last_run_timestamp": "2026-10-03T16:59:22.739Z"
    },
    {
      "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": 0.7794466007905138,
      "accuracy_std_dev": 0.015712587779687186,
      "run_count": 8,
      "total_questions": 1265,
      "avg_cost_per_task_usd": 0.05468981976284586,
      "avg_duration_per_task_ms": 1028648.5209486166,
      "avg_output_tokens_per_task": 28503.08932806324,
      "last_run_timestamp": "2026-10-03T18:12:08.757Z"
    },
    {
      "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": 0.7791806666666666,
      "accuracy_std_dev": 0.11069114794126743,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.14781284764309766,
      "avg_duration_per_task_ms": 1097204.1927609427,
      "avg_output_tokens_per_task": 42915.851851851854,
      "last_run_timestamp": "2026-10-03T17:43:01.258Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "Google",
      "accuracy": 0.777456579479769,
      "accuracy_std_dev": 0.02062598971775015,
      "run_count": 7,
      "total_questions": 1384,
      "avg_cost_per_task_usd": 0.021605648843930635,
      "avg_duration_per_task_ms": 786693.5744219653,
      "avg_output_tokens_per_task": 37390.66040462428,
      "last_run_timestamp": "2026-10-03T16:57:26.262Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.4-nano-20260317",
      "model_slug": "openai/gpt-5.4-nano",
      "model_name": "OpenAI: GPT-5.4 Nano",
      "provider_name": "auto-routing",
      "accuracy": 0.7772166666666667,
      "accuracy_std_dev": 0.003459479954495376,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.013080006734006736,
      "avg_duration_per_task_ms": 67697.39225589225,
      "avg_output_tokens_per_task": 10341.513468013469,
      "last_run_timestamp": "2026-10-01T19:12:45.924Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.4-nano-20260317",
      "model_slug": "openai/gpt-5.4-nano",
      "model_name": "OpenAI: GPT-5.4 Nano",
      "provider_name": "OpenAI",
      "accuracy": 0.77567325,
      "accuracy_std_dev": 0.010934638822910419,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.013401992424242422,
      "avg_duration_per_task_ms": 67621.48106060606,
      "avg_output_tokens_per_task": 10598.125,
      "last_run_timestamp": "2026-10-02T23:20:38.343Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.1-terminus",
      "model_slug": "deepseek/deepseek-v3.1-terminus",
      "model_name": "DeepSeek: DeepSeek V3.1 Terminus",
      "provider_name": "auto-routing",
      "accuracy": 0.7756132857142857,
      "accuracy_std_dev": 0.020236370423531182,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.015738134199134196,
      "avg_duration_per_task_ms": 439103.9264069264,
      "avg_output_tokens_per_task": 15898.891053391053,
      "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": "Crusoe",
      "accuracy": 0.774411,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.007553676767676767,
      "avg_duration_per_task_ms": 145590.66666666666,
      "avg_output_tokens_per_task": 30019.929292929293,
      "last_run_timestamp": "2026-10-03T16:43:00.637Z"
    },
    {
      "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": 0.7744108749999999,
      "accuracy_std_dev": 0.03582138896119713,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.05256629166666667,
      "avg_duration_per_task_ms": 881480.1900252525,
      "avg_output_tokens_per_task": 23763.922348484848,
      "last_run_timestamp": "2026-10-03T18:36:41.004Z"
    },
    {
      "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": 0.7741623318872016,
      "accuracy_std_dev": 0.020647498917569188,
      "run_count": 7,
      "total_questions": 1383,
      "avg_cost_per_task_usd": 0.05963888141720896,
      "avg_duration_per_task_ms": 613056.7744034707,
      "avg_output_tokens_per_task": 27371.603036876357,
      "last_run_timestamp": "2026-08-29T18:26:43.144Z"
    },
    {
      "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": 0.7738971985294117,
      "accuracy_std_dev": 0.026639499999999983,
      "run_count": 2,
      "total_questions": 272,
      "avg_cost_per_task_usd": 0.009933933823529412,
      "avg_duration_per_task_ms": 1695099,
      "avg_output_tokens_per_task": 35217.91176470588,
      "last_run_timestamp": "2026-10-03T12:01:54.630Z"
    },
    {
      "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": 0.7725740909090909,
      "accuracy_std_dev": 0.024590222227006273,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.0059262144168962345,
      "avg_duration_per_task_ms": 733314.9329660239,
      "avg_output_tokens_per_task": 37199.05142332415,
      "last_run_timestamp": "2026-10-03T19:04:58.692Z"
    },
    {
      "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": 0.7722011249999999,
      "accuracy_std_dev": 0.05713206000013804,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.10372384911616161,
      "avg_duration_per_task_ms": 556419.0063131313,
      "avg_output_tokens_per_task": 31124.368686868685,
      "last_run_timestamp": "2026-10-03T16:59:42.233Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4-sonnet-20250522",
      "model_slug": "anthropic/claude-sonnet-4",
      "model_name": "Anthropic: Claude Sonnet 4",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7714368913043479,
      "accuracy_std_dev": 0.014261246953428234,
      "run_count": 3,
      "total_questions": 552,
      "avg_cost_per_task_usd": 0.23206160869565215,
      "avg_duration_per_task_ms": 164099.60869565216,
      "avg_output_tokens_per_task": 15307.03079710145,
      "last_run_timestamp": "2026-09-01T13:33:46.786Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Mara",
      "accuracy": 0.7713930343283582,
      "accuracy_std_dev": 0.03931269277500872,
      "run_count": 6,
      "total_questions": 670,
      "avg_cost_per_task_usd": 0.012700616417910445,
      "avg_duration_per_task_ms": 53684.10149253732,
      "avg_output_tokens_per_task": 16854.058208955223,
      "last_run_timestamp": "2026-10-03T17:06:37.556Z"
    },
    {
      "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": 0.7711041709956711,
      "accuracy_std_dev": 0.019798063555812727,
      "run_count": 5,
      "total_questions": 924,
      "avg_cost_per_task_usd": 0.020921933982683983,
      "avg_duration_per_task_ms": 243233.94264069264,
      "avg_output_tokens_per_task": 27811.309523809523,
      "last_run_timestamp": "2026-10-03T16:58:21.850Z"
    },
    {
      "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": 0.7707632222222222,
      "accuracy_std_dev": 0.010632614017235168,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.014933182940516276,
      "avg_duration_per_task_ms": 152284.73961840628,
      "avg_output_tokens_per_task": 29838.40684624018,
      "last_run_timestamp": "2026-10-03T17:01:22.843Z"
    },
    {
      "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": 0.7706395810055866,
      "accuracy_std_dev": 0.012812584360697848,
      "run_count": 3,
      "total_questions": 537,
      "avg_cost_per_task_usd": 0.014063508379888267,
      "avg_duration_per_task_ms": 616967.677839851,
      "avg_output_tokens_per_task": 42632.232774674114,
      "last_run_timestamp": "2026-10-03T17:47:01.225Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.4-nano-20260317",
      "model_slug": "openai/gpt-5.4-nano",
      "model_name": "OpenAI: GPT-5.4 Nano",
      "provider_name": "Azure",
      "accuracy": 0.770202,
      "accuracy_std_dev": 0.008625232557251217,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.013833271043771043,
      "avg_duration_per_task_ms": 119359.86616161616,
      "avg_output_tokens_per_task": 10422.487373737375,
      "last_run_timestamp": "2026-10-01T19:20:04.986Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "Parasail",
      "accuracy": 0.7696007142857143,
      "accuracy_std_dev": 0.01982640472929463,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.018328969696969694,
      "avg_duration_per_task_ms": 935948.049062049,
      "avg_output_tokens_per_task": 45779.92929292929,
      "last_run_timestamp": "2026-10-03T16:16:35.793Z"
    },
    {
      "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": 0.7688791428571428,
      "accuracy_std_dev": 0.026826310461552323,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.017481934343434345,
      "avg_duration_per_task_ms": 620493.1940836941,
      "avg_output_tokens_per_task": 43809.34992784993,
      "last_run_timestamp": "2026-10-03T16:23:37.615Z"
    },
    {
      "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": "Phala",
      "accuracy": 0.7688491071428571,
      "accuracy_std_dev": 0.036143499999999995,
      "run_count": 2,
      "total_questions": 168,
      "avg_cost_per_task_usd": 0.013457690476190477,
      "avg_duration_per_task_ms": 239500.45238095237,
      "avg_output_tokens_per_task": 24312.886904761905,
      "last_run_timestamp": "2026-07-18T09:01:08.846Z"
    },
    {
      "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": 0.7688279969604862,
      "accuracy_std_dev": 0.009398481531077219,
      "run_count": 10,
      "total_questions": 1974,
      "avg_cost_per_task_usd": 0.019866108409321178,
      "avg_duration_per_task_ms": 104728.69604863222,
      "avg_output_tokens_per_task": 5941.306484295846,
      "last_run_timestamp": "2026-10-03T13:44:15.277Z"
    },
    {
      "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": 0.7685897384615384,
      "accuracy_std_dev": 0.017564000000000024,
      "run_count": 2,
      "total_questions": 260,
      "avg_cost_per_task_usd": 0.008261646153846154,
      "avg_duration_per_task_ms": 560681.2384615385,
      "avg_output_tokens_per_task": 27419.23076923077,
      "last_run_timestamp": "2026-10-03T16:10:46.858Z"
    },
    {
      "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": 0.768458045751634,
      "accuracy_std_dev": 0.007179978625401643,
      "run_count": 7,
      "total_questions": 1377,
      "avg_cost_per_task_usd": 0.01120065286855483,
      "avg_duration_per_task_ms": 926810.7356572258,
      "avg_output_tokens_per_task": 11933.642701525054,
      "last_run_timestamp": "2026-08-30T01:20:44.266Z"
    },
    {
      "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": 0.7676545740944017,
      "accuracy_std_dev": 0.04018594976550331,
      "run_count": 6,
      "total_questions": 911,
      "avg_cost_per_task_usd": 0.14126989352360045,
      "avg_duration_per_task_ms": 1023387.3852908892,
      "avg_output_tokens_per_task": 40525.553238199784,
      "last_run_timestamp": "2026-08-22T17:30:52.646Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "SambaNova",
      "accuracy": 0.76748545750332,
      "accuracy_std_dev": 0.010027388436645664,
      "run_count": 8,
      "total_questions": 1506,
      "avg_cost_per_task_usd": 0.02429856507304117,
      "avg_duration_per_task_ms": 72708.38977423639,
      "avg_output_tokens_per_task": 25460.314741035858,
      "last_run_timestamp": "2026-10-03T17:11:16.743Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4-sonnet-20250522",
      "model_slug": "anthropic/claude-sonnet-4",
      "model_name": "Anthropic: Claude Sonnet 4",
      "provider_name": "Google",
      "accuracy": 0.7672559999999999,
      "accuracy_std_dev": 0.008469894233892938,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.27037940151515155,
      "avg_duration_per_task_ms": 263755.6675084175,
      "avg_output_tokens_per_task": 17837.74579124579,
      "last_run_timestamp": "2026-09-01T13:23:02.059Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4-sonnet-20250522",
      "model_slug": "anthropic/claude-sonnet-4",
      "model_name": "Anthropic: Claude Sonnet 4",
      "provider_name": "auto-routing",
      "accuracy": 0.7668349999999999,
      "accuracy_std_dev": 0.01935999999999999,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.2670539015151515,
      "avg_duration_per_task_ms": 177095.6616161616,
      "avg_output_tokens_per_task": 17616.689393939392,
      "last_run_timestamp": "2026-09-01T13:19:59.945Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Crusoe",
      "accuracy": 0.76638602247191,
      "accuracy_std_dev": 0.0119351260850762,
      "run_count": 3,
      "total_questions": 356,
      "avg_cost_per_task_usd": 0.002961258426966292,
      "avg_duration_per_task_ms": 69864.89325842696,
      "avg_output_tokens_per_task": 3595.182584269663,
      "last_run_timestamp": "2026-08-15T14:54:48.654Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "stepfun/step-3.7-flash-20260528",
      "model_slug": "stepfun/step-3.7-flash",
      "model_name": "StepFun: Step 3.7 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.765993,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.08016306060606061,
      "avg_duration_per_task_ms": 425226.22222222225,
      "avg_output_tokens_per_task": 69822.58585858585,
      "last_run_timestamp": "2026-07-18T08:29:23.765Z"
    },
    {
      "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": 0.7659136443148689,
      "accuracy_std_dev": 0.08670660767325684,
      "run_count": 7,
      "total_questions": 1372,
      "avg_cost_per_task_usd": 0.04918383236151604,
      "avg_duration_per_task_ms": 1637065.3381924198,
      "avg_output_tokens_per_task": 18412.217201166182,
      "last_run_timestamp": "2026-10-03T18:57:53.982Z"
    },
    {
      "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": 0.7657003043478261,
      "accuracy_std_dev": 0.07140090192078836,
      "run_count": 7,
      "total_questions": 1380,
      "avg_cost_per_task_usd": 0.22668331956521737,
      "avg_duration_per_task_ms": 1257437.0898550726,
      "avg_output_tokens_per_task": 72600.84420289854,
      "last_run_timestamp": "2026-10-03T16:45:32.765Z"
    },
    {
      "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": 0.7652910342857142,
      "accuracy_std_dev": 0.032967884417073254,
      "run_count": 8,
      "total_questions": 1575,
      "avg_cost_per_task_usd": 0.11766670793650795,
      "avg_duration_per_task_ms": 799679.3631746032,
      "avg_output_tokens_per_task": 45940.76698412698,
      "last_run_timestamp": "2026-10-03T16:58:32.422Z"
    },
    {
      "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": 0.7652582599231755,
      "accuracy_std_dev": 0.017480436789719386,
      "run_count": 8,
      "total_questions": 1562,
      "avg_cost_per_task_usd": 0.008710414212548015,
      "avg_duration_per_task_ms": 425255.4129321383,
      "avg_output_tokens_per_task": 26929.24647887324,
      "last_run_timestamp": "2026-10-03T17:35:16.842Z"
    },
    {
      "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": 0.7647939808988764,
      "accuracy_std_dev": 0.09845223455671283,
      "run_count": 10,
      "total_questions": 890,
      "avg_cost_per_task_usd": 0.003642867415730337,
      "avg_duration_per_task_ms": 208810.34157303371,
      "avg_output_tokens_per_task": 15849.9202247191,
      "last_run_timestamp": "2026-10-03T16:12:33.499Z"
    },
    {
      "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": 0.7647061294117647,
      "accuracy_std_dev": 0.11983655888350603,
      "run_count": 5,
      "total_questions": 680,
      "avg_cost_per_task_usd": 0.015231964705882355,
      "avg_duration_per_task_ms": 558302.2632352941,
      "avg_output_tokens_per_task": 21406.642647058823,
      "last_run_timestamp": "2026-09-26T10:43:59.946Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "DeepInfra",
      "accuracy": 0.7642006242726517,
      "accuracy_std_dev": 0.029173475833436126,
      "run_count": 7,
      "total_questions": 1203,
      "avg_cost_per_task_usd": 0.04647462261014131,
      "avg_duration_per_task_ms": 263437.0906068163,
      "avg_output_tokens_per_task": 17506.297589359932,
      "last_run_timestamp": "2026-10-03T09:46:23.404Z"
    },
    {
      "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": 0.7626263333333334,
      "accuracy_std_dev": 0.021573990312410928,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.01664038832772166,
      "avg_duration_per_task_ms": 1086454.638047138,
      "avg_output_tokens_per_task": 48174.952861952865,
      "last_run_timestamp": "2026-10-03T17:13:23.633Z"
    },
    {
      "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": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Venice",
      "accuracy": 0.7622871058394161,
      "accuracy_std_dev": 0.08486398602541866,
      "run_count": 7,
      "total_questions": 1370,
      "avg_cost_per_task_usd": 0.13712116569343066,
      "avg_duration_per_task_ms": 1005369.0547445256,
      "avg_output_tokens_per_task": 37305.686131386865,
      "last_run_timestamp": "2026-10-03T16:47:49.659Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "stepfun/step-3.7-flash-20260528",
      "model_slug": "stepfun/step-3.7-flash",
      "model_name": "StepFun: Step 3.7 Flash",
      "provider_name": "StepFun",
      "accuracy": 0.760943,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.08179280303030302,
      "avg_duration_per_task_ms": 426340.89898989897,
      "avg_output_tokens_per_task": 70994.31818181818,
      "last_run_timestamp": "2026-07-18T08:26:55.806Z"
    },
    {
      "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": 0.760732375,
      "accuracy_std_dev": 0.007014241975037307,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.017655010101010102,
      "avg_duration_per_task_ms": 83495.94002525252,
      "avg_output_tokens_per_task": 29184.116792929293,
      "last_run_timestamp": "2026-10-03T16:55:04.036Z"
    },
    {
      "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": 0.7607321250000001,
      "accuracy_std_dev": 0.04893340233275604,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.011686217171717172,
      "avg_duration_per_task_ms": 765424.8869949495,
      "avg_output_tokens_per_task": 35637.83775252525,
      "last_run_timestamp": "2026-10-03T10:19:08.905Z"
    },
    {
      "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-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "OpenInference",
      "accuracy": 0.7605443265306122,
      "accuracy_std_dev": 0.0343317834601635,
      "run_count": 3,
      "total_questions": 490,
      "avg_cost_per_task_usd": 0.0037374408163265305,
      "avg_duration_per_task_ms": 668153.0571428571,
      "avg_output_tokens_per_task": 14206.583673469388,
      "last_run_timestamp": "2026-08-08T11:42:26.676Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "mistralai/mistral-small-2603",
      "model_slug": "mistralai/mistral-small-2603",
      "model_name": "Mistral: Mistral Small 4",
      "provider_name": "Venice",
      "accuracy": 0.759259,
      "accuracy_std_dev": 0.003367000000000009,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.018210444444444442,
      "avg_duration_per_task_ms": 254627.46717171717,
      "avg_output_tokens_per_task": 24079.401515151516,
      "last_run_timestamp": "2026-09-01T12:59:55.618Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Morph",
      "accuracy": 0.7592001072112318,
      "accuracy_std_dev": 0.11665676901706133,
      "run_count": 8,
      "total_questions": 1567,
      "avg_cost_per_task_usd": 0.17622283024888324,
      "avg_duration_per_task_ms": 1196115.2144224632,
      "avg_output_tokens_per_task": 51216.97766432674,
      "last_run_timestamp": "2026-10-03T18:23:36.655Z"
    },
    {
      "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": 0.7590187142857142,
      "accuracy_std_dev": 0.03157783272538906,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.168772468975469,
      "avg_duration_per_task_ms": 557901.2611832612,
      "avg_output_tokens_per_task": 52553.58585858586,
      "last_run_timestamp": "2026-10-03T17:03:01.289Z"
    },
    {
      "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": 0.7590111668383109,
      "accuracy_std_dev": 0.016611473377157127,
      "run_count": 5,
      "total_questions": 971,
      "avg_cost_per_task_usd": 0.004071515962924819,
      "avg_duration_per_task_ms": 4054.4541709577757,
      "avg_output_tokens_per_task": 2252.576725025747,
      "last_run_timestamp": "2026-08-29T18:58:30.919Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "NextBit",
      "accuracy": 0.758804538083538,
      "accuracy_std_dev": 0.09270215339222468,
      "run_count": 3,
      "total_questions": 407,
      "avg_cost_per_task_usd": 0.03842371253071253,
      "avg_duration_per_task_ms": 1486381.9508599509,
      "avg_output_tokens_per_task": 26977.572481572483,
      "last_run_timestamp": "2026-08-01T23:27:37.975Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-2.5-flash",
      "model_slug": "google/gemini-2.5-flash",
      "model_name": "Google: Gemini 2.5 Flash",
      "provider_name": "Google AI Studio",
      "accuracy": 0.758698,
      "accuracy_std_dev": 0.012009662443216293,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.056970961279461274,
      "avg_duration_per_task_ms": 99388.35016835017,
      "avg_output_tokens_per_task": 22699.07239057239,
      "last_run_timestamp": "2026-10-01T19:11:42.323Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "DekaLLM",
      "accuracy": 0.7575759999999999,
      "accuracy_std_dev": 0.017346370936500422,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.005004900673400673,
      "avg_duration_per_task_ms": 851788.595959596,
      "avg_output_tokens_per_task": 28068.07744107744,
      "last_run_timestamp": "2026-10-03T17:30:01.649Z"
    },
    {
      "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": 0.75757575,
      "accuracy_std_dev": 0.010030692819416823,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.01872427398989899,
      "avg_duration_per_task_ms": 177720.84722222222,
      "avg_output_tokens_per_task": 31014.384469696968,
      "last_run_timestamp": "2026-10-03T16:34:22.379Z"
    },
    {
      "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": 0.7569444375000001,
      "accuracy_std_dev": 0.012152708031385188,
      "run_count": 16,
      "total_questions": 3168,
      "avg_cost_per_task_usd": 0.018743317234848483,
      "avg_duration_per_task_ms": 181792.14551767678,
      "avg_output_tokens_per_task": 31017.58143939394,
      "last_run_timestamp": "2026-10-03T16:20:04.323Z"
    },
    {
      "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": 0.7568806935779815,
      "accuracy_std_dev": 0.10522151155327296,
      "run_count": 3,
      "total_questions": 545,
      "avg_cost_per_task_usd": 0.004833203669724771,
      "avg_duration_per_task_ms": 268953.0513761468,
      "avg_output_tokens_per_task": 24299.52110091743,
      "last_run_timestamp": "2026-10-01T16:23:05.513Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "stepfun/step-3.7-flash-20260528",
      "model_slug": "stepfun/step-3.7-flash",
      "model_name": "StepFun: Step 3.7 Flash",
      "provider_name": "DeepInfra",
      "accuracy": 0.7565450000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 191,
      "avg_cost_per_task_usd": 0.020916916230366494,
      "avg_duration_per_task_ms": 124793.11518324607,
      "avg_output_tokens_per_task": 18115.47120418848,
      "last_run_timestamp": "2026-07-18T08:26:49.504Z"
    },
    {
      "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": 0.7550891068702291,
      "accuracy_std_dev": 0.013103736509103058,
      "run_count": 8,
      "total_questions": 1572,
      "avg_cost_per_task_usd": 0.01650294465648855,
      "avg_duration_per_task_ms": 296983.9370229008,
      "avg_output_tokens_per_task": 28161.054707379135,
      "last_run_timestamp": "2026-10-03T17:05:00.424Z"
    },
    {
      "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": 0.7548401249999999,
      "accuracy_std_dev": 0.04325769864844148,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.05385063825757575,
      "avg_duration_per_task_ms": 956346.2373737374,
      "avg_output_tokens_per_task": 27778.031565656565,
      "last_run_timestamp": "2026-10-03T17:12:39.572Z"
    },
    {
      "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": 0.7545347341659232,
      "accuracy_std_dev": 0.02761225844139473,
      "run_count": 8,
      "total_questions": 1121,
      "avg_cost_per_task_usd": 0.05582876717216771,
      "avg_duration_per_task_ms": 404528.79750223015,
      "avg_output_tokens_per_task": 16141.22033898305,
      "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": "Alibaba",
      "accuracy": 0.7542087500000001,
      "accuracy_std_dev": 0.008073547791243961,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.022397893939393936,
      "avg_duration_per_task_ms": 97901.73674242424,
      "avg_output_tokens_per_task": 14784.859217171717,
      "last_run_timestamp": "2026-10-03T16:56:57.458Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "auto-routing",
      "accuracy": 0.7539683918918919,
      "accuracy_std_dev": 0.07441117537362499,
      "run_count": 8,
      "total_questions": 1554,
      "avg_cost_per_task_usd": 0.131239731016731,
      "avg_duration_per_task_ms": 1027517.648005148,
      "avg_output_tokens_per_task": 56777.77670527671,
      "last_run_timestamp": "2026-10-03T16:56:39.926Z"
    },
    {
      "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": 0.753934655143339,
      "accuracy_std_dev": 0.024433692146069302,
      "run_count": 6,
      "total_questions": 1186,
      "avg_cost_per_task_usd": 0.020737796795952786,
      "avg_duration_per_task_ms": 145659.42242833052,
      "avg_output_tokens_per_task": 11989.828836424958,
      "last_run_timestamp": "2026-08-22T17:12:08.919Z"
    },
    {
      "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": 0.7539281666666667,
      "accuracy_std_dev": 0.05140798989267677,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.0019029528619528618,
      "avg_duration_per_task_ms": 665715.8501683502,
      "avg_output_tokens_per_task": 12898.684343434343,
      "last_run_timestamp": "2026-08-22T17:47:50.063Z"
    },
    {
      "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": 0.7537277142857143,
      "accuracy_std_dev": 0.11897912112721323,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.13752438961038962,
      "avg_duration_per_task_ms": 1748776.3001443,
      "avg_output_tokens_per_task": 48295.48701298701,
      "last_run_timestamp": "2026-10-03T17:21:47.311Z"
    },
    {
      "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/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "Alibaba",
      "accuracy": 0.7504208749999999,
      "accuracy_std_dev": 0.14227753697829243,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.030610907196969697,
      "avg_duration_per_task_ms": 141563.8926767677,
      "avg_output_tokens_per_task": 23450.24873737374,
      "last_run_timestamp": "2026-10-03T16:48:25.700Z"
    },
    {
      "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": 0.7499064444444444,
      "accuracy_std_dev": 0.008289683281861894,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.01351321212121212,
      "avg_duration_per_task_ms": 98095.24747474748,
      "avg_output_tokens_per_task": 14852.885521885522,
      "last_run_timestamp": "2026-10-03T16:54:39.591Z"
    },
    {
      "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": 0.749158,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.016407893939393937,
      "avg_duration_per_task_ms": 1120280.1313131314,
      "avg_output_tokens_per_task": 58254.84848484849,
      "last_run_timestamp": "2026-10-03T11:59:33.633Z"
    },
    {
      "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": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "Groq",
      "accuracy": 0.7477631238816241,
      "accuracy_std_dev": 0.09701744760278624,
      "run_count": 8,
      "total_questions": 1453,
      "avg_cost_per_task_usd": 0.03234037852718513,
      "avg_duration_per_task_ms": 147355.79421885754,
      "avg_output_tokens_per_task": 28069.35856847901,
      "last_run_timestamp": "2026-10-03T17:57:43.271Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "stepfun/step-3.7-flash-20260528",
      "model_slug": "stepfun/step-3.7-flash",
      "model_name": "StepFun: Step 3.7 Flash",
      "provider_name": "Novita",
      "accuracy": 0.7474749999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.08105034848484848,
      "avg_duration_per_task_ms": 422601,
      "avg_output_tokens_per_task": 70367.68181818182,
      "last_run_timestamp": "2026-07-18T08:27:39.521Z"
    },
    {
      "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": 0.7474749999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.016666974747474746,
      "avg_duration_per_task_ms": 1024552.0252525252,
      "avg_output_tokens_per_task": 59525.38383838384,
      "last_run_timestamp": "2026-10-03T09:35:36.415Z"
    },
    {
      "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": 0.7474560693802035,
      "accuracy_std_dev": 0.030553086408774483,
      "run_count": 7,
      "total_questions": 1081,
      "avg_cost_per_task_usd": 0.007020049028677151,
      "avg_duration_per_task_ms": 981607.5957446808,
      "avg_output_tokens_per_task": 18100.610545790933,
      "last_run_timestamp": "2026-10-04T00:24:24.477Z"
    },
    {
      "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": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "GMICloud",
      "accuracy": 0.7465211040462427,
      "accuracy_std_dev": 0.07408730878961323,
      "run_count": 8,
      "total_questions": 1557,
      "avg_cost_per_task_usd": 0.20187897559409118,
      "avg_duration_per_task_ms": 872112.8529222865,
      "avg_output_tokens_per_task": 63704.67630057804,
      "last_run_timestamp": "2026-10-03T16:12:13.512Z"
    },
    {
      "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": 0.7461931421319796,
      "accuracy_std_dev": 0.031200414988782142,
      "run_count": 6,
      "total_questions": 1182,
      "avg_cost_per_task_usd": 0.004617404399323182,
      "avg_duration_per_task_ms": 792703.1945854484,
      "avg_output_tokens_per_task": 27906.86040609137,
      "last_run_timestamp": "2026-10-03T17:02:01.121Z"
    },
    {
      "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": 0.7457915,
      "accuracy_std_dev": 0.008417500000000022,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.014859042929292928,
      "avg_duration_per_task_ms": 987926.6818181818,
      "avg_output_tokens_per_task": 52717.79797979798,
      "last_run_timestamp": "2026-10-03T10:50:00.676Z"
    },
    {
      "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": 0.7457914444444445,
      "accuracy_std_dev": 0.00862087100666636,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.014793652076318742,
      "avg_duration_per_task_ms": 108312.31313131313,
      "avg_output_tokens_per_task": 14346.404601571268,
      "last_run_timestamp": "2026-10-03T16:43:33.673Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-thinking-2507",
      "model_slug": "qwen/qwen3-235b-a22b-thinking-2507",
      "model_name": "Qwen: Qwen3 235B A22B Thinking 2507",
      "provider_name": "Venice",
      "accuracy": 0.7455108333333333,
      "accuracy_std_dev": 0.020304992673532824,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.08322730471380473,
      "avg_duration_per_task_ms": 907303.9099326599,
      "avg_output_tokens_per_task": 23781.781144781144,
      "last_run_timestamp": "2026-08-22T17:38:59.682Z"
    },
    {
      "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": 0.7451755736842105,
      "accuracy_std_dev": 0.011659500000000045,
      "run_count": 2,
      "total_questions": 380,
      "avg_cost_per_task_usd": 0.014937544736842105,
      "avg_duration_per_task_ms": 179152.7552631579,
      "avg_output_tokens_per_task": 54052.70263157895,
      "last_run_timestamp": "2026-10-03T09:41:31.910Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2-0905",
      "model_slug": "moonshotai/kimi-k2-0905",
      "model_name": "MoonshotAI: Kimi K2 0905",
      "provider_name": "auto-routing",
      "accuracy": 0.744669,
      "accuracy_std_dev": 0.014698031999783739,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.01758458585858586,
      "avg_duration_per_task_ms": 101220.53872053872,
      "avg_output_tokens_per_task": 6276.715488215488,
      "last_run_timestamp": "2026-08-01T17:45:27.258Z"
    },
    {
      "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": 0.7446127,
      "accuracy_std_dev": 0.00984666853357015,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.006898398484848485,
      "avg_duration_per_task_ms": 165372.10808080807,
      "avg_output_tokens_per_task": 8017.161616161616,
      "last_run_timestamp": "2026-10-03T16:58:32.590Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-coder-next-2025-02-03",
      "model_slug": "qwen/qwen3-coder-next",
      "model_name": "Qwen: Qwen3 Coder Next",
      "provider_name": "Novita",
      "accuracy": 0.743897375,
      "accuracy_std_dev": 0.010561589427466657,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.022802391414141414,
      "avg_duration_per_task_ms": 99703.82891414141,
      "avg_output_tokens_per_task": 15093.268308080807,
      "last_run_timestamp": "2026-10-03T16:58:08.462Z"
    },
    {
      "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": 0.7436867500000001,
      "accuracy_std_dev": 0.05045411551288456,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.01589729419191919,
      "avg_duration_per_task_ms": 476456.76136363635,
      "avg_output_tokens_per_task": 15487.320707070707,
      "last_run_timestamp": "2026-10-04T00:29:23.413Z"
    },
    {
      "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": 0.7426978294314383,
      "accuracy_std_dev": 0.00453846587516089,
      "run_count": 8,
      "total_questions": 1495,
      "avg_cost_per_task_usd": 0.008910557190635451,
      "avg_duration_per_task_ms": 136917.0508361204,
      "avg_output_tokens_per_task": 11277.847491638797,
      "last_run_timestamp": "2026-10-03T17:03:40.665Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Nebius",
      "accuracy": 0.74242425,
      "accuracy_std_dev": 0.006838470877140611,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.005376655303030303,
      "avg_duration_per_task_ms": 143698.49747474748,
      "avg_output_tokens_per_task": 8690.27840909091,
      "last_run_timestamp": "2026-10-03T17:05:50.656Z"
    },
    {
      "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": 0.7422268568329717,
      "accuracy_std_dev": 0.051186502535695226,
      "run_count": 7,
      "total_questions": 1383,
      "avg_cost_per_task_usd": 0.006795763557483733,
      "avg_duration_per_task_ms": 1042317.6109906002,
      "avg_output_tokens_per_task": 18262.2856109906,
      "last_run_timestamp": "2026-08-29T20:11:32.700Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2-0905",
      "model_slug": "moonshotai/kimi-k2-0905",
      "model_name": "MoonshotAI: Kimi K2 0905",
      "provider_name": "Groq",
      "accuracy": 0.7415824999999999,
      "accuracy_std_dev": 0.005892500000000023,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.019016893939393938,
      "avg_duration_per_task_ms": 33403.05808080808,
      "avg_output_tokens_per_task": 6109.055555555556,
      "last_run_timestamp": "2026-08-01T17:45:49.025Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Alibaba",
      "accuracy": 0.7407407777777778,
      "accuracy_std_dev": 0.010468392689719497,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.004919860830527497,
      "avg_duration_per_task_ms": 109862.1470258137,
      "avg_output_tokens_per_task": 8024.6543209876545,
      "last_run_timestamp": "2026-10-03T16:59:36.480Z"
    },
    {
      "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": 0.7406781050847459,
      "accuracy_std_dev": 0.10340659429589846,
      "run_count": 9,
      "total_questions": 1770,
      "avg_cost_per_task_usd": 0.006412678531073445,
      "avg_duration_per_task_ms": 534746.5333333333,
      "avg_output_tokens_per_task": 27485.332768361583,
      "last_run_timestamp": "2026-10-03T10:03:14.593Z"
    },
    {
      "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": "Friendli",
      "accuracy": 0.7395381428571429,
      "accuracy_std_dev": 0.07707797586939116,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.0066781443001443,
      "avg_duration_per_task_ms": 95940.86147186147,
      "avg_output_tokens_per_task": 16425.329725829724,
      "last_run_timestamp": "2026-10-03T16:54:56.858Z"
    },
    {
      "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": 0.7392771671732522,
      "accuracy_std_dev": 0.07975374971554629,
      "run_count": 10,
      "total_questions": 1974,
      "avg_cost_per_task_usd": 0.005161332320162108,
      "avg_duration_per_task_ms": 757299.1337386018,
      "avg_output_tokens_per_task": 34332.25075987842,
      "last_run_timestamp": "2026-09-26T18:04:35.621Z"
    },
    {
      "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": 0.7388467499999999,
      "accuracy_std_dev": 0.07747406520692911,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.01745328409090909,
      "avg_duration_per_task_ms": 2275674.0732323234,
      "avg_output_tokens_per_task": 17131.02398989899,
      "last_run_timestamp": "2026-08-22T18:24:57.217Z"
    },
    {
      "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": 0.7387387252252252,
      "accuracy_std_dev": 0.018499983466208795,
      "run_count": 4,
      "total_questions": 444,
      "avg_cost_per_task_usd": 0.01493145045045045,
      "avg_duration_per_task_ms": 79529.70945945945,
      "avg_output_tokens_per_task": 8581.265765765766,
      "last_run_timestamp": "2026-10-04T00:00:36.943Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-2.5-flash",
      "model_slug": "google/gemini-2.5-flash",
      "model_name": "Google: Gemini 2.5 Flash",
      "provider_name": "Google",
      "accuracy": 0.738309,
      "accuracy_std_dev": 0.008136215650479735,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.059655349046015715,
      "avg_duration_per_task_ms": 125601.6632996633,
      "avg_output_tokens_per_task": 23853.966891133558,
      "last_run_timestamp": "2026-10-01T19:15:31.968Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Venice",
      "accuracy": 0.7379703298337709,
      "accuracy_std_dev": 0.07144210650532616,
      "run_count": 7,
      "total_questions": 1143,
      "avg_cost_per_task_usd": 0.03304302712160979,
      "avg_duration_per_task_ms": 696117.6115485565,
      "avg_output_tokens_per_task": 35929.485564304465,
      "last_run_timestamp": "2026-08-29T18:51:07.939Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "stepfun/step-3.7-flash-20260528",
      "model_slug": "stepfun/step-3.7-flash",
      "model_name": "StepFun: Step 3.7 Flash",
      "provider_name": "Ambient",
      "accuracy": 0.736111,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 168,
      "avg_cost_per_task_usd": 0.05987852380952381,
      "avg_duration_per_task_ms": 1235915.3511904762,
      "avg_output_tokens_per_task": 62102.07738095238,
      "last_run_timestamp": "2026-07-18T08:29:05.636Z"
    },
    {
      "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": 0.7345221716196138,
      "accuracy_std_dev": 0.016459468688265724,
      "run_count": 8,
      "total_questions": 1346,
      "avg_cost_per_task_usd": 0.006231184249628529,
      "avg_duration_per_task_ms": 219949.4487369985,
      "avg_output_tokens_per_task": 7708.209509658247,
      "last_run_timestamp": "2026-10-03T16:52:01.284Z"
    },
    {
      "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": 0.7344877142857142,
      "accuracy_std_dev": 0.06178190798217846,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.011764064935064936,
      "avg_duration_per_task_ms": 95375.22294372294,
      "avg_output_tokens_per_task": 5009.694083694084,
      "last_run_timestamp": "2026-10-03T10:46:59.690Z"
    },
    {
      "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": 0.7316240615384616,
      "accuracy_std_dev": 0.012678999999999996,
      "run_count": 2,
      "total_questions": 390,
      "avg_cost_per_task_usd": 0.014848610256410256,
      "avg_duration_per_task_ms": 978212.5923076923,
      "avg_output_tokens_per_task": 52658.97692307692,
      "last_run_timestamp": "2026-10-03T10:14:05.783Z"
    },
    {
      "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": 0.7316081564544914,
      "accuracy_std_dev": 0.016316209027987538,
      "run_count": 17,
      "total_questions": 3362,
      "avg_cost_per_task_usd": 0.006837803985722784,
      "avg_duration_per_task_ms": 102812.45925044616,
      "avg_output_tokens_per_task": 7135.6109458655565,
      "last_run_timestamp": "2026-10-03T16:40:17.848Z"
    },
    {
      "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": 0.7313943448275863,
      "accuracy_std_dev": 0.013515629737139094,
      "run_count": 29,
      "total_questions": 5742,
      "avg_cost_per_task_usd": 0.2044269764890282,
      "avg_duration_per_task_ms": 243437.65552072448,
      "avg_output_tokens_per_task": 38184.69731800766,
      "last_run_timestamp": "2026-10-03T09:50:25.399Z"
    },
    {
      "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": 0.7306398571428572,
      "accuracy_std_dev": 0.05292415038100302,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.015631727272727273,
      "avg_duration_per_task_ms": 854213.9992784993,
      "avg_output_tokens_per_task": 42235.36435786436,
      "last_run_timestamp": "2026-10-03T18:22:40.169Z"
    },
    {
      "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": 0.7297978999999999,
      "accuracy_std_dev": 0.016326424363282986,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.18559751616161616,
      "avg_duration_per_task_ms": 291435.0106060606,
      "avg_output_tokens_per_task": 36955.371717171714,
      "last_run_timestamp": "2026-10-03T09:40:02.418Z"
    },
    {
      "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": 0.7286196,
      "accuracy_std_dev": 0.010123398502479267,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.19021829797979797,
      "avg_duration_per_task_ms": 241955.48484848486,
      "avg_output_tokens_per_task": 37852.89646464647,
      "last_run_timestamp": "2026-10-03T09:25:15.662Z"
    },
    {
      "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": 0.727903875,
      "accuracy_std_dev": 0.014553723049768914,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.0046047815656565655,
      "avg_duration_per_task_ms": 179282.44696969696,
      "avg_output_tokens_per_task": 7813.462752525253,
      "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": 0.7276937500000001,
      "accuracy_std_dev": 0.01898605081862735,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.005756878787878787,
      "avg_duration_per_task_ms": 224182.07007575757,
      "avg_output_tokens_per_task": 7857.475378787879,
      "last_run_timestamp": "2026-10-03T17:25:00.727Z"
    },
    {
      "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": "Phala",
      "accuracy": 0.7270113724137931,
      "accuracy_std_dev": 0.06592199999999998,
      "run_count": 2,
      "total_questions": 290,
      "avg_cost_per_task_usd": 0.06701382413793103,
      "avg_duration_per_task_ms": 911772.7965517241,
      "avg_output_tokens_per_task": 29657.59655172414,
      "last_run_timestamp": "2026-07-18T16:15:33.587Z"
    },
    {
      "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": "WandB",
      "accuracy": 0.7264309999999999,
      "accuracy_std_dev": 0.01094299999999998,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.003979111111111111,
      "avg_duration_per_task_ms": 697805.6414141414,
      "avg_output_tokens_per_task": 30154.29292929293,
      "last_run_timestamp": "2026-07-18T16:28:34.215Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-2.5-flash",
      "model_slug": "google/gemini-2.5-flash",
      "model_name": "Google: Gemini 2.5 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.7261503333333333,
      "accuracy_std_dev": 0.004418546090690373,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.059926168350168355,
      "avg_duration_per_task_ms": 123042.70033670033,
      "avg_output_tokens_per_task": 23940.483164983165,
      "last_run_timestamp": "2026-10-01T19:12:41.089Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.7260702857142858,
      "accuracy_std_dev": 0.011156764145758455,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.006142652958152958,
      "avg_duration_per_task_ms": 382505.44805194804,
      "avg_output_tokens_per_task": 8147.619769119769,
      "last_run_timestamp": "2026-10-03T17:45:28.847Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "DeepInfra",
      "accuracy": 0.7260101250000001,
      "accuracy_std_dev": 0.008051806325873379,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.004004231691919192,
      "avg_duration_per_task_ms": 572174.6193181818,
      "avg_output_tokens_per_task": 8043.021464646465,
      "last_run_timestamp": "2026-10-03T17:33:20.142Z"
    },
    {
      "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": 0.7255892500000001,
      "accuracy_std_dev": 0.028594658337484976,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.008037106691919193,
      "avg_duration_per_task_ms": 265191.6982323232,
      "avg_output_tokens_per_task": 18033.603535353537,
      "last_run_timestamp": "2026-10-03T17:08:10.486Z"
    },
    {
      "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": 0.7240554670050762,
      "accuracy_std_dev": 0.06016878242540359,
      "run_count": 6,
      "total_questions": 1182,
      "avg_cost_per_task_usd": 0.014961366328257193,
      "avg_duration_per_task_ms": 1202663.8900169204,
      "avg_output_tokens_per_task": 38040.99492385787,
      "last_run_timestamp": "2026-10-03T17:20:21.640Z"
    },
    {
      "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": 0.7234887554404145,
      "accuracy_std_dev": 0.009381891562472875,
      "run_count": 10,
      "total_questions": 1930,
      "avg_cost_per_task_usd": 0.18696787461139894,
      "avg_duration_per_task_ms": 349087.42279792746,
      "avg_output_tokens_per_task": 37203.45544041451,
      "last_run_timestamp": "2026-10-03T09:27:16.866Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Krea",
      "accuracy": 0.7234044255319149,
      "accuracy_std_dev": 0.1917994784299478,
      "run_count": 3,
      "total_questions": 517,
      "avg_cost_per_task_usd": 0.033565591876208896,
      "avg_duration_per_task_ms": 242772.43907156674,
      "avg_output_tokens_per_task": 37102.30560928433,
      "last_run_timestamp": "2026-09-29T20:57:20.673Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Novita",
      "accuracy": 0.7228535,
      "accuracy_std_dev": 0.01924905807435781,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.0067263503787878785,
      "avg_duration_per_task_ms": 349070.6534090909,
      "avg_output_tokens_per_task": 26711.420454545456,
      "last_run_timestamp": "2026-10-03T17:40:21.277Z"
    },
    {
      "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": 0.7225028666666667,
      "accuracy_std_dev": 0.012643758973589382,
      "run_count": 30,
      "total_questions": 5940,
      "avg_cost_per_task_usd": 0.1973207321548822,
      "avg_duration_per_task_ms": 263893.9067340067,
      "avg_output_tokens_per_task": 36800.280303030304,
      "last_run_timestamp": "2026-10-03T09:27:45.066Z"
    },
    {
      "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": 0.7222222500000001,
      "accuracy_std_dev": 0.0717218919764914,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.02037232512626263,
      "avg_duration_per_task_ms": 1184721.3592171718,
      "avg_output_tokens_per_task": 14249.777777777777,
      "last_run_timestamp": "2026-10-03T16:57:23.086Z"
    },
    {
      "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": 0.7221168064516129,
      "accuracy_std_dev": 0.11047255085087879,
      "run_count": 8,
      "total_questions": 1581,
      "avg_cost_per_task_usd": 0.0044673187855787475,
      "avg_duration_per_task_ms": 283742.9405439595,
      "avg_output_tokens_per_task": 10988.973434535104,
      "last_run_timestamp": "2026-10-04T00:01:24.489Z"
    },
    {
      "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": "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": 0.7205388571428571,
      "accuracy_std_dev": 0.045769541832435034,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.006908489898989899,
      "avg_duration_per_task_ms": 128386.26262626263,
      "avg_output_tokens_per_task": 7669.774891774891,
      "last_run_timestamp": "2026-08-29T18:01:46.940Z"
    },
    {
      "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": 0.71923592494929,
      "accuracy_std_dev": 0.031713523289599956,
      "run_count": 5,
      "total_questions": 986,
      "avg_cost_per_task_usd": 0.01125353448275862,
      "avg_duration_per_task_ms": 528851.3123732251,
      "avg_output_tokens_per_task": 56006.78093306288,
      "last_run_timestamp": "2026-10-03T09:18:14.437Z"
    },
    {
      "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": 0.7184345000000002,
      "accuracy_std_dev": 0.00575544305418794,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.006676434343434343,
      "avg_duration_per_task_ms": 1198975.5896464647,
      "avg_output_tokens_per_task": 54074.54671717172,
      "last_run_timestamp": "2026-09-26T17:35:27.801Z"
    },
    {
      "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": 0.7183463229974161,
      "accuracy_std_dev": 0.029264511720208542,
      "run_count": 6,
      "total_questions": 1161,
      "avg_cost_per_task_usd": 0.17074779758828595,
      "avg_duration_per_task_ms": 1524939.6339362618,
      "avg_output_tokens_per_task": 41593.74849267872,
      "last_run_timestamp": "2026-08-22T16:49:30.177Z"
    },
    {
      "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": 0.71485675,
      "accuracy_std_dev": 0.013413663339576564,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.008089982323232324,
      "avg_duration_per_task_ms": 76524.31755050505,
      "avg_output_tokens_per_task": 10274.573232323231,
      "last_run_timestamp": "2026-10-03T16:40:12.993Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Darkbloom",
      "accuracy": 0.713213,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 111,
      "avg_cost_per_task_usd": 0.007665369369369369,
      "avg_duration_per_task_ms": 1401051.936936937,
      "avg_output_tokens_per_task": 27183.8018018018,
      "last_run_timestamp": "2026-10-03T09:34:00.571Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Friendli",
      "accuracy": 0.712963,
      "accuracy_std_dev": 0.007926363478922754,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.005945053872053872,
      "avg_duration_per_task_ms": 274691.60101010103,
      "avg_output_tokens_per_task": 7230.755892255892,
      "last_run_timestamp": "2026-08-01T17:52:16.799Z"
    },
    {
      "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": 0.7110947790868927,
      "accuracy_std_dev": 0.09708503564674945,
      "run_count": 7,
      "total_questions": 1358,
      "avg_cost_per_task_usd": 0.009436205449189985,
      "avg_duration_per_task_ms": 507448.52209131076,
      "avg_output_tokens_per_task": 32737.85714285714,
      "last_run_timestamp": "2026-10-03T17:30:20.154Z"
    },
    {
      "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": 0.7108066232876712,
      "accuracy_std_dev": 0.016923430604933478,
      "run_count": 5,
      "total_questions": 876,
      "avg_cost_per_task_usd": 0.06916225000000001,
      "avg_duration_per_task_ms": 1230338.2773972603,
      "avg_output_tokens_per_task": 23160.62100456621,
      "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-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": 0.7107282190293742,
      "accuracy_std_dev": 0.01626806742325285,
      "run_count": 8,
      "total_questions": 1566,
      "avg_cost_per_task_usd": 0.010758521711366539,
      "avg_duration_per_task_ms": 103410.69029374202,
      "avg_output_tokens_per_task": 9911.577266922095,
      "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": "Novita",
      "accuracy": 0.7106482500000001,
      "accuracy_std_dev": 0.007960486601803957,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.015934953282828283,
      "avg_duration_per_task_ms": 96234.6571969697,
      "avg_output_tokens_per_task": 10542.059343434343,
      "last_run_timestamp": "2026-10-03T16:46:22.205Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "GMICloud",
      "accuracy": 0.7094277999999999,
      "accuracy_std_dev": 0.0745134320251054,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.007971388888888889,
      "avg_duration_per_task_ms": 178465.76363636364,
      "avg_output_tokens_per_task": 7431.1,
      "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-vl-235b-a22b-instruct",
      "model_slug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_name": "Qwen: Qwen3 VL 235B A22B Instruct",
      "provider_name": "Novita",
      "accuracy": 0.709385625,
      "accuracy_std_dev": 0.009446333705961032,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.00881291351010101,
      "avg_duration_per_task_ms": 137334.60416666666,
      "avg_output_tokens_per_task": 5712.794191919192,
      "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-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": 0.709175125,
      "accuracy_std_dev": 0.01063905747044237,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.011125260732323233,
      "avg_duration_per_task_ms": 96740.25820707071,
      "avg_output_tokens_per_task": 10516.064393939394,
      "last_run_timestamp": "2026-10-03T16:59:33.949Z"
    },
    {
      "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": 0.70883299486008,
      "accuracy_std_dev": 0.036241367349453366,
      "run_count": 9,
      "total_questions": 1751,
      "avg_cost_per_task_usd": 0.010877087378640777,
      "avg_duration_per_task_ms": 281941.7778412336,
      "avg_output_tokens_per_task": 47824.729868646486,
      "last_run_timestamp": "2026-10-03T09:36:23.802Z"
    },
    {
      "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": 0.7070005146036162,
      "accuracy_std_dev": 0.06882289073774799,
      "run_count": 4,
      "total_questions": 719,
      "avg_cost_per_task_usd": 0.07411282058414463,
      "avg_duration_per_task_ms": 612487.7343532684,
      "avg_output_tokens_per_task": 29410.726008344922,
      "last_run_timestamp": "2026-08-15T09:29:41.209Z"
    },
    {
      "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": 0.7062016019933554,
      "accuracy_std_dev": 0.2586002901946361,
      "run_count": 8,
      "total_questions": 1505,
      "avg_cost_per_task_usd": 0.025144299667774088,
      "avg_duration_per_task_ms": 338978.96146179404,
      "avg_output_tokens_per_task": 15167.856478405316,
      "last_run_timestamp": "2026-10-04T00:21:14.629Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5-nano-2025-08-07",
      "model_slug": "openai/gpt-5-nano",
      "model_name": "OpenAI: GPT-5 Nano",
      "provider_name": "auto-routing",
      "accuracy": 0.7045455,
      "accuracy_std_dev": 0.004208500000000004,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.01336774494949495,
      "avg_duration_per_task_ms": 231791.09848484848,
      "avg_output_tokens_per_task": 33323.15909090909,
      "last_run_timestamp": "2026-10-01T19:27:22.032Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_slug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_name": "Qwen: Qwen3 VL 235B A22B Instruct",
      "provider_name": "Alibaba",
      "accuracy": 0.704545375,
      "accuracy_std_dev": 0.008501245907769926,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.006021614898989899,
      "avg_duration_per_task_ms": 108658.90909090909,
      "avg_output_tokens_per_task": 5588.438762626262,
      "last_run_timestamp": "2026-10-03T17:02:24.058Z"
    },
    {
      "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": 0.7039041742125984,
      "accuracy_std_dev": 0.06488902055802347,
      "run_count": 7,
      "total_questions": 1016,
      "avg_cost_per_task_usd": 0.03703799311023622,
      "avg_duration_per_task_ms": 230051.07480314962,
      "avg_output_tokens_per_task": 14035.937992125984,
      "last_run_timestamp": "2026-08-29T19:20:11.540Z"
    },
    {
      "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": 0.7037214026717559,
      "accuracy_std_dev": 0.050640957870974806,
      "run_count": 16,
      "total_questions": 3144,
      "avg_cost_per_task_usd": 0.008582292620865142,
      "avg_duration_per_task_ms": 446340.70642493636,
      "avg_output_tokens_per_task": 25644.754770992367,
      "last_run_timestamp": "2026-10-03T17:44:46.084Z"
    },
    {
      "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": 0.7034933750000001,
      "accuracy_std_dev": 0.013610377795431516,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.011181943813131312,
      "avg_duration_per_task_ms": 198847.17866161617,
      "avg_output_tokens_per_task": 5822.409722222223,
      "last_run_timestamp": "2026-10-03T16:59:57.847Z"
    },
    {
      "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": 0.70328275,
      "accuracy_std_dev": 0.02153875190644758,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.0019388005050505052,
      "avg_duration_per_task_ms": 636092.8409090909,
      "avg_output_tokens_per_task": 14000.35101010101,
      "last_run_timestamp": "2026-09-05T08:28:07.290Z"
    },
    {
      "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": 0.7024411666666667,
      "accuracy_std_dev": 0.14218853509503576,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.013173934343434343,
      "avg_duration_per_task_ms": 786438.8409090909,
      "avg_output_tokens_per_task": 27569.127104377105,
      "last_run_timestamp": "2026-09-22T02:20:03.900Z"
    },
    {
      "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": "Wafer",
      "accuracy": 0.70202,
      "accuracy_std_dev": 0.010101000000000027,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.012498363636363634,
      "avg_duration_per_task_ms": 411305.4646464646,
      "avg_output_tokens_per_task": 31035.28787878788,
      "last_run_timestamp": "2026-07-18T08:04:42.524Z"
    },
    {
      "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": 0.7015995,
      "accuracy_std_dev": 0.023308248459504657,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.011333077020202021,
      "avg_duration_per_task_ms": 63982.55744949495,
      "avg_output_tokens_per_task": 9378.460858585859,
      "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": 0.7013888749999999,
      "accuracy_std_dev": 0.009189646775549915,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.010544860479797979,
      "avg_duration_per_task_ms": 145490.4551767677,
      "avg_output_tokens_per_task": 9519.845328282829,
      "last_run_timestamp": "2026-10-03T16:56:42.844Z"
    },
    {
      "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": 0.6994219060693642,
      "accuracy_std_dev": 0.028501362281769377,
      "run_count": 6,
      "total_questions": 692,
      "avg_cost_per_task_usd": 0.010150841040462426,
      "avg_duration_per_task_ms": 46976.1950867052,
      "avg_output_tokens_per_task": 6570.052023121388,
      "last_run_timestamp": "2026-10-04T00:12:48.620Z"
    },
    {
      "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": 0.6960783363171356,
      "accuracy_std_dev": 0.1309593611346665,
      "run_count": 5,
      "total_questions": 782,
      "avg_cost_per_task_usd": 0.0047808593350383635,
      "avg_duration_per_task_ms": 587688.8618925831,
      "avg_output_tokens_per_task": 25488.57928388747,
      "last_run_timestamp": "2026-10-03T09:54:28.477Z"
    },
    {
      "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": 0.6956243088803089,
      "accuracy_std_dev": 0.0895741115626294,
      "run_count": 8,
      "total_questions": 1554,
      "avg_cost_per_task_usd": 0.12983634813384812,
      "avg_duration_per_task_ms": 1082525.285070785,
      "avg_output_tokens_per_task": 70424.97168597169,
      "last_run_timestamp": "2026-10-03T16:56:56.694Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3-nano-30b-a3b",
      "model_slug": "nvidia/nemotron-3-nano-30b-a3b",
      "model_name": "NVIDIA: Nemotron 3 Nano 30B A3B",
      "provider_name": "DeepInfra",
      "accuracy": 0.6948820649606299,
      "accuracy_std_dev": 0.027729042227588305,
      "run_count": 8,
      "total_questions": 1524,
      "avg_cost_per_task_usd": 0.015564356299212597,
      "avg_duration_per_task_ms": 765563.3963254594,
      "avg_output_tokens_per_task": 77634.76837270342,
      "last_run_timestamp": "2026-09-26T16:34:21.219Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_slug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_name": "Qwen: Qwen3 VL 235B A22B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.694865375,
      "accuracy_std_dev": 0.01804853225291118,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.008836173611111112,
      "avg_duration_per_task_ms": 277270.93055555556,
      "avg_output_tokens_per_task": 6352.661616161616,
      "last_run_timestamp": "2026-10-03T17:02:48.284Z"
    },
    {
      "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": 0.6931105699373695,
      "accuracy_std_dev": 0.08497842021164353,
      "run_count": 6,
      "total_questions": 958,
      "avg_cost_per_task_usd": 0.0069657160751565765,
      "avg_duration_per_task_ms": 481855.8267223382,
      "avg_output_tokens_per_task": 32574.692066805845,
      "last_run_timestamp": "2026-08-29T12:30:09.617Z"
    },
    {
      "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": 0.6929713750000001,
      "accuracy_std_dev": 0.008745132416628961,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.005507261994949495,
      "avg_duration_per_task_ms": 385008.58585858584,
      "avg_output_tokens_per_task": 6073.715277777777,
      "last_run_timestamp": "2026-10-03T17:41:44.575Z"
    },
    {
      "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": 0.6926927837837837,
      "accuracy_std_dev": 0.02593753677273029,
      "run_count": 7,
      "total_questions": 1332,
      "avg_cost_per_task_usd": 0.009898264264264264,
      "avg_duration_per_task_ms": 376272.99399399396,
      "avg_output_tokens_per_task": 45270.37087087087,
      "last_run_timestamp": "2026-10-03T09:53:13.324Z"
    },
    {
      "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": "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": 0.6896746666666668,
      "accuracy_std_dev": 0.04960523518294776,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.04329939057239058,
      "avg_duration_per_task_ms": 1345214.5555555555,
      "avg_output_tokens_per_task": 31899.07744107744,
      "last_run_timestamp": "2026-08-22T16:20:35.762Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Io Net",
      "accuracy": 0.685185,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.007298616161616161,
      "avg_duration_per_task_ms": 578052.9191919192,
      "avg_output_tokens_per_task": 17923.358585858587,
      "last_run_timestamp": "2026-07-18T08:44:31.047Z"
    },
    {
      "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": "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": 0.684238375,
      "accuracy_std_dev": 0.042521690017970536,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.01436286111111111,
      "avg_duration_per_task_ms": 255009.83333333334,
      "avg_output_tokens_per_task": 7755.36553030303,
      "last_run_timestamp": "2026-10-03T17:36:48.437Z"
    },
    {
      "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": 0.681700944587629,
      "accuracy_std_dev": 0.10151777573533613,
      "run_count": 4,
      "total_questions": 776,
      "avg_cost_per_task_usd": 0.061980561855670094,
      "avg_duration_per_task_ms": 552656.2654639175,
      "avg_output_tokens_per_task": 25817.090206185567,
      "last_run_timestamp": "2026-08-15T10:07:37.662Z"
    },
    {
      "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": 0.6815885046931408,
      "accuracy_std_dev": 0.03569859904757024,
      "run_count": 7,
      "total_questions": 1385,
      "avg_cost_per_task_usd": 0.005665397111913357,
      "avg_duration_per_task_ms": 406427.05487364624,
      "avg_output_tokens_per_task": 15523.718411552347,
      "last_run_timestamp": "2026-10-03T17:31:07.224Z"
    },
    {
      "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": 0.6762865714285715,
      "accuracy_std_dev": 0.05037246428033344,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.008260983405483407,
      "avg_duration_per_task_ms": 263120.9155844156,
      "avg_output_tokens_per_task": 42075.49062049062,
      "last_run_timestamp": "2026-10-03T09:28:39.948Z"
    },
    {
      "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": "DigitalOcean",
      "accuracy": 0.675926,
      "accuracy_std_dev": 0.009259000000000017,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.0023789772727272726,
      "avg_duration_per_task_ms": 1250422.4444444445,
      "avg_output_tokens_per_task": 14190.916666666666,
      "last_run_timestamp": "2026-07-18T09:58:41.490Z"
    },
    {
      "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": 0.6757730775095298,
      "accuracy_std_dev": 0.020105082568893876,
      "run_count": 7,
      "total_questions": 787,
      "avg_cost_per_task_usd": 0.005445952986022871,
      "avg_duration_per_task_ms": 625853.6467598475,
      "avg_output_tokens_per_task": 36252.05082592122,
      "last_run_timestamp": "2026-08-29T17:24:11.557Z"
    },
    {
      "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": 0.6757497120822621,
      "accuracy_std_dev": 0.20642627550360443,
      "run_count": 10,
      "total_questions": 1945,
      "avg_cost_per_task_usd": 0.014975519280205657,
      "avg_duration_per_task_ms": 371086.62262210797,
      "avg_output_tokens_per_task": 18143.37994858612,
      "last_run_timestamp": "2026-09-26T08:53:03.343Z"
    },
    {
      "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": 0.6750842,
      "accuracy_std_dev": 0.023641210809939493,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.005652539393939394,
      "avg_duration_per_task_ms": 500294.62828282826,
      "avg_output_tokens_per_task": 16033.422222222222,
      "last_run_timestamp": "2026-08-29T18:40:21.422Z"
    },
    {
      "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": 0.6694725555555555,
      "accuracy_std_dev": 0.032119090204740325,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.0024531571268237937,
      "avg_duration_per_task_ms": 211506.06902356903,
      "avg_output_tokens_per_task": 5166.5179573512905,
      "last_run_timestamp": "2026-10-03T16:12:07.496Z"
    },
    {
      "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "DigitalOcean",
      "accuracy": 0.6688176666666668,
      "accuracy_std_dev": 0.09003633453223205,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.0024410033670033675,
      "avg_duration_per_task_ms": 222109.77160493826,
      "avg_output_tokens_per_task": 5583.682940516273,
      "last_run_timestamp": "2026-10-03T16:12:35.932Z"
    },
    {
      "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": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Venice",
      "accuracy": 0.6673561406411582,
      "accuracy_std_dev": 0.17106562595518718,
      "run_count": 5,
      "total_questions": 967,
      "avg_cost_per_task_usd": 0.07843655946225439,
      "avg_duration_per_task_ms": 1463737.4043433298,
      "avg_output_tokens_per_task": 37688.086866597725,
      "last_run_timestamp": "2026-08-22T17:02:13.071Z"
    },
    {
      "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": "WandB",
      "accuracy": 0.6649835000000001,
      "accuracy_std_dev": 0.0016835000000000044,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.009143497474747474,
      "avg_duration_per_task_ms": 686922.851010101,
      "avg_output_tokens_per_task": 70113.30555555556,
      "last_run_timestamp": "2026-07-18T17:04:25.343Z"
    },
    {
      "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": 0.6637464899415964,
      "accuracy_std_dev": 0.013487143446978517,
      "run_count": 8,
      "total_questions": 1541,
      "avg_cost_per_task_usd": 0.007471570408825438,
      "avg_duration_per_task_ms": 501530.87605451007,
      "avg_output_tokens_per_task": 53289.73977936405,
      "last_run_timestamp": "2026-10-03T18:05:00.262Z"
    },
    {
      "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": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "Venice",
      "accuracy": 0.6622475000000001,
      "accuracy_std_dev": 0.023522840681983983,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.028767857323232324,
      "avg_duration_per_task_ms": 847663.7582070708,
      "avg_output_tokens_per_task": 19263.888257575756,
      "last_run_timestamp": "2026-10-03T16:57:14.637Z"
    },
    {
      "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": 0.6621005671232876,
      "accuracy_std_dev": 0.014348054525498073,
      "run_count": 3,
      "total_questions": 365,
      "avg_cost_per_task_usd": 0.004233909589041096,
      "avg_duration_per_task_ms": 723664.6465753425,
      "avg_output_tokens_per_task": 42200.62191780822,
      "last_run_timestamp": "2026-10-03T19:04:05.275Z"
    },
    {
      "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": 0.6616163333333333,
      "accuracy_std_dev": 0.0111670002935236,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.0024947028619528616,
      "avg_duration_per_task_ms": 94165.7398989899,
      "avg_output_tokens_per_task": 2312.0664983164984,
      "last_run_timestamp": "2026-08-29T20:54:16.286Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-chat-v3",
      "model_slug": "deepseek/deepseek-chat",
      "model_name": "DeepSeek: DeepSeek V3",
      "provider_name": "Novita",
      "accuracy": 0.661616,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.003362974747474748,
      "avg_duration_per_task_ms": 65871.15656565657,
      "avg_output_tokens_per_task": 2362.121212121212,
      "last_run_timestamp": "2026-08-01T09:30:53.884Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4.1-2025-04-14",
      "model_slug": "openai/gpt-4.1",
      "model_name": "OpenAI: GPT-4.1",
      "provider_name": "OpenAI",
      "accuracy": 0.6607744999999999,
      "accuracy_std_dev": 0.011572094721786573,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.015565636363636365,
      "avg_duration_per_task_ms": 14159.460858585859,
      "avg_output_tokens_per_task": 1751.3977272727273,
      "last_run_timestamp": "2026-10-02T23:17:48.072Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.6606341666666665,
      "accuracy_std_dev": 0.013370433711697206,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.00517740867003367,
      "avg_duration_per_task_ms": 106914.1734006734,
      "avg_output_tokens_per_task": 9753.443602693602,
      "last_run_timestamp": "2026-10-03T09:23:08.256Z"
    },
    {
      "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": "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": 0.6599325714285714,
      "accuracy_std_dev": 0.010532623440084252,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.0027436955266955266,
      "avg_duration_per_task_ms": 79836.26262626263,
      "avg_output_tokens_per_task": 2910.7698412698414,
      "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": 0.6596921428571428,
      "accuracy_std_dev": 0.011884572910513284,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.0022321969696969695,
      "avg_duration_per_task_ms": 222919.87445887446,
      "avg_output_tokens_per_task": 3156.7669552669554,
      "last_run_timestamp": "2026-08-29T18:47:21.578Z"
    },
    {
      "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": 0.6596062028241334,
      "accuracy_std_dev": 0.016648273731831805,
      "run_count": 4,
      "total_questions": 779,
      "avg_cost_per_task_usd": 0.011759379974326058,
      "avg_duration_per_task_ms": 1526360.2888318356,
      "avg_output_tokens_per_task": 50819.251604621306,
      "last_run_timestamp": "2026-10-03T17:41:49.078Z"
    },
    {
      "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": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.6571429371428571,
      "accuracy_std_dev": 0.01564259566443798,
      "run_count": 16,
      "total_questions": 3150,
      "avg_cost_per_task_usd": 0.009064113015873019,
      "avg_duration_per_task_ms": 336807.49333333335,
      "avg_output_tokens_per_task": 60106.63492063492,
      "last_run_timestamp": "2026-10-03T17:56:49.434Z"
    },
    {
      "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": 0.654882,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.01704559090909091,
      "avg_duration_per_task_ms": 1937600.9292929294,
      "avg_output_tokens_per_task": 68071.47474747474,
      "last_run_timestamp": "2026-10-01T08:41:43.744Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "Parasail",
      "accuracy": 0.6539546483050847,
      "accuracy_std_dev": 0.06028980821738355,
      "run_count": 3,
      "total_questions": 236,
      "avg_cost_per_task_usd": 0.09281549576271186,
      "avg_duration_per_task_ms": 1220256.3305084745,
      "avg_output_tokens_per_task": 32723.118644067796,
      "last_run_timestamp": "2026-08-01T18:16:39.158Z"
    },
    {
      "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": 0.6512783352272727,
      "accuracy_std_dev": 0.07301920016707937,
      "run_count": 5,
      "total_questions": 704,
      "avg_cost_per_task_usd": 0.06126106249999999,
      "avg_duration_per_task_ms": 1570715.0482954546,
      "avg_output_tokens_per_task": 22881.873579545456,
      "last_run_timestamp": "2026-08-22T17:09:42.218Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-coder-next-2025-02-03",
      "model_slug": "qwen/qwen3-coder-next",
      "model_name": "Qwen: Qwen3 Coder Next",
      "provider_name": "Ionstream",
      "accuracy": 0.6509389830985916,
      "accuracy_std_dev": 0.10781834547167748,
      "run_count": 4,
      "total_questions": 710,
      "avg_cost_per_task_usd": 0.0075063676056338026,
      "avg_duration_per_task_ms": 162639.05070422534,
      "avg_output_tokens_per_task": 9309.304225352113,
      "last_run_timestamp": "2026-08-08T16:46:11.805Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-chat-v3-0324",
      "model_slug": "deepseek/deepseek-chat-v3-0324",
      "model_name": "DeepSeek: DeepSeek V3 0324",
      "provider_name": "Crusoe",
      "accuracy": 0.650363823580786,
      "accuracy_std_dev": 0.007103203700131033,
      "run_count": 6,
      "total_questions": 1145,
      "avg_cost_per_task_usd": 0.0036315545851528385,
      "avg_duration_per_task_ms": 58503.03493449782,
      "avg_output_tokens_per_task": 2202.661135371179,
      "last_run_timestamp": "2026-08-29T16:10:45.295Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4.1-mini-2025-04-14",
      "model_slug": "openai/gpt-4.1-mini",
      "model_name": "OpenAI: GPT-4.1 Mini",
      "provider_name": "auto-routing",
      "accuracy": 0.6492706666666667,
      "accuracy_std_dev": 0.011195005920300148,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.004075954545454546,
      "avg_duration_per_task_ms": 28668.510101010103,
      "avg_output_tokens_per_task": 2354.649831649832,
      "last_run_timestamp": "2026-10-01T19:12:31.553Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.6487794999999998,
      "accuracy_std_dev": 0.12844600199597497,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.6039836092171716,
      "avg_duration_per_task_ms": 625325.3693181818,
      "avg_output_tokens_per_task": 40216.142676767675,
      "last_run_timestamp": "2026-10-03T10:37:00.593Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.5-air",
      "model_slug": "z-ai/glm-4.5-air",
      "model_name": "Z.ai: GLM 4.5 Air",
      "provider_name": "SiliconFlow",
      "accuracy": 0.6478542063492063,
      "accuracy_std_dev": 0.03133768881430511,
      "run_count": 3,
      "total_questions": 567,
      "avg_cost_per_task_usd": 0.025561564373897706,
      "avg_duration_per_task_ms": 917973.6243386243,
      "avg_output_tokens_per_task": 30424.42680776014,
      "last_run_timestamp": "2026-08-01T17:59:38.872Z"
    },
    {
      "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": 0.6478421818181818,
      "accuracy_std_dev": 0.01794361558690384,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.00699533930211203,
      "avg_duration_per_task_ms": 234218.77502295683,
      "avg_output_tokens_per_task": 11266.905417814509,
      "last_run_timestamp": "2026-10-03T10:20:18.247Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "CoreWeave",
      "accuracy": 0.6476671428571429,
      "accuracy_std_dev": 0.01801337100543905,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.007287767676767677,
      "avg_duration_per_task_ms": 701126.4617604618,
      "avg_output_tokens_per_task": 56538.52525252525,
      "last_run_timestamp": "2026-10-03T18:53:13.795Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "ibm-granite/granite-4.2-8b-20260831",
      "model_slug": "ibm-granite/granite-4.2-8b",
      "model_name": "IBM: Granite 4.2 8B",
      "provider_name": "CoreWeave",
      "accuracy": 0.647493,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 113,
      "avg_cost_per_task_usd": 0.005189893805309735,
      "avg_duration_per_task_ms": 619234.2654867256,
      "avg_output_tokens_per_task": 34365.4778761062,
      "last_run_timestamp": "2026-10-01T08:28:32.911Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Mancer 2",
      "accuracy": 0.6474267142857143,
      "accuracy_std_dev": 0.08020884466675425,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.005957715007215008,
      "avg_duration_per_task_ms": 632251.2438672439,
      "avg_output_tokens_per_task": 21165.58008658009,
      "last_run_timestamp": "2026-10-03T17:57:14.269Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4.1-mini-2025-04-14",
      "model_slug": "openai/gpt-4.1-mini",
      "model_name": "OpenAI: GPT-4.1 Mini",
      "provider_name": "OpenAI",
      "accuracy": 0.6473065000000001,
      "accuracy_std_dev": 0.009707729665065862,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.00413114393939394,
      "avg_duration_per_task_ms": 27377.72601010101,
      "avg_output_tokens_per_task": 2388.9305555555557,
      "last_run_timestamp": "2026-10-02T23:15:53.868Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.6471861428571428,
      "accuracy_std_dev": 0.010488726920277937,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.0025087236652236652,
      "avg_duration_per_task_ms": 65942.57142857143,
      "avg_output_tokens_per_task": 2709.98556998557,
      "last_run_timestamp": "2026-08-29T17:03:33.297Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-chat-v3-0324",
      "model_slug": "deepseek/deepseek-chat-v3-0324",
      "model_name": "DeepSeek: DeepSeek V3 0324",
      "provider_name": "GMICloud",
      "accuracy": 0.646465,
      "accuracy_std_dev": 0.004761657064510171,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0029047154882154884,
      "avg_duration_per_task_ms": 53981.06734006734,
      "avg_output_tokens_per_task": 2363.939393939394,
      "last_run_timestamp": "2026-08-29T16:33:11.860Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Parasail",
      "accuracy": 0.64646475,
      "accuracy_std_dev": 0.02185317805692116,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.007503073232323232,
      "avg_duration_per_task_ms": 814987.6553030303,
      "avg_output_tokens_per_task": 52529.47348484849,
      "last_run_timestamp": "2026-10-03T20:08:16.264Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4.1-2025-04-14",
      "model_slug": "openai/gpt-4.1",
      "model_name": "OpenAI: GPT-4.1",
      "provider_name": "auto-routing",
      "accuracy": 0.6464646666666667,
      "accuracy_std_dev": 0.013537798302858865,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.015368191919191919,
      "avg_duration_per_task_ms": 14060.520202020201,
      "avg_output_tokens_per_task": 1727.3535353535353,
      "last_run_timestamp": "2026-10-01T19:12:45.033Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-4-maverick-17b-128e-instruct",
      "model_slug": "meta-llama/llama-4-maverick",
      "model_name": "Meta: Llama 4 Maverick",
      "provider_name": "Parasail",
      "accuracy": 0.6459835714285715,
      "accuracy_std_dev": 0.011408025080832253,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.0032850678210678214,
      "avg_duration_per_task_ms": 40715.905483405484,
      "avg_output_tokens_per_task": 3121.487012987013,
      "last_run_timestamp": "2026-08-29T19:51:51.198Z"
    },
    {
      "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": 0.6450217142857143,
      "accuracy_std_dev": 0.013022733909966664,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.0023176284271284268,
      "avg_duration_per_task_ms": 69183.5432900433,
      "avg_output_tokens_per_task": 2807.774891774892,
      "last_run_timestamp": "2026-08-29T16:11:23.388Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "DeepInfra",
      "accuracy": 0.6447570300850229,
      "accuracy_std_dev": 0.0798361906300707,
      "run_count": 9,
      "total_questions": 1529,
      "avg_cost_per_task_usd": 0.02925614519293656,
      "avg_duration_per_task_ms": 627719.6899934597,
      "avg_output_tokens_per_task": 30710.851536952257,
      "last_run_timestamp": "2026-10-03T17:54:48.467Z"
    },
    {
      "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": "WandB",
      "accuracy": 0.6446141374045801,
      "accuracy_std_dev": 0.005257999999999985,
      "run_count": 2,
      "total_questions": 393,
      "avg_cost_per_task_usd": 0.0025083384223918577,
      "avg_duration_per_task_ms": 104571.39694656488,
      "avg_output_tokens_per_task": 8089.860050890586,
      "last_run_timestamp": "2026-07-18T08:33:31.052Z"
    },
    {
      "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": 0.6443544137931034,
      "accuracy_std_dev": 0.06155566230851553,
      "run_count": 5,
      "total_questions": 986,
      "avg_cost_per_task_usd": 0.04049388945233266,
      "avg_duration_per_task_ms": 420330.3397565923,
      "avg_output_tokens_per_task": 20151.943204868156,
      "last_run_timestamp": "2026-08-22T10:29:25.897Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "auto-routing",
      "accuracy": 0.6439909795918367,
      "accuracy_std_dev": 0.019637275869637497,
      "run_count": 9,
      "total_questions": 1764,
      "avg_cost_per_task_usd": 0.01045445804988662,
      "avg_duration_per_task_ms": 693718.4659863946,
      "avg_output_tokens_per_task": 52902.34467120181,
      "last_run_timestamp": "2026-10-03T20:22:45.900Z"
    },
    {
      "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": 0.6438963244781784,
      "accuracy_std_dev": 0.010816729929113277,
      "run_count": 8,
      "total_questions": 1581,
      "avg_cost_per_task_usd": 0.0026027520556609744,
      "avg_duration_per_task_ms": 140999.67488931055,
      "avg_output_tokens_per_task": 9253.603415559772,
      "last_run_timestamp": "2026-10-03T17:16:11.892Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4.1-2025-04-14",
      "model_slug": "openai/gpt-4.1",
      "model_name": "OpenAI: GPT-4.1",
      "provider_name": "Azure",
      "accuracy": 0.643659,
      "accuracy_std_dev": 0.015752669318774726,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.015761657407407408,
      "avg_duration_per_task_ms": 18961.11531986532,
      "avg_output_tokens_per_task": 1683.8872053872053,
      "last_run_timestamp": "2026-10-01T19:14:40.644Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "Alibaba",
      "accuracy": 0.640783,
      "accuracy_std_dev": 0.01270844444650878,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.004579357954545454,
      "avg_duration_per_task_ms": 73293.53661616161,
      "avg_output_tokens_per_task": 8611.67803030303,
      "last_run_timestamp": "2026-10-03T16:51:25.541Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4.1-mini-2025-04-14",
      "model_slug": "openai/gpt-4.1-mini",
      "model_name": "OpenAI: GPT-4.1 Mini",
      "provider_name": "Azure",
      "accuracy": 0.6405723333333334,
      "accuracy_std_dev": 0.0030349496792460334,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.004210166666666667,
      "avg_duration_per_task_ms": 30431.28367003367,
      "avg_output_tokens_per_task": 2308.9638047138046,
      "last_run_timestamp": "2026-10-01T19:13:29.560Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "Venice",
      "accuracy": 0.6404040121212122,
      "accuracy_std_dev": 0.11145785266385451,
      "run_count": 8,
      "total_questions": 1485,
      "avg_cost_per_task_usd": 0.09477065723905725,
      "avg_duration_per_task_ms": 760875.5326599326,
      "avg_output_tokens_per_task": 29720.403367003368,
      "last_run_timestamp": "2026-10-03T16:55:14.697Z"
    },
    {
      "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": 0.6386253795066414,
      "accuracy_std_dev": 0.016118987931551414,
      "run_count": 8,
      "total_questions": 1581,
      "avg_cost_per_task_usd": 0.002519017710309931,
      "avg_duration_per_task_ms": 220560.1669829222,
      "avg_output_tokens_per_task": 8172.611638203669,
      "last_run_timestamp": "2026-10-03T17:03:50.406Z"
    },
    {
      "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": 0.6384837407407408,
      "accuracy_std_dev": 0.1301056437821581,
      "run_count": 27,
      "total_questions": 5346,
      "avg_cost_per_task_usd": 0.6548667364384586,
      "avg_duration_per_task_ms": 493199.0927796483,
      "avg_output_tokens_per_task": 40750.34081556304,
      "last_run_timestamp": "2026-10-03T10:04:02.304Z"
    },
    {
      "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": 0.6381495969581749,
      "accuracy_std_dev": 0.013726909852890242,
      "run_count": 8,
      "total_questions": 1578,
      "avg_cost_per_task_usd": 0.003178458808618504,
      "avg_duration_per_task_ms": 206354.68187579213,
      "avg_output_tokens_per_task": 10325.579847908744,
      "last_run_timestamp": "2026-10-03T16:41:22.866Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "StreamLake",
      "accuracy": 0.6376733333333334,
      "accuracy_std_dev": 0.015547005006324102,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.0016130179573512905,
      "avg_duration_per_task_ms": 106654.77497194163,
      "avg_output_tokens_per_task": 8152.802469135802,
      "last_run_timestamp": "2026-10-03T16:47:32.559Z"
    },
    {
      "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": 0.6374935167427701,
      "accuracy_std_dev": 0.023026873072564586,
      "run_count": 8,
      "total_questions": 1314,
      "avg_cost_per_task_usd": 0.016053601217656012,
      "avg_duration_per_task_ms": 96215.6704718417,
      "avg_output_tokens_per_task": 53290.504566210046,
      "last_run_timestamp": "2026-10-03T17:21:59.201Z"
    },
    {
      "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": 0.6371463734317343,
      "accuracy_std_dev": 0.2220532628164637,
      "run_count": 8,
      "total_questions": 1355,
      "avg_cost_per_task_usd": 0.007668025830258302,
      "avg_duration_per_task_ms": 311815.4856088561,
      "avg_output_tokens_per_task": 12706.190405904059,
      "last_run_timestamp": "2026-10-03T16:57:17.907Z"
    },
    {
      "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": 0.6368643521865888,
      "accuracy_std_dev": 0.13454240742991402,
      "run_count": 26,
      "total_questions": 5145,
      "avg_cost_per_task_usd": 0.6621504369290573,
      "avg_duration_per_task_ms": 598729.5551020408,
      "avg_output_tokens_per_task": 41288.28415937804,
      "last_run_timestamp": "2026-10-03T10:34:21.171Z"
    },
    {
      "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": 0.636738,
      "accuracy_std_dev": 0.1325038277837034,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.6152635606060606,
      "avg_duration_per_task_ms": 631219.6868686868,
      "avg_output_tokens_per_task": 40865.88888888889,
      "last_run_timestamp": "2026-10-03T09:11:52.463Z"
    },
    {
      "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": "Alibaba",
      "accuracy": 0.635522,
      "accuracy_std_dev": 0.009259000000000017,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.013100151515151516,
      "avg_duration_per_task_ms": 444234.55555555556,
      "avg_output_tokens_per_task": 31287.63888888889,
      "last_run_timestamp": "2026-07-18T16:20:00.500Z"
    },
    {
      "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": 0.6352981489141675,
      "accuracy_std_dev": 0.02103570346434842,
      "run_count": 5,
      "total_questions": 967,
      "avg_cost_per_task_usd": 0.006571570837642193,
      "avg_duration_per_task_ms": 890942.791106515,
      "avg_output_tokens_per_task": 36274.74457083764,
      "last_run_timestamp": "2026-10-03T08:12:46.446Z"
    },
    {
      "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": 0.6352257765567765,
      "accuracy_std_dev": 0.022072905759475025,
      "run_count": 3,
      "total_questions": 546,
      "avg_cost_per_task_usd": 0.013415796703296703,
      "avg_duration_per_task_ms": 548616.1373626373,
      "avg_output_tokens_per_task": 46655.3021978022,
      "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": "Novita",
      "accuracy": 0.6349608333333333,
      "accuracy_std_dev": 0.0326935856217325,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.006947168350168351,
      "avg_duration_per_task_ms": 113160.32070707071,
      "avg_output_tokens_per_task": 9692.425505050505,
      "last_run_timestamp": "2026-10-03T10:32:36.729Z"
    },
    {
      "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": 0.6337581592356688,
      "accuracy_std_dev": 0.004703999999999986,
      "run_count": 2,
      "total_questions": 314,
      "avg_cost_per_task_usd": 0.002880643312101911,
      "avg_duration_per_task_ms": 145451.6560509554,
      "avg_output_tokens_per_task": 9370.200636942674,
      "last_run_timestamp": "2026-10-03T17:25:22.852Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.5-air",
      "model_slug": "z-ai/glm-4.5-air",
      "model_name": "Z.ai: GLM 4.5 Air",
      "provider_name": "Z.AI",
      "accuracy": 0.6337451904761905,
      "accuracy_std_dev": 0.08567913487347233,
      "run_count": 3,
      "total_questions": 567,
      "avg_cost_per_task_usd": 0.03312850793650794,
      "avg_duration_per_task_ms": 835240.3333333334,
      "avg_output_tokens_per_task": 32942.40211640212,
      "last_run_timestamp": "2026-08-01T18:05:37.080Z"
    },
    {
      "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": 0.633745,
      "accuracy_std_dev": 0.1400560587510904,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.6191078804713805,
      "avg_duration_per_task_ms": 634861.303030303,
      "avg_output_tokens_per_task": 41204.60381593715,
      "last_run_timestamp": "2026-10-03T10:40:39.138Z"
    },
    {
      "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": 0.6326063626168223,
      "accuracy_std_dev": 0.02740974656739623,
      "run_count": 9,
      "total_questions": 1605,
      "avg_cost_per_task_usd": 0.022848554517133954,
      "avg_duration_per_task_ms": 1040840.2074766355,
      "avg_output_tokens_per_task": 95001.86853582555,
      "last_run_timestamp": "2026-10-03T17:27:37.016Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3-nano-30b-a3b",
      "model_slug": "nvidia/nemotron-3-nano-30b-a3b",
      "model_name": "NVIDIA: Nemotron 3 Nano 30B A3B",
      "provider_name": "Crusoe",
      "accuracy": 0.6325426993295513,
      "accuracy_std_dev": 0.010534965676735725,
      "run_count": 10,
      "total_questions": 1939,
      "avg_cost_per_task_usd": 0.01766013512119649,
      "avg_duration_per_task_ms": 478451.96183599794,
      "avg_output_tokens_per_task": 88282.73388344508,
      "last_run_timestamp": "2026-10-03T17:08:02.072Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "inclusionai/ling-3.0-flash-fin-20260827",
      "model_slug": "inclusionai/ling-3.0-flash-fin",
      "model_name": "inclusionAI: Ling 3.0 Flash Fin",
      "provider_name": "Novita",
      "accuracy": 0.632155,
      "accuracy_std_dev": 0.005892000000000008,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.005966434343434343,
      "avg_duration_per_task_ms": 235473.01515151514,
      "avg_output_tokens_per_task": 48249.11363636364,
      "last_run_timestamp": "2026-10-01T08:24:52.614Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.5-air",
      "model_slug": "z-ai/glm-4.5-air",
      "model_name": "Z.ai: GLM 4.5 Air",
      "provider_name": "auto-routing",
      "accuracy": 0.6317296185567011,
      "accuracy_std_dev": 0.03997713513997718,
      "run_count": 3,
      "total_questions": 582,
      "avg_cost_per_task_usd": 0.02984942783505154,
      "avg_duration_per_task_ms": 817292.206185567,
      "avg_output_tokens_per_task": 36396.65120274914,
      "last_run_timestamp": "2026-08-01T18:25:56.196Z"
    },
    {
      "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": 0.6308931798152096,
      "accuracy_std_dev": 0.018213636160511713,
      "run_count": 8,
      "total_questions": 1407,
      "avg_cost_per_task_usd": 0.007495727078891258,
      "avg_duration_per_task_ms": 35794.715707178395,
      "avg_output_tokens_per_task": 12413.633262260128,
      "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": "CoreWeave",
      "accuracy": 0.6308680607902736,
      "accuracy_std_dev": 0.011456259937693471,
      "run_count": 5,
      "total_questions": 987,
      "avg_cost_per_task_usd": 0.0034108257345491387,
      "avg_duration_per_task_ms": 140057.4062816616,
      "avg_output_tokens_per_task": 11472.941236068895,
      "last_run_timestamp": "2026-08-29T18:20:19.448Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "SiliconFlow",
      "accuracy": 0.6305648888888888,
      "accuracy_std_dev": 0.016837183193056476,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.0016988282828282828,
      "avg_duration_per_task_ms": 204946.65937149271,
      "avg_output_tokens_per_task": 9301.569023569024,
      "last_run_timestamp": "2026-10-03T16:12:29.903Z"
    },
    {
      "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": 0.6293489999999999,
      "accuracy_std_dev": 0.008199388066597833,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0011419814814814814,
      "avg_duration_per_task_ms": 334863.3063973064,
      "avg_output_tokens_per_task": 11634.42760942761,
      "last_run_timestamp": "2026-10-03T17:38:50.485Z"
    },
    {
      "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/qwen3-32b-04-28",
      "model_slug": "qwen/qwen3-32b",
      "model_name": "Qwen: Qwen3 32B",
      "provider_name": "SiliconFlow",
      "accuracy": 0.628788,
      "accuracy_std_dev": 0.01430999999999999,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.017457550505050507,
      "avg_duration_per_task_ms": 1341424.7676767677,
      "avg_output_tokens_per_task": 31206.515151515152,
      "last_run_timestamp": "2026-07-18T17:46:02.370Z"
    },
    {
      "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": "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": 0.6284272857142859,
      "accuracy_std_dev": 0.0807098788319077,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.011130141414141415,
      "avg_duration_per_task_ms": 1152637.8816738818,
      "avg_output_tokens_per_task": 32863.90331890332,
      "last_run_timestamp": "2026-10-03T17:46:52.441Z"
    },
    {
      "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": 0.626793113785558,
      "accuracy_std_dev": 0.019460388437090578,
      "run_count": 7,
      "total_questions": 1371,
      "avg_cost_per_task_usd": 0.014277733041575492,
      "avg_duration_per_task_ms": 690347.9810357403,
      "avg_output_tokens_per_task": 14917.212253829322,
      "last_run_timestamp": "2026-10-03T11:06:38.068Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.5-air",
      "model_slug": "z-ai/glm-4.5-air",
      "model_name": "Z.ai: GLM 4.5 Air",
      "provider_name": "Novita",
      "accuracy": 0.6249264761904761,
      "accuracy_std_dev": 0.06315233322688879,
      "run_count": 3,
      "total_questions": 567,
      "avg_cost_per_task_usd": 0.029899758377425045,
      "avg_duration_per_task_ms": 776085.4091710758,
      "avg_output_tokens_per_task": 36770.42328042328,
      "last_run_timestamp": "2026-08-01T18:07:49.346Z"
    },
    {
      "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-04-28",
      "model_slug": "qwen/qwen3-30b-a3b",
      "model_name": "Qwen: Qwen3 30B A3B",
      "provider_name": "auto-routing",
      "accuracy": 0.6212123333333334,
      "accuracy_std_dev": 0.009622003960829715,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.00826737037037037,
      "avg_duration_per_task_ms": 370134.4276094276,
      "avg_output_tokens_per_task": 15814.858585858587,
      "last_run_timestamp": "2026-08-01T18:08:49.019Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-32b-04-28",
      "model_slug": "qwen/qwen3-32b",
      "model_name": "Qwen: Qwen3 32B",
      "provider_name": "Nebius",
      "accuracy": 0.6164022857142857,
      "accuracy_std_dev": 0.022360138497617886,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.0054965389610389616,
      "avg_duration_per_task_ms": 603772.5,
      "avg_output_tokens_per_task": 18275.68831168831,
      "last_run_timestamp": "2026-08-29T19:06:01.882Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-14b-04-28",
      "model_slug": "qwen/qwen3-14b",
      "model_name": "Qwen: Qwen3 14B",
      "provider_name": "DeepInfra",
      "accuracy": 0.6162989117147707,
      "accuracy_std_dev": 0.004859400671779274,
      "run_count": 3,
      "total_questions": 589,
      "avg_cost_per_task_usd": 0.0033007164685908323,
      "avg_duration_per_task_ms": 414472.32767402375,
      "avg_output_tokens_per_task": 13345.62818336163,
      "last_run_timestamp": "2026-10-01T16:36:31.937Z"
    },
    {
      "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": 0.6132539090909092,
      "accuracy_std_dev": 0.1299958112139524,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.6145130344352616,
      "avg_duration_per_task_ms": 628297.9467401286,
      "avg_output_tokens_per_task": 40850.65381083563,
      "last_run_timestamp": "2026-10-03T10:13:27.517Z"
    },
    {
      "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": 0.6131824185936763,
      "accuracy_std_dev": 0.020395407276319392,
      "run_count": 11,
      "total_questions": 2119,
      "avg_cost_per_task_usd": 0.008309551675318546,
      "avg_duration_per_task_ms": 519028.40207645117,
      "avg_output_tokens_per_task": 13645.85040113261,
      "last_run_timestamp": "2026-10-03T10:33:57.761Z"
    },
    {
      "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": 0.6107830459317585,
      "accuracy_std_dev": 0.004553237055930632,
      "run_count": 8,
      "total_questions": 1524,
      "avg_cost_per_task_usd": 0.004046526902887138,
      "avg_duration_per_task_ms": 40209.83464566929,
      "avg_output_tokens_per_task": 2149.975721784777,
      "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": "NextBit",
      "accuracy": 0.6102695,
      "accuracy_std_dev": 0.004208500000000004,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.00920193181818182,
      "avg_duration_per_task_ms": 945597.9292929292,
      "avg_output_tokens_per_task": 16723.31818181818,
      "last_run_timestamp": "2026-07-18T08:03:56.103Z"
    },
    {
      "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": 0.609638125,
      "accuracy_std_dev": 0.00961364604920396,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.003643919823232323,
      "avg_duration_per_task_ms": 34782.93308080808,
      "avg_output_tokens_per_task": 2151.7373737373737,
      "last_run_timestamp": "2026-10-03T17:03:29.436Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-32b-04-28",
      "model_slug": "qwen/qwen3-32b",
      "model_name": "Qwen: Qwen3 32B",
      "provider_name": "auto-routing",
      "accuracy": 0.6090626312364424,
      "accuracy_std_dev": 0.009643076431368988,
      "run_count": 7,
      "total_questions": 1383,
      "avg_cost_per_task_usd": 0.008417777295733913,
      "avg_duration_per_task_ms": 272268.59725234995,
      "avg_output_tokens_per_task": 17483.89949385394,
      "last_run_timestamp": "2026-08-29T19:52:22.642Z"
    },
    {
      "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": 0.6081999956331878,
      "accuracy_std_dev": 0.04268956350879486,
      "run_count": 7,
      "total_questions": 1374,
      "avg_cost_per_task_usd": 0.023787051673944692,
      "avg_duration_per_task_ms": 1027252.8209606986,
      "avg_output_tokens_per_task": 62021.267831149926,
      "last_run_timestamp": "2026-08-30T00:32:50.853Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "CoreWeave",
      "accuracy": 0.607744,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.003953611111111112,
      "avg_duration_per_task_ms": 57224.242424242424,
      "avg_output_tokens_per_task": 2094.1161616161617,
      "last_run_timestamp": "2026-08-01T17:44:33.510Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-30b-a3b-04-28",
      "model_slug": "qwen/qwen3-30b-a3b",
      "model_name": "Qwen: Qwen3 30B A3B",
      "provider_name": "DeepInfra",
      "accuracy": 0.606622,
      "accuracy_std_dev": 0.0007933738084913017,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.007795220538720538,
      "avg_duration_per_task_ms": 300972.74074074073,
      "avg_output_tokens_per_task": 15395.011784511784,
      "last_run_timestamp": "2026-08-01T18:16:00.374Z"
    },
    {
      "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": 0.6063702713347922,
      "accuracy_std_dev": 0.021503606331620406,
      "run_count": 7,
      "total_questions": 1371,
      "avg_cost_per_task_usd": 0.06956871261852661,
      "avg_duration_per_task_ms": 1033277.1312910285,
      "avg_output_tokens_per_task": 26710.033552151715,
      "last_run_timestamp": "2026-08-30T00:22:07.120Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-30b-a3b-04-28",
      "model_slug": "qwen/qwen3-30b-a3b",
      "model_name": "Qwen: Qwen3 30B A3B",
      "provider_name": "Alibaba",
      "accuracy": 0.6060606666666666,
      "accuracy_std_dev": 0.0013745720142000956,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.008301892255892255,
      "avg_duration_per_task_ms": 153838.79292929292,
      "avg_output_tokens_per_task": 15775.223905723906,
      "last_run_timestamp": "2026-08-01T17:51:29.206Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "WandB",
      "accuracy": 0.6060605,
      "accuracy_std_dev": 0.001683499999999949,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.004138252525252526,
      "avg_duration_per_task_ms": 50192.44191919192,
      "avg_output_tokens_per_task": 2217.2070707070707,
      "last_run_timestamp": "2026-07-18T08:26:39.830Z"
    },
    {
      "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": 0.60437725,
      "accuracy_std_dev": 0.004908375259441744,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.011036202651515153,
      "avg_duration_per_task_ms": 34065.54861111111,
      "avg_output_tokens_per_task": 2102.629419191919,
      "last_run_timestamp": "2026-10-03T17:43:44.972Z"
    },
    {
      "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": 0.604166875,
      "accuracy_std_dev": 0.006651157426296166,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.003487729797979798,
      "avg_duration_per_task_ms": 53432.55744949495,
      "avg_output_tokens_per_task": 2119.665404040404,
      "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": "DeepInfra",
      "accuracy": 0.6041191250947687,
      "accuracy_std_dev": 0.014817500782134993,
      "run_count": 7,
      "total_questions": 1319,
      "avg_cost_per_task_usd": 0.0021458051554207732,
      "avg_duration_per_task_ms": 47625.93252463988,
      "avg_output_tokens_per_task": 2007.382107657316,
      "last_run_timestamp": "2026-10-03T16:40:33.970Z"
    },
    {
      "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": 0.6004493333333333,
      "accuracy_std_dev": 0.0031744380430067634,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.014951090909090909,
      "avg_duration_per_task_ms": 287967.1666666667,
      "avg_output_tokens_per_task": 12268.760942760942,
      "last_run_timestamp": "2026-08-15T15:26:03.892Z"
    },
    {
      "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": 0.5982746250000001,
      "accuracy_std_dev": 0.011782541692452221,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.0032721849747474746,
      "avg_duration_per_task_ms": 64540.08964646464,
      "avg_output_tokens_per_task": 2086.2304292929293,
      "last_run_timestamp": "2026-10-03T16:59:44.383Z"
    },
    {
      "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": 0.5929293999999999,
      "accuracy_std_dev": 0.011437975198434383,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.0030511959595959596,
      "avg_duration_per_task_ms": 648893.6383838383,
      "avg_output_tokens_per_task": 13498.415151515152,
      "last_run_timestamp": "2026-10-02T16:16:46.227Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-2.5-flash-lite",
      "model_slug": "google/gemini-2.5-flash-lite",
      "model_name": "Google: Gemini 2.5 Flash Lite",
      "provider_name": "Google AI Studio",
      "accuracy": 0.591751,
      "accuracy_std_dev": 0.006627755628164546,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.017204867003367003,
      "avg_duration_per_task_ms": 118615.4696969697,
      "avg_output_tokens_per_task": 42873.59259259259,
      "last_run_timestamp": "2026-10-01T19:14:48.696Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-14b-04-28",
      "model_slug": "qwen/qwen3-14b",
      "model_name": "Qwen: Qwen3 14B",
      "provider_name": "auto-routing",
      "accuracy": 0.591751,
      "accuracy_std_dev": 0.0025249999999999995,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.0029208080808080807,
      "avg_duration_per_task_ms": 521963.51515151514,
      "avg_output_tokens_per_task": 11688.588383838383,
      "last_run_timestamp": "2026-10-01T16:32:22.405Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "CoreWeave",
      "accuracy": 0.5872781331360947,
      "accuracy_std_dev": 0.004687499999999956,
      "run_count": 2,
      "total_questions": 338,
      "avg_cost_per_task_usd": 0.09507673668639052,
      "avg_duration_per_task_ms": 191708.50591715975,
      "avg_output_tokens_per_task": 21455.86094674556,
      "last_run_timestamp": "2026-08-01T18:26:20.700Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-32b-04-28",
      "model_slug": "qwen/qwen3-32b",
      "model_name": "Qwen: Qwen3 32B",
      "provider_name": "Tenstorrent",
      "accuracy": 0.58642,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 162,
      "avg_cost_per_task_usd": 0.002223469135802469,
      "avg_duration_per_task_ms": 407240.81481481483,
      "avg_output_tokens_per_task": 10928.734567901234,
      "last_run_timestamp": "2026-10-03T09:17:20.735Z"
    },
    {
      "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": 0.5858372098049551,
      "accuracy_std_dev": 0.06404439817310623,
      "run_count": 11,
      "total_questions": 1897,
      "avg_cost_per_task_usd": 0.008970768581971532,
      "avg_duration_per_task_ms": 524600.9272535583,
      "avg_output_tokens_per_task": 44661.65946230891,
      "last_run_timestamp": "2026-10-03T17:04:03.802Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Novita",
      "accuracy": 0.5839199654510556,
      "accuracy_std_dev": 0.014881331030858787,
      "run_count": 8,
      "total_questions": 1563,
      "avg_cost_per_task_usd": 0.006653753678822777,
      "avg_duration_per_task_ms": 744303.1906589891,
      "avg_output_tokens_per_task": 44099.619321817016,
      "last_run_timestamp": "2026-10-03T16:12:07.492Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Cerebras",
      "accuracy": 0.5797561250000001,
      "accuracy_std_dev": 0.01826692000884044,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.010468239898989899,
      "avg_duration_per_task_ms": 11933.318813131313,
      "avg_output_tokens_per_task": 13503.182449494949,
      "last_run_timestamp": "2026-10-03T17:02:12.175Z"
    },
    {
      "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": 0.5767120715096482,
      "accuracy_std_dev": 0.20359371905450435,
      "run_count": 9,
      "total_questions": 1762,
      "avg_cost_per_task_usd": 0.023737154370034052,
      "avg_duration_per_task_ms": 751908.8337116913,
      "avg_output_tokens_per_task": 33435.202610669694,
      "last_run_timestamp": "2026-10-03T17:20:14.386Z"
    },
    {
      "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": 0.5709477142857142,
      "accuracy_std_dev": 0.027047512796739667,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.004546116883116884,
      "avg_duration_per_task_ms": 351634.9336219336,
      "avg_output_tokens_per_task": 16008.746753246753,
      "last_run_timestamp": "2026-08-29T19:16:45.297Z"
    },
    {
      "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": 0.5686779050064185,
      "accuracy_std_dev": 0.0536708857668662,
      "run_count": 6,
      "total_questions": 779,
      "avg_cost_per_task_usd": 0.053176206675224644,
      "avg_duration_per_task_ms": 695387.1527599486,
      "avg_output_tokens_per_task": 25450.709884467266,
      "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-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "SambaNova",
      "accuracy": 0.5676931173020527,
      "accuracy_std_dev": 0.02908231088142895,
      "run_count": 7,
      "total_questions": 682,
      "avg_cost_per_task_usd": 0.032412470674486805,
      "avg_duration_per_task_ms": 57640.85923753666,
      "avg_output_tokens_per_task": 6927.941348973607,
      "last_run_timestamp": "2026-10-04T00:00:34.891Z"
    },
    {
      "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": 0.5651467231270357,
      "accuracy_std_dev": 0.004022999999999999,
      "run_count": 2,
      "total_questions": 307,
      "avg_cost_per_task_usd": 0.003757250814332247,
      "avg_duration_per_task_ms": 2542703.263843648,
      "avg_output_tokens_per_task": 34539.9674267101,
      "last_run_timestamp": "2026-07-18T16:00:09.483Z"
    },
    {
      "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": 0.5637339187817259,
      "accuracy_std_dev": 0.03764953294383463,
      "run_count": 6,
      "total_questions": 1182,
      "avg_cost_per_task_usd": 0.025736050761421318,
      "avg_duration_per_task_ms": 1231284.9746192894,
      "avg_output_tokens_per_task": 65120.38747884941,
      "last_run_timestamp": "2026-08-30T00:49:28.653Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-2.5-flash-lite",
      "model_slug": "google/gemini-2.5-flash-lite",
      "model_name": "Google: Gemini 2.5 Flash Lite",
      "provider_name": "Google",
      "accuracy": 0.5505053333333333,
      "accuracy_std_dev": 0.016890890559378122,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.017397131313131312,
      "avg_duration_per_task_ms": 129348.1271043771,
      "avg_output_tokens_per_task": 43400.656565656565,
      "last_run_timestamp": "2026-10-01T19:27:01.395Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-2.5-flash-lite",
      "model_slug": "google/gemini-2.5-flash-lite",
      "model_name": "Google: Gemini 2.5 Flash Lite",
      "provider_name": "auto-routing",
      "accuracy": 0.5488216666666668,
      "accuracy_std_dev": 0.006298902038362641,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0171891531986532,
      "avg_duration_per_task_ms": 121387.12121212122,
      "avg_output_tokens_per_task": 42892.78114478115,
      "last_run_timestamp": "2026-10-01T19:15:36.386Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "DigitalOcean",
      "accuracy": 0.5482828260135134,
      "accuracy_std_dev": 0.09041648367539086,
      "run_count": 6,
      "total_questions": 1184,
      "avg_cost_per_task_usd": 0.02739165118243243,
      "avg_duration_per_task_ms": 2548856.001689189,
      "avg_output_tokens_per_task": 22585.01097972973,
      "last_run_timestamp": "2026-08-22T18:19:05.094Z"
    },
    {
      "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": 0.5468087361702126,
      "accuracy_std_dev": 0.04498858864739341,
      "run_count": 8,
      "total_questions": 1410,
      "avg_cost_per_task_usd": 0.016162113475177306,
      "avg_duration_per_task_ms": 706288.4482269504,
      "avg_output_tokens_per_task": 40872.247517730495,
      "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-2024-05-13",
      "model_slug": "openai/gpt-4o-2024-05-13",
      "model_name": "OpenAI: GPT-4o (2024-05-13)",
      "provider_name": "OpenAI",
      "accuracy": 0.542088,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.023943484848484846,
      "avg_duration_per_task_ms": 10793.646464646465,
      "avg_output_tokens_per_task": 1334.611111111111,
      "last_run_timestamp": "2026-08-01T08:53:34.329Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Google",
      "accuracy": 0.542088,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0071059848484848484,
      "avg_duration_per_task_ms": 292481.80303030304,
      "avg_output_tokens_per_task": 36717.48484848485,
      "last_run_timestamp": "2026-07-11T15:34:14.755Z"
    },
    {
      "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": 0.5409653333333333,
      "accuracy_std_dev": 0.009356358206529358,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0018076902356902358,
      "avg_duration_per_task_ms": 84884.97474747474,
      "avg_output_tokens_per_task": 1770.5521885521885,
      "last_run_timestamp": "2026-10-01T19:33:41.234Z"
    },
    {
      "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": 0.5396825971428572,
      "accuracy_std_dev": 0.025001612450040917,
      "run_count": 7,
      "total_questions": 1050,
      "avg_cost_per_task_usd": 0.09786349714285715,
      "avg_duration_per_task_ms": 941441.8342857143,
      "avg_output_tokens_per_task": 27874.138095238097,
      "last_run_timestamp": "2026-08-29T18:52:09.401Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-vl-8b-instruct",
      "model_slug": "qwen/qwen3-vl-8b-instruct",
      "model_name": "Qwen: Qwen3 VL 8B Instruct",
      "provider_name": "Alibaba",
      "accuracy": 0.5342313333333333,
      "accuracy_std_dev": 0.002099515235053612,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.005432905723905724,
      "avg_duration_per_task_ms": 120646.2558922559,
      "avg_output_tokens_per_task": 11731.540404040405,
      "last_run_timestamp": "2026-08-01T17:48:17.532Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-14b-04-28",
      "model_slug": "qwen/qwen3-14b",
      "model_name": "Qwen: Qwen3 14B",
      "provider_name": "Alibaba",
      "accuracy": 0.5331086666666667,
      "accuracy_std_dev": 0.05754091454106565,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.005119237373737374,
      "avg_duration_per_task_ms": 101325.84511784512,
      "avg_output_tokens_per_task": 5420.193602693603,
      "last_run_timestamp": "2026-10-01T16:05:56.758Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Google",
      "accuracy": 0.5309734740834386,
      "accuracy_std_dev": 0.07649629189542459,
      "run_count": 8,
      "total_questions": 1582,
      "avg_cost_per_task_usd": 0.00888451264222503,
      "avg_duration_per_task_ms": 348655.1245259166,
      "avg_output_tokens_per_task": 26392.39317319848,
      "last_run_timestamp": "2026-10-03T16:13:54.660Z"
    },
    {
      "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": 0.522126,
      "accuracy_std_dev": 0.02474951208754282,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.01514130735930736,
      "avg_duration_per_task_ms": 13491.472582972583,
      "avg_output_tokens_per_task": 1320.8239538239538,
      "last_run_timestamp": "2026-10-02T23:17:37.980Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Coder 30B A3B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.5202023333333333,
      "accuracy_std_dev": 0.0023810642345154617,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0035723956228956226,
      "avg_duration_per_task_ms": 90157.08922558922,
      "avg_output_tokens_per_task": 2721.4242424242425,
      "last_run_timestamp": "2026-08-15T08:46:42.897Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4o-2024-08-06",
      "model_slug": "openai/gpt-4o-2024-08-06",
      "model_name": "OpenAI: GPT-4o (2024-08-06)",
      "provider_name": "auto-routing",
      "accuracy": 0.5202019999999999,
      "accuracy_std_dev": 0.0033669999999999534,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.017511833333333334,
      "avg_duration_per_task_ms": 16177.118686868687,
      "avg_output_tokens_per_task": 1559.209595959596,
      "last_run_timestamp": "2026-10-01T16:04:12.703Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Coder 30B A3B Instruct",
      "provider_name": "Alibaba",
      "accuracy": 0.5190796666666666,
      "accuracy_std_dev": 0.008837261503932561,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.004016927609427609,
      "avg_duration_per_task_ms": 28916.838383838385,
      "avg_output_tokens_per_task": 2584.1296296296296,
      "last_run_timestamp": "2026-08-15T09:25:42.974Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4o",
      "model_slug": "openai/gpt-4o",
      "model_name": "OpenAI: GPT-4o",
      "provider_name": "Azure",
      "accuracy": 0.5168355,
      "accuracy_std_dev": 0.005050499999999958,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.017194939393939394,
      "avg_duration_per_task_ms": 13199.247474747475,
      "avg_output_tokens_per_task": 1523.2777777777778,
      "last_run_timestamp": "2026-10-01T16:28:34.286Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Coder 30B A3B Instruct",
      "provider_name": "DigitalOcean",
      "accuracy": 0.5157126666666666,
      "accuracy_std_dev": 0.02305568803763809,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.004710382154882156,
      "avg_duration_per_task_ms": 211036.34006734006,
      "avg_output_tokens_per_task": 2694.722222222222,
      "last_run_timestamp": "2026-08-15T09:11:29.276Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4o-2024-08-06",
      "model_slug": "openai/gpt-4o-2024-08-06",
      "model_name": "OpenAI: GPT-4o (2024-08-06)",
      "provider_name": "Azure",
      "accuracy": 0.5151519999999999,
      "accuracy_std_dev": 0.006733999999999962,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.018041959595959597,
      "avg_duration_per_task_ms": 13498.479797979799,
      "avg_output_tokens_per_task": 1611.5353535353536,
      "last_run_timestamp": "2026-10-01T16:26:08.836Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Coder 30B A3B Instruct",
      "provider_name": "SiliconFlow",
      "accuracy": 0.5134683333333333,
      "accuracy_std_dev": 0.00836132925370652,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0008388316498316498,
      "avg_duration_per_task_ms": 89120.7138047138,
      "avg_output_tokens_per_task": 2792.7188552188554,
      "last_run_timestamp": "2026-08-15T08:49:03.156Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-vl-8b-instruct",
      "model_slug": "qwen/qwen3-vl-8b-instruct",
      "model_name": "Qwen: Qwen3 VL 8B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.513468,
      "accuracy_std_dev": 0,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.006884491582491581,
      "avg_duration_per_task_ms": 176603.3367003367,
      "avg_output_tokens_per_task": 13632.439393939394,
      "last_run_timestamp": "2026-08-01T18:08:24.732Z"
    },
    {
      "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": 0.5095547757619738,
      "accuracy_std_dev": 0.07224117651877167,
      "run_count": 7,
      "total_questions": 1378,
      "avg_cost_per_task_usd": 0.003949532656023222,
      "avg_duration_per_task_ms": 149830.31567489114,
      "avg_output_tokens_per_task": 4238.851959361394,
      "last_run_timestamp": "2026-08-29T20:34:10.117Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4o-2024-05-13",
      "model_slug": "openai/gpt-4o-2024-05-13",
      "model_name": "OpenAI: GPT-4o (2024-05-13)",
      "provider_name": "auto-routing",
      "accuracy": 0.506734,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0237055303030303,
      "avg_duration_per_task_ms": 11478.515151515152,
      "avg_output_tokens_per_task": 1318.7474747474748,
      "last_run_timestamp": "2026-08-01T08:34:23.935Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4o",
      "model_slug": "openai/gpt-4o",
      "model_name": "OpenAI: GPT-4o",
      "provider_name": "auto-routing",
      "accuracy": 0.5056116666666668,
      "accuracy_std_dev": 0.007570922855827376,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.012942734006734005,
      "avg_duration_per_task_ms": 12612.308080808081,
      "avg_output_tokens_per_task": 1100.3198653198654,
      "last_run_timestamp": "2026-10-01T16:03:45.472Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "inclusionai/ling-3.0-flash-fin-20260827",
      "model_slug": "inclusionai/ling-3.0-flash-fin",
      "model_name": "inclusionAI: Ling 3.0 Flash Fin",
      "provider_name": "auto-routing",
      "accuracy": 0.505051,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.009755954545454545,
      "avg_duration_per_task_ms": 700759.0555555555,
      "avg_output_tokens_per_task": 64762.606060606064,
      "last_run_timestamp": "2026-10-01T08:24:17.043Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4o-2024-05-13",
      "model_slug": "openai/gpt-4o-2024-05-13",
      "model_name": "OpenAI: GPT-4o (2024-05-13)",
      "provider_name": "Azure",
      "accuracy": 0.505051,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.02292151515151515,
      "avg_duration_per_task_ms": 12797.353535353535,
      "avg_output_tokens_per_task": 1266.479797979798,
      "last_run_timestamp": "2026-08-01T08:47:47.980Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4.1-nano-2025-04-14",
      "model_slug": "openai/gpt-4.1-nano",
      "model_name": "OpenAI: GPT-4.1 Nano",
      "provider_name": "auto-routing",
      "accuracy": 0.5,
      "accuracy_std_dev": 0.015120700182200542,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0008362104377104377,
      "avg_duration_per_task_ms": 19874.548821548822,
      "avg_output_tokens_per_task": 1898.5521885521885,
      "last_run_timestamp": "2026-10-01T19:13:37.127Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4.1-nano-2025-04-14",
      "model_slug": "openai/gpt-4.1-nano",
      "model_name": "OpenAI: GPT-4.1 Nano",
      "provider_name": "Azure",
      "accuracy": 0.49764320000000006,
      "accuracy_std_dev": 0.01852187333290023,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.0008495191919191919,
      "avg_duration_per_task_ms": 13033.880808080808,
      "avg_output_tokens_per_task": 1851.6353535353535,
      "last_run_timestamp": "2026-10-01T19:12:29.219Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.49668788473767883,
      "accuracy_std_dev": 0.0232226329739954,
      "run_count": 7,
      "total_questions": 1258,
      "avg_cost_per_task_usd": 0.010930109697933227,
      "avg_duration_per_task_ms": 457519.1629570747,
      "avg_output_tokens_per_task": 26233.00158982512,
      "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": "OpenAI",
      "accuracy": 0.4966332500000001,
      "accuracy_std_dev": 0.014773079898501197,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.0008230315656565656,
      "avg_duration_per_task_ms": 17739.39898989899,
      "avg_output_tokens_per_task": 1863.909090909091,
      "last_run_timestamp": "2026-10-02T23:16:50.042Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.7-flash-20260119",
      "model_slug": "z-ai/glm-4.7-flash",
      "model_name": "Z.ai: GLM 4.7 Flash",
      "provider_name": "DeepInfra",
      "accuracy": 0.4962063575624083,
      "accuracy_std_dev": 0.015775167535576604,
      "run_count": 7,
      "total_questions": 1362,
      "avg_cost_per_task_usd": 0.011085070484581497,
      "avg_duration_per_task_ms": 452241.5058737151,
      "avg_output_tokens_per_task": 27649.281938325992,
      "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": "Groq",
      "accuracy": 0.49588466666666664,
      "accuracy_std_dev": 0.01555384748400071,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.0023523799102132437,
      "avg_duration_per_task_ms": 10975.11672278339,
      "avg_output_tokens_per_task": 2330.834455667789,
      "last_run_timestamp": "2026-10-03T16:58:35.370Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-vl-8b-instruct",
      "model_slug": "qwen/qwen3-vl-8b-instruct",
      "model_name": "Qwen: Qwen3 VL 8B Instruct",
      "provider_name": "Parasail",
      "accuracy": 0.4938273950617284,
      "accuracy_std_dev": 0.04192749999999998,
      "run_count": 2,
      "total_questions": 243,
      "avg_cost_per_task_usd": 0.006835131687242798,
      "avg_duration_per_task_ms": 360677.2674897119,
      "avg_output_tokens_per_task": 12627.074074074075,
      "last_run_timestamp": "2026-08-01T18:13:08.069Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "SambaNova",
      "accuracy": 0.48512888888888883,
      "accuracy_std_dev": 0.005722837707999308,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.0023719074074074076,
      "avg_duration_per_task_ms": 13827.23063973064,
      "avg_output_tokens_per_task": 2232.787878787879,
      "last_run_timestamp": "2026-10-03T16:13:49.760Z"
    },
    {
      "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": 0.48292587155963307,
      "accuracy_std_dev": 0.12738947709449225,
      "run_count": 7,
      "total_questions": 1308,
      "avg_cost_per_task_usd": 0.0038326207951070333,
      "avg_duration_per_task_ms": 105139.82721712538,
      "avg_output_tokens_per_task": 3476.9533639143733,
      "last_run_timestamp": "2026-08-29T20:57:58.443Z"
    },
    {
      "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": 0.48291236785162295,
      "accuracy_std_dev": 0.019256962836594973,
      "run_count": 10,
      "total_questions": 1941,
      "avg_cost_per_task_usd": 0.001084483771251932,
      "avg_duration_per_task_ms": 59779.49407521896,
      "avg_output_tokens_per_task": 2180.590932509016,
      "last_run_timestamp": "2026-10-03T16:56:42.850Z"
    },
    {
      "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": 0.481818,
      "accuracy_std_dev": 0.3073381612185509,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.019882339393939394,
      "avg_duration_per_task_ms": 252178.03333333333,
      "avg_output_tokens_per_task": 21432.7,
      "last_run_timestamp": "2026-08-15T15:24:35.389Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "Novita",
      "accuracy": 0.47763705063291145,
      "accuracy_std_dev": 0.016207885580790598,
      "run_count": 8,
      "total_questions": 1580,
      "avg_cost_per_task_usd": 0.0012874082278481011,
      "avg_duration_per_task_ms": 94391.31139240507,
      "avg_output_tokens_per_task": 2946.448101265823,
      "last_run_timestamp": "2026-10-03T17:16:32.671Z"
    },
    {
      "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": 0.4772725,
      "accuracy_std_dev": 0.0074500120973056095,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.002467788720538721,
      "avg_duration_per_task_ms": 42647.26262626263,
      "avg_output_tokens_per_task": 2666.861952861953,
      "last_run_timestamp": "2026-10-03T16:39:56.573Z"
    },
    {
      "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": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "Cerebras",
      "accuracy": 0.47643079999999993,
      "accuracy_std_dev": 0.009097059028059564,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.048748201010101015,
      "avg_duration_per_task_ms": 46893.51111111111,
      "avg_output_tokens_per_task": 17089.750505050506,
      "last_run_timestamp": "2026-08-15T14:46:10.940Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "Together",
      "accuracy": 0.4758696666666667,
      "accuracy_std_dev": 0.012294713037182558,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.0031561644219977557,
      "avg_duration_per_task_ms": 51856.26262626263,
      "avg_output_tokens_per_task": 2228.726150392817,
      "last_run_timestamp": "2026-10-03T17:41:38.624Z"
    },
    {
      "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": 0.4750470884396716,
      "accuracy_std_dev": 0.012160791117742913,
      "run_count": 8,
      "total_questions": 1583,
      "avg_cost_per_task_usd": 0.0013148066961465572,
      "avg_duration_per_task_ms": 47431.649399873655,
      "avg_output_tokens_per_task": 2358.5874921036007,
      "last_run_timestamp": "2026-10-03T16:14:03.229Z"
    },
    {
      "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": 0.473274125,
      "accuracy_std_dev": 0.010392731840540058,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.0014396685606060608,
      "avg_duration_per_task_ms": 66288.69570707071,
      "avg_output_tokens_per_task": 2272.647095959596,
      "last_run_timestamp": "2026-10-03T16:57:23.026Z"
    },
    {
      "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": 0.47210185714285713,
      "accuracy_std_dev": 0.012910961261419602,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.0010585887445887445,
      "avg_duration_per_task_ms": 186335.3593073593,
      "avg_output_tokens_per_task": 2390.782828282828,
      "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": "Crusoe",
      "accuracy": 0.46742483824768327,
      "accuracy_std_dev": 0.006948085005956674,
      "run_count": 6,
      "total_questions": 1187,
      "avg_cost_per_task_usd": 0.002137747262005055,
      "avg_duration_per_task_ms": 34994.02274641955,
      "avg_output_tokens_per_task": 2587.217354675653,
      "last_run_timestamp": "2026-08-29T16:31:42.504Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "WandB",
      "accuracy": 0.46632999999999997,
      "accuracy_std_dev": 0.003366999999999981,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.0023583939393939392,
      "avg_duration_per_task_ms": 36706.02525252525,
      "avg_output_tokens_per_task": 2516.760101010101,
      "last_run_timestamp": "2026-07-18T16:13:31.071Z"
    },
    {
      "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": 0.46632975000000004,
      "accuracy_std_dev": 0.020618712591660506,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.0007949242424242423,
      "avg_duration_per_task_ms": 165495.66098484848,
      "avg_output_tokens_per_task": 2233.563762626263,
      "last_run_timestamp": "2026-10-03T16:57:26.215Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen2.5-vl-72b-instruct",
      "model_slug": "qwen/qwen2.5-vl-72b-instruct",
      "model_name": "Qwen: Qwen2.5 VL 72B Instruct",
      "provider_name": "Parasail",
      "accuracy": 0.462542,
      "accuracy_std_dev": 0.005471000000000004,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.001783989898989899,
      "avg_duration_per_task_ms": 27644.777777777777,
      "avg_output_tokens_per_task": 1397.5252525252524,
      "last_run_timestamp": "2026-09-01T13:03:42.840Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "Ionstream",
      "accuracy": 0.46128619685039374,
      "accuracy_std_dev": 0.075159,
      "run_count": 2,
      "total_questions": 254,
      "avg_cost_per_task_usd": 0.00731248031496063,
      "avg_duration_per_task_ms": 1016477.8543307086,
      "avg_output_tokens_per_task": 20586.740157480315,
      "last_run_timestamp": "2026-08-08T10:11:59.748Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen-2.5-72b-instruct",
      "model_slug": "qwen/qwen-2.5-72b-instruct",
      "model_name": "Qwen2.5 72B Instruct",
      "provider_name": "Novita",
      "accuracy": 0.461279,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0011232171717171719,
      "avg_duration_per_task_ms": 136788.7626262626,
      "avg_output_tokens_per_task": 2036.2272727272727,
      "last_run_timestamp": "2026-10-01T16:27:00.375Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen-2.5-72b-instruct",
      "model_slug": "qwen/qwen-2.5-72b-instruct",
      "model_name": "Qwen2.5 72B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.45959600000000006,
      "accuracy_std_dev": 0.01346799999999998,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.0010278964646464647,
      "avg_duration_per_task_ms": 96379.95959595959,
      "avg_output_tokens_per_task": 1848.388888888889,
      "last_run_timestamp": "2026-10-01T16:47:11.191Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen2.5-vl-72b-instruct",
      "model_slug": "qwen/qwen2.5-vl-72b-instruct",
      "model_name": "Qwen: Qwen2.5 VL 72B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.4444445,
      "accuracy_std_dev": 0.015151500000000012,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.0015239469696969696,
      "avg_duration_per_task_ms": 59255.792929292926,
      "avg_output_tokens_per_task": 1592.9924242424242,
      "last_run_timestamp": "2026-09-01T13:21:35.512Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen2.5-vl-72b-instruct",
      "model_slug": "qwen/qwen2.5-vl-72b-instruct",
      "model_name": "Qwen: Qwen2.5 VL 72B Instruct",
      "provider_name": "Nebius",
      "accuracy": 0.4402355,
      "accuracy_std_dev": 0.002525500000000014,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.0013545,
      "avg_duration_per_task_ms": 68001.7196969697,
      "avg_output_tokens_per_task": 1535.1843434343434,
      "last_run_timestamp": "2026-09-01T12:41:25.955Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4o-mini",
      "model_slug": "openai/gpt-4o-mini",
      "model_name": "OpenAI: GPT-4o-mini",
      "provider_name": "auto-routing",
      "accuracy": 0.4320986666666667,
      "accuracy_std_dev": 0.00781631474687543,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0010060690235690235,
      "avg_duration_per_task_ms": 20918.510101010103,
      "avg_output_tokens_per_task": 1483.5286195286196,
      "last_run_timestamp": "2026-10-01T19:16:07.087Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4o-mini",
      "model_slug": "openai/gpt-4o-mini",
      "model_name": "OpenAI: GPT-4o-mini",
      "provider_name": "Azure",
      "accuracy": 0.4315373333333334,
      "accuracy_std_dev": 0.010588221658469791,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.0010723358585858589,
      "avg_duration_per_task_ms": 21765.176767676767,
      "avg_output_tokens_per_task": 1505.858585858586,
      "last_run_timestamp": "2026-10-01T19:05:18.003Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4o-mini",
      "model_slug": "openai/gpt-4o-mini",
      "model_name": "OpenAI: GPT-4o-mini",
      "provider_name": "OpenAI",
      "accuracy": 0.4305555,
      "accuracy_std_dev": 0.007469952225416168,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.0010262436868686867,
      "avg_duration_per_task_ms": 17602.7702020202,
      "avg_output_tokens_per_task": 1517.9292929292928,
      "last_run_timestamp": "2026-10-02T23:19:57.879Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "DigitalOcean",
      "accuracy": 0.39305708286674135,
      "accuracy_std_dev": 0.0830628121927015,
      "run_count": 5,
      "total_questions": 893,
      "avg_cost_per_task_usd": 0.015135358342665176,
      "avg_duration_per_task_ms": 3112776.771556551,
      "avg_output_tokens_per_task": 15727.527435610302,
      "last_run_timestamp": "2026-08-22T17:37:26.215Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "tencent/hy-mt2-30b-a3b-20260521",
      "model_slug": "tencent/hy-mt2-30b-a3b",
      "model_name": "Tencent: Hy-MT2-30B-A3B",
      "provider_name": "Reka",
      "accuracy": 0.388889,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.00023147979797979797,
      "avg_duration_per_task_ms": 8604.353535353535,
      "avg_output_tokens_per_task": 710.540404040404,
      "last_run_timestamp": "2026-09-24T06:00:03.613Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "amazon/nova-micro-v1",
      "model_slug": "amazon/nova-micro-v1",
      "model_name": "Amazon: Nova Micro 1.0",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.38720499999999997,
      "accuracy_std_dev": 0.010647388881786938,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.00015740101010101008,
      "avg_duration_per_task_ms": 6591.242424242424,
      "avg_output_tokens_per_task": 937.0363636363636,
      "last_run_timestamp": "2026-10-01T19:39:32.831Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-r1-0528",
      "model_slug": "deepseek/deepseek-r1-0528",
      "model_name": "DeepSeek: R1 0528",
      "provider_name": "Novita",
      "accuracy": 0.382155,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.05791383333333333,
      "avg_duration_per_task_ms": 1031309.4747474748,
      "avg_output_tokens_per_task": 22994.131313131315,
      "last_run_timestamp": "2026-10-03T19:39:23.691Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen-2.5-7b-instruct",
      "model_slug": "qwen/qwen-2.5-7b-instruct",
      "model_name": "Qwen: Qwen2.5 7B Instruct",
      "provider_name": "Phala",
      "accuracy": 0.36363599999999996,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.00041437878787878784,
      "avg_duration_per_task_ms": 34981.92424242424,
      "avg_output_tokens_per_task": 1665.6666666666667,
      "last_run_timestamp": "2026-08-01T08:55:19.318Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Reka",
      "accuracy": 0.346801,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.008616373737373737,
      "avg_duration_per_task_ms": 1179181.5303030303,
      "avg_output_tokens_per_task": 32995.232323232325,
      "last_run_timestamp": "2026-10-03T00:19:24.096Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.3422008461538461,
      "accuracy_std_dev": 0.08129102269931164,
      "run_count": 8,
      "total_questions": 1560,
      "avg_cost_per_task_usd": 0.00332221217948718,
      "avg_duration_per_task_ms": 35770.848076923074,
      "avg_output_tokens_per_task": 1391.8583333333333,
      "last_run_timestamp": "2026-10-03T16:11:17.907Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen-2.5-7b-instruct",
      "model_slug": "qwen/qwen-2.5-7b-instruct",
      "model_name": "Qwen: Qwen2.5 7B Instruct",
      "provider_name": "Together",
      "accuracy": 0.3417510000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0007054343434343434,
      "avg_duration_per_task_ms": 14064.868686868687,
      "avg_output_tokens_per_task": 1539,
      "last_run_timestamp": "2026-08-01T08:06:23.600Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Nebius",
      "accuracy": 0.338384,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.06806519191919191,
      "avg_duration_per_task_ms": 227007.78787878787,
      "avg_output_tokens_per_task": 16827.81313131313,
      "last_run_timestamp": "2026-07-11T08:07:17.820Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "mistralai/mistral-nemo",
      "model_slug": "mistralai/mistral-nemo",
      "model_name": "Mistral: Mistral Nemo",
      "provider_name": "Mistral",
      "accuracy": 0.333333,
      "accuracy_std_dev": 0.01683499999999999,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.00014776262626262624,
      "avg_duration_per_task_ms": 9430.126262626263,
      "avg_output_tokens_per_task": 516.290404040404,
      "last_run_timestamp": "2026-07-18T09:09:45.839Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "mistralai/mistral-nemo",
      "model_slug": "mistralai/mistral-nemo",
      "model_name": "Mistral: Mistral Nemo",
      "provider_name": "auto-routing",
      "accuracy": 0.3299665,
      "accuracy_std_dev": 0.008417500000000022,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.00011648232323232323,
      "avg_duration_per_task_ms": 12253.176767676769,
      "avg_output_tokens_per_task": 504.4191919191919,
      "last_run_timestamp": "2026-07-18T09:08:52.364Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen-2.5-7b-instruct",
      "model_slug": "qwen/qwen-2.5-7b-instruct",
      "model_name": "Qwen: Qwen2.5 7B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.328283,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0005318181818181818,
      "avg_duration_per_task_ms": 36388.32323232323,
      "avg_output_tokens_per_task": 2041.050505050505,
      "last_run_timestamp": "2026-08-01T08:36:11.581Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "mistralai/mistral-nemo",
      "model_slug": "mistralai/mistral-nemo",
      "model_name": "Mistral: Mistral Nemo",
      "provider_name": "Novita",
      "accuracy": 0.32468100546448087,
      "accuracy_std_dev": 0.019489000000000006,
      "run_count": 2,
      "total_questions": 366,
      "avg_cost_per_task_usd": 0.00011272131147540984,
      "avg_duration_per_task_ms": 50027.79234972678,
      "avg_output_tokens_per_task": 481.05464480874315,
      "last_run_timestamp": "2026-07-18T08:00:15.379Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "mistralai/mistral-nemo",
      "model_slug": "mistralai/mistral-nemo",
      "model_name": "Mistral: Mistral Nemo",
      "provider_name": "DeepInfra",
      "accuracy": 0.310606,
      "accuracy_std_dev": 0.009259000000000017,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.000030025252525252528,
      "avg_duration_per_task_ms": 13752.287878787878,
      "avg_output_tokens_per_task": 429.8787878787879,
      "last_run_timestamp": "2026-07-18T08:00:15.193Z"
    },
    {
      "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": 0.2895623333333333,
      "accuracy_std_dev": 0.00363692996297092,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.00006115319865319864,
      "avg_duration_per_task_ms": 9548.159932659933,
      "avg_output_tokens_per_task": 578.2474747474747,
      "last_run_timestamp": "2026-10-01T16:00:58.332Z"
    },
    {
      "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": 0.28577437499999997,
      "accuracy_std_dev": 0.01740932382329581,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.0014133573232323234,
      "avg_duration_per_task_ms": 87435.43244949495,
      "avg_output_tokens_per_task": 18401.633838383837,
      "last_run_timestamp": "2026-10-03T17:14:12.203Z"
    },
    {
      "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": 0.2765752857142857,
      "accuracy_std_dev": 0.06033612304585108,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.00880667316017316,
      "avg_duration_per_task_ms": 555602.5541125541,
      "avg_output_tokens_per_task": 21788.924242424244,
      "last_run_timestamp": "2026-10-03T17:50:31.478Z"
    },
    {
      "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": 0.27641208165057063,
      "accuracy_std_dev": 0.013856526257206988,
      "run_count": 6,
      "total_questions": 1139,
      "avg_cost_per_task_usd": 0.007524540825285337,
      "avg_duration_per_task_ms": 300907.412642669,
      "avg_output_tokens_per_task": 33503.4363476734,
      "last_run_timestamp": "2026-10-03T17:43:23.676Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "mistralai/mistral-nemo",
      "model_slug": "mistralai/mistral-nemo",
      "model_name": "Mistral: Mistral Nemo",
      "provider_name": "DekaLLM",
      "accuracy": 0.275862,
      "accuracy_std_dev": 0.01589199999999999,
      "run_count": 2,
      "total_questions": 348,
      "avg_cost_per_task_usd": 0.000023160919540229885,
      "avg_duration_per_task_ms": 13795.669540229885,
      "avg_output_tokens_per_task": 286.0431034482759,
      "last_run_timestamp": "2026-07-18T08:04:03.449Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Coder 30B A3B Instruct",
      "provider_name": "Novita",
      "accuracy": 0.275533,
      "accuracy_std_dev": 0.009152576176501717,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0007405521885521885,
      "avg_duration_per_task_ms": 48360.410774410775,
      "avg_output_tokens_per_task": 2532.1531986531986,
      "last_run_timestamp": "2026-08-15T09:28:07.208Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.1-8b-instruct",
      "model_slug": "meta-llama/llama-3.1-8b-instruct",
      "model_name": "Meta: Llama 3.1 8B Instruct",
      "provider_name": "Groq",
      "accuracy": 0.27409525,
      "accuracy_std_dev": 0.01585954213360209,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.0020373964646464644,
      "avg_duration_per_task_ms": 44270.92613636364,
      "avg_output_tokens_per_task": 25522.758838383837,
      "last_run_timestamp": "2026-10-03T16:55:38.344Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "sao10k/l3-lunaris-8b",
      "model_slug": "sao10k/l3-lunaris-8b",
      "model_name": "Sao10K: Llama 3 8B Lunaris",
      "provider_name": "Novita",
      "accuracy": 0.2739361560283688,
      "accuracy_std_dev": 0.008153796102974918,
      "run_count": 3,
      "total_questions": 564,
      "avg_cost_per_task_usd": 0.00006610283687943263,
      "avg_duration_per_task_ms": 8879.929078014184,
      "avg_output_tokens_per_task": 551.2074468085107,
      "last_run_timestamp": "2026-10-01T16:01:04.877Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "sao10k/l3-lunaris-8b",
      "model_slug": "sao10k/l3-lunaris-8b",
      "model_name": "Sao10K: Llama 3 8B Lunaris",
      "provider_name": "auto-routing",
      "accuracy": 0.273148,
      "accuracy_std_dev": 0.006113657170630353,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.00006268181818181819,
      "avg_duration_per_task_ms": 9553.181818181818,
      "avg_output_tokens_per_task": 586.439393939394,
      "last_run_timestamp": "2026-10-01T20:43:57.173Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.27053954314720813,
      "accuracy_std_dev": 0.018384550162205943,
      "run_count": 9,
      "total_questions": 1773,
      "avg_cost_per_task_usd": 0.0003133981951494642,
      "avg_duration_per_task_ms": 375042.1866892273,
      "avg_output_tokens_per_task": 7590.988155668359,
      "last_run_timestamp": "2026-10-03T16:18:32.756Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "sao10k/l3-lunaris-8b",
      "model_slug": "sao10k/l3-lunaris-8b",
      "model_name": "Sao10K: Llama 3 8B Lunaris",
      "provider_name": "Parasail",
      "accuracy": 0.2700681612244898,
      "accuracy_std_dev": 0.016752437809465226,
      "run_count": 5,
      "total_questions": 980,
      "avg_cost_per_task_usd": 0.00005762346938775511,
      "avg_duration_per_task_ms": 9734.870408163266,
      "avg_output_tokens_per_task": 576.8010204081633,
      "last_run_timestamp": "2026-10-01T20:47:49.128Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.1-8b-instruct",
      "model_slug": "meta-llama/llama-3.1-8b-instruct",
      "model_name": "Meta: Llama 3.1 8B Instruct",
      "provider_name": "WandB",
      "accuracy": 0.25757599999999997,
      "accuracy_std_dev": 0.003366999999999981,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.001321409090909091,
      "avg_duration_per_task_ms": 39858.34595959596,
      "avg_output_tokens_per_task": 5199.055555555556,
      "last_run_timestamp": "2026-07-18T08:26:34.893Z"
    },
    {
      "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": 0.2504677777777778,
      "accuracy_std_dev": 0.01889361411093281,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.00039258585858585853,
      "avg_duration_per_task_ms": 66188.20258136925,
      "avg_output_tokens_per_task": 7528.086419753086,
      "last_run_timestamp": "2026-10-03T16:12:07.494Z"
    },
    {
      "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": 0.23833585714285715,
      "accuracy_std_dev": 0.09800656715959478,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.001642277056277056,
      "avg_duration_per_task_ms": 571156.3896103896,
      "avg_output_tokens_per_task": 5738.018037518037,
      "last_run_timestamp": "2026-08-29T17:16:14.290Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.2-3b-instruct",
      "model_slug": "meta-llama/llama-3.2-3b-instruct",
      "model_name": "Meta: Llama 3.2 3B Instruct",
      "provider_name": "Parasail",
      "accuracy": 0.21734899999999996,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 171,
      "avg_cost_per_task_usd": 0.0003523450292397661,
      "avg_duration_per_task_ms": 72727.70760233919,
      "avg_output_tokens_per_task": 4269.023391812865,
      "last_run_timestamp": "2026-08-01T08:05:56.858Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-chat-v3-0324",
      "model_slug": "deepseek/deepseek-chat-v3-0324",
      "model_name": "DeepSeek: DeepSeek V3 0324",
      "provider_name": "Novita",
      "accuracy": 0.18302071428571431,
      "accuracy_std_dev": 0.10704161330157577,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.00148023088023088,
      "avg_duration_per_task_ms": 47434.371572871576,
      "avg_output_tokens_per_task": 1278.541847041847,
      "last_run_timestamp": "2026-08-29T18:43:09.114Z"
    },
    {
      "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": 0.1780305,
      "accuracy_std_dev": 0.018939499999999998,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.019786199494949495,
      "avg_duration_per_task_ms": 1989363.3636363635,
      "avg_output_tokens_per_task": 109668.40909090909,
      "last_run_timestamp": "2026-10-01T10:42:24.053Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.2-3b-instruct",
      "model_slug": "meta-llama/llama-3.2-3b-instruct",
      "model_name": "Meta: Llama 3.2 3B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.116162,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0001537979797979798,
      "avg_duration_per_task_ms": 29722.909090909092,
      "avg_output_tokens_per_task": 1836.7929292929293,
      "last_run_timestamp": "2026-08-01T09:03:02.970Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.1-flash-lite-image-20260630",
      "model_slug": "google/gemini-3.1-flash-lite-image",
      "model_name": "Google: Nano Banana 2 Lite (Gemini 3.1 Flash Lite Image)",
      "provider_name": "Google AI Studio",
      "accuracy": 0.1161615,
      "accuracy_std_dev": 0.0016835000000000044,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.017850737373737373,
      "avg_duration_per_task_ms": 46853.982323232325,
      "avg_output_tokens_per_task": 11776.444444444445,
      "last_run_timestamp": "2026-10-01T16:00:58.389Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.1-flash-lite-image-20260630",
      "model_slug": "google/gemini-3.1-flash-lite-image",
      "model_name": "Google: Nano Banana 2 Lite (Gemini 3.1 Flash Lite Image)",
      "provider_name": "Google",
      "accuracy": 0.112795,
      "accuracy_std_dev": 0.006733999999999997,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.018038613636363634,
      "avg_duration_per_task_ms": 21125.79292929293,
      "avg_output_tokens_per_task": 11795.222222222223,
      "last_run_timestamp": "2026-10-01T16:00:58.364Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.1-flash-lite-image-20260630",
      "model_slug": "google/gemini-3.1-flash-lite-image",
      "model_name": "Google: Nano Banana 2 Lite (Gemini 3.1 Flash Lite Image)",
      "provider_name": "auto-routing",
      "accuracy": 0.102694,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.017984406565656568,
      "avg_duration_per_task_ms": 21554.590909090908,
      "avg_output_tokens_per_task": 11812.823232323231,
      "last_run_timestamp": "2026-10-01T16:02:13.023Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.060606,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.000040161616161616165,
      "avg_duration_per_task_ms": 3671.1060606060605,
      "avg_output_tokens_per_task": 93.72222222222223,
      "last_run_timestamp": "2026-08-01T08:05:34.450Z"
    },
    {
      "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": 0.002245,
      "accuracy_std_dev": 0.0007933738084913065,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.003512888888888889,
      "avg_duration_per_task_ms": 175891.9090909091,
      "avg_output_tokens_per_task": 3312.4141414141413,
      "last_run_timestamp": "2026-08-22T17:29:06.480Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "SiliconFlow",
      "accuracy": 0.001684,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.0034544065656565656,
      "avg_duration_per_task_ms": 164367.952020202,
      "avg_output_tokens_per_task": 3253.931818181818,
      "last_run_timestamp": "2026-08-15T14:54:58.071Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "BaseTen",
      "accuracy": 0.822917,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 48,
      "avg_cost_per_task_usd": 0.0087319375,
      "avg_duration_per_task_ms": 31283.875,
      "avg_output_tokens_per_task": 9903.5625,
      "last_run_timestamp": "2026-09-26T08:05:12.798Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Baidu",
      "accuracy": 0.806667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.03810672,
      "avg_duration_per_task_ms": 107015.82,
      "avg_output_tokens_per_task": 17830.82,
      "last_run_timestamp": "2026-10-03T08:11:26.752Z"
    },
    {
      "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": 0.8051851111111111,
      "accuracy_std_dev": 0.020068562864143984,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.87284366,
      "avg_duration_per_task_ms": 135007.57555555555,
      "avg_output_tokens_per_task": 5778.728888888889,
      "last_run_timestamp": "2026-10-03T08:06:43.563Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-5-fable-20260609",
      "model_slug": "anthropic/claude-fable-5",
      "model_name": "Anthropic: Claude Fable 5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.8,
      "accuracy_std_dev": 0.033110611481316166,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.7322859599999999,
      "avg_duration_per_task_ms": 146639.07333333333,
      "avg_output_tokens_per_task": 5472.34,
      "last_run_timestamp": "2026-08-15T08:09:14.229Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "Xiaomi",
      "accuracy": 0.8,
      "accuracy_std_dev": 0.006667000000000034,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.053672830000000005,
      "avg_duration_per_task_ms": 853001.35,
      "avg_output_tokens_per_task": 37564.6,
      "last_run_timestamp": "2026-10-03T08:15:57.209Z"
    },
    {
      "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": 0.7985185555555555,
      "accuracy_std_dev": 0.0278934726458882,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.9265346288888889,
      "avg_duration_per_task_ms": 128415.08888888889,
      "avg_output_tokens_per_task": 5765.7555555555555,
      "last_run_timestamp": "2026-10-03T08:17:58.914Z"
    },
    {
      "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": "Claude Platform on AWS",
      "accuracy": 0.7952381428571429,
      "accuracy_std_dev": 0.02038185356374333,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.8750597885714285,
      "avg_duration_per_task_ms": 147569.17142857143,
      "avg_output_tokens_per_task": 5729.828571428571,
      "last_run_timestamp": "2026-10-03T08:21:31.946Z"
    },
    {
      "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": 0.793333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.020424579999999998,
      "avg_duration_per_task_ms": 413179.4,
      "avg_output_tokens_per_task": 22096.3,
      "last_run_timestamp": "2026-10-03T08:38:04.173Z"
    },
    {
      "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": 0.7925926666666666,
      "accuracy_std_dev": 0.01616105838118283,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.8220114666666667,
      "avg_duration_per_task_ms": 137835.87333333332,
      "avg_output_tokens_per_task": 5751.835555555555,
      "last_run_timestamp": "2026-10-03T08:17:58.906Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-fable-5.1-20260831",
      "model_slug": "anthropic/claude-fable-5.1",
      "model_name": "Anthropic: Claude Fable 5.1",
      "provider_name": "auto-routing",
      "accuracy": 0.79166675,
      "accuracy_std_dev": 0.012801703448662584,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.71616581,
      "avg_duration_per_task_ms": 111830.815,
      "avg_output_tokens_per_task": 6900.955,
      "last_run_timestamp": "2026-10-01T08:03:01.529Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-fable-5.1-20260831",
      "model_slug": "anthropic/claude-fable-5.1",
      "model_name": "Anthropic: Claude Fable 5.1",
      "provider_name": "Anthropic",
      "accuracy": 0.7906666,
      "accuracy_std_dev": 0.009977824303925176,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.814870324,
      "avg_duration_per_task_ms": 182126.652,
      "avg_output_tokens_per_task": 7046.908,
      "last_run_timestamp": "2026-10-01T08:05:22.427Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "Parasail",
      "accuracy": 0.7904040101010101,
      "accuracy_std_dev": 0.014478087690972844,
      "run_count": 4,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.07222263636363636,
      "avg_duration_per_task_ms": 476697.67171717173,
      "avg_output_tokens_per_task": 9418.479797979799,
      "last_run_timestamp": "2026-08-01T17:45:43.989Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-5-fable-20260609",
      "model_slug": "anthropic/claude-fable-5",
      "model_name": "Anthropic: Claude Fable 5",
      "provider_name": "Google",
      "accuracy": 0.7896079411764706,
      "accuracy_std_dev": 0.020702777175085597,
      "run_count": 17,
      "total_questions": 850,
      "avg_cost_per_task_usd": 0.9446278929411763,
      "avg_duration_per_task_ms": 97552.96117647059,
      "avg_output_tokens_per_task": 5626.15294117647,
      "last_run_timestamp": "2026-10-03T08:25:13.024Z"
    },
    {
      "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": 0.7883334999999999,
      "accuracy_std_dev": 0.018333500000000003,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.09102503,
      "avg_duration_per_task_ms": 1512645.25,
      "avg_output_tokens_per_task": 43012.63,
      "last_run_timestamp": "2026-10-03T08:06:43.541Z"
    },
    {
      "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": "WandB",
      "accuracy": 0.786667,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.14780512,
      "avg_duration_per_task_ms": 461958.93,
      "avg_output_tokens_per_task": 11472.92,
      "last_run_timestamp": "2026-07-18T08:04:04.764Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "Darkbloom",
      "accuracy": 0.7862320000000002,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 46,
      "avg_cost_per_task_usd": 0.029582717391304348,
      "avg_duration_per_task_ms": 2435445.304347826,
      "avg_output_tokens_per_task": 108721.58695652174,
      "last_run_timestamp": "2026-10-03T16:12:07.492Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Crusoe",
      "accuracy": 0.7859532642140468,
      "accuracy_std_dev": 0.022060806735148097,
      "run_count": 6,
      "total_questions": 299,
      "avg_cost_per_task_usd": 0.028781180602006688,
      "avg_duration_per_task_ms": 346700.6989966555,
      "avg_output_tokens_per_task": 11874.802675585284,
      "last_run_timestamp": "2026-08-29T17:06:21.366Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-opus-5-20260723",
      "model_slug": "anthropic/claude-opus-5",
      "model_name": "Anthropic: Claude Opus 5",
      "provider_name": "auto-routing",
      "accuracy": 0.7859258888888889,
      "accuracy_std_dev": 0.023611001599746027,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.4867972777777778,
      "avg_duration_per_task_ms": 127433.78,
      "avg_output_tokens_per_task": 7132.633333333333,
      "last_run_timestamp": "2026-10-03T08:09:09.780Z"
    },
    {
      "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": 0.7857143728222997,
      "accuracy_std_dev": 0.022636034639976627,
      "run_count": 6,
      "total_questions": 287,
      "avg_cost_per_task_usd": 0.11757857142857142,
      "avg_duration_per_task_ms": 73705.68641114983,
      "avg_output_tokens_per_task": 8241.627177700348,
      "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": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Crusoe",
      "accuracy": 0.785,
      "accuracy_std_dev": 0.009860245661239886,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.08538871499999999,
      "avg_duration_per_task_ms": 41332.83,
      "avg_output_tokens_per_task": 4686.75,
      "last_run_timestamp": "2026-10-03T08:21:29.529Z"
    },
    {
      "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": 0.7847225,
      "accuracy_std_dev": 0.020833500000000005,
      "run_count": 2,
      "total_questions": 96,
      "avg_cost_per_task_usd": 0.11493318749999999,
      "avg_duration_per_task_ms": 147359.25,
      "avg_output_tokens_per_task": 14387.041666666666,
      "last_run_timestamp": "2026-10-01T16:00:58.391Z"
    },
    {
      "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": 0.784471297188755,
      "accuracy_std_dev": 0.02769278737577711,
      "run_count": 5,
      "total_questions": 249,
      "avg_cost_per_task_usd": 0.09832240562248996,
      "avg_duration_per_task_ms": 406639.9036144578,
      "avg_output_tokens_per_task": 17335.586345381525,
      "last_run_timestamp": "2026-09-26T08:23:09.250Z"
    },
    {
      "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": 0.7835159358372455,
      "accuracy_std_dev": 0.020461293571987414,
      "run_count": 13,
      "total_questions": 639,
      "avg_cost_per_task_usd": 0.08511225195618151,
      "avg_duration_per_task_ms": 86820.21752738654,
      "avg_output_tokens_per_task": 4360.830985915493,
      "last_run_timestamp": "2026-10-03T08:39:04.230Z"
    },
    {
      "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": 0.7833639159049359,
      "accuracy_std_dev": 0.011285272864949418,
      "run_count": 11,
      "total_questions": 547,
      "avg_cost_per_task_usd": 0.10846238025594149,
      "avg_duration_per_task_ms": 142469.36563071297,
      "avg_output_tokens_per_task": 11291.52285191956,
      "last_run_timestamp": "2026-10-03T08:33:36.127Z"
    },
    {
      "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": "WandB",
      "accuracy": 0.7833334999999999,
      "accuracy_std_dev": 0.01666650000000003,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.02978465,
      "avg_duration_per_task_ms": 411185.87,
      "avg_output_tokens_per_task": 14644.16,
      "last_run_timestamp": "2026-07-18T08:03:40.888Z"
    },
    {
      "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": "ModelRun",
      "accuracy": 0.7833334999999999,
      "accuracy_std_dev": 0.01666650000000003,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.06834454,
      "avg_duration_per_task_ms": 91367.71,
      "avg_output_tokens_per_task": 11815.81,
      "last_run_timestamp": "2026-07-18T08:04:03.432Z"
    },
    {
      "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": 0.7813334,
      "accuracy_std_dev": 0.022861647653657863,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.016082664,
      "avg_duration_per_task_ms": 178213.516,
      "avg_output_tokens_per_task": 11864.648,
      "last_run_timestamp": "2026-08-08T08:29:14.821Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-opus-5-20260723",
      "model_slug": "anthropic/claude-opus-5",
      "model_name": "Anthropic: Claude Opus 5",
      "provider_name": "Azure",
      "accuracy": 0.7807408333333333,
      "accuracy_std_dev": 0.022758947352268594,
      "run_count": 18,
      "total_questions": 900,
      "avg_cost_per_task_usd": 0.47270033222222213,
      "avg_duration_per_task_ms": 109166.16777777778,
      "avg_output_tokens_per_task": 7230.745555555555,
      "last_run_timestamp": "2026-10-03T08:13:54.015Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "amazon/nova-micro-v1",
      "model_slug": "amazon/nova-micro-v1",
      "model_name": "Amazon: Nova Micro 1.0",
      "provider_name": "auto-routing",
      "accuracy": 0.7800003333333333,
      "accuracy_std_dev": 0.018856180831641284,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 1.2757919799999997,
      "avg_duration_per_task_ms": 102078.39333333333,
      "avg_output_tokens_per_task": 5685.653333333334,
      "last_run_timestamp": "2026-10-01T19:04:24.509Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "Phala",
      "accuracy": 0.7800001249999999,
      "accuracy_std_dev": 0.02538593223833576,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.18096821249999998,
      "avg_duration_per_task_ms": 702726.9025,
      "avg_output_tokens_per_task": 29032.6575,
      "last_run_timestamp": "2026-10-03T16:06:04.623Z"
    },
    {
      "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": 0.78,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.12438624000000001,
      "avg_duration_per_task_ms": 511620.62,
      "avg_output_tokens_per_task": 12374.38,
      "last_run_timestamp": "2026-10-03T13:23:27.233Z"
    },
    {
      "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": 0.78,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.01813584,
      "avg_duration_per_task_ms": 915938.12,
      "avg_output_tokens_per_task": 36481,
      "last_run_timestamp": "2026-10-03T08:17:09.266Z"
    },
    {
      "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": "Phala",
      "accuracy": 0.78,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.12014696000000001,
      "avg_duration_per_task_ms": 363719.42,
      "avg_output_tokens_per_task": 9566.28,
      "last_run_timestamp": "2026-07-18T16:00:09.009Z"
    },
    {
      "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": 0.7790475714285714,
      "accuracy_std_dev": 0.02348347998583042,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.4780597828571429,
      "avg_duration_per_task_ms": 128145.17428571428,
      "avg_output_tokens_per_task": 7892.805714285714,
      "last_run_timestamp": "2026-10-03T08:15:57.205Z"
    },
    {
      "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": 0.7786668,
      "accuracy_std_dev": 0.02508661263223875,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.095489364,
      "avg_duration_per_task_ms": 83534.308,
      "avg_output_tokens_per_task": 4737.02,
      "last_run_timestamp": "2026-10-03T08:24:07.566Z"
    },
    {
      "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": 0.778333375,
      "accuracy_std_dev": 0.020749956493553795,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.1860782475,
      "avg_duration_per_task_ms": 191849.3575,
      "avg_output_tokens_per_task": 9582.0075,
      "last_run_timestamp": "2026-10-03T16:10:17.831Z"
    },
    {
      "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": 0.77833325,
      "accuracy_std_dev": 0.011902275693643617,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.19232103999999997,
      "avg_duration_per_task_ms": 75116.48,
      "avg_output_tokens_per_task": 6244.94,
      "last_run_timestamp": "2026-10-03T08:17:58.904Z"
    },
    {
      "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": 0.777857107142857,
      "accuracy_std_dev": 0.024498453027457068,
      "run_count": 28,
      "total_questions": 1400,
      "avg_cost_per_task_usd": 0.5571781464285714,
      "avg_duration_per_task_ms": 116452.10642857142,
      "avg_output_tokens_per_task": 7231.447142857143,
      "last_run_timestamp": "2026-10-03T08:36:29.351Z"
    },
    {
      "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": "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": 0.7777776666666667,
      "accuracy_std_dev": 0.021998866415239553,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.08069113666666665,
      "avg_duration_per_task_ms": 131032.13666666667,
      "avg_output_tokens_per_task": 4578.13,
      "last_run_timestamp": "2026-10-03T08:38:05.155Z"
    },
    {
      "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": 0.7772839629629629,
      "accuracy_std_dev": 0.02231524453767886,
      "run_count": 27,
      "total_questions": 1350,
      "avg_cost_per_task_usd": 0.4874687607407408,
      "avg_duration_per_task_ms": 125767.80296296296,
      "avg_output_tokens_per_task": 7179.893333333333,
      "last_run_timestamp": "2026-10-03T08:32:03.356Z"
    },
    {
      "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": 0.7770371111111111,
      "accuracy_std_dev": 0.02456751957110331,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.06889805555555557,
      "avg_duration_per_task_ms": 301183.13333333336,
      "avg_output_tokens_per_task": 14999.031111111111,
      "last_run_timestamp": "2026-10-03T16:18:22.812Z"
    },
    {
      "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": 0.776803125730994,
      "accuracy_std_dev": 0.042092438184140796,
      "run_count": 7,
      "total_questions": 342,
      "avg_cost_per_task_usd": 0.1097883888888889,
      "avg_duration_per_task_ms": 2869054.269005848,
      "avg_output_tokens_per_task": 25251.207602339182,
      "last_run_timestamp": "2026-10-03T16:10:49.371Z"
    },
    {
      "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": 0.7766669999999999,
      "accuracy_std_dev": 0.010000000000000009,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.18507052999999998,
      "avg_duration_per_task_ms": 70446.4,
      "avg_output_tokens_per_task": 6214.06,
      "last_run_timestamp": "2026-10-03T08:13:12.599Z"
    },
    {
      "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": 0.7766667,
      "accuracy_std_dev": 0.02955231519542927,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.043956857999999994,
      "avg_duration_per_task_ms": 213316.834,
      "avg_output_tokens_per_task": 9541.228,
      "last_run_timestamp": "2026-10-03T16:15:21.561Z"
    },
    {
      "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": 0.7766666666666666,
      "accuracy_std_dev": 0.011385435794128478,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.18964581,
      "avg_duration_per_task_ms": 65753.47666666667,
      "avg_output_tokens_per_task": 6306.6466666666665,
      "last_run_timestamp": "2026-10-03T08:23:24.811Z"
    },
    {
      "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": 0.7766666666666666,
      "accuracy_std_dev": 0.026874233841275463,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.15658131666666666,
      "avg_duration_per_task_ms": 581092.8233333334,
      "avg_output_tokens_per_task": 12459.696666666667,
      "last_run_timestamp": "2026-10-03T08:17:09.274Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "DeepSeek",
      "accuracy": 0.7766665,
      "accuracy_std_dev": 0.019148513142887433,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.008114253333333333,
      "avg_duration_per_task_ms": 195047.99,
      "avg_output_tokens_per_task": 21425.716666666667,
      "last_run_timestamp": "2026-08-08T08:30:38.880Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "InferenceNet",
      "accuracy": 0.7766665,
      "accuracy_std_dev": 0.016666499999999973,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.017346550000000002,
      "avg_duration_per_task_ms": 335231.13,
      "avg_output_tokens_per_task": 20113.88,
      "last_run_timestamp": "2026-10-03T08:17:48.643Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-opus-5-20260723",
      "model_slug": "anthropic/claude-opus-5",
      "model_name": "Anthropic: Claude Opus 5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.7760001,
      "accuracy_std_dev": 0.021333425001391598,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.44147494200000004,
      "avg_duration_per_task_ms": 131608.992,
      "avg_output_tokens_per_task": 7283.092,
      "last_run_timestamp": "2026-10-03T08:36:29.563Z"
    },
    {
      "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": 0.776,
      "accuracy_std_dev": 0.012364752775530953,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.074485596,
      "avg_duration_per_task_ms": 47627.076,
      "avg_output_tokens_per_task": 4707.808,
      "last_run_timestamp": "2026-10-03T08:37:28.530Z"
    },
    {
      "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": 0.7755376129032258,
      "accuracy_std_dev": 0.01731650665001458,
      "run_count": 5,
      "total_questions": 248,
      "avg_cost_per_task_usd": 0.1579919798387097,
      "avg_duration_per_task_ms": 333040.58467741933,
      "avg_output_tokens_per_task": 20363.45564516129,
      "last_run_timestamp": "2026-10-03T08:32:03.329Z"
    },
    {
      "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": 0.775,
      "accuracy_std_dev": 0.009860245661239886,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.029526945,
      "avg_duration_per_task_ms": 114391.885,
      "avg_output_tokens_per_task": 11970.57,
      "last_run_timestamp": "2026-10-03T16:15:04.036Z"
    },
    {
      "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": 0.7748298673469387,
      "accuracy_std_dev": 0.0364724490240236,
      "run_count": 10,
      "total_questions": 490,
      "avg_cost_per_task_usd": 0.08125569795918368,
      "avg_duration_per_task_ms": 1108829.7081632654,
      "avg_output_tokens_per_task": 26616.873469387756,
      "last_run_timestamp": "2026-10-03T16:05:34.190Z"
    },
    {
      "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": 0.7748147777777777,
      "accuracy_std_dev": 0.029275825597155435,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.023741562222222224,
      "avg_duration_per_task_ms": 304660.43555555557,
      "avg_output_tokens_per_task": 11954.117777777778,
      "last_run_timestamp": "2026-10-03T16:14:24.154Z"
    },
    {
      "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": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Cloudflare",
      "accuracy": 0.7744441666666667,
      "accuracy_std_dev": 0.015112691911289488,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.09511451000000001,
      "avg_duration_per_task_ms": 133903.96,
      "avg_output_tokens_per_task": 4573.283333333334,
      "last_run_timestamp": "2026-10-03T08:23:53.940Z"
    },
    {
      "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": 0.7744105858585859,
      "accuracy_std_dev": 0.019115499999999952,
      "run_count": 2,
      "total_questions": 99,
      "avg_cost_per_task_usd": 0.24473854545454546,
      "avg_duration_per_task_ms": 51240.79797979798,
      "avg_output_tokens_per_task": 4631.454545454545,
      "last_run_timestamp": "2026-10-03T08:37:57.937Z"
    },
    {
      "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": 0.7742857142857142,
      "accuracy_std_dev": 0.017612503291811795,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.12337239714285714,
      "avg_duration_per_task_ms": 82299.41428571429,
      "avg_output_tokens_per_task": 4508.834285714286,
      "last_run_timestamp": "2026-10-03T08:19:27.959Z"
    },
    {
      "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": "anthropic/claude-opus-5-20260723",
      "model_slug": "anthropic/claude-opus-5",
      "model_name": "Anthropic: Claude Opus 5",
      "provider_name": "Anthropic",
      "accuracy": 0.7733333333333333,
      "accuracy_std_dev": 0.02914423052490644,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.4720031333333333,
      "avg_duration_per_task_ms": 115130.24444444444,
      "avg_output_tokens_per_task": 7206.9,
      "last_run_timestamp": "2026-10-03T08:16:37.919Z"
    },
    {
      "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": 0.77333325,
      "accuracy_std_dev": 0.017638436566794497,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.4344371716666668,
      "avg_duration_per_task_ms": 95528.38,
      "avg_output_tokens_per_task": 5662.913333333333,
      "last_run_timestamp": "2026-10-03T08:06:28.791Z"
    },
    {
      "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": 0.7733332,
      "accuracy_std_dev": 0.004216264811417815,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.007032796,
      "avg_duration_per_task_ms": 161320.244,
      "avg_output_tokens_per_task": 16830.132,
      "last_run_timestamp": "2026-08-08T08:27:33.509Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-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": "stepfun/step-3.7-flash-20260528",
      "model_slug": "stepfun/step-3.7-flash",
      "model_name": "StepFun: Step 3.7 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.773333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.02012318,
      "avg_duration_per_task_ms": 229027.46,
      "avg_output_tokens_per_task": 10940.04,
      "last_run_timestamp": "2026-07-18T08:04:46.147Z"
    },
    {
      "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-4.8-opus-20260528",
      "model_slug": "anthropic/claude-opus-4.8",
      "model_name": "Anthropic: Claude Opus 4.8",
      "provider_name": "Azure",
      "accuracy": 0.7729411176470589,
      "accuracy_std_dev": 0.02923599978246472,
      "run_count": 17,
      "total_questions": 850,
      "avg_cost_per_task_usd": 0.5054271811764706,
      "avg_duration_per_task_ms": 128787.48470588235,
      "avg_output_tokens_per_task": 7945.302352941177,
      "last_run_timestamp": "2026-10-03T08:15:30.352Z"
    },
    {
      "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": 0.77277775,
      "accuracy_std_dev": 0.019333070721456368,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.4942197533333333,
      "avg_duration_per_task_ms": 128626.26333333334,
      "avg_output_tokens_per_task": 8065.555,
      "last_run_timestamp": "2026-10-03T08:20:22.745Z"
    },
    {
      "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": 0.7727776666666667,
      "accuracy_std_dev": 0.03324520758278134,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.11460699333333334,
      "avg_duration_per_task_ms": 172737.43666666668,
      "avg_output_tokens_per_task": 9388.596666666666,
      "last_run_timestamp": "2026-10-03T16:16:11.149Z"
    },
    {
      "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": 0.7726666,
      "accuracy_std_dev": 0.029544518575194266,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.222121268,
      "avg_duration_per_task_ms": 606766.992,
      "avg_output_tokens_per_task": 12995.352,
      "last_run_timestamp": "2026-10-03T08:11:04.620Z"
    },
    {
      "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": 0.7722221666666667,
      "accuracy_std_dev": 0.0225805089137857,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.009277203333333336,
      "avg_duration_per_task_ms": 211225.32333333333,
      "avg_output_tokens_per_task": 4258.733333333334,
      "last_run_timestamp": "2026-10-03T08:53:37.283Z"
    },
    {
      "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": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7718518888888889,
      "accuracy_std_dev": 0.028765211996609332,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.09609024444444446,
      "avg_duration_per_task_ms": 233464.92444444445,
      "avg_output_tokens_per_task": 11786.324444444444,
      "last_run_timestamp": "2026-10-03T16:04:46.698Z"
    },
    {
      "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": 0.7718518888888889,
      "accuracy_std_dev": 0.020313041215733758,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.1761100022222222,
      "avg_duration_per_task_ms": 382781.3977777778,
      "avg_output_tokens_per_task": 27598.153333333332,
      "last_run_timestamp": "2026-10-03T16:19:10.407Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.7-opus-20260416",
      "model_slug": "anthropic/claude-opus-4.7",
      "model_name": "Anthropic: Claude Opus 4.7",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7718518333333333,
      "accuracy_std_dev": 0.03201514259473338,
      "run_count": 18,
      "total_questions": 900,
      "avg_cost_per_task_usd": 0.43984256000000005,
      "avg_duration_per_task_ms": 80156.35111111111,
      "avg_output_tokens_per_task": 5674.794444444445,
      "last_run_timestamp": "2026-10-03T08:11:53.173Z"
    },
    {
      "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": 0.77166675,
      "accuracy_std_dev": 0.027233420032884263,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.08857231500000001,
      "avg_duration_per_task_ms": 275121.275,
      "avg_output_tokens_per_task": 19864.29,
      "last_run_timestamp": "2026-10-03T08:38:07.639Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "AtlasCloud",
      "accuracy": 0.77166675,
      "accuracy_std_dev": 0.02020715632362706,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.09734501,
      "avg_duration_per_task_ms": 129281.66,
      "avg_output_tokens_per_task": 5411.39,
      "last_run_timestamp": "2026-10-03T08:23:24.809Z"
    },
    {
      "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": 0.77166675,
      "accuracy_std_dev": 0.009860203404975986,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.11018350999999998,
      "avg_duration_per_task_ms": 71664.89,
      "avg_output_tokens_per_task": 4808.815,
      "last_run_timestamp": "2026-10-03T08:21:31.953Z"
    },
    {
      "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": 0.7711112222222222,
      "accuracy_std_dev": 0.021773400922418973,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.5318819755555556,
      "avg_duration_per_task_ms": 174829.24,
      "avg_output_tokens_per_task": 10576.931111111111,
      "last_run_timestamp": "2026-10-03T08:10:31.433Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Friendli",
      "accuracy": 0.7711111666666667,
      "accuracy_std_dev": 0.016629684897963515,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.06979356333333334,
      "avg_duration_per_task_ms": 35038.676666666666,
      "avg_output_tokens_per_task": 4742.786666666667,
      "last_run_timestamp": "2026-10-03T08:30:16.613Z"
    },
    {
      "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": 0.7711111111111111,
      "accuracy_std_dev": 0.03771242713673055,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.11559718000000001,
      "avg_duration_per_task_ms": 294861.4888888889,
      "avg_output_tokens_per_task": 12304.771111111111,
      "last_run_timestamp": "2026-10-03T16:16:31.766Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.1-20251113",
      "model_slug": "openai/gpt-5.1",
      "model_name": "OpenAI: GPT-5.1",
      "provider_name": "auto-routing",
      "accuracy": 0.771111,
      "accuracy_std_dev": 0.04085521560339636,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.4149127733333333,
      "avg_duration_per_task_ms": 472514.2266666667,
      "avg_output_tokens_per_task": 36470.113333333335,
      "last_run_timestamp": "2026-10-01T16:00:45.071Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "Wafer",
      "accuracy": 0.7706668,
      "accuracy_std_dev": 0.010832125062055002,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.100171232,
      "avg_duration_per_task_ms": 273464.56,
      "avg_output_tokens_per_task": 22455.776,
      "last_run_timestamp": "2026-10-03T08:37:40.434Z"
    },
    {
      "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": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Parasail",
      "accuracy": 0.7704762857142857,
      "accuracy_std_dev": 0.03627188525453795,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.07425714857142857,
      "avg_duration_per_task_ms": 258403.66,
      "avg_output_tokens_per_task": 11728.06,
      "last_run_timestamp": "2026-10-03T08:42:46.753Z"
    },
    {
      "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": 0.7700001666666666,
      "accuracy_std_dev": 0.02168818589629438,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.08111510333333333,
      "avg_duration_per_task_ms": 263647.49333333335,
      "avg_output_tokens_per_task": 13148.976666666667,
      "last_run_timestamp": "2026-10-03T08:16:12.045Z"
    },
    {
      "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": 0.77,
      "accuracy_std_dev": 0.0033330000000000304,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.18774403,
      "avg_duration_per_task_ms": 78191.08,
      "avg_output_tokens_per_task": 6339.04,
      "last_run_timestamp": "2026-10-03T08:11:23.532Z"
    },
    {
      "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": 0.77,
      "accuracy_std_dev": 0.017950332712793934,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.10653246,
      "avg_duration_per_task_ms": 360324.605,
      "avg_output_tokens_per_task": 20634.68,
      "last_run_timestamp": "2026-10-03T08:06:28.791Z"
    },
    {
      "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": "Io Net",
      "accuracy": 0.77,
      "accuracy_std_dev": 0.023332999999999993,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.35428131,
      "avg_duration_per_task_ms": 261243.27,
      "avg_output_tokens_per_task": 8792.41,
      "last_run_timestamp": "2026-07-18T08:04:09.215Z"
    },
    {
      "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": 0.7696296666666667,
      "accuracy_std_dev": 0.03444242845677408,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.04467215555555556,
      "avg_duration_per_task_ms": 206092.15333333332,
      "avg_output_tokens_per_task": 9425.135555555555,
      "last_run_timestamp": "2026-10-03T16:14:17.312Z"
    },
    {
      "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": 0.7696295555555556,
      "accuracy_std_dev": 0.019654008621998214,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.4308894377777778,
      "avg_duration_per_task_ms": 100253.29111111112,
      "avg_output_tokens_per_task": 5601.184444444444,
      "last_run_timestamp": "2026-10-03T08:11:39.222Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Alibaba",
      "accuracy": 0.7695238571428571,
      "accuracy_std_dev": 0.020113036578502115,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.13556214,
      "avg_duration_per_task_ms": 309797.7,
      "avg_output_tokens_per_task": 19536.211428571427,
      "last_run_timestamp": "2026-10-03T08:11:53.201Z"
    },
    {
      "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": 0.7695237142857143,
      "accuracy_std_dev": 0.01326512827695538,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.10013912285714285,
      "avg_duration_per_task_ms": 333739.5742857143,
      "avg_output_tokens_per_task": 20222.222857142857,
      "last_run_timestamp": "2026-10-03T08:17:48.665Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.5-opus-20251124",
      "model_slug": "anthropic/claude-opus-4.5",
      "model_name": "Anthropic: Claude Opus 4.5",
      "provider_name": "Anthropic",
      "accuracy": 0.7694445,
      "accuracy_std_dev": 0.02899006747864978,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.551655145,
      "avg_duration_per_task_ms": 183576.22666666665,
      "avg_output_tokens_per_task": 10538.358333333334,
      "last_run_timestamp": "2026-10-03T08:10:31.444Z"
    },
    {
      "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": 0.76916675,
      "accuracy_std_dev": 0.021327567581360528,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.1697008125,
      "avg_duration_per_task_ms": 726758.755,
      "avg_output_tokens_per_task": 27638.9,
      "last_run_timestamp": "2026-10-03T16:14:18.634Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.5-flash-lite-20260721",
      "model_slug": "google/gemini-3.5-flash-lite",
      "model_name": "Google: Gemini 3.5 Flash Lite",
      "provider_name": "auto-routing",
      "accuracy": 0.7688886666666668,
      "accuracy_std_dev": 0.021998731726069008,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.09783254,
      "avg_duration_per_task_ms": 102292.53333333334,
      "avg_output_tokens_per_task": 23933.02,
      "last_run_timestamp": "2026-10-01T19:04:42.463Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Io Net",
      "accuracy": 0.7688886666666668,
      "accuracy_std_dev": 0.011331263428624758,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.0562409,
      "avg_duration_per_task_ms": 124558.21333333333,
      "avg_output_tokens_per_task": 4508.253333333333,
      "last_run_timestamp": "2026-09-26T08:24:47.398Z"
    },
    {
      "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": 0.768707,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 49,
      "avg_cost_per_task_usd": 0.18590030612244896,
      "avg_duration_per_task_ms": 362695.9183673469,
      "avg_output_tokens_per_task": 6540.102040816327,
      "last_run_timestamp": "2026-09-29T21:00:05.041Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "CoreWeave",
      "accuracy": 0.7684563154362417,
      "accuracy_std_dev": 0.006017520216464235,
      "run_count": 3,
      "total_questions": 149,
      "avg_cost_per_task_usd": 0.1581116510067114,
      "avg_duration_per_task_ms": 537705.5637583892,
      "avg_output_tokens_per_task": 10979.71812080537,
      "last_run_timestamp": "2026-08-08T08:27:49.545Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "BaseTen",
      "accuracy": 0.768421252631579,
      "accuracy_std_dev": 0.0038080000000000336,
      "run_count": 2,
      "total_questions": 95,
      "avg_cost_per_task_usd": 0.06421121052631579,
      "avg_duration_per_task_ms": 69460.44210526315,
      "avg_output_tokens_per_task": 5883.873684210526,
      "last_run_timestamp": "2026-07-18T08:06:26.187Z"
    },
    {
      "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": 0.7683335,
      "accuracy_std_dev": 0.017873086533948198,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.14871092,
      "avg_duration_per_task_ms": 286909.625,
      "avg_output_tokens_per_task": 13559.755,
      "last_run_timestamp": "2026-10-03T16:12:29.902Z"
    },
    {
      "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": 0.7683335,
      "accuracy_std_dev": 0.01190259075789804,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.15200416500000002,
      "avg_duration_per_task_ms": 286481.905,
      "avg_output_tokens_per_task": 19800.085,
      "last_run_timestamp": "2026-10-03T08:10:31.466Z"
    },
    {
      "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": 0.7683331250000001,
      "accuracy_std_dev": 0.02254616560214563,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0784358,
      "avg_duration_per_task_ms": 226865.8275,
      "avg_output_tokens_per_task": 9339.8125,
      "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-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "GMICloud",
      "accuracy": 0.767778,
      "accuracy_std_dev": 0.009557997105391226,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.055267166666666666,
      "avg_duration_per_task_ms": 921928.7333333333,
      "avg_output_tokens_per_task": 38360.306666666664,
      "last_run_timestamp": "2026-10-03T08:15:16.206Z"
    },
    {
      "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": 0.7677777777777778,
      "accuracy_std_dev": 0.03175926811456523,
      "run_count": 18,
      "total_questions": 900,
      "avg_cost_per_task_usd": 0.5635850022222223,
      "avg_duration_per_task_ms": 180702.17222222223,
      "avg_output_tokens_per_task": 10553.803333333333,
      "last_run_timestamp": "2026-10-03T08:12:06.145Z"
    },
    {
      "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": 0.7676188571428572,
      "accuracy_std_dev": 0.04456902618708767,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.13690867142857144,
      "avg_duration_per_task_ms": 207208.99714285714,
      "avg_output_tokens_per_task": 7510.56,
      "last_run_timestamp": "2026-10-03T08:11:26.753Z"
    },
    {
      "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": 0.7673375906040268,
      "accuracy_std_dev": 0.01364674745945795,
      "run_count": 3,
      "total_questions": 149,
      "avg_cost_per_task_usd": 0.02802279865771812,
      "avg_duration_per_task_ms": 90972.77852348993,
      "avg_output_tokens_per_task": 17409.81879194631,
      "last_run_timestamp": "2026-10-03T08:20:48.319Z"
    },
    {
      "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": 0.7668161793721974,
      "accuracy_std_dev": 0.039499788335422885,
      "run_count": 9,
      "total_questions": 446,
      "avg_cost_per_task_usd": 0.028219390134529145,
      "avg_duration_per_task_ms": 483397.735426009,
      "avg_output_tokens_per_task": 13354.013452914798,
      "last_run_timestamp": "2026-10-03T16:10:32.037Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "Novita",
      "accuracy": 0.7666669999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.0337152,
      "avg_duration_per_task_ms": 192245.9,
      "avg_output_tokens_per_task": 17702.16,
      "last_run_timestamp": "2026-09-01T12:41:31.535Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Reka",
      "accuracy": 0.7666669999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.025223939999999997,
      "avg_duration_per_task_ms": 165137.1,
      "avg_output_tokens_per_task": 14240.82,
      "last_run_timestamp": "2026-09-15T00:00:05.846Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Novita",
      "accuracy": 0.7666669999999999,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01215033,
      "avg_duration_per_task_ms": 551566.9,
      "avg_output_tokens_per_task": 23566.32,
      "last_run_timestamp": "2026-10-03T08:18:05.266Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Parasail",
      "accuracy": 0.7666669999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.0912162,
      "avg_duration_per_task_ms": 191018.42,
      "avg_output_tokens_per_task": 9260.76,
      "last_run_timestamp": "2026-08-04T03:12:10.013Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "NextBit",
      "accuracy": 0.76666675,
      "accuracy_std_dev": 0.024267078752653787,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.036940722499999995,
      "avg_duration_per_task_ms": 209470.17,
      "avg_output_tokens_per_task": 17358.2575,
      "last_run_timestamp": "2026-10-03T08:39:48.671Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.1-pro-preview-20260219",
      "model_slug": "google/gemini-3.1-pro-preview",
      "model_name": "Google: Gemini 3.1 Pro Preview",
      "provider_name": "Google",
      "accuracy": 0.7666666666666667,
      "accuracy_std_dev": 0.014401440259764175,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.3460653,
      "avg_duration_per_task_ms": 170899.55333333334,
      "avg_output_tokens_per_task": 12611.153333333334,
      "last_run_timestamp": "2026-10-01T19:04:40.361Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Mistral",
      "accuracy": 0.7666666666666667,
      "accuracy_std_dev": 0.020727473271535585,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.06962397,
      "avg_duration_per_task_ms": 57268.18666666667,
      "avg_output_tokens_per_task": 4623.046666666667,
      "last_run_timestamp": "2026-10-03T08:43:42.470Z"
    },
    {
      "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": 0.7666666153846154,
      "accuracy_std_dev": 0.020085283979679725,
      "run_count": 26,
      "total_questions": 1300,
      "avg_cost_per_task_usd": 0.5291991723076925,
      "avg_duration_per_task_ms": 121619.19230769231,
      "avg_output_tokens_per_task": 8086.011538461538,
      "last_run_timestamp": "2026-10-03T08:16:24.268Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Parasail",
      "accuracy": 0.7666666,
      "accuracy_std_dev": 0.009428137559454682,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.081377272,
      "avg_duration_per_task_ms": 89133.128,
      "avg_output_tokens_per_task": 4543.572,
      "last_run_timestamp": "2026-10-03T08:30:16.616Z"
    },
    {
      "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": 0.7666666,
      "accuracy_std_dev": 0.018856133692780202,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.061894704,
      "avg_duration_per_task_ms": 48771.608,
      "avg_output_tokens_per_task": 4659.024,
      "last_run_timestamp": "2026-10-03T08:19:53.303Z"
    },
    {
      "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": 0.7666665714285714,
      "accuracy_std_dev": 0.01951809905012813,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.08535362285714285,
      "avg_duration_per_task_ms": 334395.88857142854,
      "avg_output_tokens_per_task": 19919.711428571427,
      "last_run_timestamp": "2026-10-03T08:40:08.943Z"
    },
    {
      "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": 0.7661904285714286,
      "accuracy_std_dev": 0.01872337959311179,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.019808859999999998,
      "avg_duration_per_task_ms": 510174.85142857145,
      "avg_output_tokens_per_task": 13353.934285714286,
      "last_run_timestamp": "2026-08-29T16:47:05.637Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7659258888888889,
      "accuracy_std_dev": 0.030541586931935746,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.16052984222222222,
      "avg_duration_per_task_ms": 227188.16666666666,
      "avg_output_tokens_per_task": 9718.513333333334,
      "last_run_timestamp": "2026-10-03T16:12:07.499Z"
    },
    {
      "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": 0.765833375,
      "accuracy_std_dev": 0.020933096139471957,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0119665625,
      "avg_duration_per_task_ms": 276764.21,
      "avg_output_tokens_per_task": 13967.9875,
      "last_run_timestamp": "2026-10-03T16:20:06.645Z"
    },
    {
      "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": 0.7653333999999999,
      "accuracy_std_dev": 0.032496087971323535,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.030119820000000002,
      "avg_duration_per_task_ms": 258326.172,
      "avg_output_tokens_per_task": 18849.548,
      "last_run_timestamp": "2026-10-03T08:13:52.063Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Io Net",
      "accuracy": 0.7653059591836735,
      "accuracy_std_dev": 0.005416499999999991,
      "run_count": 2,
      "total_questions": 98,
      "avg_cost_per_task_usd": 0.02433298979591837,
      "avg_duration_per_task_ms": 682809.275510204,
      "avg_output_tokens_per_task": 13469.224489795919,
      "last_run_timestamp": "2026-08-01T08:05:58.651Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Baidu",
      "accuracy": 0.7651852222222222,
      "accuracy_std_dev": 0.0219489393759333,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.05072059555555555,
      "avg_duration_per_task_ms": 180142.03555555554,
      "avg_output_tokens_per_task": 11751.844444444445,
      "last_run_timestamp": "2026-10-03T16:16:48.795Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Alibaba",
      "accuracy": 0.765,
      "accuracy_std_dev": 0.036930670783239215,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.032793829999999996,
      "avg_duration_per_task_ms": 164299.645,
      "avg_output_tokens_per_task": 15398.3,
      "last_run_timestamp": "2026-10-03T08:19:00.261Z"
    },
    {
      "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": 0.765,
      "accuracy_std_dev": 0.024664481886915866,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0358974525,
      "avg_duration_per_task_ms": 421646.1775,
      "avg_output_tokens_per_task": 13842.23,
      "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": "z-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": 0.765,
      "accuracy_std_dev": 0.028819341500110663,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.011177185,
      "avg_duration_per_task_ms": 267604.5325,
      "avg_output_tokens_per_task": 6349.9575,
      "last_run_timestamp": "2026-10-03T08:39:56.920Z"
    },
    {
      "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": 0.7646153076923077,
      "accuracy_std_dev": 0.017615963863166925,
      "run_count": 13,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.019299193846153844,
      "avg_duration_per_task_ms": 239187.3676923077,
      "avg_output_tokens_per_task": 17117.927692307694,
      "last_run_timestamp": "2026-10-03T08:37:24.250Z"
    },
    {
      "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": 0.7644446666666667,
      "accuracy_std_dev": 0.020122943113658996,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.4976620088888889,
      "avg_duration_per_task_ms": 132296.16222222222,
      "avg_output_tokens_per_task": 8163.442222222222,
      "last_run_timestamp": "2026-10-03T08:17:09.269Z"
    },
    {
      "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": 0.7644445,
      "accuracy_std_dev": 0.021999102116147078,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.06149954333333334,
      "avg_duration_per_task_ms": 120941.48333333334,
      "avg_output_tokens_per_task": 4754.513333333333,
      "last_run_timestamp": "2026-10-03T08:41:33.560Z"
    },
    {
      "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": 0.7644445,
      "accuracy_std_dev": 0.03281078940303024,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.08256609000000001,
      "avg_duration_per_task_ms": 48194.92,
      "avg_output_tokens_per_task": 4847.44,
      "last_run_timestamp": "2026-10-03T08:29:04.086Z"
    },
    {
      "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": 0.7644444444444445,
      "accuracy_std_dev": 0.025141456592255085,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.03985802,
      "avg_duration_per_task_ms": 244206.85555555555,
      "avg_output_tokens_per_task": 12186.455555555556,
      "last_run_timestamp": "2026-10-03T16:07:44.182Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Relace",
      "accuracy": 0.7644443333333334,
      "accuracy_std_dev": 0.00831485358994506,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.029735113333333334,
      "avg_duration_per_task_ms": 54092.806666666664,
      "avg_output_tokens_per_task": 4732.093333333333,
      "last_run_timestamp": "2026-10-03T08:37:59.129Z"
    },
    {
      "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": 0.7641026153846153,
      "accuracy_std_dev": 0.016021001176525294,
      "run_count": 13,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.017410869230769233,
      "avg_duration_per_task_ms": 197270.92,
      "avg_output_tokens_per_task": 19327.204615384617,
      "last_run_timestamp": "2026-10-03T08:29:06.422Z"
    },
    {
      "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": 0.7640449662921348,
      "accuracy_std_dev": 0.01812749711295652,
      "run_count": 9,
      "total_questions": 445,
      "avg_cost_per_task_usd": 0.06108097303370787,
      "avg_duration_per_task_ms": 254446.68764044944,
      "avg_output_tokens_per_task": 11309.25168539326,
      "last_run_timestamp": "2026-10-03T08:11:26.726Z"
    },
    {
      "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": 0.764,
      "accuracy_std_dev": 0.02585427538338681,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.050899012,
      "avg_duration_per_task_ms": 154368.784,
      "avg_output_tokens_per_task": 4603.632,
      "last_run_timestamp": "2026-10-03T08:39:30.463Z"
    },
    {
      "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": 0.7637362417582418,
      "accuracy_std_dev": 0.023909098570340986,
      "run_count": 11,
      "total_questions": 546,
      "avg_cost_per_task_usd": 0.03624780952380952,
      "avg_duration_per_task_ms": 557719.2857142857,
      "avg_output_tokens_per_task": 12893.100732600733,
      "last_run_timestamp": "2026-10-03T16:10:49.376Z"
    },
    {
      "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": 0.7637214417670682,
      "accuracy_std_dev": 0.03933839868627089,
      "run_count": 10,
      "total_questions": 498,
      "avg_cost_per_task_usd": 0.11078959437751006,
      "avg_duration_per_task_ms": 857713.8875502008,
      "avg_output_tokens_per_task": 10189.773092369478,
      "last_run_timestamp": "2026-10-03T16:09:05.689Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "DeepInfra",
      "accuracy": 0.7636985890410959,
      "accuracy_std_dev": 0.021292961012190532,
      "run_count": 3,
      "total_questions": 146,
      "avg_cost_per_task_usd": 0.059730445205479446,
      "avg_duration_per_task_ms": 156883.1780821918,
      "avg_output_tokens_per_task": 9476.835616438357,
      "last_run_timestamp": "2026-10-03T16:18:52.059Z"
    },
    {
      "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": 0.7633335,
      "accuracy_std_dev": 0.015275361427802623,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.028111930000000004,
      "avg_duration_per_task_ms": 550867.43,
      "avg_output_tokens_per_task": 16398.445,
      "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": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "GMICloud",
      "accuracy": 0.7633335,
      "accuracy_std_dev": 0.01666650000000003,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01259681,
      "avg_duration_per_task_ms": 657803.07,
      "avg_output_tokens_per_task": 25182.87,
      "last_run_timestamp": "2026-10-03T08:05:49.905Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "DeepInfra",
      "accuracy": 0.7633335,
      "accuracy_std_dev": 0.027284651770730018,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.049680514999999995,
      "avg_duration_per_task_ms": 101520.115,
      "avg_output_tokens_per_task": 4912.065,
      "last_run_timestamp": "2026-10-03T08:32:05.590Z"
    },
    {
      "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": 0.763333,
      "accuracy_std_dev": 0.010000000000000009,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.02542324,
      "avg_duration_per_task_ms": 138965.85,
      "avg_output_tokens_per_task": 13863.01,
      "last_run_timestamp": "2026-10-03T16:12:10.916Z"
    },
    {
      "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": "WandB",
      "accuracy": 0.763333,
      "accuracy_std_dev": 0.010000000000000009,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.10827190999999999,
      "avg_duration_per_task_ms": 51051.39,
      "avg_output_tokens_per_task": 12753.93,
      "last_run_timestamp": "2026-07-18T08:03:41.319Z"
    },
    {
      "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": 0.763333,
      "accuracy_std_dev": 0.010000000000000009,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.06322074000000001,
      "avg_duration_per_task_ms": 968065,
      "avg_output_tokens_per_task": 43847.04,
      "last_run_timestamp": "2026-10-03T08:20:05.772Z"
    },
    {
      "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": "DeepInfra",
      "accuracy": 0.763333,
      "accuracy_std_dev": 0.010000000000000009,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.043428839999999996,
      "avg_duration_per_task_ms": 343116.57,
      "avg_output_tokens_per_task": 9247.74,
      "last_run_timestamp": "2026-07-18T16:00:10.793Z"
    },
    {
      "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": 0.7631872962962963,
      "accuracy_std_dev": 0.03277405103846979,
      "run_count": 6,
      "total_questions": 297,
      "avg_cost_per_task_usd": 0.08760998316498315,
      "avg_duration_per_task_ms": 291835.16835016833,
      "avg_output_tokens_per_task": 9985.114478114478,
      "last_run_timestamp": "2026-10-03T16:17:43.913Z"
    },
    {
      "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": 0.7631579766081871,
      "accuracy_std_dev": 0.021601269198350176,
      "run_count": 7,
      "total_questions": 342,
      "avg_cost_per_task_usd": 0.025081707602339184,
      "avg_duration_per_task_ms": 429534.44444444444,
      "avg_output_tokens_per_task": 12901.88596491228,
      "last_run_timestamp": "2026-10-03T16:09:19.685Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "NextBit",
      "accuracy": 0.7628571428571429,
      "accuracy_std_dev": 0.023328310442945685,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.11986269428571429,
      "avg_duration_per_task_ms": 415507.14,
      "avg_output_tokens_per_task": 18962.734285714287,
      "last_run_timestamp": "2026-10-03T08:33:31.578Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Novita",
      "accuracy": 0.7627686612903226,
      "accuracy_std_dev": 0.02051209766552413,
      "run_count": 5,
      "total_questions": 248,
      "avg_cost_per_task_usd": 0.02082460887096774,
      "avg_duration_per_task_ms": 113248.77822580645,
      "avg_output_tokens_per_task": 14257.657258064517,
      "last_run_timestamp": "2026-10-03T08:37:24.028Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "CoreWeave",
      "accuracy": 0.7626666,
      "accuracy_std_dev": 0.03227647784749755,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.146569104,
      "avg_duration_per_task_ms": 373419.58,
      "avg_output_tokens_per_task": 24831.152,
      "last_run_timestamp": "2026-08-29T08:26:04.467Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Moonshot AI",
      "accuracy": 0.7626487366071428,
      "accuracy_std_dev": 0.021906600123881723,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.10720718303571429,
      "avg_duration_per_task_ms": 355012.6004464286,
      "avg_output_tokens_per_task": 11827.252232142857,
      "last_run_timestamp": "2026-10-03T16:15:36.234Z"
    },
    {
      "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": 0.7625,
      "accuracy_std_dev": 0.01823840575543818,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.08220987750000001,
      "avg_duration_per_task_ms": 289030.8525,
      "avg_output_tokens_per_task": 15510.735,
      "last_run_timestamp": "2026-10-03T16:22:04.066Z"
    },
    {
      "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": 0.762381,
      "accuracy_std_dev": 0.013179314517184024,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.12107822285714286,
      "avg_duration_per_task_ms": 467340.5485714286,
      "avg_output_tokens_per_task": 21773.494285714285,
      "last_run_timestamp": "2026-10-03T08:54:39.113Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.7622223333333333,
      "accuracy_std_dev": 0.011330936570096763,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.03153908666666666,
      "avg_duration_per_task_ms": 312202.32,
      "avg_output_tokens_per_task": 39820.306666666664,
      "last_run_timestamp": "2026-10-03T08:09:38.144Z"
    },
    {
      "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": 0.7621211818181818,
      "accuracy_std_dev": 0.026372448546363295,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.05366102181818182,
      "avg_duration_per_task_ms": 93039.47818181818,
      "avg_output_tokens_per_task": 4281.278181818182,
      "last_run_timestamp": "2026-10-03T08:43:08.437Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Chutes",
      "accuracy": 0.7619049795918367,
      "accuracy_std_dev": 0.01847199999999999,
      "run_count": 2,
      "total_questions": 98,
      "avg_cost_per_task_usd": 0.1408940612244898,
      "avg_duration_per_task_ms": 345711.98979591834,
      "avg_output_tokens_per_task": 8276.57142857143,
      "last_run_timestamp": "2026-08-01T08:06:00.902Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7619048571428572,
      "accuracy_std_dev": 0.017353196366158268,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.01491982857142857,
      "avg_duration_per_task_ms": 401525.7,
      "avg_output_tokens_per_task": 15852.922857142858,
      "last_run_timestamp": "2026-10-03T16:17:41.694Z"
    },
    {
      "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": 0.7619047142857143,
      "accuracy_std_dev": 0.020067932249624297,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.026417691428571432,
      "avg_duration_per_task_ms": 160543.33428571429,
      "avg_output_tokens_per_task": 8753.874285714286,
      "last_run_timestamp": "2026-10-03T16:17:22.242Z"
    },
    {
      "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": 0.7619047142857143,
      "accuracy_std_dev": 0.03611523004698906,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.07077762857142857,
      "avg_duration_per_task_ms": 309481.20857142855,
      "avg_output_tokens_per_task": 12537.965714285714,
      "last_run_timestamp": "2026-10-03T08:21:29.527Z"
    },
    {
      "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": 0.761666625,
      "accuracy_std_dev": 0.017559268684497514,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.08635762500000001,
      "avg_duration_per_task_ms": 193062.185,
      "avg_output_tokens_per_task": 11589.4775,
      "last_run_timestamp": "2026-10-03T16:12:13.479Z"
    },
    {
      "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": 0.7614815555555555,
      "accuracy_std_dev": 0.026063039543516672,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.12853624,
      "avg_duration_per_task_ms": 270459.68,
      "avg_output_tokens_per_task": 11691.066666666668,
      "last_run_timestamp": "2026-10-03T16:12:29.878Z"
    },
    {
      "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": 0.7613333999999999,
      "accuracy_std_dev": 0.019955635240202208,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.054660096000000005,
      "avg_duration_per_task_ms": 124144.4,
      "avg_output_tokens_per_task": 4448.496,
      "last_run_timestamp": "2026-10-03T08:45:16.467Z"
    },
    {
      "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": 0.7613333999999999,
      "accuracy_std_dev": 0.022469788112930676,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.06195499599999999,
      "avg_duration_per_task_ms": 143749.252,
      "avg_output_tokens_per_task": 6410.492,
      "last_run_timestamp": "2026-10-03T08:22:27.259Z"
    },
    {
      "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": 0.7610256923076923,
      "accuracy_std_dev": 0.021218684472687738,
      "run_count": 26,
      "total_questions": 1300,
      "avg_cost_per_task_usd": 0.589651923076923,
      "avg_duration_per_task_ms": 120803.91923076923,
      "avg_output_tokens_per_task": 8142.596923076923,
      "last_run_timestamp": "2026-10-03T08:38:25.156Z"
    },
    {
      "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": 0.7609522857142857,
      "accuracy_std_dev": 0.0335097115932266,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.43242590285714283,
      "avg_duration_per_task_ms": 94424.92,
      "avg_output_tokens_per_task": 5553.762857142857,
      "last_run_timestamp": "2026-10-03T08:20:09.018Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7607406666666667,
      "accuracy_std_dev": 0.04314159352643342,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.07637585555555555,
      "avg_duration_per_task_ms": 225464.4111111111,
      "avg_output_tokens_per_task": 9525.68,
      "last_run_timestamp": "2026-10-03T16:14:20.292Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "Baidu",
      "accuracy": 0.7607406666666667,
      "accuracy_std_dev": 0.014553267185679581,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.056603166666666656,
      "avg_duration_per_task_ms": 229343.16444444444,
      "avg_output_tokens_per_task": 10057.042222222222,
      "last_run_timestamp": "2026-10-03T16:05:34.170Z"
    },
    {
      "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": 0.7606059090909091,
      "accuracy_std_dev": 0.02517882197928376,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.16901538363636365,
      "avg_duration_per_task_ms": 622317.2672727272,
      "avg_output_tokens_per_task": 38851.54727272727,
      "last_run_timestamp": "2026-10-03T08:17:06.941Z"
    },
    {
      "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": 0.7604938148148149,
      "accuracy_std_dev": 0.020602075859156715,
      "run_count": 27,
      "total_questions": 1350,
      "avg_cost_per_task_usd": 0.656406777037037,
      "avg_duration_per_task_ms": 73567.76444444444,
      "avg_output_tokens_per_task": 5646.465185185185,
      "last_run_timestamp": "2026-10-03T08:42:35.471Z"
    },
    {
      "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": 0.7604588825910932,
      "accuracy_std_dev": 0.019116824324139206,
      "run_count": 5,
      "total_questions": 247,
      "avg_cost_per_task_usd": 0.16499732388663968,
      "avg_duration_per_task_ms": 837460.5141700405,
      "avg_output_tokens_per_task": 31365.983805668016,
      "last_run_timestamp": "2026-10-03T16:08:33.546Z"
    },
    {
      "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": 0.7600502512562815,
      "accuracy_std_dev": 0.021538150907169366,
      "run_count": 8,
      "total_questions": 398,
      "avg_cost_per_task_usd": 0.04400205025125628,
      "avg_duration_per_task_ms": 537418.8140703518,
      "avg_output_tokens_per_task": 22676.286432160803,
      "last_run_timestamp": "2026-10-03T16:15:26.048Z"
    },
    {
      "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": 0.7600001249999999,
      "accuracy_std_dev": 0.03349974295213884,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0116075025,
      "avg_duration_per_task_ms": 155751.2425,
      "avg_output_tokens_per_task": 14262.745,
      "last_run_timestamp": "2026-10-03T16:12:33.521Z"
    },
    {
      "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": 0.7600001111111111,
      "accuracy_std_dev": 0.02012312716720875,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.16286194444444443,
      "avg_duration_per_task_ms": 435512.70222222223,
      "avg_output_tokens_per_task": 9533.81111111111,
      "last_run_timestamp": "2026-10-03T16:16:10.656Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.6-opus-20260205",
      "model_slug": "anthropic/claude-opus-4.6",
      "model_name": "Anthropic: Claude Opus 4.6",
      "provider_name": "Anthropic",
      "accuracy": 0.76,
      "accuracy_std_dev": 0.02735228340559523,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.5190067183333332,
      "avg_duration_per_task_ms": 221047.60166666665,
      "avg_output_tokens_per_task": 10200.993333333334,
      "last_run_timestamp": "2026-10-03T08:20:09.019Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Ionstream",
      "accuracy": 0.76,
      "accuracy_std_dev": 0.0066669999999999785,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01747962,
      "avg_duration_per_task_ms": 589736.28,
      "avg_output_tokens_per_task": 13675.16,
      "last_run_timestamp": "2026-08-01T08:06:01.204Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "StreamLake",
      "accuracy": 0.76,
      "accuracy_std_dev": 0.020870148806369355,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.0067541,
      "avg_duration_per_task_ms": 260092.14,
      "avg_output_tokens_per_task": 15338.358,
      "last_run_timestamp": "2026-10-03T08:38:03.907Z"
    },
    {
      "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-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Azure",
      "accuracy": 0.76,
      "accuracy_std_dev": 0.013984245106547586,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.16696927200000003,
      "avg_duration_per_task_ms": 223419.404,
      "avg_output_tokens_per_task": 11293.668,
      "last_run_timestamp": "2026-10-03T16:10:46.857Z"
    },
    {
      "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": 0.76,
      "accuracy_std_dev": 0.026666999999999996,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.12531742999999998,
      "avg_duration_per_task_ms": 312784.77,
      "avg_output_tokens_per_task": 20756.6,
      "last_run_timestamp": "2026-10-03T08:29:42.404Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Decart",
      "accuracy": 0.76,
      "accuracy_std_dev": 0.01333300000000004,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.015769889999999998,
      "avg_duration_per_task_ms": 1152795.83,
      "avg_output_tokens_per_task": 36264.12,
      "last_run_timestamp": "2026-10-03T08:16:24.271Z"
    },
    {
      "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": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "GMICloud",
      "accuracy": 0.76,
      "accuracy_std_dev": 0.019626323972325197,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.02215426,
      "avg_duration_per_task_ms": 122620.84,
      "avg_output_tokens_per_task": 8751.653333333334,
      "last_run_timestamp": "2026-10-03T16:07:50.330Z"
    },
    {
      "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": "ModelRun",
      "accuracy": 0.76,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.0333696,
      "avg_duration_per_task_ms": 112327.2,
      "avg_output_tokens_per_task": 8636.08,
      "last_run_timestamp": "2026-07-11T08:05:22.288Z"
    },
    {
      "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": 0.76,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.30272262,
      "avg_duration_per_task_ms": 355584.56,
      "avg_output_tokens_per_task": 10613.82,
      "last_run_timestamp": "2026-10-03T18:42:35.851Z"
    },
    {
      "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": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "DekaLLM",
      "accuracy": 0.76,
      "accuracy_std_dev": 0.01333300000000004,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.06203801500000001,
      "avg_duration_per_task_ms": 146531.195,
      "avg_output_tokens_per_task": 14707.395,
      "last_run_timestamp": "2026-10-03T08:15:00.912Z"
    },
    {
      "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": 0.76,
      "accuracy_std_dev": 0.02177351432359967,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.03884535333333333,
      "avg_duration_per_task_ms": 362781.7866666667,
      "avg_output_tokens_per_task": 12784.353333333333,
      "last_run_timestamp": "2026-10-03T08:09:52.969Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "Novita",
      "accuracy": 0.76,
      "accuracy_std_dev": 0.01333300000000004,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.07523408,
      "avg_duration_per_task_ms": 676626.08,
      "avg_output_tokens_per_task": 16285.62,
      "last_run_timestamp": "2026-10-03T08:15:42.490Z"
    },
    {
      "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": "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-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "WandB",
      "accuracy": 0.76,
      "accuracy_std_dev": 0.020000000000000018,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.12663513999999998,
      "avg_duration_per_task_ms": 73292.09,
      "avg_output_tokens_per_task": 9837.08,
      "last_run_timestamp": "2026-07-18T08:04:45.990Z"
    },
    {
      "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": "Baidu",
      "accuracy": 0.76,
      "accuracy_std_dev": 1.1102230246251565e-16,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.05899852666666667,
      "avg_duration_per_task_ms": 72162.04666666666,
      "avg_output_tokens_per_task": 4660.633333333333,
      "last_run_timestamp": "2026-10-03T08:13:12.618Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "SiliconFlow",
      "accuracy": 0.76,
      "accuracy_std_dev": 0.020547857418718854,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.011870755,
      "avg_duration_per_task_ms": 198678.775,
      "avg_output_tokens_per_task": 4474.35,
      "last_run_timestamp": "2026-10-03T08:39:24.758Z"
    },
    {
      "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": 0.76,
      "accuracy_std_dev": 0.026666999999999996,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.014824000000000002,
      "avg_duration_per_task_ms": 335920.77,
      "avg_output_tokens_per_task": 4448.51,
      "last_run_timestamp": "2026-09-12T08:11:26.740Z"
    },
    {
      "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": 0.7599998888888889,
      "accuracy_std_dev": 0.031894974255600735,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.08708899555555556,
      "avg_duration_per_task_ms": 223739.0377777778,
      "avg_output_tokens_per_task": 9920.264444444445,
      "last_run_timestamp": "2026-10-03T16:09:05.689Z"
    },
    {
      "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": "InferenceNet",
      "accuracy": 0.7599998333333334,
      "accuracy_std_dev": 0.01333333333437501,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.03796188999999999,
      "avg_duration_per_task_ms": 83545.53666666667,
      "avg_output_tokens_per_task": 4558.626666666667,
      "last_run_timestamp": "2026-10-03T08:35:08.461Z"
    },
    {
      "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": 0.7595079060402684,
      "accuracy_std_dev": 0.0216087995861768,
      "run_count": 6,
      "total_questions": 298,
      "avg_cost_per_task_usd": 0.1067814966442953,
      "avg_duration_per_task_ms": 41164.906040268455,
      "avg_output_tokens_per_task": 4181.802013422819,
      "last_run_timestamp": "2026-10-03T21:42:23.046Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Sail Research",
      "accuracy": 0.7594443333333334,
      "accuracy_std_dev": 0.015265178644731146,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.008844093333333332,
      "avg_duration_per_task_ms": 361189.67333333334,
      "avg_output_tokens_per_task": 14258.985,
      "last_run_timestamp": "2026-10-03T08:39:28.580Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.5-flash-lite-20260721",
      "model_slug": "google/gemini-3.5-flash-lite",
      "model_name": "Google: Gemini 3.5 Flash Lite",
      "provider_name": "Google",
      "accuracy": 0.7593333,
      "accuracy_std_dev": 0.016452043441773433,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.083344808,
      "avg_duration_per_task_ms": 106115.21,
      "avg_output_tokens_per_task": 23741.31,
      "last_run_timestamp": "2026-10-01T19:05:47.886Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.7592592222222222,
      "accuracy_std_dev": 0.020714267497322178,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.41923416888888887,
      "avg_duration_per_task_ms": 97992.62444444445,
      "avg_output_tokens_per_task": 5633.653333333334,
      "last_run_timestamp": "2026-10-03T08:15:30.343Z"
    },
    {
      "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": 0.75916675,
      "accuracy_std_dev": 0.03187424291159712,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.072068195,
      "avg_duration_per_task_ms": 294630.755,
      "avg_output_tokens_per_task": 11469.0825,
      "last_run_timestamp": "2026-10-03T16:21:33.429Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.6-flash-20260721",
      "model_slug": "google/gemini-3.6-flash",
      "model_name": "Google: Gemini 3.6 Flash",
      "provider_name": "Google",
      "accuracy": 0.7590477142857143,
      "accuracy_std_dev": 0.013997181448055748,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.24529245999999996,
      "avg_duration_per_task_ms": 151053.55142857143,
      "avg_output_tokens_per_task": 23224.54,
      "last_run_timestamp": "2026-10-01T19:04:58.090Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "CoreWeave",
      "accuracy": 0.7590477142857143,
      "accuracy_std_dev": 0.034992667843071215,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.08880086285714285,
      "avg_duration_per_task_ms": 68112.38857142857,
      "avg_output_tokens_per_task": 12261.222857142857,
      "last_run_timestamp": "2026-10-03T16:15:11.803Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Cloudflare",
      "accuracy": 0.7588384343434343,
      "accuracy_std_dev": 0.024611570108132058,
      "run_count": 8,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.20684339646464647,
      "avg_duration_per_task_ms": 493166.0580808081,
      "avg_output_tokens_per_task": 20296.997474747473,
      "last_run_timestamp": "2026-10-03T08:19:53.308Z"
    },
    {
      "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": 0.7586668,
      "accuracy_std_dev": 0.01654634971708262,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.071547176,
      "avg_duration_per_task_ms": 127553.92,
      "avg_output_tokens_per_task": 4587.528,
      "last_run_timestamp": "2026-10-03T09:01:39.121Z"
    },
    {
      "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": 0.7585185555555556,
      "accuracy_std_dev": 0.03232214986356059,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.01401343777777778,
      "avg_duration_per_task_ms": 177576.9088888889,
      "avg_output_tokens_per_task": 14965.488888888889,
      "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": "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": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Novita",
      "accuracy": 0.758333375,
      "accuracy_std_dev": 0.019649939923174698,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.09441886249999999,
      "avg_duration_per_task_ms": 355424.545,
      "avg_output_tokens_per_task": 12862.18,
      "last_run_timestamp": "2026-10-03T16:13:59.491Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.1-20251113",
      "model_slug": "openai/gpt-5.1",
      "model_name": "OpenAI: GPT-5.1",
      "provider_name": "OpenAI",
      "accuracy": 0.7583332500000001,
      "accuracy_std_dev": 0.005527984369324845,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.40697005500000005,
      "avg_duration_per_task_ms": 402356.405,
      "avg_output_tokens_per_task": 35595.7,
      "last_run_timestamp": "2026-10-02T23:12:57.355Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Crusoe",
      "accuracy": 0.7583331666666667,
      "accuracy_std_dev": 0.013844473427047428,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.09591996333333334,
      "avg_duration_per_task_ms": 214673.9,
      "avg_output_tokens_per_task": 9689.54,
      "last_run_timestamp": "2026-08-29T16:31:41.554Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.7580951428571429,
      "accuracy_std_dev": 0.025125616384357164,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.5482415914285714,
      "avg_duration_per_task_ms": 178558.36285714287,
      "avg_output_tokens_per_task": 10788.502857142857,
      "last_run_timestamp": "2026-10-03T08:08:31.999Z"
    },
    {
      "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": 0.758,
      "accuracy_std_dev": 0.03709447398198281,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.079319272,
      "avg_duration_per_task_ms": 620750.224,
      "avg_output_tokens_per_task": 10895.356,
      "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": "qwen/qwen3.5-122b-a10b-20260224",
      "model_slug": "qwen/qwen3.5-122b-a10b",
      "model_name": "Qwen: Qwen3.5-122B-A10B",
      "provider_name": "DeepInfra",
      "accuracy": 0.7579964141414142,
      "accuracy_std_dev": 0.023533015514113357,
      "run_count": 8,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.20713992929292932,
      "avg_duration_per_task_ms": 664979.7727272727,
      "avg_output_tokens_per_task": 53099.67929292929,
      "last_run_timestamp": "2026-09-26T08:05:12.736Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Novita",
      "accuracy": 0.7577778333333334,
      "accuracy_std_dev": 0.024545289459260582,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.026035279999999997,
      "avg_duration_per_task_ms": 169995.66166666665,
      "avg_output_tokens_per_task": 18134.881666666668,
      "last_run_timestamp": "2026-10-03T08:33:35.373Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3-flash-preview-20251217",
      "model_slug": "google/gemini-3-flash-preview",
      "model_name": "Google: Gemini 3 Flash Preview",
      "provider_name": "Google AI Studio",
      "accuracy": 0.7577776666666667,
      "accuracy_std_dev": 0.04629468145358481,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.12420428666666666,
      "avg_duration_per_task_ms": 116279.86,
      "avg_output_tokens_per_task": 19029.046666666665,
      "last_run_timestamp": "2026-10-01T19:05:00.104Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.1-pro-preview-20260219",
      "model_slug": "google/gemini-3.1-pro-preview",
      "model_name": "Google: Gemini 3.1 Pro Preview",
      "provider_name": "auto-routing",
      "accuracy": 0.7577776666666667,
      "accuracy_std_dev": 0.017497908167041637,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.35916456,
      "avg_duration_per_task_ms": 165675.65333333332,
      "avg_output_tokens_per_task": 12758.866666666667,
      "last_run_timestamp": "2026-10-01T19:06:01.702Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "AkashML",
      "accuracy": 0.7576270508474576,
      "accuracy_std_dev": 0.017540967140446465,
      "run_count": 6,
      "total_questions": 295,
      "avg_cost_per_task_usd": 0.11473522711864406,
      "avg_duration_per_task_ms": 198579.0440677966,
      "avg_output_tokens_per_task": 7218.661016949152,
      "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": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "InferenceNet",
      "accuracy": 0.757619,
      "accuracy_std_dev": 0.014553057204587645,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.3625878914285714,
      "avg_duration_per_task_ms": 337462.2485714286,
      "avg_output_tokens_per_task": 10492.708571428571,
      "last_run_timestamp": "2026-10-03T08:37:29.554Z"
    },
    {
      "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": 0.757437664845173,
      "accuracy_std_dev": 0.028522387468022964,
      "run_count": 11,
      "total_questions": 549,
      "avg_cost_per_task_usd": 0.08139836976320584,
      "avg_duration_per_task_ms": 124217.46448087432,
      "avg_output_tokens_per_task": 12954.497267759563,
      "last_run_timestamp": "2026-10-03T16:24:57.626Z"
    },
    {
      "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": 0.7573627951807229,
      "accuracy_std_dev": 0.03546880209832297,
      "run_count": 10,
      "total_questions": 498,
      "avg_cost_per_task_usd": 0.1904817610441767,
      "avg_duration_per_task_ms": 475270.8333333333,
      "avg_output_tokens_per_task": 28088.13855421687,
      "last_run_timestamp": "2026-10-03T16:15:58.694Z"
    },
    {
      "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": "anthropic/claude-4.5-opus-20251124",
      "model_slug": "anthropic/claude-opus-4.5",
      "model_name": "Anthropic: Claude Opus 4.5",
      "provider_name": "Google",
      "accuracy": 0.757037,
      "accuracy_std_dev": 0.028822348820933333,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.5493601955555556,
      "avg_duration_per_task_ms": 192839.5377777778,
      "avg_output_tokens_per_task": 10676.25111111111,
      "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": "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": 0.7566669999999999,
      "accuracy_std_dev": 0.010000000000000009,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.17767661,
      "avg_duration_per_task_ms": 84052.21,
      "avg_output_tokens_per_task": 6226.55,
      "last_run_timestamp": "2026-10-03T08:17:58.907Z"
    },
    {
      "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": 0.7566669999999999,
      "accuracy_std_dev": 0.030000000000000027,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.02875953,
      "avg_duration_per_task_ms": 155469.3,
      "avg_output_tokens_per_task": 15776.48,
      "last_run_timestamp": "2026-10-03T08:38:11.307Z"
    },
    {
      "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": 0.7566667777777778,
      "accuracy_std_dev": 0.031583837092826035,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.07302445555555555,
      "avg_duration_per_task_ms": 400612.2711111111,
      "avg_output_tokens_per_task": 9214.242222222223,
      "last_run_timestamp": "2026-10-03T16:19:32.420Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Reka",
      "accuracy": 0.756525067114094,
      "accuracy_std_dev": 0.028844409691384666,
      "run_count": 9,
      "total_questions": 447,
      "avg_cost_per_task_usd": 0.0700060134228188,
      "avg_duration_per_task_ms": 265527.27293064876,
      "avg_output_tokens_per_task": 11642.608501118568,
      "last_run_timestamp": "2026-10-03T13:30:10.047Z"
    },
    {
      "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": 0.7563528834080718,
      "accuracy_std_dev": 0.046239678716204095,
      "run_count": 9,
      "total_questions": 446,
      "avg_cost_per_task_usd": 0.07944193273542602,
      "avg_duration_per_task_ms": 387708.71524663677,
      "avg_output_tokens_per_task": 12769.977578475336,
      "last_run_timestamp": "2026-10-03T16:20:30.141Z"
    },
    {
      "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": 0.7561905714285714,
      "accuracy_std_dev": 0.0243929472000011,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.1949313942857143,
      "avg_duration_per_task_ms": 124043.37428571428,
      "avg_output_tokens_per_task": 7684.874285714286,
      "last_run_timestamp": "2026-10-03T08:17:09.273Z"
    },
    {
      "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": 0.7561538461538462,
      "accuracy_std_dev": 0.031169392555582855,
      "run_count": 26,
      "total_questions": 1300,
      "avg_cost_per_task_usd": 0.2008702830769231,
      "avg_duration_per_task_ms": 120980.52153846154,
      "avg_output_tokens_per_task": 7797.5623076923075,
      "last_run_timestamp": "2026-10-03T08:16:37.922Z"
    },
    {
      "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": 0.756,
      "accuracy_std_dev": 0.02048314212224288,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.022952103999999997,
      "avg_duration_per_task_ms": 125535.02,
      "avg_output_tokens_per_task": 15581.832,
      "last_run_timestamp": "2026-10-03T08:16:37.919Z"
    },
    {
      "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": 0.756,
      "accuracy_std_dev": 0.021745967699782896,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.006672712000000001,
      "avg_duration_per_task_ms": 144287.16,
      "avg_output_tokens_per_task": 4265.224,
      "last_run_timestamp": "2026-10-03T08:12:06.166Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "BaseTen",
      "accuracy": 0.7557079178082191,
      "accuracy_std_dev": 0.01607886717264481,
      "run_count": 3,
      "total_questions": 146,
      "avg_cost_per_task_usd": 0.2608585205479452,
      "avg_duration_per_task_ms": 125025.47260273973,
      "avg_output_tokens_per_task": 7152.054794520548,
      "last_run_timestamp": "2026-10-03T08:05:09.514Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "DeepInfra",
      "accuracy": 0.7557075582191781,
      "accuracy_std_dev": 0.014317957738642594,
      "run_count": 6,
      "total_questions": 292,
      "avg_cost_per_task_usd": 0.03211151712328767,
      "avg_duration_per_task_ms": 2014288.493150685,
      "avg_output_tokens_per_task": 47550.393835616436,
      "last_run_timestamp": "2026-10-03T16:09:32.725Z"
    },
    {
      "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": 0.7555555555555555,
      "accuracy_std_dev": 0.02266234163693451,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.34417582888888887,
      "avg_duration_per_task_ms": 219931.23555555556,
      "avg_output_tokens_per_task": 11911.208888888888,
      "last_run_timestamp": "2026-10-03T08:23:27.050Z"
    },
    {
      "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": 0.7555555,
      "accuracy_std_dev": 0.03119039105103792,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.0069151533333333326,
      "avg_duration_per_task_ms": 128985.71333333333,
      "avg_output_tokens_per_task": 4275.633333333333,
      "last_run_timestamp": "2026-10-03T08:51:15.407Z"
    },
    {
      "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": 0.755,
      "accuracy_std_dev": 0.009860245661239886,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.07160828000000001,
      "avg_duration_per_task_ms": 331234.47,
      "avg_output_tokens_per_task": 13566.125,
      "last_run_timestamp": "2026-10-03T08:43:45.086Z"
    },
    {
      "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": 0.7549019380804953,
      "accuracy_std_dev": 0.022500075722423546,
      "run_count": 13,
      "total_questions": 646,
      "avg_cost_per_task_usd": 0.020211260061919506,
      "avg_duration_per_task_ms": 487189.79102167184,
      "avg_output_tokens_per_task": 17152.354489164085,
      "last_run_timestamp": "2026-10-03T08:34:10.987Z"
    },
    {
      "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": 0.7546668,
      "accuracy_std_dev": 0.012927338819726197,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.19784133199999998,
      "avg_duration_per_task_ms": 328733.652,
      "avg_output_tokens_per_task": 12530.504,
      "last_run_timestamp": "2026-10-03T08:11:26.748Z"
    },
    {
      "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": 0.7546666,
      "accuracy_std_dev": 0.0308113572151569,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.025647999999999997,
      "avg_duration_per_task_ms": 294521.988,
      "avg_output_tokens_per_task": 13284.328,
      "last_run_timestamp": "2026-10-03T16:12:20.035Z"
    },
    {
      "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": 0.7545098823529411,
      "accuracy_std_dev": 0.035662554927909464,
      "run_count": 17,
      "total_questions": 850,
      "avg_cost_per_task_usd": 0.19809168117647058,
      "avg_duration_per_task_ms": 103146.24235294118,
      "avg_output_tokens_per_task": 7666.7011764705885,
      "last_run_timestamp": "2026-10-03T08:20:09.021Z"
    },
    {
      "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": 0.7545043243243243,
      "accuracy_std_dev": 0.0263378748741976,
      "run_count": 6,
      "total_questions": 296,
      "avg_cost_per_task_usd": 0.023569141891891895,
      "avg_duration_per_task_ms": 316460.4864864865,
      "avg_output_tokens_per_task": 4132.989864864865,
      "last_run_timestamp": "2026-10-03T08:34:09.439Z"
    },
    {
      "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": 0.7544445,
      "accuracy_std_dev": 0.02678215501504689,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.08861642333333332,
      "avg_duration_per_task_ms": 293005.13333333336,
      "avg_output_tokens_per_task": 12023.27,
      "last_run_timestamp": "2026-10-03T08:37:04.257Z"
    },
    {
      "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": 0.7544445,
      "accuracy_std_dev": 0.025067718163340406,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.008447913333333333,
      "avg_duration_per_task_ms": 189202.61333333334,
      "avg_output_tokens_per_task": 4437.176666666666,
      "last_run_timestamp": "2026-10-03T08:32:50.838Z"
    },
    {
      "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": 0.7544443333333334,
      "accuracy_std_dev": 0.02786659578340269,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.008521756666666666,
      "avg_duration_per_task_ms": 190718.40666666668,
      "avg_output_tokens_per_task": 4439.933333333333,
      "last_run_timestamp": "2026-10-03T08:37:24.517Z"
    },
    {
      "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": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "GMICloud",
      "accuracy": 0.7542855714285714,
      "accuracy_std_dev": 0.03175878188495785,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.010189677142857144,
      "avg_duration_per_task_ms": 226644.26285714286,
      "avg_output_tokens_per_task": 15464.714285714286,
      "last_run_timestamp": "2026-10-03T16:05:34.199Z"
    },
    {
      "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": 0.75416675,
      "accuracy_std_dev": 0.02012119874504252,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.052130175,
      "avg_duration_per_task_ms": 124686.1225,
      "avg_output_tokens_per_task": 4528.42,
      "last_run_timestamp": "2026-10-03T13:20:27.251Z"
    },
    {
      "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": 0.7541132888482633,
      "accuracy_std_dev": 0.033056453595789424,
      "run_count": 11,
      "total_questions": 547,
      "avg_cost_per_task_usd": 0.1956822632541134,
      "avg_duration_per_task_ms": 586732.365630713,
      "avg_output_tokens_per_task": 27848.411334552104,
      "last_run_timestamp": "2026-10-03T08:06:43.538Z"
    },
    {
      "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": 0.754074,
      "accuracy_std_dev": 0.039182303332556186,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.19633914666666663,
      "avg_duration_per_task_ms": 104140.85111111111,
      "avg_output_tokens_per_task": 7619.131111111111,
      "last_run_timestamp": "2026-10-03T08:06:43.541Z"
    },
    {
      "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": 0.7540738888888889,
      "accuracy_std_dev": 0.021646890628368387,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.07777800888888889,
      "avg_duration_per_task_ms": 250139.2111111111,
      "avg_output_tokens_per_task": 9830.866666666667,
      "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/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "Parasail",
      "accuracy": 0.754059358166189,
      "accuracy_std_dev": 0.029275519612321662,
      "run_count": 7,
      "total_questions": 349,
      "avg_cost_per_task_usd": 0.16745791404011462,
      "avg_duration_per_task_ms": 461873.38681948424,
      "avg_output_tokens_per_task": 23575.730659025787,
      "last_run_timestamp": "2026-08-29T17:19:03.369Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "CoreWeave",
      "accuracy": 0.754,
      "accuracy_std_dev": 0.021385259722528513,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.022165832,
      "avg_duration_per_task_ms": 128895.106,
      "avg_output_tokens_per_task": 17780.99,
      "last_run_timestamp": "2026-10-03T08:11:39.228Z"
    },
    {
      "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": 0.754,
      "accuracy_std_dev": 0.02048284919633984,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.08895154,
      "avg_duration_per_task_ms": 301180.94,
      "avg_output_tokens_per_task": 9633.452,
      "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": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Morph",
      "accuracy": 0.753914899328859,
      "accuracy_std_dev": 0.040895046741628756,
      "run_count": 9,
      "total_questions": 447,
      "avg_cost_per_task_usd": 0.07723322147651007,
      "avg_duration_per_task_ms": 229349.62863534674,
      "avg_output_tokens_per_task": 14123.306487695749,
      "last_run_timestamp": "2026-10-03T08:50:19.608Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Together",
      "accuracy": 0.7538975456570156,
      "accuracy_std_dev": 0.023333857993043206,
      "run_count": 9,
      "total_questions": 449,
      "avg_cost_per_task_usd": 0.013686612472160357,
      "avg_duration_per_task_ms": 214524.49888641425,
      "avg_output_tokens_per_task": 16219.260579064588,
      "last_run_timestamp": "2026-10-03T08:34:35.934Z"
    },
    {
      "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": 0.7538888333333333,
      "accuracy_std_dev": 0.019140560322838583,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.025353155,
      "avg_duration_per_task_ms": 294929.6716666667,
      "avg_output_tokens_per_task": 17480.86,
      "last_run_timestamp": "2026-10-03T08:34:34.719Z"
    },
    {
      "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": 0.7533335,
      "accuracy_std_dev": 0.00666650000000002,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.06322421,
      "avg_duration_per_task_ms": 228589.91,
      "avg_output_tokens_per_task": 10641.07,
      "last_run_timestamp": "2026-10-03T16:12:33.507Z"
    },
    {
      "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": 0.7533335,
      "accuracy_std_dev": 0.01805358280000584,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.15314664666666666,
      "avg_duration_per_task_ms": 312626.37,
      "avg_output_tokens_per_task": 20191.006666666668,
      "last_run_timestamp": "2026-10-03T08:43:11.974Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "x-ai/grok-4.6-20260810",
      "model_slug": "x-ai/grok-4.6",
      "model_name": "SpaceXAI: Grok 4.6",
      "provider_name": "auto-routing",
      "accuracy": 0.7533335,
      "accuracy_std_dev": 0.00666650000000002,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.33087047999999997,
      "avg_duration_per_task_ms": 241177.17,
      "avg_output_tokens_per_task": 11982.77,
      "last_run_timestamp": "2026-10-01T16:01:02.676Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Venice",
      "accuracy": 0.7533333999999999,
      "accuracy_std_dev": 0.015776240928687683,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.138680804,
      "avg_duration_per_task_ms": 204157.96,
      "avg_output_tokens_per_task": 8340.072,
      "last_run_timestamp": "2026-10-03T08:52:57.109Z"
    },
    {
      "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": 0.7533333333333333,
      "accuracy_std_dev": 0.027755626724844277,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.5350316133333334,
      "avg_duration_per_task_ms": 188978.14666666667,
      "avg_output_tokens_per_task": 10298.875555555556,
      "last_run_timestamp": "2026-10-03T08:09:23.860Z"
    },
    {
      "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": 0.7533333333333333,
      "accuracy_std_dev": 0.02739730544739352,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.19939614444444445,
      "avg_duration_per_task_ms": 129460.44888888889,
      "avg_output_tokens_per_task": 7850.873333333333,
      "last_run_timestamp": "2026-10-03T08:12:20.019Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7533333333333333,
      "accuracy_std_dev": 0.027216688781374985,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.02417726666666667,
      "avg_duration_per_task_ms": 450171.92,
      "avg_output_tokens_per_task": 14030.733333333334,
      "last_run_timestamp": "2026-10-03T08:13:54.014Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3-flash-preview-20251217",
      "model_slug": "google/gemini-3-flash-preview",
      "model_name": "Google: Gemini 3 Flash Preview",
      "provider_name": "Google",
      "accuracy": 0.7533333333333333,
      "accuracy_std_dev": 0.028803034820811656,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.11451421333333334,
      "avg_duration_per_task_ms": 146387.56666666668,
      "avg_output_tokens_per_task": 17546.113333333335,
      "last_run_timestamp": "2026-10-01T19:05:47.885Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Morph",
      "accuracy": 0.7533333333333333,
      "accuracy_std_dev": 0.014401903168987377,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.021082753333333332,
      "avg_duration_per_task_ms": 915986.2133333334,
      "avg_output_tokens_per_task": 13268.026666666667,
      "last_run_timestamp": "2026-08-15T08:15:52.453Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.4-20260305",
      "model_slug": "openai/gpt-5.4",
      "model_name": "OpenAI: GPT-5.4",
      "provider_name": "auto-routing",
      "accuracy": 0.7533333333333333,
      "accuracy_std_dev": 0.005443174461857933,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.29158895333333334,
      "avg_duration_per_task_ms": 171961.5,
      "avg_output_tokens_per_task": 10696.2,
      "last_run_timestamp": "2026-10-01T19:04:36.169Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "Darkbloom",
      "accuracy": 0.7533333333333333,
      "accuracy_std_dev": 0.018856180831641284,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.06088320666666667,
      "avg_duration_per_task_ms": 1112758.12,
      "avg_output_tokens_per_task": 52398.04,
      "last_run_timestamp": "2026-10-03T16:14:15.492Z"
    },
    {
      "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": 0.7533333,
      "accuracy_std_dev": 0.023475765282733617,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.021858692000000002,
      "avg_duration_per_task_ms": 119951.828,
      "avg_output_tokens_per_task": 15133.982,
      "last_run_timestamp": "2026-10-03T08:40:00.936Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Venice",
      "accuracy": 0.7533332500000001,
      "accuracy_std_dev": 0.014907082584714572,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.243283165,
      "avg_duration_per_task_ms": 338011.175,
      "avg_output_tokens_per_task": 17182.585,
      "last_run_timestamp": "2026-10-03T08:32:03.359Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3-flash-preview-20251217",
      "model_slug": "google/gemini-3-flash-preview",
      "model_name": "Google: Gemini 3 Flash Preview",
      "provider_name": "auto-routing",
      "accuracy": 0.753333,
      "accuracy_std_dev": 0.032659863237109024,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.12209446666666665,
      "avg_duration_per_task_ms": 147223.81333333332,
      "avg_output_tokens_per_task": 19831.306666666667,
      "last_run_timestamp": "2026-10-01T19:04:26.702Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.8-flash-20260902",
      "model_slug": "google/gemini-3.8-flash",
      "model_name": "Google: Gemini 3.8 Flash",
      "provider_name": "Google",
      "accuracy": 0.753333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.2515843,
      "avg_duration_per_task_ms": 421006.06,
      "avg_output_tokens_per_task": 32340.48,
      "last_run_timestamp": "2026-10-01T08:00:24.867Z"
    },
    {
      "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": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "GMICloud",
      "accuracy": 0.753333,
      "accuracy_std_dev": 0.020000000000000018,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.25651981,
      "avg_duration_per_task_ms": 445171.22,
      "avg_output_tokens_per_task": 25359.93,
      "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": "stepfun/step-3.7-flash-20260528",
      "model_slug": "stepfun/step-3.7-flash",
      "model_name": "StepFun: Step 3.7 Flash",
      "provider_name": "DeepInfra",
      "accuracy": 0.753333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.0196027,
      "avg_duration_per_task_ms": 76385.02,
      "avg_output_tokens_per_task": 9950,
      "last_run_timestamp": "2026-07-18T08:04:02.230Z"
    },
    {
      "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": 0.753333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.07618375999999999,
      "avg_duration_per_task_ms": 574082.66,
      "avg_output_tokens_per_task": 14524.72,
      "last_run_timestamp": "2026-10-03T08:19:27.962Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "CoreWeave",
      "accuracy": 0.753333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.10139794,
      "avg_duration_per_task_ms": 119629.34,
      "avg_output_tokens_per_task": 9506.02,
      "last_run_timestamp": "2026-08-01T17:46:04.356Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Ionstream",
      "accuracy": 0.753333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.09723594,
      "avg_duration_per_task_ms": 178772.1,
      "avg_output_tokens_per_task": 8310.58,
      "last_run_timestamp": "2026-08-01T08:05:58.203Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "DeepInfra",
      "accuracy": 0.752923926900585,
      "accuracy_std_dev": 0.02558177145020674,
      "run_count": 7,
      "total_questions": 342,
      "avg_cost_per_task_usd": 0.006783909356725145,
      "avg_duration_per_task_ms": 266624.59064327483,
      "avg_output_tokens_per_task": 11932.982456140351,
      "last_run_timestamp": "2026-08-29T16:33:29.856Z"
    },
    {
      "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": 0.7528571428571429,
      "accuracy_std_dev": 0.043949114567478,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.1576592571428571,
      "avg_duration_per_task_ms": 349242.77428571426,
      "avg_output_tokens_per_task": 22490.9,
      "last_run_timestamp": "2026-10-03T08:08:18.114Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.5-20260423",
      "model_slug": "openai/gpt-5.5",
      "model_name": "OpenAI: GPT-5.5",
      "provider_name": "Azure",
      "accuracy": 0.7525925555555556,
      "accuracy_std_dev": 0.03021638651133634,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.5157013133333334,
      "avg_duration_per_task_ms": 178548.1511111111,
      "avg_output_tokens_per_task": 7818.084444444445,
      "last_run_timestamp": "2026-10-01T19:05:47.886Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-sonnet-5-20260630",
      "model_slug": "anthropic/claude-sonnet-5",
      "model_name": "Anthropic: Claude Sonnet 5",
      "provider_name": "Google",
      "accuracy": 0.7523456666666667,
      "accuracy_std_dev": 0.03581868589830568,
      "run_count": 27,
      "total_questions": 1350,
      "avg_cost_per_task_usd": 0.22564089555555555,
      "avg_duration_per_task_ms": 98845.40814814816,
      "avg_output_tokens_per_task": 7557.676296296297,
      "last_run_timestamp": "2026-10-03T08:43:57.161Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.1-20251113",
      "model_slug": "openai/gpt-5.1",
      "model_name": "OpenAI: GPT-5.1",
      "provider_name": "Azure",
      "accuracy": 0.7522221666666666,
      "accuracy_std_dev": 0.01511278994115765,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.44058496666666663,
      "avg_duration_per_task_ms": 458389.25,
      "avg_output_tokens_per_task": 35681.92333333333,
      "last_run_timestamp": "2026-10-01T16:01:13.647Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Alibaba",
      "accuracy": 0.752,
      "accuracy_std_dev": 0.015434478002187193,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.036405735999999994,
      "avg_duration_per_task_ms": 202702.796,
      "avg_output_tokens_per_task": 18338.232,
      "last_run_timestamp": "2026-09-26T08:01:18.110Z"
    },
    {
      "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": 0.7519045714285715,
      "accuracy_std_dev": 0.03226166192184128,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.016922877142857146,
      "avg_duration_per_task_ms": 77003.04285714286,
      "avg_output_tokens_per_task": 4241.571428571428,
      "last_run_timestamp": "2026-10-03T08:48:31.438Z"
    },
    {
      "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": 0.751852,
      "accuracy_std_dev": 0.03366914875008805,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.021374442222222224,
      "avg_duration_per_task_ms": 173854.87555555557,
      "avg_output_tokens_per_task": 14192.557777777778,
      "last_run_timestamp": "2026-10-03T08:32:03.332Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "Chutes",
      "accuracy": 0.7516778389261746,
      "accuracy_std_dev": 0.008511827118127408,
      "run_count": 3,
      "total_questions": 149,
      "avg_cost_per_task_usd": 0.046181536912751675,
      "avg_duration_per_task_ms": 406586.1208053691,
      "avg_output_tokens_per_task": 7644.885906040268,
      "last_run_timestamp": "2026-07-25T08:07:26.285Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "CoreWeave",
      "accuracy": 0.7511113333333332,
      "accuracy_std_dev": 0.014740562999952068,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.14565638333333333,
      "avg_duration_per_task_ms": 139837.15,
      "avg_output_tokens_per_task": 20029.903333333332,
      "last_run_timestamp": "2026-10-03T08:32:03.353Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.5-20260423",
      "model_slug": "openai/gpt-5.5",
      "model_name": "OpenAI: GPT-5.5",
      "provider_name": "auto-routing",
      "accuracy": 0.7511113333333332,
      "accuracy_std_dev": 0.02199873172606905,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.48754586,
      "avg_duration_per_task_ms": 141651.29333333333,
      "avg_output_tokens_per_task": 7913.286666666667,
      "last_run_timestamp": "2026-10-01T19:04:40.362Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "AtlasCloud",
      "accuracy": 0.751111,
      "accuracy_std_dev": 0.023726809342082964,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.014194133333333334,
      "avg_duration_per_task_ms": 218507.32444444444,
      "avg_output_tokens_per_task": 15286.804444444444,
      "last_run_timestamp": "2026-10-03T16:07:05.244Z"
    },
    {
      "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": 0.7510256923076923,
      "accuracy_std_dev": 0.024103851303776277,
      "run_count": 26,
      "total_questions": 1300,
      "avg_cost_per_task_usd": 0.3703844146153846,
      "avg_duration_per_task_ms": 173913.4323076923,
      "avg_output_tokens_per_task": 11840.180769230768,
      "last_run_timestamp": "2026-10-03T08:17:09.274Z"
    },
    {
      "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": 0.7510038393574298,
      "accuracy_std_dev": 0.038821325135033696,
      "run_count": 10,
      "total_questions": 498,
      "avg_cost_per_task_usd": 0.13673703614457833,
      "avg_duration_per_task_ms": 427247.5200803213,
      "avg_output_tokens_per_task": 19689.73092369478,
      "last_run_timestamp": "2026-10-03T16:14:16.547Z"
    },
    {
      "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": 0.7508331250000001,
      "accuracy_std_dev": 0.02331842529759192,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.115564,
      "avg_duration_per_task_ms": 319983.2525,
      "avg_output_tokens_per_task": 9741.9925,
      "last_run_timestamp": "2026-10-03T16:14:01.845Z"
    },
    {
      "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": 0.7506265313283209,
      "accuracy_std_dev": 0.029800800965536136,
      "run_count": 8,
      "total_questions": 399,
      "avg_cost_per_task_usd": 0.06770081704260651,
      "avg_duration_per_task_ms": 350619.045112782,
      "avg_output_tokens_per_task": 10643.185463659147,
      "last_run_timestamp": "2026-08-29T17:16:16.996Z"
    },
    {
      "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": 0.7505960686695279,
      "accuracy_std_dev": 0.014417471897786718,
      "run_count": 14,
      "total_questions": 699,
      "avg_cost_per_task_usd": 0.05111362374821173,
      "avg_duration_per_task_ms": 282613.5150214592,
      "avg_output_tokens_per_task": 30005.542203147354,
      "last_run_timestamp": "2026-10-03T16:08:50.062Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Fireworks",
      "accuracy": 0.7505555,
      "accuracy_std_dev": 0.03274012621004996,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.02124632666666667,
      "avg_duration_per_task_ms": 354797.86666666664,
      "avg_output_tokens_per_task": 18494.39,
      "last_run_timestamp": "2026-09-19T08:13:24.062Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "BaseTen",
      "accuracy": 0.75,
      "accuracy_std_dev": 0.0033330000000000304,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.28014248,
      "avg_duration_per_task_ms": 223020.56,
      "avg_output_tokens_per_task": 18715.38,
      "last_run_timestamp": "2026-09-12T08:12:13.731Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.5-20260423",
      "model_slug": "openai/gpt-5.5",
      "model_name": "OpenAI: GPT-5.5",
      "provider_name": "OpenAI",
      "accuracy": 0.75,
      "accuracy_std_dev": 0.04384306609373936,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.50077502,
      "avg_duration_per_task_ms": 139265.845,
      "avg_output_tokens_per_task": 7986.47,
      "last_run_timestamp": "2026-10-02T23:08:08.058Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Venice",
      "accuracy": 0.75,
      "accuracy_std_dev": 0.016666999999999987,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01952235,
      "avg_duration_per_task_ms": 2267441.92,
      "avg_output_tokens_per_task": 19667.19,
      "last_run_timestamp": "2026-10-03T08:13:52.062Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "Z.AI",
      "accuracy": 0.75,
      "accuracy_std_dev": 0.03766314634582015,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.04774656,
      "avg_duration_per_task_ms": 476512.7733333333,
      "avg_output_tokens_per_task": 9148.09,
      "last_run_timestamp": "2026-08-29T16:10:33.029Z"
    },
    {
      "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": 0.7495237142857143,
      "accuracy_std_dev": 0.027338579222746164,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.11086225999999999,
      "avg_duration_per_task_ms": 273444.3714285714,
      "avg_output_tokens_per_task": 12185.84,
      "last_run_timestamp": "2026-10-03T08:09:23.862Z"
    },
    {
      "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": 0.7493332,
      "accuracy_std_dev": 0.011623875814890659,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.177717976,
      "avg_duration_per_task_ms": 166159.384,
      "avg_output_tokens_per_task": 12813.752,
      "last_run_timestamp": "2026-10-03T08:12:20.018Z"
    },
    {
      "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": 0.749228462962963,
      "accuracy_std_dev": 0.019134704306557976,
      "run_count": 13,
      "total_questions": 648,
      "avg_cost_per_task_usd": 0.006892067901234569,
      "avg_duration_per_task_ms": 186483.29166666666,
      "avg_output_tokens_per_task": 12568.354938271605,
      "last_run_timestamp": "2026-10-03T08:37:24.837Z"
    },
    {
      "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": 0.749166625,
      "accuracy_std_dev": 0.02787621821076839,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.021932149999999997,
      "avg_duration_per_task_ms": 451561.2275,
      "avg_output_tokens_per_task": 15251.1375,
      "last_run_timestamp": "2026-10-03T16:20:50.532Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "NextBit",
      "accuracy": 0.7490367167630058,
      "accuracy_std_dev": 0.02179500605877504,
      "run_count": 7,
      "total_questions": 346,
      "avg_cost_per_task_usd": 0.011687421965317918,
      "avg_duration_per_task_ms": 205541.59826589594,
      "avg_output_tokens_per_task": 4094.841040462428,
      "last_run_timestamp": "2026-10-03T08:38:03.473Z"
    },
    {
      "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": 0.748889,
      "accuracy_std_dev": 0.019626097517789383,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.24296280222222222,
      "avg_duration_per_task_ms": 310167.38,
      "avg_output_tokens_per_task": 9997.322222222223,
      "last_run_timestamp": "2026-10-03T08:39:17.397Z"
    },
    {
      "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": 0.7488888888888889,
      "accuracy_std_dev": 0.050965958138620863,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.07998689333333331,
      "avg_duration_per_task_ms": 240287.86666666667,
      "avg_output_tokens_per_task": 9722.404444444444,
      "last_run_timestamp": "2026-10-03T16:12:22.573Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.5-flash-20260519",
      "model_slug": "google/gemini-3.5-flash",
      "model_name": "Google: Gemini 3.5 Flash",
      "provider_name": "Google",
      "accuracy": 0.7488737702702704,
      "accuracy_std_dev": 0.035681240826082414,
      "run_count": 6,
      "total_questions": 296,
      "avg_cost_per_task_usd": 0.3075426081081082,
      "avg_duration_per_task_ms": 149494.7972972973,
      "avg_output_tokens_per_task": 20582.41554054054,
      "last_run_timestamp": "2026-10-01T19:05:00.071Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "Darkbloom",
      "accuracy": 0.7485381177944861,
      "accuracy_std_dev": 0.021900710861750612,
      "run_count": 8,
      "total_questions": 399,
      "avg_cost_per_task_usd": 0.028614501253132835,
      "avg_duration_per_task_ms": 631491.8922305765,
      "avg_output_tokens_per_task": 22158.218045112782,
      "last_run_timestamp": "2026-10-03T16:12:26.817Z"
    },
    {
      "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-sonnet-20260217",
      "model_slug": "anthropic/claude-sonnet-4.6",
      "model_name": "Anthropic: Claude Sonnet 4.6",
      "provider_name": "Anthropic",
      "accuracy": 0.7483332500000001,
      "accuracy_std_dev": 0.02803765820916874,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.338020715,
      "avg_duration_per_task_ms": 209359.69166666668,
      "avg_output_tokens_per_task": 11805.883333333333,
      "last_run_timestamp": "2026-10-03T08:02:27.539Z"
    },
    {
      "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": 0.7481482222222222,
      "accuracy_std_dev": 0.03186030551146598,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.5489024666666666,
      "avg_duration_per_task_ms": 183789.9622222222,
      "avg_output_tokens_per_task": 10558.302222222223,
      "last_run_timestamp": "2026-10-03T08:38:04.046Z"
    },
    {
      "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": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "Parasail",
      "accuracy": 0.7478787272727273,
      "accuracy_std_dev": 0.021801232475965104,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.03953649272727272,
      "avg_duration_per_task_ms": 299791.2454545455,
      "avg_output_tokens_per_task": 24986.36,
      "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": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Together",
      "accuracy": 0.7477778333333334,
      "accuracy_std_dev": 0.0267822656461116,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.07834820999999999,
      "avg_duration_per_task_ms": 157847.65,
      "avg_output_tokens_per_task": 8613.366666666667,
      "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": "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": 0.7477776666666667,
      "accuracy_std_dev": 0.011166340517625106,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.17740497333333335,
      "avg_duration_per_task_ms": 310994.1566666667,
      "avg_output_tokens_per_task": 10783.98,
      "last_run_timestamp": "2026-10-03T08:17:48.657Z"
    },
    {
      "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": 0.7477776666666667,
      "accuracy_std_dev": 0.016063016338436865,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.07122743666666666,
      "avg_duration_per_task_ms": 145473.51333333334,
      "avg_output_tokens_per_task": 4717.823333333334,
      "last_run_timestamp": "2026-10-03T08:39:53.772Z"
    },
    {
      "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": 0.7477678928571428,
      "accuracy_std_dev": 0.025597104175528147,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.0058840625,
      "avg_duration_per_task_ms": 283531.1651785714,
      "avg_output_tokens_per_task": 12883.198660714286,
      "last_run_timestamp": "2026-10-03T16:13:42.336Z"
    },
    {
      "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": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "SambaNova",
      "accuracy": 0.7474747407407407,
      "accuracy_std_dev": 0.061334593171164214,
      "run_count": 6,
      "total_questions": 297,
      "avg_cost_per_task_usd": 0.07039795622895623,
      "avg_duration_per_task_ms": 90683.95286195286,
      "avg_output_tokens_per_task": 12893.693602693602,
      "last_run_timestamp": "2026-08-15T14:12:46.929Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.6-opus-20260205",
      "model_slug": "anthropic/claude-opus-4.6",
      "model_name": "Anthropic: Claude Opus 4.6",
      "provider_name": "auto-routing",
      "accuracy": 0.7474074444444444,
      "accuracy_std_dev": 0.025809327224909778,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.4897610111111111,
      "avg_duration_per_task_ms": 194827.99555555556,
      "avg_output_tokens_per_task": 10055.273333333333,
      "last_run_timestamp": "2026-10-03T08:17:06.939Z"
    },
    {
      "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": 0.74722225,
      "accuracy_std_dev": 0.048339794907379365,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.08423352666666667,
      "avg_duration_per_task_ms": 189496.69333333333,
      "avg_output_tokens_per_task": 8052.658333333334,
      "last_run_timestamp": "2026-10-03T16:25:24.079Z"
    },
    {
      "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": 0.7470370000000001,
      "accuracy_std_dev": 0.028671978645437867,
      "run_count": 18,
      "total_questions": 900,
      "avg_cost_per_task_usd": 0.5621734177777777,
      "avg_duration_per_task_ms": 202509.61777777778,
      "avg_output_tokens_per_task": 9856.906666666666,
      "last_run_timestamp": "2026-10-03T08:21:15.324Z"
    },
    {
      "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": 0.7469343143812709,
      "accuracy_std_dev": 0.029603801347968823,
      "run_count": 6,
      "total_questions": 299,
      "avg_cost_per_task_usd": 0.06045377926421405,
      "avg_duration_per_task_ms": 1570213.8294314381,
      "avg_output_tokens_per_task": 8606.695652173914,
      "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": "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": 0.746667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.1787217,
      "avg_duration_per_task_ms": 84226.38,
      "avg_output_tokens_per_task": 6412.78,
      "last_run_timestamp": "2026-10-03T08:15:16.203Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "stepfun/step-3.7-flash-20260528",
      "model_slug": "stepfun/step-3.7-flash",
      "model_name": "StepFun: Step 3.7 Flash",
      "provider_name": "StepFun",
      "accuracy": 0.746667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.02698528,
      "avg_duration_per_task_ms": 147907.84,
      "avg_output_tokens_per_task": 11065.38,
      "last_run_timestamp": "2026-07-18T08:04:04.303Z"
    },
    {
      "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": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Baidu",
      "accuracy": 0.7466666666666667,
      "accuracy_std_dev": 0.027034372269390698,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.015178880000000002,
      "avg_duration_per_task_ms": 145373.45333333334,
      "avg_output_tokens_per_task": 16435.604444444445,
      "last_run_timestamp": "2026-10-03T08:25:27.328Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.6-flash-20260721",
      "model_slug": "google/gemini-3.6-flash",
      "model_name": "Google: Gemini 3.6 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.7466666666666667,
      "accuracy_std_dev": 0.023726809342082992,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.21412848666666662,
      "avg_duration_per_task_ms": 142371.99333333335,
      "avg_output_tokens_per_task": 22756.126666666667,
      "last_run_timestamp": "2026-10-01T19:04:26.676Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-sol-20260709",
      "model_slug": "openai/gpt-5.6-sol",
      "model_name": "OpenAI: GPT-5.6 Sol",
      "provider_name": "auto-routing",
      "accuracy": 0.7466666666666667,
      "accuracy_std_dev": 0.028803034820811656,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.18399340666666666,
      "avg_duration_per_task_ms": 116870.40666666666,
      "avg_output_tokens_per_task": 5117.2266666666665,
      "last_run_timestamp": "2026-10-01T19:04:55.696Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Parasail",
      "accuracy": 0.746666625,
      "accuracy_std_dev": 0.05695993231197147,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.019652835,
      "avg_duration_per_task_ms": 219535.91,
      "avg_output_tokens_per_task": 12359.6225,
      "last_run_timestamp": "2026-10-03T16:19:59.478Z"
    },
    {
      "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": 0.746666625,
      "accuracy_std_dev": 0.03349969319985448,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.087697375,
      "avg_duration_per_task_ms": 258190.44,
      "avg_output_tokens_per_task": 12176.68,
      "last_run_timestamp": "2026-10-03T08:20:48.321Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta/muse-glimmer-30b-20260810",
      "model_slug": "meta/muse-glimmer-30b",
      "model_name": "Meta: Muse Glimmer 30B",
      "provider_name": "auto-routing",
      "accuracy": 0.7466665714285714,
      "accuracy_std_dev": 0.03380611384628165,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.03743281428571429,
      "avg_duration_per_task_ms": 197093.65714285715,
      "avg_output_tokens_per_task": 13601.3,
      "last_run_timestamp": "2026-10-03T08:15:57.213Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-opus-5.5-20260921",
      "model_slug": "anthropic/claude-opus-5.5",
      "model_name": "Anthropic: Claude Opus 5.5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7466665555555556,
      "accuracy_std_dev": 0.012957518436208841,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.3578504977777778,
      "avg_duration_per_task_ms": 113715.46888888889,
      "avg_output_tokens_per_task": 7110.575555555555,
      "last_run_timestamp": "2026-10-03T08:17:21.944Z"
    },
    {
      "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": 0.7466665,
      "accuracy_std_dev": 0.013333499999999998,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.037271729999999996,
      "avg_duration_per_task_ms": 168918.11,
      "avg_output_tokens_per_task": 24042.94,
      "last_run_timestamp": "2026-10-03T08:36:42.222Z"
    },
    {
      "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": "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": 0.7466665,
      "accuracy_std_dev": 0.02854502616189613,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.17205141333333332,
      "avg_duration_per_task_ms": 322261.24,
      "avg_output_tokens_per_task": 11142.893333333333,
      "last_run_timestamp": "2026-10-03T08:20:41.479Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Sail Research",
      "accuracy": 0.7466665,
      "accuracy_std_dev": 0.019999888889274717,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.06283585333333332,
      "avg_duration_per_task_ms": 361191.69,
      "avg_output_tokens_per_task": 8869.28,
      "last_run_timestamp": "2026-08-29T08:25:32.049Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Venice",
      "accuracy": 0.7466421794871795,
      "accuracy_std_dev": 0.021108433938566915,
      "run_count": 11,
      "total_questions": 546,
      "avg_cost_per_task_usd": 0.14117064285714287,
      "avg_duration_per_task_ms": 562123.7032967033,
      "avg_output_tokens_per_task": 27465.01098901099,
      "last_run_timestamp": "2026-10-03T16:05:51.295Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Wafer",
      "accuracy": 0.7463769999999998,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 46,
      "avg_cost_per_task_usd": 0.05924082608695652,
      "avg_duration_per_task_ms": 289784.32608695654,
      "avg_output_tokens_per_task": 9531.695652173914,
      "last_run_timestamp": "2026-08-15T14:23:29.125Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7462312663316583,
      "accuracy_std_dev": 0.0675814732785547,
      "run_count": 8,
      "total_questions": 398,
      "avg_cost_per_task_usd": 0.24451848241206028,
      "avg_duration_per_task_ms": 427951.13567839196,
      "avg_output_tokens_per_task": 27213.93216080402,
      "last_run_timestamp": "2026-10-03T16:11:17.912Z"
    },
    {
      "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": 0.7461904285714286,
      "accuracy_std_dev": 0.01987481114273,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.034756851428571434,
      "avg_duration_per_task_ms": 433579.28571428574,
      "avg_output_tokens_per_task": 9627.26857142857,
      "last_run_timestamp": "2026-08-22T16:09:51.985Z"
    },
    {
      "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": 0.745926,
      "accuracy_std_dev": 0.02095136179610079,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.3237732711111111,
      "avg_duration_per_task_ms": 198551.82222222222,
      "avg_output_tokens_per_task": 11463.67111111111,
      "last_run_timestamp": "2026-10-03T08:12:20.076Z"
    },
    {
      "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": 0.745926,
      "accuracy_std_dev": 0.019737876245094522,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.32719486666666664,
      "avg_duration_per_task_ms": 562449.7333333333,
      "avg_output_tokens_per_task": 25798.042222222222,
      "last_run_timestamp": "2026-10-03T16:09:05.690Z"
    },
    {
      "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": 0.7459258888888889,
      "accuracy_std_dev": 0.029555403324169362,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.07304336,
      "avg_duration_per_task_ms": 258879.6311111111,
      "avg_output_tokens_per_task": 9877.288888888888,
      "last_run_timestamp": "2026-10-03T16:21:25.001Z"
    },
    {
      "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": 0.7458334999999999,
      "accuracy_std_dev": 0.02933281723172869,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.28571915,
      "avg_duration_per_task_ms": 523715.0475,
      "avg_output_tokens_per_task": 9580.605,
      "last_run_timestamp": "2026-10-03T16:04:32.682Z"
    },
    {
      "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": 0.7457142857142857,
      "accuracy_std_dev": 0.040383631987350975,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.011250988571428571,
      "avg_duration_per_task_ms": 1210664.6628571427,
      "avg_output_tokens_per_task": 11628.308571428572,
      "last_run_timestamp": "2026-10-03T16:13:56.390Z"
    },
    {
      "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": 0.7455556666666666,
      "accuracy_std_dev": 0.02506781419913237,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.3528521766666667,
      "avg_duration_per_task_ms": 231482.565,
      "avg_output_tokens_per_task": 10841.498333333333,
      "last_run_timestamp": "2026-10-03T08:06:43.559Z"
    },
    {
      "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": 0.7453332,
      "accuracy_std_dev": 0.019043753132195366,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.014976084,
      "avg_duration_per_task_ms": 412623.44,
      "avg_output_tokens_per_task": 16235.352,
      "last_run_timestamp": "2026-10-03T08:13:54.001Z"
    },
    {
      "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": 0.7451852222222222,
      "accuracy_std_dev": 0.03648480000029778,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.09990454222222223,
      "avg_duration_per_task_ms": 432298.65777777776,
      "avg_output_tokens_per_task": 9726.897777777778,
      "last_run_timestamp": "2026-10-03T16:05:51.299Z"
    },
    {
      "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": 0.7451851111111111,
      "accuracy_std_dev": 0.023471128692475918,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.01492189111111111,
      "avg_duration_per_task_ms": 122706.42666666667,
      "avg_output_tokens_per_task": 15563.857777777777,
      "last_run_timestamp": "2026-10-03T13:25:39.404Z"
    },
    {
      "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": 0.745,
      "accuracy_std_dev": 0.0050000000000000044,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.010689059999999998,
      "avg_duration_per_task_ms": 257202.02,
      "avg_output_tokens_per_task": 15869.95,
      "last_run_timestamp": "2026-10-03T08:16:12.037Z"
    },
    {
      "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": 0.745,
      "accuracy_std_dev": 0.015184184024833212,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.055364632500000004,
      "avg_duration_per_task_ms": 219991,
      "avg_output_tokens_per_task": 9220.305,
      "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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Venice",
      "accuracy": 0.745,
      "accuracy_std_dev": 0.046597238036890865,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.12023054166666668,
      "avg_duration_per_task_ms": 155795.78166666668,
      "avg_output_tokens_per_task": 9438.261666666667,
      "last_run_timestamp": "2026-10-03T16:12:33.493Z"
    },
    {
      "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": 0.745,
      "accuracy_std_dev": 0.019649399202011284,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.019549249999999997,
      "avg_duration_per_task_ms": 83357.6,
      "avg_output_tokens_per_task": 4168.05,
      "last_run_timestamp": "2026-10-03T08:30:42.966Z"
    },
    {
      "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": 0.7448975714285715,
      "accuracy_std_dev": 0.011805499999999969,
      "run_count": 2,
      "total_questions": 98,
      "avg_cost_per_task_usd": 0.26889934693877554,
      "avg_duration_per_task_ms": 528124.887755102,
      "avg_output_tokens_per_task": 13421.112244897959,
      "last_run_timestamp": "2026-10-03T08:42:17.390Z"
    },
    {
      "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": "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": 0.7447619999999999,
      "accuracy_std_dev": 0.026836107941140986,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.3472446428571429,
      "avg_duration_per_task_ms": 201649.75142857144,
      "avg_output_tokens_per_task": 11858.474285714286,
      "last_run_timestamp": "2026-10-03T08:17:06.935Z"
    },
    {
      "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": "Darkbloom",
      "accuracy": 0.744681,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 47,
      "avg_cost_per_task_usd": 0.02665859574468085,
      "avg_duration_per_task_ms": 538486.1063829787,
      "avg_output_tokens_per_task": 9135.659574468085,
      "last_run_timestamp": "2026-10-03T08:11:53.173Z"
    },
    {
      "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": 0.7444445,
      "accuracy_std_dev": 0.03235589896113743,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.007050623333333332,
      "avg_duration_per_task_ms": 202444.91666666666,
      "avg_output_tokens_per_task": 4335.703333333333,
      "last_run_timestamp": "2026-10-03T08:42:41.346Z"
    },
    {
      "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": 0.7444444444444445,
      "accuracy_std_dev": 0.028803274848959327,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.5183121866666667,
      "avg_duration_per_task_ms": 226393.7577777778,
      "avg_output_tokens_per_task": 10022.293333333333,
      "last_run_timestamp": "2026-10-03T08:15:30.348Z"
    },
    {
      "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": "auto-routing",
      "accuracy": 0.7444444444444445,
      "accuracy_std_dev": 0.01987615980186154,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.011055831111111109,
      "avg_duration_per_task_ms": 226361.5911111111,
      "avg_output_tokens_per_task": 14386.48,
      "last_run_timestamp": "2026-10-03T16:11:33.008Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "amazon/nova-micro-v1",
      "model_slug": "amazon/nova-micro-v1",
      "model_name": "Amazon: Nova Micro 1.0",
      "provider_name": "Anthropic",
      "accuracy": 0.7444443333333334,
      "accuracy_std_dev": 0.00831485358994506,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 1.2319787333333332,
      "avg_duration_per_task_ms": 98531.56,
      "avg_output_tokens_per_task": 5524.42,
      "last_run_timestamp": "2026-10-01T19:04:26.700Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-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": 0.7444443333333334,
      "accuracy_std_dev": 0.00831485358994506,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.34168689999999996,
      "avg_duration_per_task_ms": 130878.06,
      "avg_output_tokens_per_task": 7080.526666666667,
      "last_run_timestamp": "2026-10-03T08:24:43.704Z"
    },
    {
      "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-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "GMICloud",
      "accuracy": 0.7444443333333334,
      "accuracy_std_dev": 0.02845816509467109,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.00811217,
      "avg_duration_per_task_ms": 250640.82333333333,
      "avg_output_tokens_per_task": 5287.056666666666,
      "last_run_timestamp": "2026-10-03T08:42:16.119Z"
    },
    {
      "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": 0.7441947955056181,
      "accuracy_std_dev": 0.013929720069329346,
      "run_count": 9,
      "total_questions": 445,
      "avg_cost_per_task_usd": 0.10463494606741572,
      "avg_duration_per_task_ms": 943614.8921348314,
      "avg_output_tokens_per_task": 21227.357303370787,
      "last_run_timestamp": "2026-09-26T08:24:43.106Z"
    },
    {
      "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": 0.744166625,
      "accuracy_std_dev": 0.029237937922917463,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.1196885575,
      "avg_duration_per_task_ms": 407890.7825,
      "avg_output_tokens_per_task": 9306.4675,
      "last_run_timestamp": "2026-10-03T16:14:05.348Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Ambient",
      "accuracy": 0.7439998000000001,
      "accuracy_std_dev": 0.025508095447524134,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.010447856000000002,
      "avg_duration_per_task_ms": 427694.796,
      "avg_output_tokens_per_task": 12591.384,
      "last_run_timestamp": "2026-08-29T09:31:13.940Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Io Net",
      "accuracy": 0.743738,
      "accuracy_std_dev": 0.01688456971344439,
      "run_count": 7,
      "total_questions": 346,
      "avg_cost_per_task_usd": 0.00938970520231214,
      "avg_duration_per_task_ms": 347914.32658959535,
      "avg_output_tokens_per_task": 3877.916184971098,
      "last_run_timestamp": "2026-10-03T08:46:08.334Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7437036666666667,
      "accuracy_std_dev": 0.03863221183876941,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.11224634666666668,
      "avg_duration_per_task_ms": 284243.8644444444,
      "avg_output_tokens_per_task": 9107.922222222222,
      "last_run_timestamp": "2026-10-03T16:18:39.856Z"
    },
    {
      "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": 0.74338923575419,
      "accuracy_std_dev": 0.032960776721153426,
      "run_count": 18,
      "total_questions": 895,
      "avg_cost_per_task_usd": 0.11096166927374301,
      "avg_duration_per_task_ms": 107590.68156424581,
      "avg_output_tokens_per_task": 8138.489385474861,
      "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-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "AkashML",
      "accuracy": 0.7433335,
      "accuracy_std_dev": 0.0033334999999999892,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.02746015,
      "avg_duration_per_task_ms": 758204.27,
      "avg_output_tokens_per_task": 13329.06,
      "last_run_timestamp": "2026-07-18T08:03:38.106Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "x-ai/grok-4.6-20260810",
      "model_slug": "x-ai/grok-4.6",
      "model_name": "SpaceXAI: Grok 4.6",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7433335,
      "accuracy_std_dev": 0.043333500000000025,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.64823893,
      "avg_duration_per_task_ms": 264133.83,
      "avg_output_tokens_per_task": 10511,
      "last_run_timestamp": "2026-10-01T16:00:37.003Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Cloudflare",
      "accuracy": 0.7433333333333333,
      "accuracy_std_dev": 0.017532868606008445,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.012350623333333333,
      "avg_duration_per_task_ms": 125234.99333333333,
      "avg_output_tokens_per_task": 4198.8,
      "last_run_timestamp": "2026-10-03T08:23:09.816Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-opus-5.5-20260921",
      "model_slug": "anthropic/claude-opus-5.5",
      "model_name": "Anthropic: Claude Opus 5.5",
      "provider_name": "auto-routing",
      "accuracy": 0.7433330000000001,
      "accuracy_std_dev": 0.010000000000000009,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.34579436,
      "avg_duration_per_task_ms": 133736.57,
      "avg_output_tokens_per_task": 7063.35,
      "last_run_timestamp": "2026-10-03T08:17:21.926Z"
    },
    {
      "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": 0.7432883288590605,
      "accuracy_std_dev": 0.0384259139109129,
      "run_count": 6,
      "total_questions": 298,
      "avg_cost_per_task_usd": 0.058029342281879205,
      "avg_duration_per_task_ms": 234207.55369127516,
      "avg_output_tokens_per_task": 7397.003355704698,
      "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": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "auto-routing",
      "accuracy": 0.7430303636363637,
      "accuracy_std_dev": 0.04248479617949489,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.13434207818181818,
      "avg_duration_per_task_ms": 839557.9927272727,
      "avg_output_tokens_per_task": 32597.374545454546,
      "last_run_timestamp": "2026-10-03T16:09:32.721Z"
    },
    {
      "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": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Ionstream",
      "accuracy": 0.7425475772357722,
      "accuracy_std_dev": 0.026593049538554243,
      "run_count": 5,
      "total_questions": 246,
      "avg_cost_per_task_usd": 0.05760138211382114,
      "avg_duration_per_task_ms": 427993.5081300813,
      "avg_output_tokens_per_task": 12666.333333333334,
      "last_run_timestamp": "2026-10-03T08:41:56.167Z"
    },
    {
      "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": 0.74250025,
      "accuracy_std_dev": 0.024874694303397962,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.01637653,
      "avg_duration_per_task_ms": 283609.14,
      "avg_output_tokens_per_task": 16652.845,
      "last_run_timestamp": "2026-10-03T08:46:48.915Z"
    },
    {
      "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": 0.742500125,
      "accuracy_std_dev": 0.029802515508080434,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.07630773249999999,
      "avg_duration_per_task_ms": 525311.875,
      "avg_output_tokens_per_task": 11683.695,
      "last_run_timestamp": "2026-10-03T16:12:13.504Z"
    },
    {
      "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": 0.74239996,
      "accuracy_std_dev": 0.0339871010313972,
      "run_count": 25,
      "total_questions": 1250,
      "avg_cost_per_task_usd": 0.3750872199999999,
      "avg_duration_per_task_ms": 202842.4192,
      "avg_output_tokens_per_task": 11926.788,
      "last_run_timestamp": "2026-10-03T08:42:18.933Z"
    },
    {
      "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": 0.742267824742268,
      "accuracy_std_dev": 0.022978500000000013,
      "run_count": 2,
      "total_questions": 97,
      "avg_cost_per_task_usd": 0.011093752577319588,
      "avg_duration_per_task_ms": 1767717.5051546392,
      "avg_output_tokens_per_task": 17128.134020618556,
      "last_run_timestamp": "2026-10-03T08:17:21.950Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.1-pro-preview-20260219",
      "model_slug": "google/gemini-3.1-pro-preview",
      "model_name": "Google: Gemini 3.1 Pro Preview",
      "provider_name": "Google AI Studio",
      "accuracy": 0.7422223333333333,
      "accuracy_std_dev": 0.006285707880227596,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.38184122000000004,
      "avg_duration_per_task_ms": 149071.25333333333,
      "avg_output_tokens_per_task": 12794.173333333334,
      "last_run_timestamp": "2026-10-01T19:04:55.722Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "inclusionai/ling-3.0-flash-20260723",
      "model_slug": "inclusionai/ling-3.0-flash",
      "model_name": "inclusionAI: Ling 3.0 Flash",
      "provider_name": "Novita",
      "accuracy": 0.7422223333333333,
      "accuracy_std_dev": 0.003142853940113798,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.0028365266666666666,
      "avg_duration_per_task_ms": 68565.14666666667,
      "avg_output_tokens_per_task": 11675.026666666667,
      "last_run_timestamp": "2026-10-01T16:01:02.671Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "Phala",
      "accuracy": 0.7422222222222222,
      "accuracy_std_dev": 0.023306958640323044,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.08909326444444444,
      "avg_duration_per_task_ms": 804074.62,
      "avg_output_tokens_per_task": 37512.27111111111,
      "last_run_timestamp": "2026-10-03T16:15:37.291Z"
    },
    {
      "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": 0.7422221111111111,
      "accuracy_std_dev": 0.02012307808597685,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.009145446666666668,
      "avg_duration_per_task_ms": 281815.08222222223,
      "avg_output_tokens_per_task": 4191.742222222222,
      "last_run_timestamp": "2026-10-03T08:37:33.027Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7421122716049383,
      "accuracy_std_dev": 0.04949233809146621,
      "run_count": 5,
      "total_questions": 243,
      "avg_cost_per_task_usd": 0.07618280658436215,
      "avg_duration_per_task_ms": 605963.6255144033,
      "avg_output_tokens_per_task": 7686.909465020576,
      "last_run_timestamp": "2026-08-29T17:13:58.312Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "Xiaomi",
      "accuracy": 0.7419176153846153,
      "accuracy_std_dev": 0.04044592450077292,
      "run_count": 12,
      "total_questions": 598,
      "avg_cost_per_task_usd": 0.02214324247491639,
      "avg_duration_per_task_ms": 326050.33277591976,
      "avg_output_tokens_per_task": 12612.583612040135,
      "last_run_timestamp": "2026-10-03T08:11:53.178Z"
    },
    {
      "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": 0.7419047142857143,
      "accuracy_std_dev": 0.02753696505589889,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.34922034571428573,
      "avg_duration_per_task_ms": 221709.42857142858,
      "avg_output_tokens_per_task": 10867.765714285715,
      "last_run_timestamp": "2026-10-03T08:22:39.770Z"
    },
    {
      "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": 0.7417787623318386,
      "accuracy_std_dev": 0.02093398684990749,
      "run_count": 9,
      "total_questions": 446,
      "avg_cost_per_task_usd": 0.11271321748878924,
      "avg_duration_per_task_ms": 104918.57623318386,
      "avg_output_tokens_per_task": 9898.578475336322,
      "last_run_timestamp": "2026-10-04T00:00:36.942Z"
    },
    {
      "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": 0.7416667499999999,
      "accuracy_std_dev": 0.03032220949959782,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.08285687,
      "avg_duration_per_task_ms": 201628.3825,
      "avg_output_tokens_per_task": 10019.7275,
      "last_run_timestamp": "2026-10-03T16:14:23.910Z"
    },
    {
      "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": 0.741666625,
      "accuracy_std_dev": 0.015545511144197728,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.00868724,
      "avg_duration_per_task_ms": 150133.0975,
      "avg_output_tokens_per_task": 13397.195,
      "last_run_timestamp": "2026-10-03T16:14:52.862Z"
    },
    {
      "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": 0.7416665,
      "accuracy_std_dev": 0.015898969219732467,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.011650780000000001,
      "avg_duration_per_task_ms": 547605.185,
      "avg_output_tokens_per_task": 5560.365,
      "last_run_timestamp": "2026-10-03T08:36:39.549Z"
    },
    {
      "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": 0.7415591841620627,
      "accuracy_std_dev": 0.038947192468013606,
      "run_count": 11,
      "total_questions": 543,
      "avg_cost_per_task_usd": 0.12226556721915285,
      "avg_duration_per_task_ms": 510809.66114180477,
      "avg_output_tokens_per_task": 24893.696132596684,
      "last_run_timestamp": "2026-10-03T08:18:33.162Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Cloudflare",
      "accuracy": 0.7414814444444444,
      "accuracy_std_dev": 0.022172731311285703,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.07712953555555555,
      "avg_duration_per_task_ms": 256540.3088888889,
      "avg_output_tokens_per_task": 11203.284444444444,
      "last_run_timestamp": "2026-10-03T16:11:33.007Z"
    },
    {
      "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": 0.7413600133779265,
      "accuracy_std_dev": 0.021148441080582975,
      "run_count": 6,
      "total_questions": 299,
      "avg_cost_per_task_usd": 0.06112101672240802,
      "avg_duration_per_task_ms": 177751.98996655518,
      "avg_output_tokens_per_task": 7034.244147157191,
      "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": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Cloudflare",
      "accuracy": 0.7411111666666667,
      "accuracy_std_dev": 0.025361580257130863,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.019523156666666666,
      "avg_duration_per_task_ms": 217853.99333333335,
      "avg_output_tokens_per_task": 12622.736666666666,
      "last_run_timestamp": "2026-08-29T16:56:37.261Z"
    },
    {
      "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": 0.7411111666666667,
      "accuracy_std_dev": 0.03430067666687576,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.4197743266666667,
      "avg_duration_per_task_ms": 38304.543333333335,
      "avg_output_tokens_per_task": 9197.25,
      "last_run_timestamp": "2026-08-15T14:16:11.865Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "Venice",
      "accuracy": 0.7407407777777778,
      "accuracy_std_dev": 0.03670968370812554,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.04790690888888889,
      "avg_duration_per_task_ms": 506951.23333333334,
      "avg_output_tokens_per_task": 9305.406666666666,
      "last_run_timestamp": "2026-10-04T00:00:36.971Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "auto-routing",
      "accuracy": 0.7407406666666667,
      "accuracy_std_dev": 0.020714235711488657,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.07330832000000001,
      "avg_duration_per_task_ms": 331512.53777777776,
      "avg_output_tokens_per_task": 12794.486666666666,
      "last_run_timestamp": "2026-10-03T16:14:15.894Z"
    },
    {
      "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": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "auto-routing",
      "accuracy": 0.7405555833333334,
      "accuracy_std_dev": 0.026311338096146348,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.023486161666666668,
      "avg_duration_per_task_ms": 159438.6,
      "avg_output_tokens_per_task": 7518.025,
      "last_run_timestamp": "2026-10-03T16:10:46.859Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-sol-20260709",
      "model_slug": "openai/gpt-5.6-sol",
      "model_name": "OpenAI: GPT-5.6 Sol",
      "provider_name": "Azure",
      "accuracy": 0.7403703333333334,
      "accuracy_std_dev": 0.030325141571529955,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.3409674066666667,
      "avg_duration_per_task_ms": 128710.22,
      "avg_output_tokens_per_task": 5246.528888888889,
      "last_run_timestamp": "2026-10-01T19:05:16.562Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "auto-routing",
      "accuracy": 0.7403273080357142,
      "accuracy_std_dev": 0.029611212101141125,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.029858823660714285,
      "avg_duration_per_task_ms": 549473.7120535715,
      "avg_output_tokens_per_task": 14542.716517857143,
      "last_run_timestamp": "2026-10-04T00:00:28.210Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "GMICloud",
      "accuracy": 0.7400002,
      "accuracy_std_dev": 0.01264900523203309,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.034869184,
      "avg_duration_per_task_ms": 129959.044,
      "avg_output_tokens_per_task": 16349.368,
      "last_run_timestamp": "2026-10-03T08:21:01.017Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-opus-5.5-20260921",
      "model_slug": "anthropic/claude-opus-5.5",
      "model_name": "Anthropic: Claude Opus 5.5",
      "provider_name": "Anthropic",
      "accuracy": 0.74,
      "accuracy_std_dev": 0.014401594564491825,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.3455118266666667,
      "avg_duration_per_task_ms": 100747.32666666666,
      "avg_output_tokens_per_task": 6984.213333333333,
      "last_run_timestamp": "2026-10-03T08:20:22.717Z"
    },
    {
      "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": "Fireworks",
      "accuracy": 0.74,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.2696536,
      "avg_duration_per_task_ms": 310259.6,
      "avg_output_tokens_per_task": 11435.58,
      "last_run_timestamp": "2026-07-18T08:04:29.408Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "DekaLLM",
      "accuracy": 0.74,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.0077757799999999995,
      "avg_duration_per_task_ms": 547010.54,
      "avg_output_tokens_per_task": 14311.64,
      "last_run_timestamp": "2026-10-03T08:35:48.617Z"
    },
    {
      "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": "Ambient",
      "accuracy": 0.74,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.05110574,
      "avg_duration_per_task_ms": 158677.44,
      "avg_output_tokens_per_task": 8249.06,
      "last_run_timestamp": "2026-07-18T16:00:09.002Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-sol-pro-20260709",
      "model_slug": "openai/gpt-5.6-sol-pro",
      "model_name": "OpenAI: GPT-5.6 Sol Pro",
      "provider_name": "OpenAI",
      "accuracy": 0.74,
      "accuracy_std_dev": 0.044555500139339316,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 1.0935073933333332,
      "avg_duration_per_task_ms": 295552.94,
      "avg_output_tokens_per_task": 22860.126666666667,
      "last_run_timestamp": "2026-10-02T23:08:29.026Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "stepfun/step-3.7-flash-20260528",
      "model_slug": "stepfun/step-3.7-flash",
      "model_name": "StepFun: Step 3.7 Flash",
      "provider_name": "Novita",
      "accuracy": 0.74,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.02561014,
      "avg_duration_per_task_ms": 147210.62,
      "avg_output_tokens_per_task": 10787.18,
      "last_run_timestamp": "2026-07-18T08:04:42.215Z"
    },
    {
      "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": 0.7399999166666666,
      "accuracy_std_dev": 0.04127600414982522,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.09576565499999999,
      "avg_duration_per_task_ms": 151409.00166666668,
      "avg_output_tokens_per_task": 9037.948333333334,
      "last_run_timestamp": "2026-10-03T16:15:37.342Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Together",
      "accuracy": 0.7399998333333334,
      "accuracy_std_dev": 0.02108185106844802,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.015687136666666667,
      "avg_duration_per_task_ms": 84185.48333333334,
      "avg_output_tokens_per_task": 4443.856666666667,
      "last_run_timestamp": "2026-10-03T08:49:53.707Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Wafer",
      "accuracy": 0.7396624683544304,
      "accuracy_std_dev": 0.02347519769052809,
      "run_count": 8,
      "total_questions": 395,
      "avg_cost_per_task_usd": 0.019082217721518985,
      "avg_duration_per_task_ms": 310670.8101265823,
      "avg_output_tokens_per_task": 15967.20506329114,
      "last_run_timestamp": "2026-10-03T08:37:28.611Z"
    },
    {
      "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": 0.7396296666666666,
      "accuracy_std_dev": 0.019177150471908547,
      "run_count": 18,
      "total_questions": 900,
      "avg_cost_per_task_usd": 0.48443006888888884,
      "avg_duration_per_task_ms": 673286.5122222222,
      "avg_output_tokens_per_task": 9807.494444444445,
      "last_run_timestamp": "2026-10-03T08:24:07.563Z"
    },
    {
      "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": 0.7394443333333334,
      "accuracy_std_dev": 0.024260716100771285,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.023661249999999998,
      "avg_duration_per_task_ms": 387738.6583333333,
      "avg_output_tokens_per_task": 12362.963333333333,
      "last_run_timestamp": "2026-10-03T08:16:24.263Z"
    },
    {
      "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": 0.7392593333333333,
      "accuracy_std_dev": 0.014553267185679581,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.015606891111111108,
      "avg_duration_per_task_ms": 184036.77333333335,
      "avg_output_tokens_per_task": 4234.36,
      "last_run_timestamp": "2026-10-03T13:17:20.799Z"
    },
    {
      "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": 0.7391550136986301,
      "accuracy_std_dev": 0.013576518152268981,
      "run_count": 6,
      "total_questions": 292,
      "avg_cost_per_task_usd": 0.0136173698630137,
      "avg_duration_per_task_ms": 57786.109589041094,
      "avg_output_tokens_per_task": 3732.6301369863013,
      "last_run_timestamp": "2026-10-03T08:21:31.962Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "CoreWeave",
      "accuracy": 0.738889,
      "accuracy_std_dev": 0.017812306747489716,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.022993023333333338,
      "avg_duration_per_task_ms": 282964.01666666666,
      "avg_output_tokens_per_task": 13071.093333333334,
      "last_run_timestamp": "2026-08-29T16:33:29.042Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-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": 0.7386668,
      "accuracy_std_dev": 0.009798067838099509,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.021031684000000002,
      "avg_duration_per_task_ms": 190250.892,
      "avg_output_tokens_per_task": 4221.716,
      "last_run_timestamp": "2026-10-03T08:46:01.665Z"
    },
    {
      "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": 0.7385184444444445,
      "accuracy_std_dev": 0.03154897514486578,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.03057054888888889,
      "avg_duration_per_task_ms": 269640.66222222225,
      "avg_output_tokens_per_task": 17038.204444444444,
      "last_run_timestamp": "2026-10-03T08:35:45.438Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.4-20260305",
      "model_slug": "openai/gpt-5.4",
      "model_name": "OpenAI: GPT-5.4",
      "provider_name": "Azure",
      "accuracy": 0.7385184444444445,
      "accuracy_std_dev": 0.017992647486688756,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.3103866155555556,
      "avg_duration_per_task_ms": 191198.25333333333,
      "avg_output_tokens_per_task": 10879.702222222222,
      "last_run_timestamp": "2026-10-01T19:05:18.005Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Phala",
      "accuracy": 0.7383335,
      "accuracy_std_dev": 0.03261724152193133,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.12250923999999998,
      "avg_duration_per_task_ms": 419416.9975,
      "avg_output_tokens_per_task": 11580.0775,
      "last_run_timestamp": "2026-10-03T16:34:12.025Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "AkashML",
      "accuracy": 0.7382865983935742,
      "accuracy_std_dev": 0.027461257286584682,
      "run_count": 5,
      "total_questions": 249,
      "avg_cost_per_task_usd": 0.07498479919678715,
      "avg_duration_per_task_ms": 91080.29718875502,
      "avg_output_tokens_per_task": 4367.955823293173,
      "last_run_timestamp": "2026-09-26T08:11:26.495Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-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": 0.7381758445945945,
      "accuracy_std_dev": 0.015122946755951125,
      "run_count": 6,
      "total_questions": 296,
      "avg_cost_per_task_usd": 0.020533699324324327,
      "avg_duration_per_task_ms": 105846.9831081081,
      "avg_output_tokens_per_task": 3919.314189189189,
      "last_run_timestamp": "2026-10-03T08:06:57.005Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Modal",
      "accuracy": 0.737500125,
      "accuracy_std_dev": 0.01854046812675925,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.386765285,
      "avg_duration_per_task_ms": 179778.535,
      "avg_output_tokens_per_task": 10745.6125,
      "last_run_timestamp": "2026-10-03T08:40:00.236Z"
    },
    {
      "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": 0.7375,
      "accuracy_std_dev": 0.016645620047027417,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.08416762750000001,
      "avg_duration_per_task_ms": 304561.8525,
      "avg_output_tokens_per_task": 13455.715,
      "last_run_timestamp": "2026-10-03T16:10:32.007Z"
    },
    {
      "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": 0.7374686954887218,
      "accuracy_std_dev": 0.031950697455106813,
      "run_count": 11,
      "total_questions": 532,
      "avg_cost_per_task_usd": 0.0578316992481203,
      "avg_duration_per_task_ms": 124019.84022556391,
      "avg_output_tokens_per_task": 7638.883458646616,
      "last_run_timestamp": "2026-10-03T16:12:10.890Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Moonshot AI",
      "accuracy": 0.7374528407310704,
      "accuracy_std_dev": 0.026611110651845488,
      "run_count": 23,
      "total_questions": 1149,
      "avg_cost_per_task_usd": 0.12236416971279374,
      "avg_duration_per_task_ms": 174480.13751087902,
      "avg_output_tokens_per_task": 9112.760661444734,
      "last_run_timestamp": "2026-10-03T16:13:53.559Z"
    },
    {
      "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": 0.7373735454545455,
      "accuracy_std_dev": 0.022590007983343003,
      "run_count": 4,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0463935,
      "avg_duration_per_task_ms": 782908.3939393939,
      "avg_output_tokens_per_task": 14148.70202020202,
      "last_run_timestamp": "2026-10-04T00:00:36.967Z"
    },
    {
      "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": 0.7371428571428571,
      "accuracy_std_dev": 0.033210782885032,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.4983071857142858,
      "avg_duration_per_task_ms": 191159.05714285714,
      "avg_output_tokens_per_task": 9902.92,
      "last_run_timestamp": "2026-10-03T08:21:29.528Z"
    },
    {
      "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": 0.736667,
      "accuracy_std_dev": 0.010000000000000009,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.37755702999999996,
      "avg_duration_per_task_ms": 287090.13,
      "avg_output_tokens_per_task": 9995.37,
      "last_run_timestamp": "2026-10-03T08:10:45.812Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-sol-pro-20260709",
      "model_slug": "openai/gpt-5.6-sol-pro",
      "model_name": "OpenAI: GPT-5.6 Sol Pro",
      "provider_name": "auto-routing",
      "accuracy": 0.736667,
      "accuracy_std_dev": 0.02999999999999997,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 1.2879983000000002,
      "avg_duration_per_task_ms": 310547.73,
      "avg_output_tokens_per_task": 22913.4,
      "last_run_timestamp": "2026-10-01T16:00:34.660Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "SiliconFlow",
      "accuracy": 0.736667,
      "accuracy_std_dev": 0.010000000000000009,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.06833741,
      "avg_duration_per_task_ms": 266166.56,
      "avg_output_tokens_per_task": 13745.94,
      "last_run_timestamp": "2026-10-03T08:19:53.305Z"
    },
    {
      "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": 0.7366668333333333,
      "accuracy_std_dev": 0.023333261905247794,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.35079288666666664,
      "avg_duration_per_task_ms": 89188.23666666666,
      "avg_output_tokens_per_task": 6914.113333333334,
      "last_run_timestamp": "2026-10-03T08:12:46.438Z"
    },
    {
      "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": "GMICloud",
      "accuracy": 0.7366667499999999,
      "accuracy_std_dev": 0.0328295429291591,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.10887684250000002,
      "avg_duration_per_task_ms": 323590.875,
      "avg_output_tokens_per_task": 18425.0325,
      "last_run_timestamp": "2026-10-03T08:19:27.960Z"
    },
    {
      "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": "AtlasCloud",
      "accuracy": 0.736666625,
      "accuracy_std_dev": 0.03915774721219767,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0416237425,
      "avg_duration_per_task_ms": 215980.8525,
      "avg_output_tokens_per_task": 20025.32,
      "last_run_timestamp": "2026-10-03T08:21:01.017Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "z-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": 0.7366665,
      "accuracy_std_dev": 0.023333357143440252,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.019450683333333333,
      "avg_duration_per_task_ms": 216938.31,
      "avg_output_tokens_per_task": 4564.706666666667,
      "last_run_timestamp": "2026-10-03T08:34:35.340Z"
    },
    {
      "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": 0.7365900057471265,
      "accuracy_std_dev": 0.030044845419567656,
      "run_count": 7,
      "total_questions": 348,
      "avg_cost_per_task_usd": 0.09514021264367815,
      "avg_duration_per_task_ms": 383890.3362068966,
      "avg_output_tokens_per_task": 11244.488505747126,
      "last_run_timestamp": "2026-10-03T16:13:47.481Z"
    },
    {
      "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": 0.7361110833333333,
      "accuracy_std_dev": 0.03144161010099605,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.030371235,
      "avg_duration_per_task_ms": 428092.15,
      "avg_output_tokens_per_task": 13168.441666666668,
      "last_run_timestamp": "2026-10-03T08:16:37.922Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "Together",
      "accuracy": 0.736,
      "accuracy_std_dev": 0.01717892766152766,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.23393872400000001,
      "avg_duration_per_task_ms": 250693.228,
      "avg_output_tokens_per_task": 12185.72,
      "last_run_timestamp": "2026-10-03T08:11:53.179Z"
    },
    {
      "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": 0.736,
      "accuracy_std_dev": 0.006798758063058283,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.091687124,
      "avg_duration_per_task_ms": 467869.472,
      "avg_output_tokens_per_task": 12583.56,
      "last_run_timestamp": "2026-10-03T08:09:52.943Z"
    },
    {
      "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": 0.7358334999999999,
      "accuracy_std_dev": 0.025097105854062156,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.015653559999999997,
      "avg_duration_per_task_ms": 644029.88,
      "avg_output_tokens_per_task": 23956.73,
      "last_run_timestamp": "2026-10-03T08:19:27.959Z"
    },
    {
      "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": 0.73583325,
      "accuracy_std_dev": 0.04026825845424036,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0081671125,
      "avg_duration_per_task_ms": 333111.9,
      "avg_output_tokens_per_task": 15655.8425,
      "last_run_timestamp": "2026-10-03T08:38:11.814Z"
    },
    {
      "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": 0.7355556666666666,
      "accuracy_std_dev": 0.012570472951413688,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.13620342666666665,
      "avg_duration_per_task_ms": 101042.81333333334,
      "avg_output_tokens_per_task": 8769.966666666667,
      "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": "anthropic/claude-opus-5.5-20260921",
      "model_slug": "anthropic/claude-opus-5.5",
      "model_name": "Anthropic: Claude Opus 5.5",
      "provider_name": "Google",
      "accuracy": 0.7355554444444444,
      "accuracy_std_dev": 0.012171531247693556,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.3527382688888889,
      "avg_duration_per_task_ms": 87063.18222222223,
      "avg_output_tokens_per_task": 7051.828888888889,
      "last_run_timestamp": "2026-10-03T08:18:17.834Z"
    },
    {
      "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": 0.7355554166666667,
      "accuracy_std_dev": 0.0214879236838522,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.007812348333333334,
      "avg_duration_per_task_ms": 338040.58666666667,
      "avg_output_tokens_per_task": 14032.848333333333,
      "last_run_timestamp": "2026-10-03T08:21:29.543Z"
    },
    {
      "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": 0.735194730964467,
      "accuracy_std_dev": 0.02636439034355808,
      "run_count": 4,
      "total_questions": 197,
      "avg_cost_per_task_usd": 0.11292960406091371,
      "avg_duration_per_task_ms": 484722.62436548225,
      "avg_output_tokens_per_task": 29978.527918781725,
      "last_run_timestamp": "2026-10-03T08:25:13.022Z"
    },
    {
      "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": 0.7351928099173552,
      "accuracy_std_dev": 0.026445667029780136,
      "run_count": 10,
      "total_questions": 484,
      "avg_cost_per_task_usd": 0.019751508264462807,
      "avg_duration_per_task_ms": 93530.47314049587,
      "avg_output_tokens_per_task": 6382.072314049587,
      "last_run_timestamp": "2026-10-03T16:12:29.900Z"
    },
    {
      "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": 0.7349998750000001,
      "accuracy_std_dev": 0.029767504188449773,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.1274803,
      "avg_duration_per_task_ms": 258225.3975,
      "avg_output_tokens_per_task": 10070.7475,
      "last_run_timestamp": "2026-10-03T16:12:24.690Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.5-flash-lite-20260721",
      "model_slug": "google/gemini-3.5-flash-lite",
      "model_name": "Google: Gemini 3.5 Flash Lite",
      "provider_name": "Google AI Studio",
      "accuracy": 0.7349997500000001,
      "accuracy_std_dev": 0.002886895683515417,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.12295462999999998,
      "avg_duration_per_task_ms": 89986.35,
      "avg_output_tokens_per_task": 23350.75,
      "last_run_timestamp": "2026-10-01T19:05:46.177Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Parasail",
      "accuracy": 0.7346666000000001,
      "accuracy_std_dev": 0.025785270559759498,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.382410584,
      "avg_duration_per_task_ms": 202893.328,
      "avg_output_tokens_per_task": 8639.172,
      "last_run_timestamp": "2026-10-03T08:18:05.263Z"
    },
    {
      "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": 0.7345953452115812,
      "accuracy_std_dev": 0.020421322845629582,
      "run_count": 9,
      "total_questions": 449,
      "avg_cost_per_task_usd": 0.01083965478841871,
      "avg_duration_per_task_ms": 494548.15144766145,
      "avg_output_tokens_per_task": 14266.481069042316,
      "last_run_timestamp": "2026-10-03T08:38:09.636Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7343749866071428,
      "accuracy_std_dev": 0.03726027594412176,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.08364443973214285,
      "avg_duration_per_task_ms": 526591.8638392857,
      "avg_output_tokens_per_task": 14298.44642857143,
      "last_run_timestamp": "2026-10-03T16:14:31.194Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Alibaba",
      "accuracy": 0.7342569118387909,
      "accuracy_std_dev": 0.03174336773937983,
      "run_count": 8,
      "total_questions": 397,
      "avg_cost_per_task_usd": 0.07517244332493701,
      "avg_duration_per_task_ms": 212080.59697733,
      "avg_output_tokens_per_task": 8222.277078085643,
      "last_run_timestamp": "2026-10-03T16:14:02.962Z"
    },
    {
      "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": 0.7340742222222222,
      "accuracy_std_dev": 0.04842648211642921,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.07961510222222222,
      "avg_duration_per_task_ms": 198839.9311111111,
      "avg_output_tokens_per_task": 8248.502222222221,
      "last_run_timestamp": "2026-09-19T08:12:24.315Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "auto-routing",
      "accuracy": 0.7340741111111111,
      "accuracy_std_dev": 0.04903445388804455,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.00916024,
      "avg_duration_per_task_ms": 334894.9577777778,
      "avg_output_tokens_per_task": 15526.06,
      "last_run_timestamp": "2026-10-03T08:11:53.180Z"
    },
    {
      "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": 0.73375,
      "accuracy_std_dev": 0.03190438406003161,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.07429103249999999,
      "avg_duration_per_task_ms": 245485.365,
      "avg_output_tokens_per_task": 11375.32,
      "last_run_timestamp": "2026-10-03T16:17:46.210Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.5-flash-20260519",
      "model_slug": "google/gemini-3.5-flash",
      "model_name": "Google: Gemini 3.5 Flash",
      "provider_name": "Google AI Studio",
      "accuracy": 0.7333336666666668,
      "accuracy_std_dev": 0.03771236166328251,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.44470873333333333,
      "avg_duration_per_task_ms": 130557.96,
      "avg_output_tokens_per_task": 20380.893333333333,
      "last_run_timestamp": "2026-10-01T19:04:52.658Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "ModelRun",
      "accuracy": 0.7333334444444444,
      "accuracy_std_dev": 0.020846304112139443,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.06912064444444445,
      "avg_duration_per_task_ms": 144871.33777777778,
      "avg_output_tokens_per_task": 9759.733333333334,
      "last_run_timestamp": "2026-10-03T16:19:07.736Z"
    },
    {
      "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": "Phala",
      "accuracy": 0.73333325,
      "accuracy_std_dev": 0.012472414008021857,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.025873325,
      "avg_duration_per_task_ms": 147169.53,
      "avg_output_tokens_per_task": 18134.845,
      "last_run_timestamp": "2026-10-03T08:34:36.496Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-sol-20260709",
      "model_slug": "openai/gpt-5.6-sol",
      "model_name": "OpenAI: GPT-5.6 Sol",
      "provider_name": "OpenAI",
      "accuracy": 0.73333325,
      "accuracy_std_dev": 0.012472414008021857,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.17039939499999998,
      "avg_duration_per_task_ms": 114141.445,
      "avg_output_tokens_per_task": 5083.04,
      "last_run_timestamp": "2026-10-02T23:08:37.533Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "DigitalOcean",
      "accuracy": 0.73333325,
      "accuracy_std_dev": 0.02815296839922036,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.06546306166666667,
      "avg_duration_per_task_ms": 280345.87666666665,
      "avg_output_tokens_per_task": 9846.026666666667,
      "last_run_timestamp": "2026-10-03T16:07:50.336Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "NextBit",
      "accuracy": 0.7333330000000001,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01186947,
      "avg_duration_per_task_ms": 176514.74,
      "avg_output_tokens_per_task": 14201.61,
      "last_run_timestamp": "2026-08-29T16:50:07.898Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7333330000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.058005,
      "avg_duration_per_task_ms": 191243.62,
      "avg_output_tokens_per_task": 30812.32,
      "last_run_timestamp": "2026-09-12T08:09:12.993Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "CoreWeave",
      "accuracy": 0.7333330000000001,
      "accuracy_std_dev": 0.020000000000000018,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.04628359,
      "avg_duration_per_task_ms": 154349.4,
      "avg_output_tokens_per_task": 38415.97,
      "last_run_timestamp": "2026-10-03T08:10:45.808Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.2-20251211",
      "model_slug": "openai/gpt-5.2",
      "model_name": "OpenAI: GPT-5.2",
      "provider_name": "auto-routing",
      "accuracy": 0.7333330000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.2412942,
      "avg_duration_per_task_ms": 207134.84,
      "avg_output_tokens_per_task": 12538.34,
      "last_run_timestamp": "2026-08-01T08:06:12.346Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Inceptron",
      "accuracy": 0.7331108173719375,
      "accuracy_std_dev": 0.03143352561191731,
      "run_count": 9,
      "total_questions": 449,
      "avg_cost_per_task_usd": 0.07030102004454343,
      "avg_duration_per_task_ms": 234100.18485523385,
      "avg_output_tokens_per_task": 11575.886414253897,
      "last_run_timestamp": "2026-10-03T16:23:41.889Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "ModelRun",
      "accuracy": 0.7328571428571429,
      "accuracy_std_dev": 0.02503509736024079,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.06745117142857142,
      "avg_duration_per_task_ms": 76562.92857142857,
      "avg_output_tokens_per_task": 8496.594285714285,
      "last_run_timestamp": "2026-09-26T08:08:19.642Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-122b-a10b-20260224",
      "model_slug": "qwen/qwen3.5-122b-a10b",
      "model_name": "Qwen: Qwen3.5-122B-A10B",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7327273636363636,
      "accuracy_std_dev": 0.10615232467139141,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.14514397454545452,
      "avg_duration_per_task_ms": 616412.5018181818,
      "avg_output_tokens_per_task": 27320.474545454545,
      "last_run_timestamp": "2026-10-03T08:15:42.485Z"
    },
    {
      "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": 0.7325925555555556,
      "accuracy_std_dev": 0.02443313150262758,
      "run_count": 18,
      "total_questions": 900,
      "avg_cost_per_task_usd": 0.3673924444444444,
      "avg_duration_per_task_ms": 230445.1311111111,
      "avg_output_tokens_per_task": 10853.041111111112,
      "last_run_timestamp": "2026-10-03T08:12:59.963Z"
    },
    {
      "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": 0.732381,
      "accuracy_std_dev": 0.026530205265481322,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.07102172,
      "avg_duration_per_task_ms": 255607.41142857142,
      "avg_output_tokens_per_task": 7988.711428571429,
      "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": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "StreamLake",
      "accuracy": 0.7320002,
      "accuracy_std_dev": 0.03435759143129796,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.016945259999999997,
      "avg_duration_per_task_ms": 153216.504,
      "avg_output_tokens_per_task": 14926.164,
      "last_run_timestamp": "2026-10-03T08:34:50.552Z"
    },
    {
      "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": 0.73166675,
      "accuracy_std_dev": 0.01280170344866263,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.021476995,
      "avg_duration_per_task_ms": 208483.135,
      "avg_output_tokens_per_task": 8547.82,
      "last_run_timestamp": "2026-10-03T16:05:47.173Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.4-20260305",
      "model_slug": "openai/gpt-5.4",
      "model_name": "OpenAI: GPT-5.4",
      "provider_name": "OpenAI",
      "accuracy": 0.73166675,
      "accuracy_std_dev": 0.015184321229067165,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.29879598,
      "avg_duration_per_task_ms": 177541.565,
      "avg_output_tokens_per_task": 10870.785,
      "last_run_timestamp": "2026-10-02T23:13:07.333Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-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": 0.7315437382550335,
      "accuracy_std_dev": 0.03437878873504547,
      "run_count": 3,
      "total_questions": 149,
      "avg_cost_per_task_usd": 0.014048604026845638,
      "avg_duration_per_task_ms": 181440.27516778524,
      "avg_output_tokens_per_task": 3805.51677852349,
      "last_run_timestamp": "2026-10-03T08:39:48.962Z"
    },
    {
      "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": 0.7315152727272727,
      "accuracy_std_dev": 0.033162046414239345,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.009558798181818184,
      "avg_duration_per_task_ms": 364447.56727272726,
      "avg_output_tokens_per_task": 14249.385454545454,
      "last_run_timestamp": "2026-10-03T08:10:45.809Z"
    },
    {
      "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": 0.7311827741935484,
      "accuracy_std_dev": 0.030765855212556675,
      "run_count": 5,
      "total_questions": 248,
      "avg_cost_per_task_usd": 0.13947222983870966,
      "avg_duration_per_task_ms": 675891.3467741936,
      "avg_output_tokens_per_task": 27076.72177419355,
      "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": "openai/gpt-5.6-terra-20260709",
      "model_slug": "openai/gpt-5.6-terra",
      "model_name": "OpenAI: GPT-5.6 Terra",
      "provider_name": "auto-routing",
      "accuracy": 0.7311113333333332,
      "accuracy_std_dev": 0.011331263428624758,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.10536965333333334,
      "avg_duration_per_task_ms": 88766.42666666667,
      "avg_output_tokens_per_task": 4017.0533333333333,
      "last_run_timestamp": "2026-10-01T19:04:40.360Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Venice",
      "accuracy": 0.7311110833333333,
      "accuracy_std_dev": 0.02884609437300635,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.02860928166666667,
      "avg_duration_per_task_ms": 200182.31333333332,
      "avg_output_tokens_per_task": 9301.73,
      "last_run_timestamp": "2026-10-03T16:24:32.945Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.5-flash-20260519",
      "model_slug": "google/gemini-3.5-flash",
      "model_name": "Google: Gemini 3.5 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.7311110000000001,
      "accuracy_std_dev": 0.022662199996175744,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.3988358066666667,
      "avg_duration_per_task_ms": 172929.82666666666,
      "avg_output_tokens_per_task": 27148.5,
      "last_run_timestamp": "2026-10-01T19:04:40.362Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Makora",
      "accuracy": 0.7308334999999999,
      "accuracy_std_dev": 0.023731723962872993,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.043652735000000005,
      "avg_duration_per_task_ms": 134583.575,
      "avg_output_tokens_per_task": 27937.025,
      "last_run_timestamp": "2026-10-03T08:15:00.917Z"
    },
    {
      "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": 0.7308334999999999,
      "accuracy_std_dev": 0.018162178345396796,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.064435845,
      "avg_duration_per_task_ms": 500666.49,
      "avg_output_tokens_per_task": 9013.95,
      "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": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "Novita",
      "accuracy": 0.730833375,
      "accuracy_std_dev": 0.03178700271705992,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.04844423749999999,
      "avg_duration_per_task_ms": 358219.625,
      "avg_output_tokens_per_task": 10201.3225,
      "last_run_timestamp": "2026-10-03T16:15:19.157Z"
    },
    {
      "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": 0.7305555,
      "accuracy_std_dev": 0.02316726484899186,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.07086156666666667,
      "avg_duration_per_task_ms": 634585.7766666666,
      "avg_output_tokens_per_task": 8337.83,
      "last_run_timestamp": "2026-10-03T16:16:13.260Z"
    },
    {
      "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": 0.7303030909090908,
      "accuracy_std_dev": 0.037644147929729685,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.05068959090909091,
      "avg_duration_per_task_ms": 151393.85454545455,
      "avg_output_tokens_per_task": 8568.572727272727,
      "last_run_timestamp": "2026-10-03T16:21:53.588Z"
    },
    {
      "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": 0.7300587002518891,
      "accuracy_std_dev": 0.027075268354798905,
      "run_count": 8,
      "total_questions": 397,
      "avg_cost_per_task_usd": 0.37323221158690173,
      "avg_duration_per_task_ms": 401791.3123425693,
      "avg_output_tokens_per_task": 9265.304785894206,
      "last_run_timestamp": "2026-10-03T08:39:26.414Z"
    },
    {
      "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": 0.7300002499999999,
      "accuracy_std_dev": 0.024037135622563252,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.01993718,
      "avg_duration_per_task_ms": 483493.9275,
      "avg_output_tokens_per_task": 14630.11,
      "last_run_timestamp": "2026-08-29T16:31:42.782Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Crusoe",
      "accuracy": 0.7300001666666667,
      "accuracy_std_dev": 0.03605552816336059,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.11662036,
      "avg_duration_per_task_ms": 82058.93666666666,
      "avg_output_tokens_per_task": 9160.97,
      "last_run_timestamp": "2026-09-12T08:12:28.966Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Alibaba",
      "accuracy": 0.73,
      "accuracy_std_dev": 0.036666636365138984,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.3388476266666666,
      "avg_duration_per_task_ms": 404706.78,
      "avg_output_tokens_per_task": 9319.38,
      "last_run_timestamp": "2026-10-03T08:15:30.345Z"
    },
    {
      "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": 0.7299997500000001,
      "accuracy_std_dev": 0.02768883649934574,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.04588337249999999,
      "avg_duration_per_task_ms": 177085.435,
      "avg_output_tokens_per_task": 8314.0825,
      "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": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "StreamLake",
      "accuracy": 0.7299516521739131,
      "accuracy_std_dev": 0.017388525374334593,
      "run_count": 7,
      "total_questions": 345,
      "avg_cost_per_task_usd": 0.020714037681159418,
      "avg_duration_per_task_ms": 279014.0231884058,
      "avg_output_tokens_per_task": 10794.56811594203,
      "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": "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": 0.7295238571428572,
      "accuracy_std_dev": 0.034215143421598336,
      "run_count": 21,
      "total_questions": 1050,
      "avg_cost_per_task_usd": 0.3555786895238095,
      "avg_duration_per_task_ms": 206423.8238095238,
      "avg_output_tokens_per_task": 10296.156190476191,
      "last_run_timestamp": "2026-10-03T08:16:52.730Z"
    },
    {
      "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": 0.7294444166666667,
      "accuracy_std_dev": 0.02379831213713252,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.02310246,
      "avg_duration_per_task_ms": 89720.39666666667,
      "avg_output_tokens_per_task": 7429.111666666667,
      "last_run_timestamp": "2026-10-03T16:13:58.374Z"
    },
    {
      "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": 0.7293334,
      "accuracy_std_dev": 0.013727396775791089,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.21736060000000001,
      "avg_duration_per_task_ms": 395127.28,
      "avg_output_tokens_per_task": 9443.484,
      "last_run_timestamp": "2026-10-03T08:37:59.299Z"
    },
    {
      "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": 0.7293186532846716,
      "accuracy_std_dev": 0.029832182888078592,
      "run_count": 11,
      "total_questions": 548,
      "avg_cost_per_task_usd": 0.042413810218978096,
      "avg_duration_per_task_ms": 231320.98175182482,
      "avg_output_tokens_per_task": 9945.337591240876,
      "last_run_timestamp": "2026-10-03T16:16:50.387Z"
    },
    {
      "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": "Phala",
      "accuracy": 0.729167,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 48,
      "avg_cost_per_task_usd": 0.044939625000000004,
      "avg_duration_per_task_ms": 321431,
      "avg_output_tokens_per_task": 27916.854166666668,
      "last_run_timestamp": "2026-07-11T08:03:26.385Z"
    },
    {
      "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": 0.7290600769230768,
      "accuracy_std_dev": 0.010018378923258995,
      "run_count": 4,
      "total_questions": 195,
      "avg_cost_per_task_usd": 0.1270635282051282,
      "avg_duration_per_task_ms": 193981.8564102564,
      "avg_output_tokens_per_task": 10184.36923076923,
      "last_run_timestamp": "2026-10-03T16:07:34.513Z"
    },
    {
      "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": 0.728889,
      "accuracy_std_dev": 0.052305021830285715,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.024965873333333333,
      "avg_duration_per_task_ms": 476087.61333333334,
      "avg_output_tokens_per_task": 18580.853333333333,
      "last_run_timestamp": "2026-10-03T08:29:04.085Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7284848181818182,
      "accuracy_std_dev": 0.02815423218248235,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.07946297454545455,
      "avg_duration_per_task_ms": 222298.18181818182,
      "avg_output_tokens_per_task": 8681.263636363636,
      "last_run_timestamp": "2026-10-03T16:14:07.111Z"
    },
    {
      "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": 0.72833325,
      "accuracy_std_dev": 0.022298511411919422,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.044024975,
      "avg_duration_per_task_ms": 136603.385,
      "avg_output_tokens_per_task": 25927.71,
      "last_run_timestamp": "2026-10-03T08:49:47.265Z"
    },
    {
      "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": 0.7277779166666666,
      "accuracy_std_dev": 0.032413125105061814,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.022190256666666672,
      "avg_duration_per_task_ms": 117748.965,
      "avg_output_tokens_per_task": 6863.575,
      "last_run_timestamp": "2026-10-03T16:15:50.542Z"
    },
    {
      "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": 0.7275785560538117,
      "accuracy_std_dev": 0.09500029906437658,
      "run_count": 9,
      "total_questions": 446,
      "avg_cost_per_task_usd": 0.07257157174887892,
      "avg_duration_per_task_ms": 157267.63677130046,
      "avg_output_tokens_per_task": 14011.852017937219,
      "last_run_timestamp": "2026-10-03T16:16:30.145Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "BaseTen",
      "accuracy": 0.72750025,
      "accuracy_std_dev": 0.014790298059454364,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.013181285000000001,
      "avg_duration_per_task_ms": 37840.265,
      "avg_output_tokens_per_task": 3828.01,
      "last_run_timestamp": "2026-10-03T08:06:57.006Z"
    },
    {
      "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": 0.7275,
      "accuracy_std_dev": 0.011273272229038026,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.25819418250000004,
      "avg_duration_per_task_ms": 292629.7475,
      "avg_output_tokens_per_task": 10249.3025,
      "last_run_timestamp": "2026-09-26T08:16:17.361Z"
    },
    {
      "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": 0.7273567338262477,
      "accuracy_std_dev": 0.034097962965258875,
      "run_count": 11,
      "total_questions": 541,
      "avg_cost_per_task_usd": 0.10647174121996304,
      "avg_duration_per_task_ms": 340266.1959334566,
      "avg_output_tokens_per_task": 15129.060998151572,
      "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": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Morph",
      "accuracy": 0.7271506693548386,
      "accuracy_std_dev": 0.030421326219611147,
      "run_count": 5,
      "total_questions": 248,
      "avg_cost_per_task_usd": 0.13253313306451614,
      "avg_duration_per_task_ms": 725346.5120967742,
      "avg_output_tokens_per_task": 29741.665322580644,
      "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": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "GMICloud",
      "accuracy": 0.7271427142857143,
      "accuracy_std_dev": 0.03945770950094827,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.00689044,
      "avg_duration_per_task_ms": 394490.1285714286,
      "avg_output_tokens_per_task": 14238.282857142856,
      "last_run_timestamp": "2026-10-03T08:19:00.264Z"
    },
    {
      "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": 0.7270693489932887,
      "accuracy_std_dev": 0.056280655844911175,
      "run_count": 6,
      "total_questions": 298,
      "avg_cost_per_task_usd": 0.2920954026845638,
      "avg_duration_per_task_ms": 417994.4194630872,
      "avg_output_tokens_per_task": 10865.342281879195,
      "last_run_timestamp": "2026-10-03T08:36:28.640Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.2-20251211",
      "model_slug": "openai/gpt-5.2",
      "model_name": "OpenAI: GPT-5.2",
      "provider_name": "OpenAI",
      "accuracy": 0.726667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.27023132,
      "avg_duration_per_task_ms": 228964.68,
      "avg_output_tokens_per_task": 14058.24,
      "last_run_timestamp": "2026-08-01T08:05:59.194Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-astra-pro-20260903",
      "model_slug": "openai/gpt-6-astra-pro",
      "model_name": "OpenAI: GPT-6 Astra Pro",
      "provider_name": "auto-routing",
      "accuracy": 0.726667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 2.91364046,
      "avg_duration_per_task_ms": 331183.04,
      "avg_output_tokens_per_task": 12422.4,
      "last_run_timestamp": "2026-10-01T08:02:52.271Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6.1-sol-20260929",
      "model_slug": "openai/gpt-6.1-sol",
      "model_name": "OpenAI: GPT-6.1 Sol",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.726667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.11300302000000001,
      "avg_duration_per_task_ms": 353321.78,
      "avg_output_tokens_per_task": 3658.14,
      "last_run_timestamp": "2026-10-03T00:00:06.432Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.7-plus-20260602",
      "model_slug": "qwen/qwen3.7-plus",
      "model_name": "Qwen: Qwen3.7 Plus",
      "provider_name": "Alibaba",
      "accuracy": 0.726667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.05448868,
      "avg_duration_per_task_ms": 350679.6,
      "avg_output_tokens_per_task": 17629,
      "last_run_timestamp": "2026-10-01T08:04:38.753Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.7266667272727273,
      "accuracy_std_dev": 0.03959384198697136,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.1237469890909091,
      "avg_duration_per_task_ms": 184460.7309090909,
      "avg_output_tokens_per_task": 12090.834545454545,
      "last_run_timestamp": "2026-10-03T16:12:35.768Z"
    },
    {
      "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": 0.7266666666666667,
      "accuracy_std_dev": 0.050625568127402036,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.17683866,
      "avg_duration_per_task_ms": 285023.12333333335,
      "avg_output_tokens_per_task": 10017.94,
      "last_run_timestamp": "2026-10-03T08:23:53.937Z"
    },
    {
      "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": 0.7266666363636364,
      "accuracy_std_dev": 0.028284335530700075,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.03399551090909091,
      "avg_duration_per_task_ms": 161533.11272727273,
      "avg_output_tokens_per_task": 8657.467272727272,
      "last_run_timestamp": "2026-10-03T16:08:50.062Z"
    },
    {
      "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": 0.726666625,
      "accuracy_std_dev": 0.03000002777889339,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.034730545,
      "avg_duration_per_task_ms": 333782.2075,
      "avg_output_tokens_per_task": 8351.1,
      "last_run_timestamp": "2026-08-29T16:11:10.137Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "DeepInfra",
      "accuracy": 0.7266666,
      "accuracy_std_dev": 0.017888692893557094,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.46400111200000005,
      "avg_duration_per_task_ms": 797376.568,
      "avg_output_tokens_per_task": 9519.38,
      "last_run_timestamp": "2026-10-03T08:14:34.220Z"
    },
    {
      "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": 0.7266665,
      "accuracy_std_dev": 0.021343838560577626,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.050907680000000004,
      "avg_duration_per_task_ms": 398195.5025,
      "avg_output_tokens_per_task": 14890.9375,
      "last_run_timestamp": "2026-10-04T00:00:34.887Z"
    },
    {
      "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": 0.7266665,
      "accuracy_std_dev": 0.033665214491370776,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.027464944999999998,
      "avg_duration_per_task_ms": 167583.24,
      "avg_output_tokens_per_task": 16249.64,
      "last_run_timestamp": "2026-10-03T08:11:53.177Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.4-20260305",
      "model_slug": "openai/gpt-5.4",
      "model_name": "OpenAI: GPT-5.4",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7266665,
      "accuracy_std_dev": 0.03162256578378802,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.325251515,
      "avg_duration_per_task_ms": 113433.705,
      "avg_output_tokens_per_task": 10925.445,
      "last_run_timestamp": "2026-10-01T19:04:58.091Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Fireworks",
      "accuracy": 0.7264311717171718,
      "accuracy_std_dev": 0.023522929859766603,
      "run_count": 4,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.04122346464646465,
      "avg_duration_per_task_ms": 213769.9797979798,
      "avg_output_tokens_per_task": 25500.383838383837,
      "last_run_timestamp": "2026-10-03T08:37:32.589Z"
    },
    {
      "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": 0.7260604545454545,
      "accuracy_std_dev": 0.023860693807498716,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.07064769636363637,
      "avg_duration_per_task_ms": 199404.25636363638,
      "avg_output_tokens_per_task": 8425.005454545455,
      "last_run_timestamp": "2026-10-03T16:25:05.402Z"
    },
    {
      "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": 0.7259357914438503,
      "accuracy_std_dev": 0.03183481835768294,
      "run_count": 15,
      "total_questions": 748,
      "avg_cost_per_task_usd": 0.1275639679144385,
      "avg_duration_per_task_ms": 104736.3101604278,
      "avg_output_tokens_per_task": 9200.958556149733,
      "last_run_timestamp": "2026-10-03T16:09:05.691Z"
    },
    {
      "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": 0.725926,
      "accuracy_std_dev": 0.02656370968487314,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.05504480444444444,
      "avg_duration_per_task_ms": 317296.9066666667,
      "avg_output_tokens_per_task": 11442.59111111111,
      "last_run_timestamp": "2026-10-03T16:05:51.299Z"
    },
    {
      "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": 0.7255555,
      "accuracy_std_dev": 0.026643554800676796,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.07362210166666668,
      "avg_duration_per_task_ms": 249674.02666666667,
      "avg_output_tokens_per_task": 8499.056666666667,
      "last_run_timestamp": "2026-10-03T16:12:07.465Z"
    },
    {
      "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": 0.7255554166666667,
      "accuracy_std_dev": 0.021227793045354233,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.07783504166666667,
      "avg_duration_per_task_ms": 257422.64666666667,
      "avg_output_tokens_per_task": 9069.785,
      "last_run_timestamp": "2026-10-03T16:06:47.444Z"
    },
    {
      "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": 0.725244179916318,
      "accuracy_std_dev": 0.026125398745282327,
      "run_count": 5,
      "total_questions": 239,
      "avg_cost_per_task_usd": 0.13789057740585775,
      "avg_duration_per_task_ms": 590927.3305439331,
      "avg_output_tokens_per_task": 27054.07949790795,
      "last_run_timestamp": "2026-10-03T16:07:34.514Z"
    },
    {
      "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": 0.7248775772058824,
      "accuracy_std_dev": 0.06361727543977472,
      "run_count": 11,
      "total_questions": 544,
      "avg_cost_per_task_usd": 0.1465529356617647,
      "avg_duration_per_task_ms": 1298187.9650735294,
      "avg_output_tokens_per_task": 27211.007352941175,
      "last_run_timestamp": "2026-10-03T16:07:44.271Z"
    },
    {
      "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": 0.7247618571428571,
      "accuracy_std_dev": 0.03947509103012603,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.06044658285714286,
      "avg_duration_per_task_ms": 560702.9542857143,
      "avg_output_tokens_per_task": 14126.951428571429,
      "last_run_timestamp": "2026-08-29T17:14:29.092Z"
    },
    {
      "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": 0.7247618571428571,
      "accuracy_std_dev": 0.04843106328860957,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.04376539714285715,
      "avg_duration_per_task_ms": 289052.61714285717,
      "avg_output_tokens_per_task": 14151.488571428572,
      "last_run_timestamp": "2026-10-03T08:24:26.924Z"
    },
    {
      "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": 0.7244445555555555,
      "accuracy_std_dev": 0.024944550855006874,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.05747211333333333,
      "avg_duration_per_task_ms": 311689.84222222224,
      "avg_output_tokens_per_task": 9750.984444444444,
      "last_run_timestamp": "2026-10-04T00:00:36.973Z"
    },
    {
      "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": 0.7237036666666666,
      "accuracy_std_dev": 0.03501115714867033,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.3328492511111112,
      "avg_duration_per_task_ms": 204813.59333333332,
      "avg_output_tokens_per_task": 10311.322222222223,
      "last_run_timestamp": "2026-10-03T08:11:53.180Z"
    },
    {
      "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": 0.7237035555555555,
      "accuracy_std_dev": 0.03371788875396401,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.29463488,
      "avg_duration_per_task_ms": 344302.6911111111,
      "avg_output_tokens_per_task": 9396.468888888889,
      "last_run_timestamp": "2026-10-03T08:36:28.986Z"
    },
    {
      "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": 0.7235555333333333,
      "accuracy_std_dev": 0.036907479386282784,
      "run_count": 15,
      "total_questions": 750,
      "avg_cost_per_task_usd": 0.05777671999999999,
      "avg_duration_per_task_ms": 117310.35733333333,
      "avg_output_tokens_per_task": 8612.536,
      "last_run_timestamp": "2026-10-03T16:13:38.527Z"
    },
    {
      "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": 0.723445004784689,
      "accuracy_std_dev": 0.03240803794328395,
      "run_count": 21,
      "total_questions": 1045,
      "avg_cost_per_task_usd": 0.1143725071770335,
      "avg_duration_per_task_ms": 143879.64210526316,
      "avg_output_tokens_per_task": 8375.931100478469,
      "last_run_timestamp": "2026-10-03T16:12:20.066Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Phala",
      "accuracy": 0.7233335,
      "accuracy_std_dev": 0.023333500000000007,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.02063872,
      "avg_duration_per_task_ms": 638594.61,
      "avg_output_tokens_per_task": 13867.61,
      "last_run_timestamp": "2026-08-29T16:33:36.025Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2-thinking-20251106",
      "model_slug": "moonshotai/kimi-k2-thinking",
      "model_name": "MoonshotAI: Kimi K2 Thinking",
      "provider_name": "Novita",
      "accuracy": 0.7233335,
      "accuracy_std_dev": 0.016666800000300002,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.052515885000000005,
      "avg_duration_per_task_ms": 281245.84,
      "avg_output_tokens_per_task": 8472.745,
      "last_run_timestamp": "2026-08-01T17:38:23.802Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-luna-20260709",
      "model_slug": "openai/gpt-5.6-luna",
      "model_name": "OpenAI: GPT-5.6 Luna",
      "provider_name": "OpenAI",
      "accuracy": 0.72333325,
      "accuracy_std_dev": 0.01374358436462263,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.011429815,
      "avg_duration_per_task_ms": 101515.345,
      "avg_output_tokens_per_task": 5183.985,
      "last_run_timestamp": "2026-10-02T23:07:05.152Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "SambaNova",
      "accuracy": 0.7233330000000001,
      "accuracy_std_dev": 0.010000000000000009,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.12502769,
      "avg_duration_per_task_ms": 371456.84,
      "avg_output_tokens_per_task": 16328.88,
      "last_run_timestamp": "2026-10-03T16:15:26.163Z"
    },
    {
      "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": 0.7226666,
      "accuracy_std_dev": 0.02619600308902105,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.065411432,
      "avg_duration_per_task_ms": 217655.416,
      "avg_output_tokens_per_task": 9279.2,
      "last_run_timestamp": "2026-10-03T16:12:07.493Z"
    },
    {
      "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": 0.7225,
      "accuracy_std_dev": 0.02665364265442906,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.05203516,
      "avg_duration_per_task_ms": 275661.6275,
      "avg_output_tokens_per_task": 8631.19,
      "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": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Chutes",
      "accuracy": 0.7224997500000001,
      "accuracy_std_dev": 0.03430864201170749,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0677995325,
      "avg_duration_per_task_ms": 399563.9925,
      "avg_output_tokens_per_task": 11360.565,
      "last_run_timestamp": "2026-10-03T16:10:02.630Z"
    },
    {
      "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": "z-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": 0.7222223333333333,
      "accuracy_std_dev": 0.031427125187575726,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.009912693333333332,
      "avg_duration_per_task_ms": 115398.83333333333,
      "avg_output_tokens_per_task": 5230.286666666667,
      "last_run_timestamp": "2026-10-03T08:11:53.182Z"
    },
    {
      "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": 0.7222221493624772,
      "accuracy_std_dev": 0.036194051658678325,
      "run_count": 11,
      "total_questions": 549,
      "avg_cost_per_task_usd": 0.10862861566484516,
      "avg_duration_per_task_ms": 283523.6338797814,
      "avg_output_tokens_per_task": 10296.255009107468,
      "last_run_timestamp": "2026-10-03T16:20:53.360Z"
    },
    {
      "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": 0.7222221428571428,
      "accuracy_std_dev": 0.024312166707227047,
      "run_count": 3,
      "total_questions": 147,
      "avg_cost_per_task_usd": 0.046834047619047615,
      "avg_duration_per_task_ms": 58108.6462585034,
      "avg_output_tokens_per_task": 13243.625850340137,
      "last_run_timestamp": "2026-09-29T21:00:05.013Z"
    },
    {
      "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": 0.722222,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 48,
      "avg_cost_per_task_usd": 0.019033145833333334,
      "avg_duration_per_task_ms": 65993.54166666667,
      "avg_output_tokens_per_task": 3355.9791666666665,
      "last_run_timestamp": "2026-09-12T08:12:22.607Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Makora",
      "accuracy": 0.7210882448979592,
      "accuracy_std_dev": 0.0076780598851532815,
      "run_count": 5,
      "total_questions": 245,
      "avg_cost_per_task_usd": 0.009489710204081632,
      "avg_duration_per_task_ms": 145499.9755102041,
      "avg_output_tokens_per_task": 16260.81224489796,
      "last_run_timestamp": "2026-09-26T08:06:43.933Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "Friendli",
      "accuracy": 0.7207407777777778,
      "accuracy_std_dev": 0.017340394259120186,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.10771476888888888,
      "avg_duration_per_task_ms": 342540.08444444445,
      "avg_output_tokens_per_task": 15143.33111111111,
      "last_run_timestamp": "2026-10-04T00:00:34.897Z"
    },
    {
      "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": 0.01790079352541406,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.04994019555555556,
      "avg_duration_per_task_ms": 379059.3222222222,
      "avg_output_tokens_per_task": 9310.02,
      "last_run_timestamp": "2026-10-03T16:12:08.817Z"
    },
    {
      "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": 0.7207406666666667,
      "accuracy_std_dev": 0.027477417091131398,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.3540747222222222,
      "avg_duration_per_task_ms": 237714.0088888889,
      "avg_output_tokens_per_task": 10830.737777777778,
      "last_run_timestamp": "2026-10-03T08:11:26.721Z"
    },
    {
      "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": 0.7207406666666667,
      "accuracy_std_dev": 0.0246344746014379,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.05808530888888888,
      "avg_duration_per_task_ms": 328138.63333333336,
      "avg_output_tokens_per_task": 9714.76888888889,
      "last_run_timestamp": "2026-10-04T00:00:28.270Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-sol-pro-20260709",
      "model_slug": "openai/gpt-5.6-sol-pro",
      "model_name": "OpenAI: GPT-5.6 Sol Pro",
      "provider_name": "Azure",
      "accuracy": 0.7205384141414142,
      "accuracy_std_dev": 0.051112796809483045,
      "run_count": 4,
      "total_questions": 198,
      "avg_cost_per_task_usd": 1.6761988181818182,
      "avg_duration_per_task_ms": 313695.66161616164,
      "avg_output_tokens_per_task": 23404.97474747475,
      "last_run_timestamp": "2026-10-01T16:01:13.647Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7202819926470587,
      "accuracy_std_dev": 0.041371140936807896,
      "run_count": 11,
      "total_questions": 544,
      "avg_cost_per_task_usd": 0.02139444301470588,
      "avg_duration_per_task_ms": 101681.90625,
      "avg_output_tokens_per_task": 6783.099264705882,
      "last_run_timestamp": "2026-10-03T16:04:32.681Z"
    },
    {
      "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": "Baidu",
      "accuracy": 0.72,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.03553636,
      "avg_duration_per_task_ms": 263015.66,
      "avg_output_tokens_per_task": 14384.46,
      "last_run_timestamp": "2026-07-18T16:00:09.315Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Io Net",
      "accuracy": 0.72,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.12671924,
      "avg_duration_per_task_ms": 566714.78,
      "avg_output_tokens_per_task": 16714.02,
      "last_run_timestamp": "2026-09-28T18:00:05.892Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Together",
      "accuracy": 0.72,
      "accuracy_std_dev": 0.021081956479416224,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.03251169,
      "avg_duration_per_task_ms": 105065.11,
      "avg_output_tokens_per_task": 22739.925,
      "last_run_timestamp": "2026-10-03T09:03:15.742Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.6-flash-20260721",
      "model_slug": "google/gemini-3.6-flash",
      "model_name": "Google: Gemini 3.6 Flash",
      "provider_name": "Google AI Studio",
      "accuracy": 0.72,
      "accuracy_std_dev": 0.03231762505197435,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.36211610499999997,
      "avg_duration_per_task_ms": 150104.23,
      "avg_output_tokens_per_task": 23123.4,
      "last_run_timestamp": "2026-10-01T19:04:55.721Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "inclusionai/ling-3.0-flash-vl-20260910",
      "model_slug": "inclusionai/ling-3.0-flash-vl",
      "model_name": "inclusionAI: Ling 3.0 Flash VL",
      "provider_name": "Novita",
      "accuracy": 0.72,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.00317077,
      "avg_duration_per_task_ms": 128719.23,
      "avg_output_tokens_per_task": 15118.03,
      "last_run_timestamp": "2026-10-01T08:00:27.045Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3-ultra-550b-a55b-20260604",
      "model_slug": "nvidia/nemotron-3-ultra-550b-a55b",
      "model_name": "NVIDIA: Nemotron 3 Ultra",
      "provider_name": "BaseTen",
      "accuracy": 0.72,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.15154258,
      "avg_duration_per_task_ms": 105755.46,
      "avg_output_tokens_per_task": 13133.32,
      "last_run_timestamp": "2026-08-04T03:12:01.476Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.7-max-20260520",
      "model_slug": "qwen/qwen3.7-max",
      "model_name": "Qwen: Qwen3.7 Max",
      "provider_name": "AtlasCloud",
      "accuracy": 0.72,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.402901,
      "avg_duration_per_task_ms": 196016.4,
      "avg_output_tokens_per_task": 16256.52,
      "last_run_timestamp": "2026-09-27T00:00:03.103Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "Parasail",
      "accuracy": 0.72,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.028463379999999996,
      "avg_duration_per_task_ms": 583244.64,
      "avg_output_tokens_per_task": 43866.54,
      "last_run_timestamp": "2026-08-08T08:27:52.152Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Crusoe",
      "accuracy": 0.72,
      "accuracy_std_dev": 0.0066669999999999785,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.037352619999999996,
      "avg_duration_per_task_ms": 206767.2,
      "avg_output_tokens_per_task": 11957.14,
      "last_run_timestamp": "2026-09-26T08:10:43.574Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.1-flash-lite-20260507",
      "model_slug": "google/gemini-3.1-flash-lite",
      "model_name": "Google: Gemini 3.1 Flash Lite",
      "provider_name": "auto-routing",
      "accuracy": 0.7199996666666667,
      "accuracy_std_dev": 0.009428090415820642,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.09322211333333333,
      "avg_duration_per_task_ms": 149908.11333333334,
      "avg_output_tokens_per_task": 49346.81333333333,
      "last_run_timestamp": "2026-10-01T19:04:52.649Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "StreamLake",
      "accuracy": 0.7194444166666667,
      "accuracy_std_dev": 0.03371537501511324,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.03395878166666666,
      "avg_duration_per_task_ms": 181860.895,
      "avg_output_tokens_per_task": 9239.94,
      "last_run_timestamp": "2026-10-03T16:08:50.063Z"
    },
    {
      "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": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Relace",
      "accuracy": 0.7193332,
      "accuracy_std_dev": 0.02332387046268266,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.011747176000000002,
      "avg_duration_per_task_ms": 276612.664,
      "avg_output_tokens_per_task": 17492.112,
      "last_run_timestamp": "2026-10-03T08:34:32.707Z"
    },
    {
      "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": 0.7192592222222223,
      "accuracy_std_dev": 0.019485908525437762,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.02609890888888889,
      "avg_duration_per_task_ms": 230277.4688888889,
      "avg_output_tokens_per_task": 10113.222222222223,
      "last_run_timestamp": "2026-10-03T16:12:13.501Z"
    },
    {
      "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": 0.7191666250000001,
      "accuracy_std_dev": 0.02633896432539395,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.300338065,
      "avg_duration_per_task_ms": 406405.855,
      "avg_output_tokens_per_task": 9208.345,
      "last_run_timestamp": "2026-10-03T08:11:53.141Z"
    },
    {
      "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": 0.7191665,
      "accuracy_std_dev": 0.008291511728870679,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.046859950000000004,
      "avg_duration_per_task_ms": 203049.32,
      "avg_output_tokens_per_task": 32440.35,
      "last_run_timestamp": "2026-10-03T08:38:29.322Z"
    },
    {
      "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": 0.7188888333333333,
      "accuracy_std_dev": 0.026504201858175022,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.3197561666666667,
      "avg_duration_per_task_ms": 193449.66,
      "avg_output_tokens_per_task": 11623.543333333333,
      "last_run_timestamp": "2026-10-03T08:06:43.539Z"
    },
    {
      "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": 0.7188888333333333,
      "accuracy_std_dev": 0.03297943429581869,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.11799081000000002,
      "avg_duration_per_task_ms": 673829.7433333333,
      "avg_output_tokens_per_task": 29526.56,
      "last_run_timestamp": "2026-10-03T08:37:23.062Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-sol-20260709",
      "model_slug": "openai/gpt-5.6-sol",
      "model_name": "OpenAI: GPT-5.6 Sol",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7188551919191919,
      "accuracy_std_dev": 0.027525589315898408,
      "run_count": 4,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.29904142424242425,
      "avg_duration_per_task_ms": 154139.7323232323,
      "avg_output_tokens_per_task": 5036.570707070707,
      "last_run_timestamp": "2026-10-01T19:05:12.160Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "DeepInfra",
      "accuracy": 0.7186897976878613,
      "accuracy_std_dev": 0.027015736117292515,
      "run_count": 7,
      "total_questions": 346,
      "avg_cost_per_task_usd": 0.02169187283236994,
      "avg_duration_per_task_ms": 314303.23988439306,
      "avg_output_tokens_per_task": 7575.85549132948,
      "last_run_timestamp": "2026-08-29T16:48:46.265Z"
    },
    {
      "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": 0.718667,
      "accuracy_std_dev": 0.015999999999999966,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.012890876,
      "avg_duration_per_task_ms": 931505.22,
      "avg_output_tokens_per_task": 16944.504,
      "last_run_timestamp": "2026-10-01T09:55:07.922Z"
    },
    {
      "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": 0.7183906609195403,
      "accuracy_std_dev": 0.025974738950584678,
      "run_count": 7,
      "total_questions": 348,
      "avg_cost_per_task_usd": 0.06343264655172413,
      "avg_duration_per_task_ms": 497586.9425287356,
      "avg_output_tokens_per_task": 14530.419540229885,
      "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": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "StreamLake",
      "accuracy": 0.7183333749999999,
      "accuracy_std_dev": 0.031578880513475675,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.12769575749999998,
      "avg_duration_per_task_ms": 325285.93,
      "avg_output_tokens_per_task": 10098.865,
      "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": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Venice",
      "accuracy": 0.71833325,
      "accuracy_std_dev": 0.022298399296530245,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.039712805,
      "avg_duration_per_task_ms": 138975.945,
      "avg_output_tokens_per_task": 15594.74,
      "last_run_timestamp": "2026-10-03T08:38:07.682Z"
    },
    {
      "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": 0.7180392352941176,
      "accuracy_std_dev": 0.014599556014721458,
      "run_count": 17,
      "total_questions": 850,
      "avg_cost_per_task_usd": 0.037821898823529404,
      "avg_duration_per_task_ms": 192987.32,
      "avg_output_tokens_per_task": 9953.656470588236,
      "last_run_timestamp": "2026-10-03T16:15:22.425Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Together",
      "accuracy": 0.7175757272727272,
      "accuracy_std_dev": 0.03698453502624358,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.10178546363636364,
      "avg_duration_per_task_ms": 116835.80727272727,
      "avg_output_tokens_per_task": 9163.8,
      "last_run_timestamp": "2026-10-03T16:12:20.041Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-terra-20260709",
      "model_slug": "openai/gpt-5.6-terra",
      "model_name": "OpenAI: GPT-5.6 Terra",
      "provider_name": "Azure",
      "accuracy": 0.7170371111111111,
      "accuracy_std_dev": 0.016956299841680886,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.12728777111111111,
      "avg_duration_per_task_ms": 101323.44222222222,
      "avg_output_tokens_per_task": 3885.56,
      "last_run_timestamp": "2026-10-01T19:04:52.651Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Relace",
      "accuracy": 0.716667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.015077799999999999,
      "avg_duration_per_task_ms": 476878.72,
      "avg_output_tokens_per_task": 26097.54,
      "last_run_timestamp": "2026-09-28T09:00:02.615Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.7166665000000001,
      "accuracy_std_dev": 0.0033334999999999892,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.06700715,
      "avg_duration_per_task_ms": 992682.5,
      "avg_output_tokens_per_task": 40930.55,
      "last_run_timestamp": "2026-10-03T08:22:39.765Z"
    },
    {
      "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": 0.7163635454545455,
      "accuracy_std_dev": 0.14986481732266926,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.06114290181818182,
      "avg_duration_per_task_ms": 168295.4890909091,
      "avg_output_tokens_per_task": 20095.172727272726,
      "last_run_timestamp": "2026-10-03T16:07:44.277Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.2-exp",
      "model_slug": "deepseek/deepseek-v3.2-exp",
      "model_name": "DeepSeek: DeepSeek V3.2 Exp",
      "provider_name": "auto-routing",
      "accuracy": 0.7162161756756756,
      "accuracy_std_dev": 0.0136238728300322,
      "run_count": 3,
      "total_questions": 148,
      "avg_cost_per_task_usd": 0.0890948108108108,
      "avg_duration_per_task_ms": 1143483.75,
      "avg_output_tokens_per_task": 17745.06081081081,
      "last_run_timestamp": "2026-08-01T17:38:07.305Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "CoreWeave",
      "accuracy": 0.7155646528925619,
      "accuracy_std_dev": 0.07922982153103719,
      "run_count": 5,
      "total_questions": 242,
      "avg_cost_per_task_usd": 0.11734726446280992,
      "avg_duration_per_task_ms": 254637.0785123967,
      "avg_output_tokens_per_task": 42006.561983471074,
      "last_run_timestamp": "2026-08-29T17:18:52.408Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "OpenInference",
      "accuracy": 0.7155556666666666,
      "accuracy_std_dev": 0.040855070557058414,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.012743326666666667,
      "avg_duration_per_task_ms": 1224143.2066666668,
      "avg_output_tokens_per_task": 18669.653333333332,
      "last_run_timestamp": "2026-08-29T09:20:05.864Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Relace",
      "accuracy": 0.7155553333333334,
      "accuracy_std_dev": 0.029979285029648195,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.13937678,
      "avg_duration_per_task_ms": 263819.3466666667,
      "avg_output_tokens_per_task": 10311.9,
      "last_run_timestamp": "2026-10-03T16:18:50.927Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "Novita",
      "accuracy": 0.7154882272727273,
      "accuracy_std_dev": 0.03158570170163035,
      "run_count": 8,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.06011247222222223,
      "avg_duration_per_task_ms": 488450.73484848486,
      "avg_output_tokens_per_task": 14170.416666666666,
      "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": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "Mara",
      "accuracy": 0.715488212121212,
      "accuracy_std_dev": 0.013403360380889571,
      "run_count": 4,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.04391666666666666,
      "avg_duration_per_task_ms": 58944.35858585859,
      "avg_output_tokens_per_task": 8110.530303030303,
      "last_run_timestamp": "2026-08-08T08:31:07.742Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "CoreWeave",
      "accuracy": 0.715238,
      "accuracy_std_dev": 0.036811758695907394,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.0628034857142857,
      "avg_duration_per_task_ms": 75005.29714285715,
      "avg_output_tokens_per_task": 9080.965714285714,
      "last_run_timestamp": "2026-09-26T08:24:48.840Z"
    },
    {
      "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": 0.7151513636363637,
      "accuracy_std_dev": 0.042553816410993,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.09130638,
      "avg_duration_per_task_ms": 220131.52363636365,
      "avg_output_tokens_per_task": 8503.710909090909,
      "last_run_timestamp": "2026-10-03T16:13:56.534Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-terra-20260709",
      "model_slug": "openai/gpt-5.6-terra",
      "model_name": "OpenAI: GPT-5.6 Terra",
      "provider_name": "OpenAI",
      "accuracy": 0.715,
      "accuracy_std_dev": 0.029202378139459823,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.09407665500000001,
      "avg_duration_per_task_ms": 83159.84,
      "avg_output_tokens_per_task": 3660.855,
      "last_run_timestamp": "2026-10-02T23:10:36.500Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.1-flash-lite-20260507",
      "model_slug": "google/gemini-3.1-flash-lite",
      "model_name": "Google: Gemini 3.1 Flash Lite",
      "provider_name": "Google",
      "accuracy": 0.7148146666666667,
      "accuracy_std_dev": 0.031075936024161484,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.09082762666666666,
      "avg_duration_per_task_ms": 156907.88888888888,
      "avg_output_tokens_per_task": 48272.70222222222,
      "last_run_timestamp": "2026-10-01T19:06:49.017Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Parasail",
      "accuracy": 0.7145455454545454,
      "accuracy_std_dev": 0.03220115495208104,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.11228732,
      "avg_duration_per_task_ms": 99294.4490909091,
      "avg_output_tokens_per_task": 9738.792727272727,
      "last_run_timestamp": "2026-10-03T16:12:20.063Z"
    },
    {
      "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": 0.71416675,
      "accuracy_std_dev": 0.03290334118577473,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0400061,
      "avg_duration_per_task_ms": 354559.8625,
      "avg_output_tokens_per_task": 9774.86,
      "last_run_timestamp": "2026-10-03T16:14:22.302Z"
    },
    {
      "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": 0.714166625,
      "accuracy_std_dev": 0.0311691275781722,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0320175475,
      "avg_duration_per_task_ms": 366948.7475,
      "avg_output_tokens_per_task": 5303.03,
      "last_run_timestamp": "2026-10-03T08:46:44.356Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2-thinking-20251106",
      "model_slug": "moonshotai/kimi-k2-thinking",
      "model_name": "MoonshotAI: Kimi K2 Thinking",
      "provider_name": "auto-routing",
      "accuracy": 0.7133333333333334,
      "accuracy_std_dev": 0.014401903168987377,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.05460591333333333,
      "avg_duration_per_task_ms": 289287.8333333333,
      "avg_output_tokens_per_task": 8799.16,
      "last_run_timestamp": "2026-08-01T17:38:27.943Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "inclusionai/ling-3.0-flash-vl-20260910",
      "model_slug": "inclusionai/ling-3.0-flash-vl",
      "model_name": "inclusionAI: Ling 3.0 Flash VL",
      "provider_name": "DeepInfra",
      "accuracy": 0.713333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.006318460000000001,
      "avg_duration_per_task_ms": 234030.92,
      "avg_output_tokens_per_task": 14773.36,
      "last_run_timestamp": "2026-10-01T08:05:57.971Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "GMICloud",
      "accuracy": 0.713333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.09868048,
      "avg_duration_per_task_ms": 318461,
      "avg_output_tokens_per_task": 11794.24,
      "last_run_timestamp": "2026-08-29T17:18:09.782Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "DekaLLM",
      "accuracy": 0.713333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.0723164,
      "avg_duration_per_task_ms": 346522.38,
      "avg_output_tokens_per_task": 59152.72,
      "last_run_timestamp": "2026-09-12T16:00:24.399Z"
    },
    {
      "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": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "Novita",
      "accuracy": 0.7130950357142857,
      "accuracy_std_dev": 0.06532407345092787,
      "run_count": 3,
      "total_questions": 140,
      "avg_cost_per_task_usd": 0.18180849999999998,
      "avg_duration_per_task_ms": 300783.1928571428,
      "avg_output_tokens_per_task": 20383.22857142857,
      "last_run_timestamp": "2026-10-03T16:20:06.515Z"
    },
    {
      "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": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7123811428571428,
      "accuracy_std_dev": 0.0279293098601685,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.04820127142857143,
      "avg_duration_per_task_ms": 1005664.88,
      "avg_output_tokens_per_task": 8548.157142857142,
      "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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Cloudflare",
      "accuracy": 0.7121428571428572,
      "accuracy_std_dev": 0.03208647153253476,
      "run_count": 14,
      "total_questions": 700,
      "avg_cost_per_task_usd": 0.13476882285714287,
      "avg_duration_per_task_ms": 139447.84285714285,
      "avg_output_tokens_per_task": 8900.078571428572,
      "last_run_timestamp": "2026-10-03T16:12:20.071Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5-mini-2025-08-07",
      "model_slug": "openai/gpt-5-mini",
      "model_name": "OpenAI: GPT-5 Mini",
      "provider_name": "OpenAI",
      "accuracy": 0.712121101010101,
      "accuracy_std_dev": 0.029111223892169177,
      "run_count": 4,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.12218062626262625,
      "avg_duration_per_task_ms": 537818.2474747475,
      "avg_output_tokens_per_task": 55422.06060606061,
      "last_run_timestamp": "2026-10-02T23:27:09.304Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-luna-pro-20260709",
      "model_slug": "openai/gpt-5.6-luna-pro",
      "model_name": "OpenAI: GPT-5.6 Luna Pro",
      "provider_name": "OpenAI",
      "accuracy": 0.71166675,
      "accuracy_std_dev": 0.019075733568790994,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.07509325,
      "avg_duration_per_task_ms": 322731.32,
      "avg_output_tokens_per_task": 33483.435,
      "last_run_timestamp": "2026-10-02T23:12:53.530Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "Venice",
      "accuracy": 0.7115566504297995,
      "accuracy_std_dev": 0.03261413456694845,
      "run_count": 7,
      "total_questions": 349,
      "avg_cost_per_task_usd": 0.053420421203438394,
      "avg_duration_per_task_ms": 958453.7363896848,
      "avg_output_tokens_per_task": 13435.779369627508,
      "last_run_timestamp": "2026-10-04T00:00:34.891Z"
    },
    {
      "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": 0.7114285,
      "accuracy_std_dev": 0.029753450229103078,
      "run_count": 14,
      "total_questions": 700,
      "avg_cost_per_task_usd": 0.11265073714285717,
      "avg_duration_per_task_ms": 89786.91571428571,
      "avg_output_tokens_per_task": 10363.881428571429,
      "last_run_timestamp": "2026-10-03T16:12:29.882Z"
    },
    {
      "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": 0.7113343768844221,
      "accuracy_std_dev": 0.031502220499778535,
      "run_count": 12,
      "total_questions": 597,
      "avg_cost_per_task_usd": 0.40403888609715255,
      "avg_duration_per_task_ms": 235745.01172529312,
      "avg_output_tokens_per_task": 9241.058626465661,
      "last_run_timestamp": "2026-10-03T08:37:32.473Z"
    },
    {
      "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": 0.7111853238731218,
      "accuracy_std_dev": 0.034402559579759295,
      "run_count": 12,
      "total_questions": 599,
      "avg_cost_per_task_usd": 0.06973980801335558,
      "avg_duration_per_task_ms": 265076.1986644407,
      "avg_output_tokens_per_task": 8782.440734557596,
      "last_run_timestamp": "2026-10-03T16:18:22.330Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-luna-pro-20260709",
      "model_slug": "openai/gpt-5.6-luna-pro",
      "model_name": "OpenAI: GPT-5.6 Luna Pro",
      "provider_name": "auto-routing",
      "accuracy": 0.7111113333333333,
      "accuracy_std_dev": 0.011331263428624743,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.07186812,
      "avg_duration_per_task_ms": 348295.26,
      "avg_output_tokens_per_task": 33945.17333333333,
      "last_run_timestamp": "2026-10-01T16:00:47.532Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.7109990505050505,
      "accuracy_std_dev": 0.0725087554956009,
      "run_count": 12,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.04814752861952862,
      "avg_duration_per_task_ms": 285404.8703703704,
      "avg_output_tokens_per_task": 27622.824915824916,
      "last_run_timestamp": "2026-10-03T16:03:45.633Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-luna-20260709",
      "model_slug": "openai/gpt-5.6-luna",
      "model_name": "OpenAI: GPT-5.6 Luna",
      "provider_name": "Azure",
      "accuracy": 0.7103705555555555,
      "accuracy_std_dev": 0.030809822424065814,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.03036840888888889,
      "avg_duration_per_task_ms": 105409.86666666667,
      "avg_output_tokens_per_task": 5076.06,
      "last_run_timestamp": "2026-10-01T19:05:16.550Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "Venice",
      "accuracy": 0.7103704444444444,
      "accuracy_std_dev": 0.016662593743079532,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.07196275999999999,
      "avg_duration_per_task_ms": 276773.88,
      "avg_output_tokens_per_task": 10708.504444444445,
      "last_run_timestamp": "2026-10-03T16:18:18.265Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-terra-20260709",
      "model_slug": "openai/gpt-5.6-terra",
      "model_name": "OpenAI: GPT-5.6 Terra",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.71,
      "accuracy_std_dev": 0.01795033271279393,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.12391315,
      "avg_duration_per_task_ms": 76197.77,
      "avg_output_tokens_per_task": 4134.025,
      "last_run_timestamp": "2026-10-01T19:04:38.304Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Venice",
      "accuracy": 0.7099996666666667,
      "accuracy_std_dev": 0.03144660377352204,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.24028676333333332,
      "avg_duration_per_task_ms": 296408.56666666665,
      "avg_output_tokens_per_task": 20949.47,
      "last_run_timestamp": "2026-10-03T08:39:24.637Z"
    },
    {
      "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": 0.709697,
      "accuracy_std_dev": 0.02890091937197594,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.00848128909090909,
      "avg_duration_per_task_ms": 316978.2872727273,
      "avg_output_tokens_per_task": 14218.061818181819,
      "last_run_timestamp": "2026-10-03T08:22:39.768Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.2-exp",
      "model_slug": "deepseek/deepseek-v3.2-exp",
      "model_name": "DeepSeek: DeepSeek V3.2 Exp",
      "provider_name": "AtlasCloud",
      "accuracy": 0.709459418918919,
      "accuracy_std_dev": 0.03963897380384895,
      "run_count": 3,
      "total_questions": 148,
      "avg_cost_per_task_usd": 0.0811334054054054,
      "avg_duration_per_task_ms": 456032.14864864864,
      "avg_output_tokens_per_task": 13522.72972972973,
      "last_run_timestamp": "2026-08-01T17:38:10.889Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "BaseTen",
      "accuracy": 0.7091114702886249,
      "accuracy_std_dev": 0.03584624085968895,
      "run_count": 24,
      "total_questions": 1178,
      "avg_cost_per_task_usd": 0.09626076570458404,
      "avg_duration_per_task_ms": 76008.29456706283,
      "avg_output_tokens_per_task": 8032.944821731749,
      "last_run_timestamp": "2026-10-03T16:16:10.215Z"
    },
    {
      "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": 0.7074906876404494,
      "accuracy_std_dev": 0.05861604916933776,
      "run_count": 9,
      "total_questions": 445,
      "avg_cost_per_task_usd": 0.05871908539325842,
      "avg_duration_per_task_ms": 920700.0044943821,
      "avg_output_tokens_per_task": 62280.74157303371,
      "last_run_timestamp": "2026-10-03T16:09:19.692Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "inclusionai/ling-3.0-flash-fin-20260827",
      "model_slug": "inclusionai/ling-3.0-flash-fin",
      "model_name": "inclusionAI: Ling 3.0 Flash Fin",
      "provider_name": "auto-routing",
      "accuracy": 0.707483,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 49,
      "avg_cost_per_task_usd": 0.004842469387755102,
      "avg_duration_per_task_ms": 161558.55102040817,
      "avg_output_tokens_per_task": 17024.020408163266,
      "last_run_timestamp": "2026-10-01T08:01:49.153Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Nebius",
      "accuracy": 0.706667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.3541182,
      "avg_duration_per_task_ms": 1646927.84,
      "avg_output_tokens_per_task": 24534.74,
      "last_run_timestamp": "2026-09-10T00:00:02.948Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "WandB",
      "accuracy": 0.706667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.08614948,
      "avg_duration_per_task_ms": 52133.28,
      "avg_output_tokens_per_task": 7649.3,
      "last_run_timestamp": "2026-07-18T08:04:46.262Z"
    },
    {
      "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": 0.706667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.22511376,
      "avg_duration_per_task_ms": 2937291.8,
      "avg_output_tokens_per_task": 55713.12,
      "last_run_timestamp": "2026-10-03T08:06:57.007Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "ModelRun",
      "accuracy": 0.706667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.1838298,
      "avg_duration_per_task_ms": 329811.9,
      "avg_output_tokens_per_task": 9540.84,
      "last_run_timestamp": "2026-08-15T00:16:18.405Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-astra-pro-20260903",
      "model_slug": "openai/gpt-6-astra-pro",
      "model_name": "OpenAI: GPT-6 Astra Pro",
      "provider_name": "OpenAI",
      "accuracy": 0.7066669999999999,
      "accuracy_std_dev": 0.019999999999999962,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 3.04321209,
      "avg_duration_per_task_ms": 317897.29,
      "avg_output_tokens_per_task": 13272.53,
      "last_run_timestamp": "2026-10-02T23:13:16.494Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5-2025-08-07",
      "model_slug": "openai/gpt-5",
      "model_name": "OpenAI: GPT-5",
      "provider_name": "OpenAI",
      "accuracy": 0.7066666666666667,
      "accuracy_std_dev": 0.005443174461857933,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.5994524266666668,
      "avg_duration_per_task_ms": 636541.5466666666,
      "avg_output_tokens_per_task": 54354.86,
      "last_run_timestamp": "2026-10-02T23:22:41.519Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "CoreWeave",
      "accuracy": 0.7066666363636364,
      "accuracy_std_dev": 0.03527669560198404,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.08133863090909091,
      "avg_duration_per_task_ms": 74950.49272727272,
      "avg_output_tokens_per_task": 8535.285454545454,
      "last_run_timestamp": "2026-10-03T16:09:05.688Z"
    },
    {
      "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": 0.7063491754385965,
      "accuracy_std_dev": 0.03950815373122365,
      "run_count": 8,
      "total_questions": 399,
      "avg_cost_per_task_usd": 0.13163428320802006,
      "avg_duration_per_task_ms": 73963.29824561403,
      "avg_output_tokens_per_task": 10257.41604010025,
      "last_run_timestamp": "2026-08-29T17:11:32.461Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5-2025-08-07",
      "model_slug": "openai/gpt-5",
      "model_name": "OpenAI: GPT-5",
      "provider_name": "Azure",
      "accuracy": 0.7056452258064517,
      "accuracy_std_dev": 0.012405393094940622,
      "run_count": 5,
      "total_questions": 248,
      "avg_cost_per_task_usd": 0.6654760282258064,
      "avg_duration_per_task_ms": 686538.5927419355,
      "avg_output_tokens_per_task": 55521.342741935485,
      "last_run_timestamp": "2026-10-01T16:01:02.669Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "WandB",
      "accuracy": 0.7044674329896906,
      "accuracy_std_dev": 0.02710499999999999,
      "run_count": 2,
      "total_questions": 97,
      "avg_cost_per_task_usd": 0.11629732989690722,
      "avg_duration_per_task_ms": 255742.56701030929,
      "avg_output_tokens_per_task": 22402.164948453606,
      "last_run_timestamp": "2026-07-18T08:00:15.297Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.1-terminus",
      "model_slug": "deepseek/deepseek-v3.1-terminus",
      "model_name": "DeepSeek: DeepSeek V3.1 Terminus",
      "provider_name": "auto-routing",
      "accuracy": 0.7039998000000001,
      "accuracy_std_dev": 0.028782815862246673,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.059492744,
      "avg_duration_per_task_ms": 531413.344,
      "avg_output_tokens_per_task": 14170.264,
      "last_run_timestamp": "2026-08-29T16:22:37.842Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Phala",
      "accuracy": 0.7039650806451613,
      "accuracy_std_dev": 0.09864976994620919,
      "run_count": 10,
      "total_questions": 496,
      "avg_cost_per_task_usd": 0.18442565725806453,
      "avg_duration_per_task_ms": 1083811.0504032257,
      "avg_output_tokens_per_task": 49448.979838709674,
      "last_run_timestamp": "2026-10-03T16:07:05.218Z"
    },
    {
      "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-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "Z.AI",
      "accuracy": 0.7029631111111111,
      "accuracy_std_dev": 0.04275812214836289,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.06788864444444445,
      "avg_duration_per_task_ms": 338428.77777777775,
      "avg_output_tokens_per_task": 10021.08,
      "last_run_timestamp": "2026-10-04T00:00:36.971Z"
    },
    {
      "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": 0.7029628888888889,
      "accuracy_std_dev": 0.0630196289675675,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.2980291577777778,
      "avg_duration_per_task_ms": 402620.02222222224,
      "avg_output_tokens_per_task": 9612.011111111111,
      "last_run_timestamp": "2026-10-03T08:12:46.442Z"
    },
    {
      "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": 0.7028571428571428,
      "accuracy_std_dev": 0.027569581319317287,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.042805977142857145,
      "avg_duration_per_task_ms": 188835.22,
      "avg_output_tokens_per_task": 8445.88,
      "last_run_timestamp": "2026-08-29T17:03:30.001Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-sol-20260922",
      "model_slug": "openai/gpt-6-sol",
      "model_name": "OpenAI: GPT-6 Sol",
      "provider_name": "OpenAI",
      "accuracy": 0.7022223333333333,
      "accuracy_std_dev": 0.011330936570096727,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.13357544,
      "avg_duration_per_task_ms": 105586.74666666667,
      "avg_output_tokens_per_task": 5260.826666666667,
      "last_run_timestamp": "2026-10-02T23:07:17.250Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "SiliconFlow",
      "accuracy": 0.70166675,
      "accuracy_std_dev": 0.13101104385198792,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0399951825,
      "avg_duration_per_task_ms": 280949.7025,
      "avg_output_tokens_per_task": 10610.9325,
      "last_run_timestamp": "2026-10-03T16:12:22.581Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5-mini-2025-08-07",
      "model_slug": "openai/gpt-5-mini",
      "model_name": "OpenAI: GPT-5 Mini",
      "provider_name": "Azure",
      "accuracy": 0.701111,
      "accuracy_std_dev": 0.011811220865487748,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.13736028333333333,
      "avg_duration_per_task_ms": 494538.36333333334,
      "avg_output_tokens_per_task": 57774.223333333335,
      "last_run_timestamp": "2026-10-01T19:05:16.547Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Novita",
      "accuracy": 0.7001918563218391,
      "accuracy_std_dev": 0.019518025856805777,
      "run_count": 7,
      "total_questions": 348,
      "avg_cost_per_task_usd": 0.01988737068965517,
      "avg_duration_per_task_ms": 465847.8132183908,
      "avg_output_tokens_per_task": 9664.399425287356,
      "last_run_timestamp": "2026-10-03T16:05:51.293Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-luna-20260709",
      "model_slug": "openai/gpt-5.6-luna",
      "model_name": "OpenAI: GPT-5.6 Luna",
      "provider_name": "auto-routing",
      "accuracy": 0.7000003333333334,
      "accuracy_std_dev": 0.009428090415820642,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.010915493333333333,
      "avg_duration_per_task_ms": 115952.82,
      "avg_output_tokens_per_task": 5345.16,
      "last_run_timestamp": "2026-10-01T19:04:36.191Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5-mini-2025-08-07",
      "model_slug": "openai/gpt-5-mini",
      "model_name": "OpenAI: GPT-5 Mini",
      "provider_name": "auto-routing",
      "accuracy": 0.7000003333333332,
      "accuracy_std_dev": 0.024944382578492925,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.12734470666666667,
      "avg_duration_per_task_ms": 591806.9066666666,
      "avg_output_tokens_per_task": 57532.753333333334,
      "last_run_timestamp": "2026-10-01T19:04:42.411Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Morph",
      "accuracy": 0.7000002,
      "accuracy_std_dev": 0.02828436552868031,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.142434276,
      "avg_duration_per_task_ms": 459844.396,
      "avg_output_tokens_per_task": 11247.26,
      "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": "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": "openai/gpt-5.2-20251211",
      "model_slug": "openai/gpt-5.2",
      "model_name": "OpenAI: GPT-5.2",
      "provider_name": "Azure",
      "accuracy": 0.7,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.31292506,
      "avg_duration_per_task_ms": 261751.42,
      "avg_output_tokens_per_task": 14948.8,
      "last_run_timestamp": "2026-08-01T08:06:08.193Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6.1-sol-20260929",
      "model_slug": "openai/gpt-6.1-sol",
      "model_name": "OpenAI: GPT-6.1 Sol",
      "provider_name": "auto-routing",
      "accuracy": 0.7,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.1012818,
      "avg_duration_per_task_ms": 144103.72,
      "avg_output_tokens_per_task": 3638.14,
      "last_run_timestamp": "2026-10-01T08:03:00.822Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-flash-20260826",
      "model_slug": "qwen/qwen3.8-flash",
      "model_name": "Qwen: Qwen3.8 Flash",
      "provider_name": "Alibaba",
      "accuracy": 0.7,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.019795359999999998,
      "avg_duration_per_task_ms": 396615.94,
      "avg_output_tokens_per_task": 23218.96,
      "last_run_timestamp": "2026-10-01T08:03:06.120Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.5-air",
      "model_slug": "z-ai/glm-4.5-air",
      "model_name": "Z.ai: GLM 4.5 Air",
      "provider_name": "auto-routing",
      "accuracy": 0.7,
      "accuracy_std_dev": 0.01440159456449179,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.017548586666666668,
      "avg_duration_per_task_ms": 296799.9066666667,
      "avg_output_tokens_per_task": 10272.013333333334,
      "last_run_timestamp": "2026-08-01T17:39:06.362Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "Phala",
      "accuracy": 0.7,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.14202821999999998,
      "avg_duration_per_task_ms": 375738.88,
      "avg_output_tokens_per_task": 8552.48,
      "last_run_timestamp": "2026-07-11T15:34:16.332Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Near AI",
      "accuracy": 0.7,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.0112321,
      "avg_duration_per_task_ms": 596563.34,
      "avg_output_tokens_per_task": 4097.46,
      "last_run_timestamp": "2026-09-12T09:21:58.569Z"
    },
    {
      "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": 0.699275,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 46,
      "avg_cost_per_task_usd": 0.02935723913043478,
      "avg_duration_per_task_ms": 277118.347826087,
      "avg_output_tokens_per_task": 12706.934782608696,
      "last_run_timestamp": "2026-10-03T16:14:03.920Z"
    },
    {
      "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": 0.6986666,
      "accuracy_std_dev": 0.007483243510671025,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.04888268,
      "avg_duration_per_task_ms": 227205.572,
      "avg_output_tokens_per_task": 6367.684,
      "last_run_timestamp": "2026-08-22T08:43:02.347Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "DeepInfra",
      "accuracy": 0.6983967174348699,
      "accuracy_std_dev": 0.025880910808740845,
      "run_count": 10,
      "total_questions": 499,
      "avg_cost_per_task_usd": 0.03737082565130261,
      "avg_duration_per_task_ms": 243092.8116232465,
      "avg_output_tokens_per_task": 9769.018036072144,
      "last_run_timestamp": "2026-08-29T16:45:30.973Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-luna-pro-20260709",
      "model_slug": "openai/gpt-5.6-luna-pro",
      "model_name": "OpenAI: GPT-5.6 Luna Pro",
      "provider_name": "Azure",
      "accuracy": 0.6977778333333333,
      "accuracy_std_dev": 0.017915990515148434,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.21024033333333336,
      "avg_duration_per_task_ms": 335029.48333333334,
      "avg_output_tokens_per_task": 34166.46666666667,
      "last_run_timestamp": "2026-10-01T16:01:04.829Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "auto-routing",
      "accuracy": 0.6977776666666667,
      "accuracy_std_dev": 0.032659953959959374,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.025667106666666665,
      "avg_duration_per_task_ms": 221633.35777777777,
      "avg_output_tokens_per_task": 9824.533333333333,
      "last_run_timestamp": "2026-10-03T16:10:17.811Z"
    },
    {
      "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": 0.6971726205357143,
      "accuracy_std_dev": 0.03692683540345123,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.045649031250000006,
      "avg_duration_per_task_ms": 440975.8839285714,
      "avg_output_tokens_per_task": 9165.5,
      "last_run_timestamp": "2026-10-03T16:26:46.014Z"
    },
    {
      "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": 0.6971425714285715,
      "accuracy_std_dev": 0.021412676644969954,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.05162463142857143,
      "avg_duration_per_task_ms": 387031.3257142857,
      "avg_output_tokens_per_task": 9269.174285714285,
      "last_run_timestamp": "2026-10-03T16:12:07.463Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-astra-20260903",
      "model_slug": "openai/gpt-6-astra",
      "model_name": "OpenAI: GPT-6 Astra",
      "provider_name": "Azure",
      "accuracy": 0.6966665000000001,
      "accuracy_std_dev": 0.0033334999999999892,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.62488186,
      "avg_duration_per_task_ms": 152081.62,
      "avg_output_tokens_per_task": 3927.45,
      "last_run_timestamp": "2026-10-01T08:05:41.499Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-sol-20260922",
      "model_slug": "openai/gpt-6-sol",
      "model_name": "OpenAI: GPT-6 Sol",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.6966665000000001,
      "accuracy_std_dev": 0.023333500000000007,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.14710045000000002,
      "avg_duration_per_task_ms": 163276.98,
      "avg_output_tokens_per_task": 5053.91,
      "last_run_timestamp": "2026-10-01T08:02:59.984Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Reka",
      "accuracy": 0.6958335,
      "accuracy_std_dev": 0.03781341986980283,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.132227115,
      "avg_duration_per_task_ms": 205681.995,
      "avg_output_tokens_per_task": 9407.695,
      "last_run_timestamp": "2026-10-01T22:50:26.018Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "AtlasCloud",
      "accuracy": 0.6951851111111111,
      "accuracy_std_dev": 0.07109563585519382,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.05191746444444445,
      "avg_duration_per_task_ms": 304269.29333333333,
      "avg_output_tokens_per_task": 8106.5888888888885,
      "last_run_timestamp": "2026-10-03T16:13:53.555Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Venice",
      "accuracy": 0.6936709012658228,
      "accuracy_std_dev": 0.02885537658495161,
      "run_count": 8,
      "total_questions": 395,
      "avg_cost_per_task_usd": 0.017572048101265822,
      "avg_duration_per_task_ms": 298381.053164557,
      "avg_output_tokens_per_task": 9343.311392405063,
      "last_run_timestamp": "2026-08-29T17:28:21.827Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "AtlasCloud",
      "accuracy": 0.6933334999999999,
      "accuracy_std_dev": 0.04422165130905898,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.029941567500000002,
      "avg_duration_per_task_ms": 308894.385,
      "avg_output_tokens_per_task": 10124.7675,
      "last_run_timestamp": "2026-10-03T16:07:02.038Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "InferenceNet",
      "accuracy": 0.6933334999999999,
      "accuracy_std_dev": 0.013333499999999998,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.04339576,
      "avg_duration_per_task_ms": 106333.2,
      "avg_output_tokens_per_task": 10497.62,
      "last_run_timestamp": "2026-10-03T16:12:20.058Z"
    },
    {
      "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": "z-ai/glm-4.5-air",
      "model_slug": "z-ai/glm-4.5-air",
      "model_name": "Z.ai: GLM 4.5 Air",
      "provider_name": "Z.AI",
      "accuracy": 0.6933333333333334,
      "accuracy_std_dev": 0.010886757164351354,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.020424833333333333,
      "avg_duration_per_task_ms": 454611.69333333336,
      "avg_output_tokens_per_task": 10363.686666666666,
      "last_run_timestamp": "2026-08-01T17:38:51.399Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.693333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.01596628,
      "avg_duration_per_task_ms": 345497.04,
      "avg_output_tokens_per_task": 18466.26,
      "last_run_timestamp": "2026-10-03T16:11:33.011Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "Ionstream",
      "accuracy": 0.693333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.015592,
      "avg_duration_per_task_ms": 1084747.56,
      "avg_output_tokens_per_task": 17745.48,
      "last_run_timestamp": "2026-07-31T03:44:20.825Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-sol-20260922",
      "model_slug": "openai/gpt-6-sol",
      "model_name": "OpenAI: GPT-6 Sol",
      "provider_name": "auto-routing",
      "accuracy": 0.693333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.12999802000000002,
      "avg_duration_per_task_ms": 96250.54,
      "avg_output_tokens_per_task": 4982.06,
      "last_run_timestamp": "2026-10-01T08:05:35.867Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "stepfun/step-3.7-flash-20260528",
      "model_slug": "stepfun/step-3.7-flash",
      "model_name": "StepFun: Step 3.7 Flash",
      "provider_name": "Ambient",
      "accuracy": 0.693333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.02187568,
      "avg_duration_per_task_ms": 236451.76,
      "avg_output_tokens_per_task": 11760.74,
      "last_run_timestamp": "2026-07-18T08:04:03.248Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.2-exp",
      "model_slug": "deepseek/deepseek-v3.2-exp",
      "model_name": "DeepSeek: DeepSeek V3.2 Exp",
      "provider_name": "SiliconFlow",
      "accuracy": 0.6933329999999999,
      "accuracy_std_dev": 0,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.08326614666666667,
      "avg_duration_per_task_ms": 757569.06,
      "avg_output_tokens_per_task": 14889.873333333333,
      "last_run_timestamp": "2026-08-01T17:38:16.661Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Mistral",
      "accuracy": 0.6929165,
      "accuracy_std_dev": 0.02385937097609657,
      "run_count": 16,
      "total_questions": 800,
      "avg_cost_per_task_usd": 0.10850378125,
      "avg_duration_per_task_ms": 77457.025,
      "avg_output_tokens_per_task": 9370.97875,
      "last_run_timestamp": "2026-10-03T16:16:14.856Z"
    },
    {
      "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": 0.6925926777777779,
      "accuracy_std_dev": 0.05766701577287278,
      "run_count": 11,
      "total_questions": 540,
      "avg_cost_per_task_usd": 0.031123783333333335,
      "avg_duration_per_task_ms": 963670.3796296297,
      "avg_output_tokens_per_task": 42037.30925925926,
      "last_run_timestamp": "2026-10-03T16:10:32.034Z"
    },
    {
      "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": 0.692340005050505,
      "accuracy_std_dev": 0.01636465616379931,
      "run_count": 8,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.02100930303030303,
      "avg_duration_per_task_ms": 803131.053030303,
      "avg_output_tokens_per_task": 19211.46212121212,
      "last_run_timestamp": "2026-10-03T16:07:44.275Z"
    },
    {
      "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": 0.6920472487309645,
      "accuracy_std_dev": 0.10126504037171156,
      "run_count": 8,
      "total_questions": 394,
      "avg_cost_per_task_usd": 0.0741332766497462,
      "avg_duration_per_task_ms": 201116.1852791878,
      "avg_output_tokens_per_task": 25341.34771573604,
      "last_run_timestamp": "2026-10-03T16:12:22.576Z"
    },
    {
      "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": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Io Net",
      "accuracy": 0.68999975,
      "accuracy_std_dev": 0.03605553586742955,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.113432665,
      "avg_duration_per_task_ms": 675363.205,
      "avg_output_tokens_per_task": 27403.155,
      "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": "openai/gpt-5-2025-08-07",
      "model_slug": "openai/gpt-5",
      "model_name": "OpenAI: GPT-5",
      "provider_name": "auto-routing",
      "accuracy": 0.6891892297297297,
      "accuracy_std_dev": 0.03170486152346707,
      "run_count": 3,
      "total_questions": 148,
      "avg_cost_per_task_usd": 0.6622049662162162,
      "avg_duration_per_task_ms": 777791.8783783783,
      "avg_output_tokens_per_task": 59355.4527027027,
      "last_run_timestamp": "2026-10-01T16:01:02.610Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.1",
      "model_slug": "minimax/minimax-m2.1",
      "model_name": "MiniMax: MiniMax M2.1",
      "provider_name": "Minimax",
      "accuracy": 0.6891667499999999,
      "accuracy_std_dev": 0.05343089124221587,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.029582090000000002,
      "avg_duration_per_task_ms": 175383.2625,
      "avg_output_tokens_per_task": 10636.405,
      "last_run_timestamp": "2026-08-01T17:38:12.859Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "Phala",
      "accuracy": 0.6885714285714286,
      "accuracy_std_dev": 0.029913183528610335,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.19170214285714285,
      "avg_duration_per_task_ms": 1896431.7657142857,
      "avg_output_tokens_per_task": 13131.025714285714,
      "last_run_timestamp": "2026-10-04T00:00:37.132Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Friendli",
      "accuracy": 0.688485,
      "accuracy_std_dev": 0.07169463814489388,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.11140403636363637,
      "avg_duration_per_task_ms": 69132.01818181819,
      "avg_output_tokens_per_task": 8823.258181818182,
      "last_run_timestamp": "2026-10-03T16:27:27.989Z"
    },
    {
      "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": 0.6872895959595959,
      "accuracy_std_dev": 0.028668953856916365,
      "run_count": 8,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.03334008585858586,
      "avg_duration_per_task_ms": 764612.6868686868,
      "avg_output_tokens_per_task": 19659.290404040403,
      "last_run_timestamp": "2026-10-03T16:26:51.484Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "inclusionai/ling-3.0-flash-20260723",
      "model_slug": "inclusionai/ling-3.0-flash",
      "model_name": "inclusionAI: Ling 3.0 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.686667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.00219582,
      "avg_duration_per_task_ms": 78121.92,
      "avg_output_tokens_per_task": 11659.46,
      "last_run_timestamp": "2026-10-01T16:00:49.880Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.5-air",
      "model_slug": "z-ai/glm-4.5-air",
      "model_name": "Z.ai: GLM 4.5 Air",
      "provider_name": "SiliconFlow",
      "accuracy": 0.6866666666666666,
      "accuracy_std_dev": 0.014401903168987411,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.04308245333333333,
      "avg_duration_per_task_ms": 485038.98,
      "avg_output_tokens_per_task": 10019.313333333334,
      "last_run_timestamp": "2026-08-01T17:38:21.896Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "DigitalOcean",
      "accuracy": 0.6866666,
      "accuracy_std_dev": 0.08753423203433042,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.009446468,
      "avg_duration_per_task_ms": 231025.728,
      "avg_output_tokens_per_task": 4090.644,
      "last_run_timestamp": "2026-10-03T08:37:57.576Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2-thinking-20251106",
      "model_slug": "moonshotai/kimi-k2-thinking",
      "model_name": "MoonshotAI: Kimi K2 Thinking",
      "provider_name": "Google",
      "accuracy": 0.6858332500000001,
      "accuracy_std_dev": 0.009538074291359878,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.11306765,
      "avg_duration_per_task_ms": 138784.995,
      "avg_output_tokens_per_task": 8808.785,
      "last_run_timestamp": "2026-08-01T17:38:46.618Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-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": "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": 0.685,
      "accuracy_std_dev": 0.025110843165851677,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0074074375,
      "avg_duration_per_task_ms": 254406.3375,
      "avg_output_tokens_per_task": 11400.4,
      "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": "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": 0.6846240067340067,
      "accuracy_std_dev": 0.017472339200544633,
      "run_count": 6,
      "total_questions": 297,
      "avg_cost_per_task_usd": 0.01174172053872054,
      "avg_duration_per_task_ms": 245540.93602693602,
      "avg_output_tokens_per_task": 9163.602693602694,
      "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": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "AtlasCloud",
      "accuracy": 0.6844445555555555,
      "accuracy_std_dev": 0.052774778862868674,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.09775986666666667,
      "avg_duration_per_task_ms": 262551.32,
      "avg_output_tokens_per_task": 10512.555555555555,
      "last_run_timestamp": "2026-10-03T16:11:17.903Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Minimax",
      "accuracy": 0.6841667499999999,
      "accuracy_std_dev": 0.042023533421018985,
      "run_count": 16,
      "total_questions": 800,
      "avg_cost_per_task_usd": 0.03507941999999999,
      "avg_duration_per_task_ms": 165936.86375,
      "avg_output_tokens_per_task": 10189.80625,
      "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": 0.6838585558739255,
      "accuracy_std_dev": 0.02128582066608127,
      "run_count": 7,
      "total_questions": 349,
      "avg_cost_per_task_usd": 0.006118083094555874,
      "avg_duration_per_task_ms": 239133.92550143268,
      "avg_output_tokens_per_task": 9429.15759312321,
      "last_run_timestamp": "2026-08-30T00:01:09.439Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6.1-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": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "OpenInference",
      "accuracy": 0.68333325,
      "accuracy_std_dev": 0.06368309125024867,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.05480554,
      "avg_duration_per_task_ms": 1657352.435,
      "avg_output_tokens_per_task": 54132.445,
      "last_run_timestamp": "2026-10-03T08:39:18.140Z"
    },
    {
      "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": "Morph",
      "accuracy": 0.683333,
      "accuracy_std_dev": 0.010000000000000009,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.05961793,
      "avg_duration_per_task_ms": 370273.51,
      "avg_output_tokens_per_task": 10007.47,
      "last_run_timestamp": "2026-07-18T08:04:08.785Z"
    },
    {
      "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": "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": 0.6820807398843931,
      "accuracy_std_dev": 0.018427811270036426,
      "run_count": 7,
      "total_questions": 346,
      "avg_cost_per_task_usd": 0.015085965317919076,
      "avg_duration_per_task_ms": 243924.5809248555,
      "avg_output_tokens_per_task": 11146.326589595375,
      "last_run_timestamp": "2026-08-30T00:01:09.469Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-luna-20260709",
      "model_slug": "openai/gpt-5.6-luna",
      "model_name": "OpenAI: GPT-5.6 Luna",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.6816665000000001,
      "accuracy_std_dev": 0.02327362599059285,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.013877145,
      "avg_duration_per_task_ms": 70751.9,
      "avg_output_tokens_per_task": 5304.49,
      "last_run_timestamp": "2026-10-01T19:04:52.652Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "thinkingmachines/inkling-20260715",
      "model_slug": "thinkingmachines/inkling",
      "model_name": "Thinking Machines: Inkling",
      "provider_name": "Together",
      "accuracy": 0.6813334,
      "accuracy_std_dev": 0.0026667999999999913,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.08704178,
      "avg_duration_per_task_ms": 82752.556,
      "avg_output_tokens_per_task": 3576.016,
      "last_run_timestamp": "2026-10-01T16:01:04.859Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-luna-20260922",
      "model_slug": "openai/gpt-6-luna",
      "model_name": "OpenAI: GPT-6 Luna",
      "provider_name": "Azure",
      "accuracy": 0.6811111666666666,
      "accuracy_std_dev": 0.008088865998326882,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.00903582,
      "avg_duration_per_task_ms": 105469.92666666667,
      "avg_output_tokens_per_task": 8459.866666666667,
      "last_run_timestamp": "2026-10-01T08:05:22.516Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4-sonnet-20250522",
      "model_slug": "anthropic/claude-sonnet-4",
      "model_name": "Anthropic: Claude Sonnet 4",
      "provider_name": "auto-routing",
      "accuracy": 0.68,
      "accuracy_std_dev": 0.019999999999999962,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.33301777,
      "avg_duration_per_task_ms": 171524.99,
      "avg_output_tokens_per_task": 8948.71,
      "last_run_timestamp": "2026-09-01T12:41:24.616Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-astra-20260903",
      "model_slug": "openai/gpt-6-astra",
      "model_name": "OpenAI: GPT-6 Astra",
      "provider_name": "auto-routing",
      "accuracy": 0.68,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.58024224,
      "avg_duration_per_task_ms": 141487.98,
      "avg_output_tokens_per_task": 3731.58,
      "last_run_timestamp": "2026-10-01T08:05:19.702Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-luna-20260922",
      "model_slug": "openai/gpt-6-luna",
      "model_name": "OpenAI: GPT-6 Luna",
      "provider_name": "OpenAI",
      "accuracy": 0.68,
      "accuracy_std_dev": 0.03810292105162893,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.00883898,
      "avg_duration_per_task_ms": 114396.12,
      "avg_output_tokens_per_task": 8841.28,
      "last_run_timestamp": "2026-10-02T23:08:44.329Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.6791665,
      "accuracy_std_dev": 0.04377175966876361,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.020768927500000003,
      "avg_duration_per_task_ms": 425360.89,
      "avg_output_tokens_per_task": 20325.145,
      "last_run_timestamp": "2026-10-03T16:11:33.008Z"
    },
    {
      "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": 0.6785186666666666,
      "accuracy_std_dev": 0.019570227381861923,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.00874342888888889,
      "avg_duration_per_task_ms": 244185.2288888889,
      "avg_output_tokens_per_task": 10575.206666666667,
      "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": "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-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Friendli",
      "accuracy": 0.677500125,
      "accuracy_std_dev": 0.09463012829225889,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.08997287749999999,
      "avg_duration_per_task_ms": 87500.14,
      "avg_output_tokens_per_task": 8449.92,
      "last_run_timestamp": "2026-10-03T16:14:25.374Z"
    },
    {
      "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": "DeepInfra",
      "accuracy": 0.676667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.0194754,
      "avg_duration_per_task_ms": 233472.04,
      "avg_output_tokens_per_task": 10655.96,
      "last_run_timestamp": "2026-07-11T15:34:15.373Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.2-exp",
      "model_slug": "deepseek/deepseek-v3.2-exp",
      "model_name": "DeepSeek: DeepSeek V3.2 Exp",
      "provider_name": "Novita",
      "accuracy": 0.6766665000000001,
      "accuracy_std_dev": 0.0033335000000000448,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.09087626,
      "avg_duration_per_task_ms": 1373764.79,
      "avg_output_tokens_per_task": 16227.07,
      "last_run_timestamp": "2026-07-18T16:00:08.839Z"
    },
    {
      "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-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Nebius",
      "accuracy": 0.675675972972973,
      "accuracy_std_dev": 0.02284471405428887,
      "run_count": 3,
      "total_questions": 148,
      "avg_cost_per_task_usd": 0.04042408108108108,
      "avg_duration_per_task_ms": 518660.3783783784,
      "avg_output_tokens_per_task": 20990.527027027027,
      "last_run_timestamp": "2026-09-26T08:11:26.490Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-astra-20260903",
      "model_slug": "openai/gpt-6-astra",
      "model_name": "OpenAI: GPT-6 Astra",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.6733334999999999,
      "accuracy_std_dev": 0.00666650000000002,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.64995528,
      "avg_duration_per_task_ms": 85231.23,
      "avg_output_tokens_per_task": 3880.06,
      "last_run_timestamp": "2026-10-01T08:05:26.145Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "thinkingmachines/inkling-20260715",
      "model_slug": "thinkingmachines/inkling",
      "model_name": "Thinking Machines: Inkling",
      "provider_name": "DeepInfra",
      "accuracy": 0.6733334999999999,
      "accuracy_std_dev": 0.00666650000000002,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.06914647,
      "avg_duration_per_task_ms": 138661.04,
      "avg_output_tokens_per_task": 2568.53,
      "last_run_timestamp": "2026-10-01T16:01:04.851Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.1-flash-lite-20260507",
      "model_slug": "google/gemini-3.1-flash-lite",
      "model_name": "Google: Gemini 3.1 Flash Lite",
      "provider_name": "Google AI Studio",
      "accuracy": 0.6733333333333333,
      "accuracy_std_dev": 0.02372680934208296,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.11501549333333332,
      "avg_duration_per_task_ms": 145439.36666666667,
      "avg_output_tokens_per_task": 56059.84,
      "last_run_timestamp": "2026-10-01T19:05:16.548Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Nebius",
      "accuracy": 0.6733333333333333,
      "accuracy_std_dev": 0.05192585882668568,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.22792725999999996,
      "avg_duration_per_task_ms": 692249.9666666667,
      "avg_output_tokens_per_task": 10865.973333333333,
      "last_run_timestamp": "2026-10-03T16:12:13.503Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.2-chat-20251211",
      "model_slug": "openai/gpt-5.2-chat",
      "model_name": "OpenAI: GPT-5.2 Chat",
      "provider_name": "auto-routing",
      "accuracy": 0.673333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.1116277,
      "avg_duration_per_task_ms": 79582.44,
      "avg_output_tokens_per_task": 4280.42,
      "last_run_timestamp": "2026-08-01T08:05:57.504Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-luna-20260922",
      "model_slug": "openai/gpt-6-luna",
      "model_name": "OpenAI: GPT-6 Luna",
      "provider_name": "auto-routing",
      "accuracy": 0.673333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.00928508,
      "avg_duration_per_task_ms": 121838.98,
      "avg_output_tokens_per_task": 8898.44,
      "last_run_timestamp": "2026-10-01T08:00:22.469Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "DigitalOcean",
      "accuracy": 0.6715151818181818,
      "accuracy_std_dev": 0.0638870956593358,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.06997897636363637,
      "avg_duration_per_task_ms": 726161.1018181818,
      "avg_output_tokens_per_task": 18626.429090909092,
      "last_run_timestamp": "2026-10-03T08:12:20.016Z"
    },
    {
      "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": 0.671111,
      "accuracy_std_dev": 0.034282673881196074,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.1170691888888889,
      "avg_duration_per_task_ms": 122970.47333333333,
      "avg_output_tokens_per_task": 12647.033333333333,
      "last_run_timestamp": "2026-10-03T08:12:32.364Z"
    },
    {
      "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": 0.67083325,
      "accuracy_std_dev": 0.019346895388860186,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.017078424999999998,
      "avg_duration_per_task_ms": 126227.985,
      "avg_output_tokens_per_task": 21265.575,
      "last_run_timestamp": "2026-10-03T08:15:42.485Z"
    },
    {
      "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": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "DeepInfra",
      "accuracy": 0.67,
      "accuracy_std_dev": 0.037267605832679934,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.02244593,
      "avg_duration_per_task_ms": 202743.04,
      "avg_output_tokens_per_task": 31421.765,
      "last_run_timestamp": "2026-10-03T08:12:46.440Z"
    },
    {
      "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": "DigitalOcean",
      "accuracy": 0.67,
      "accuracy_std_dev": 0.02999999999999997,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.014700359999999999,
      "avg_duration_per_task_ms": 1773933.08,
      "avg_output_tokens_per_task": 23989.78,
      "last_run_timestamp": "2026-07-18T08:04:08.007Z"
    },
    {
      "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": 0.6692307307692308,
      "accuracy_std_dev": 0.031054260799641166,
      "run_count": 26,
      "total_questions": 1300,
      "avg_cost_per_task_usd": 0.12451571153846153,
      "avg_duration_per_task_ms": 119043.02846153846,
      "avg_output_tokens_per_task": 12553.901538461538,
      "last_run_timestamp": "2026-10-03T08:32:52.484Z"
    },
    {
      "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": 0.669204852791878,
      "accuracy_std_dev": 0.0421484109961455,
      "run_count": 8,
      "total_questions": 394,
      "avg_cost_per_task_usd": 0.023811266497461927,
      "avg_duration_per_task_ms": 266573.3654822335,
      "avg_output_tokens_per_task": 17577.88578680203,
      "last_run_timestamp": "2026-10-03T16:10:49.374Z"
    },
    {
      "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": 0.6690613965517241,
      "accuracy_std_dev": 0.061175733864975976,
      "run_count": 7,
      "total_questions": 348,
      "avg_cost_per_task_usd": 0.023726439655172418,
      "avg_duration_per_task_ms": 462159.5632183908,
      "avg_output_tokens_per_task": 16750.870689655174,
      "last_run_timestamp": "2026-10-03T16:16:54.867Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.5-air",
      "model_slug": "z-ai/glm-4.5-air",
      "model_name": "Z.ai: GLM 4.5 Air",
      "provider_name": "Novita",
      "accuracy": 0.66833325,
      "accuracy_std_dev": 0.023746444179866155,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.016393715,
      "avg_duration_per_task_ms": 283331.04,
      "avg_output_tokens_per_task": 10626.035,
      "last_run_timestamp": "2026-08-01T17:38:31.690Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-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": "Phala",
      "accuracy": 0.666667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.03810232,
      "avg_duration_per_task_ms": 661152.42,
      "avg_output_tokens_per_task": 27538.94,
      "last_run_timestamp": "2026-07-11T15:34:14.992Z"
    },
    {
      "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": 0.6658330000000001,
      "accuracy_std_dev": 0.014790199457749011,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.02195716,
      "avg_duration_per_task_ms": 162018.975,
      "avg_output_tokens_per_task": 16412.775,
      "last_run_timestamp": "2026-10-03T16:06:47.455Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "thinkingmachines/inkling-20260715",
      "model_slug": "thinkingmachines/inkling",
      "model_name": "Thinking Machines: Inkling",
      "provider_name": "BaseTen",
      "accuracy": 0.6655251643835617,
      "accuracy_std_dev": 0.050919510401766024,
      "run_count": 3,
      "total_questions": 146,
      "avg_cost_per_task_usd": 0.0655772808219178,
      "avg_duration_per_task_ms": 50910.90410958904,
      "avg_output_tokens_per_task": 3110.958904109589,
      "last_run_timestamp": "2026-09-12T08:11:26.919Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Friendli",
      "accuracy": 0.663333375,
      "accuracy_std_dev": 0.022360549338832768,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.029954157500000002,
      "avg_duration_per_task_ms": 239902.565,
      "avg_output_tokens_per_task": 11150.49,
      "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": "anthropic/claude-4-sonnet-20250522",
      "model_slug": "anthropic/claude-sonnet-4",
      "model_name": "Anthropic: Claude Sonnet 4",
      "provider_name": "Google",
      "accuracy": 0.6622221666666667,
      "accuracy_std_dev": 0.0198761101108564,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.31710583,
      "avg_duration_per_task_ms": 192864.10333333333,
      "avg_output_tokens_per_task": 8814.376666666667,
      "last_run_timestamp": "2026-09-01T13:04:20.137Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.5-haiku-20251001",
      "model_slug": "anthropic/claude-haiku-4.5",
      "model_name": "Anthropic: Claude Haiku 4.5",
      "provider_name": "Anthropic",
      "accuracy": 0.6603704444444444,
      "accuracy_std_dev": 0.02063470181843689,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.11393844222222221,
      "avg_duration_per_task_ms": 124032.99777777778,
      "avg_output_tokens_per_task": 12195,
      "last_run_timestamp": "2026-10-03T08:07:11.726Z"
    },
    {
      "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": 0.66,
      "accuracy_std_dev": 0.04000011666732984,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.018718389999999998,
      "avg_duration_per_task_ms": 348125.166,
      "avg_output_tokens_per_task": 19960.056,
      "last_run_timestamp": "2026-10-03T16:07:05.219Z"
    },
    {
      "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": 0.66,
      "accuracy_std_dev": 0.055243555012729276,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.022544044444444442,
      "avg_duration_per_task_ms": 206377.97777777776,
      "avg_output_tokens_per_task": 9087.275555555556,
      "last_run_timestamp": "2026-10-03T16:13:50.327Z"
    },
    {
      "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": "WandB",
      "accuracy": 0.66,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.00861218,
      "avg_duration_per_task_ms": 379136.4,
      "avg_output_tokens_per_task": 24201,
      "last_run_timestamp": "2026-07-18T16:00:10.533Z"
    },
    {
      "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": 0.66,
      "accuracy_std_dev": 0.08092060123218178,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.07575899333333333,
      "avg_duration_per_task_ms": 264060.94444444444,
      "avg_output_tokens_per_task": 24785.226666666666,
      "last_run_timestamp": "2026-10-03T16:03:45.631Z"
    },
    {
      "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": 0.659524,
      "accuracy_std_dev": 0.0937819106087553,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.058310568571428564,
      "avg_duration_per_task_ms": 524817.9057142857,
      "avg_output_tokens_per_task": 17039.157142857144,
      "last_run_timestamp": "2026-10-03T16:12:24.719Z"
    },
    {
      "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": 0.6580951428571429,
      "accuracy_std_dev": 0.021884097914947755,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.022360888571428574,
      "avg_duration_per_task_ms": 259146.80571428573,
      "avg_output_tokens_per_task": 9927.357142857143,
      "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": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "auto-routing",
      "accuracy": 0.65666675,
      "accuracy_std_dev": 0.024720818937436124,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.02259439,
      "avg_duration_per_task_ms": 296251.1175,
      "avg_output_tokens_per_task": 10399.475,
      "last_run_timestamp": "2026-08-29T16:11:34.887Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "DigitalOcean",
      "accuracy": 0.6566666250000001,
      "accuracy_std_dev": 0.017638208213545233,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.026674842500000004,
      "avg_duration_per_task_ms": 368698.9125,
      "avg_output_tokens_per_task": 10712.57,
      "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": "inclusionai/ling-3.0-flash-20260723",
      "model_slug": "inclusionai/ling-3.0-flash",
      "model_name": "inclusionAI: Ling 3.0 Flash",
      "provider_name": "DeepInfra",
      "accuracy": 0.6566665,
      "accuracy_std_dev": 0.0833335,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.009373,
      "avg_duration_per_task_ms": 175154.56,
      "avg_output_tokens_per_task": 15943.33,
      "last_run_timestamp": "2026-10-01T16:01:04.874Z"
    },
    {
      "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": 0.6566665,
      "accuracy_std_dev": 0.03666649999999999,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.008972560000000001,
      "avg_duration_per_task_ms": 569999.06,
      "avg_output_tokens_per_task": 21009.53,
      "last_run_timestamp": "2026-10-03T16:14:02.460Z"
    },
    {
      "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": 0.6560493333333334,
      "accuracy_std_dev": 0.029495577726286477,
      "run_count": 27,
      "total_questions": 1350,
      "avg_cost_per_task_usd": 0.12169699629629627,
      "avg_duration_per_task_ms": 120224.82148148149,
      "avg_output_tokens_per_task": 12512.395555555555,
      "last_run_timestamp": "2026-10-03T08:13:54.011Z"
    },
    {
      "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": 0.6557788844221106,
      "accuracy_std_dev": 0.01812503680513505,
      "run_count": 4,
      "total_questions": 199,
      "avg_cost_per_task_usd": 0.017334296482412057,
      "avg_duration_per_task_ms": 354124.29145728645,
      "avg_output_tokens_per_task": 22535.824120603014,
      "last_run_timestamp": "2026-10-03T08:21:59.532Z"
    },
    {
      "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": 0.6555554666666666,
      "accuracy_std_dev": 0.010407070153866542,
      "run_count": 4,
      "total_questions": 195,
      "avg_cost_per_task_usd": 0.014745425641025642,
      "avg_duration_per_task_ms": 278015.8512820513,
      "avg_output_tokens_per_task": 20205.384615384617,
      "last_run_timestamp": "2026-10-03T08:06:43.564Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "inclusionai/ling-3.0-flash-fin-20260827",
      "model_slug": "inclusionai/ling-3.0-flash-fin",
      "model_name": "inclusionAI: Ling 3.0 Flash Fin",
      "provider_name": "Novita",
      "accuracy": 0.6533335,
      "accuracy_std_dev": 0.00666650000000002,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.00488981,
      "avg_duration_per_task_ms": 111736.2,
      "avg_output_tokens_per_task": 17377.29,
      "last_run_timestamp": "2026-10-01T08:04:38.584Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.4-nano-20260317",
      "model_slug": "openai/gpt-5.4-nano",
      "model_name": "OpenAI: GPT-5.4 Nano",
      "provider_name": "auto-routing",
      "accuracy": 0.6533333333333333,
      "accuracy_std_dev": 0.023726809342082957,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.03210012,
      "avg_duration_per_task_ms": 143919.89333333334,
      "avg_output_tokens_per_task": 17353.273333333334,
      "last_run_timestamp": "2026-10-01T19:04:28.503Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-luna-20260922",
      "model_slug": "openai/gpt-6-luna",
      "model_name": "OpenAI: GPT-6 Luna",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.653333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.00923216,
      "avg_duration_per_task_ms": 145753.66,
      "avg_output_tokens_per_task": 8192.38,
      "last_run_timestamp": "2026-09-23T06:00:08.549Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.4-nano-20260317",
      "model_slug": "openai/gpt-5.4-nano",
      "model_name": "OpenAI: GPT-5.4 Nano",
      "provider_name": "OpenAI",
      "accuracy": 0.65166675,
      "accuracy_std_dev": 0.012801703448662586,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.031589305,
      "avg_duration_per_task_ms": 139715.76,
      "avg_output_tokens_per_task": 17049.6,
      "last_run_timestamp": "2026-10-02T23:09:54.584Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "thinkingmachines/inkling-small-20260730",
      "model_slug": "thinkingmachines/inkling-small",
      "model_name": "Thinking Machines: Inkling Small",
      "provider_name": "Together",
      "accuracy": 0.65166675,
      "accuracy_std_dev": 0.04506177549972369,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.04474220500000001,
      "avg_duration_per_task_ms": 212693.005,
      "avg_output_tokens_per_task": 3236.785,
      "last_run_timestamp": "2026-09-12T08:12:16.067Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "auto-routing",
      "accuracy": 0.6503832528735632,
      "accuracy_std_dev": 0.09824690755999416,
      "run_count": 7,
      "total_questions": 348,
      "avg_cost_per_task_usd": 0.07506860057471264,
      "avg_duration_per_task_ms": 844732.393678161,
      "avg_output_tokens_per_task": 47265.1724137931,
      "last_run_timestamp": "2026-10-03T16:13:53.353Z"
    },
    {
      "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": 0.649166625,
      "accuracy_std_dev": 0.029238002051856674,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.021406732500000004,
      "avg_duration_per_task_ms": 233528.235,
      "avg_output_tokens_per_task": 10579.235,
      "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": "AkashML",
      "accuracy": 0.6485233670886075,
      "accuracy_std_dev": 0.02680153729616082,
      "run_count": 8,
      "total_questions": 395,
      "avg_cost_per_task_usd": 0.011709784810126583,
      "avg_duration_per_task_ms": 286395.1898734177,
      "avg_output_tokens_per_task": 16354.617721518987,
      "last_run_timestamp": "2026-10-03T16:05:34.164Z"
    },
    {
      "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": 0.6483331666666667,
      "accuracy_std_dev": 0.0255133770299469,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.022298810000000002,
      "avg_duration_per_task_ms": 289277.4166666667,
      "avg_output_tokens_per_task": 10734.94,
      "last_run_timestamp": "2026-08-15T14:15:19.430Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.5-haiku-20251001",
      "model_slug": "anthropic/claude-haiku-4.5",
      "model_name": "Anthropic: Claude Haiku 4.5",
      "provider_name": "Azure",
      "accuracy": 0.6481482222222222,
      "accuracy_std_dev": 0.02470134810436142,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.11660464888888887,
      "avg_duration_per_task_ms": 145744.04888888888,
      "avg_output_tokens_per_task": 12598.522222222222,
      "last_run_timestamp": "2026-10-03T08:13:37.943Z"
    },
    {
      "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": 0.6481482020202021,
      "accuracy_std_dev": 0.0054059166891471805,
      "run_count": 4,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.009114631313131312,
      "avg_duration_per_task_ms": 604721.6666666666,
      "avg_output_tokens_per_task": 19746.29292929293,
      "last_run_timestamp": "2026-08-22T16:10:30.348Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.4-mini-20260317",
      "model_slug": "openai/gpt-5.4-mini",
      "model_name": "OpenAI: GPT-5.4 Mini",
      "provider_name": "Azure",
      "accuracy": 0.6477778333333334,
      "accuracy_std_dev": 0.03430091421724631,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.09916395333333333,
      "avg_duration_per_task_ms": 153090.92333333334,
      "avg_output_tokens_per_task": 12859.136666666667,
      "last_run_timestamp": "2026-10-01T19:04:42.436Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "AtlasCloud",
      "accuracy": 0.6465774285714286,
      "accuracy_std_dev": 0.10466666390033022,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.09237277678571428,
      "avg_duration_per_task_ms": 238254.9263392857,
      "avg_output_tokens_per_task": 25810.723214285714,
      "last_run_timestamp": "2026-10-03T16:12:24.719Z"
    },
    {
      "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": 0.646566095477387,
      "accuracy_std_dev": 0.12986162562313927,
      "run_count": 8,
      "total_questions": 398,
      "avg_cost_per_task_usd": 0.04628975628140704,
      "avg_duration_per_task_ms": 2127959.7989949747,
      "avg_output_tokens_per_task": 71164.3944723618,
      "last_run_timestamp": "2026-10-03T08:42:41.257Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Krea",
      "accuracy": 0.6455556666666666,
      "accuracy_std_dev": 0.09964144484545016,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.11177372666666668,
      "avg_duration_per_task_ms": 710817.72,
      "avg_output_tokens_per_task": 87845.5,
      "last_run_timestamp": "2026-09-29T19:45:03.655Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Parasail",
      "accuracy": 0.6454141073825505,
      "accuracy_std_dev": 0.01972979797722779,
      "run_count": 3,
      "total_questions": 149,
      "avg_cost_per_task_usd": 0.018687657718120806,
      "avg_duration_per_task_ms": 171442.07382550335,
      "avg_output_tokens_per_task": 9143.295302013423,
      "last_run_timestamp": "2026-08-01T17:40:52.342Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Novita",
      "accuracy": 0.644166625,
      "accuracy_std_dev": 0.02184215629772792,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.024566542500000007,
      "avg_duration_per_task_ms": 157744.9625,
      "avg_output_tokens_per_task": 10696.215,
      "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": "openai/gpt-5.4-nano-20260317",
      "model_slug": "openai/gpt-5.4-nano",
      "model_name": "OpenAI: GPT-5.4 Nano",
      "provider_name": "Azure",
      "accuracy": 0.6433335,
      "accuracy_std_dev": 0.014272415770172448,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.031060726666666667,
      "avg_duration_per_task_ms": 229987.72666666665,
      "avg_output_tokens_per_task": 16358.803333333333,
      "last_run_timestamp": "2026-10-01T19:05:16.543Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "OpenInference",
      "accuracy": 0.6433333333333333,
      "accuracy_std_dev": 0.14455548467015086,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.10531282,
      "avg_duration_per_task_ms": 1453976.54,
      "avg_output_tokens_per_task": 18878.353333333333,
      "last_run_timestamp": "2026-08-08T08:21:31.992Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.1",
      "model_slug": "minimax/minimax-m2.1",
      "model_name": "MiniMax: MiniMax M2.1",
      "provider_name": "auto-routing",
      "accuracy": 0.64333325,
      "accuracy_std_dev": 0.024267181772663682,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.024900534999999998,
      "avg_duration_per_task_ms": 168518.28,
      "avg_output_tokens_per_task": 11257.595,
      "last_run_timestamp": "2026-08-01T17:38:10.364Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.6415409547738694,
      "accuracy_std_dev": 0.019239053075710364,
      "run_count": 4,
      "total_questions": 199,
      "avg_cost_per_task_usd": 0.012467824120603016,
      "avg_duration_per_task_ms": 963027.175879397,
      "avg_output_tokens_per_task": 22096.447236180906,
      "last_run_timestamp": "2026-10-03T16:01:59.425Z"
    },
    {
      "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": 0.6411111111111111,
      "accuracy_std_dev": 0.028458355974395857,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.032843908888888884,
      "avg_duration_per_task_ms": 84121.66888888889,
      "avg_output_tokens_per_task": 20372.67111111111,
      "last_run_timestamp": "2026-10-03T16:20:43.625Z"
    },
    {
      "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": "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": 0.64,
      "accuracy_std_dev": 0.025647213408087847,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.218415728,
      "avg_duration_per_task_ms": 1219489.072,
      "avg_output_tokens_per_task": 31443.888,
      "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": "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": 0.63916675,
      "accuracy_std_dev": 0.03546318317900269,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.014215240000000002,
      "avg_duration_per_task_ms": 1323984.41,
      "avg_output_tokens_per_task": 23296.155,
      "last_run_timestamp": "2026-10-03T08:02:52.530Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.4-mini-20260317",
      "model_slug": "openai/gpt-5.4-mini",
      "model_name": "OpenAI: GPT-5.4 Mini",
      "provider_name": "auto-routing",
      "accuracy": 0.6377776666666667,
      "accuracy_std_dev": 0.04228062043326973,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.08944130666666666,
      "avg_duration_per_task_ms": 118157.82666666666,
      "avg_output_tokens_per_task": 12612.513333333334,
      "last_run_timestamp": "2026-10-01T19:04:42.468Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Mara",
      "accuracy": 0.637681,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 46,
      "avg_cost_per_task_usd": 0.04568002173913044,
      "avg_duration_per_task_ms": 70889.17391304347,
      "avg_output_tokens_per_task": 8168.934782608696,
      "last_run_timestamp": "2026-08-01T17:41:39.392Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "auto-routing",
      "accuracy": 0.63416675,
      "accuracy_std_dev": 0.030630012062314006,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.01972235,
      "avg_duration_per_task_ms": 313247.5575,
      "avg_output_tokens_per_task": 21107.5425,
      "last_run_timestamp": "2026-10-03T16:05:02.801Z"
    },
    {
      "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": 0.6336735020408163,
      "accuracy_std_dev": 0.10052381662193295,
      "run_count": 10,
      "total_questions": 490,
      "avg_cost_per_task_usd": 0.02281844081632653,
      "avg_duration_per_task_ms": 906837.293877551,
      "avg_output_tokens_per_task": 30461.944897959183,
      "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-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "Google",
      "accuracy": 0.6314878027681661,
      "accuracy_std_dev": 0.07985108320370836,
      "run_count": 6,
      "total_questions": 289,
      "avg_cost_per_task_usd": 0.14447499307958475,
      "avg_duration_per_task_ms": 358430.0657439446,
      "avg_output_tokens_per_task": 10974.453287197231,
      "last_run_timestamp": "2026-08-29T16:44:17.688Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "CoreWeave",
      "accuracy": 0.63,
      "accuracy_std_dev": 0.023333000000000048,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.07672618,
      "avg_duration_per_task_ms": 231628.06,
      "avg_output_tokens_per_task": 10580.58,
      "last_run_timestamp": "2026-08-01T17:40:48.165Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "NextBit",
      "accuracy": 0.63,
      "accuracy_std_dev": 0.016666999999999987,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.07387748,
      "avg_duration_per_task_ms": 1456099.47,
      "avg_output_tokens_per_task": 24632.34,
      "last_run_timestamp": "2026-07-18T16:00:08.935Z"
    },
    {
      "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": 0.627499875,
      "accuracy_std_dev": 0.020121193568210006,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.028656730000000002,
      "avg_duration_per_task_ms": 222093.8825,
      "avg_output_tokens_per_task": 10287.135,
      "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": "DeepInfra",
      "accuracy": 0.6256345228426395,
      "accuracy_std_dev": 0.03217519695525384,
      "run_count": 16,
      "total_questions": 788,
      "avg_cost_per_task_usd": 0.020946473350253805,
      "avg_duration_per_task_ms": 314363.8718274112,
      "avg_output_tokens_per_task": 19844.57614213198,
      "last_run_timestamp": "2026-10-03T16:12:29.900Z"
    },
    {
      "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": 0.6247562046783627,
      "accuracy_std_dev": 0.10453632978607418,
      "run_count": 7,
      "total_questions": 342,
      "avg_cost_per_task_usd": 0.43554279824561404,
      "avg_duration_per_task_ms": 2351365.956140351,
      "avg_output_tokens_per_task": 74116.19883040935,
      "last_run_timestamp": "2026-10-03T08:06:05.443Z"
    },
    {
      "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": 0.6245014239316239,
      "accuracy_std_dev": 0.06171917805736228,
      "run_count": 12,
      "total_questions": 585,
      "avg_cost_per_task_usd": 0.022623485470085473,
      "avg_duration_per_task_ms": 249317.92307692306,
      "avg_output_tokens_per_task": 24519.533333333333,
      "last_run_timestamp": "2026-10-03T16:12:24.693Z"
    },
    {
      "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": 0.624,
      "accuracy_std_dev": 0.012365076271499492,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.11120451600000002,
      "avg_duration_per_task_ms": 725043.5,
      "avg_output_tokens_per_task": 32381.232,
      "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-5.4-mini-20260317",
      "model_slug": "openai/gpt-5.4-mini",
      "model_name": "OpenAI: GPT-5.4 Mini",
      "provider_name": "OpenAI",
      "accuracy": 0.6216667499999999,
      "accuracy_std_dev": 0.04997217837144482,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.08547625,
      "avg_duration_per_task_ms": 106037.93,
      "avg_output_tokens_per_task": 11996.815,
      "last_run_timestamp": "2026-10-02T23:10:27.235Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "GMICloud",
      "accuracy": 0.6211111666666667,
      "accuracy_std_dev": 0.18460855717654465,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.13787135666666667,
      "avg_duration_per_task_ms": 390996.2266666667,
      "avg_output_tokens_per_task": 8010.466666666666,
      "last_run_timestamp": "2026-10-03T16:12:26.818Z"
    },
    {
      "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": 0.6211110000000001,
      "accuracy_std_dev": 0.041171085383636265,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.09511168999999998,
      "avg_duration_per_task_ms": 55650.66,
      "avg_output_tokens_per_task": 3586.0333333333333,
      "last_run_timestamp": "2026-10-03T16:14:18.044Z"
    },
    {
      "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": 0.6208576842105263,
      "accuracy_std_dev": 0.07290857624841833,
      "run_count": 7,
      "total_questions": 342,
      "avg_cost_per_task_usd": 0.08742642982456139,
      "avg_duration_per_task_ms": 1045323.0116959064,
      "avg_output_tokens_per_task": 62138.06140350877,
      "last_run_timestamp": "2026-10-03T16:14:02.971Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.1",
      "model_slug": "minimax/minimax-m2.1",
      "model_name": "MiniMax: MiniMax M2.1",
      "provider_name": "Novita",
      "accuracy": 0.6199997500000001,
      "accuracy_std_dev": 0.019999916667013916,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.02507946,
      "avg_duration_per_task_ms": 173259.145,
      "avg_output_tokens_per_task": 11472.255,
      "last_run_timestamp": "2026-08-01T17:38:10.202Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.6187741034482758,
      "accuracy_std_dev": 0.01093895001490416,
      "run_count": 7,
      "total_questions": 348,
      "avg_cost_per_task_usd": 0.11070118965517242,
      "avg_duration_per_task_ms": 713990.3045977012,
      "avg_output_tokens_per_task": 33324.629310344826,
      "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Mara",
      "accuracy": 0.6180555,
      "accuracy_std_dev": 0.00694450000000002,
      "run_count": 2,
      "total_questions": 96,
      "avg_cost_per_task_usd": 0.025814302083333334,
      "avg_duration_per_task_ms": 76875.0625,
      "avg_output_tokens_per_task": 17301.635416666668,
      "last_run_timestamp": "2026-08-29T16:33:38.021Z"
    },
    {
      "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": 0.6166667499999999,
      "accuracy_std_dev": 0.06236093417507071,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.037265545000000004,
      "avg_duration_per_task_ms": 382273.955,
      "avg_output_tokens_per_task": 20068.4325,
      "last_run_timestamp": "2026-10-03T16:12:10.894Z"
    },
    {
      "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": 0.6158248141414141,
      "accuracy_std_dev": 0.2285695922821756,
      "run_count": 10,
      "total_questions": 495,
      "avg_cost_per_task_usd": 0.020359361616161618,
      "avg_duration_per_task_ms": 340617.86262626265,
      "avg_output_tokens_per_task": 10409.925252525252,
      "last_run_timestamp": "2026-09-26T08:14:40.334Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2-0905",
      "model_slug": "moonshotai/kimi-k2-0905",
      "model_name": "MoonshotAI: Kimi K2 0905",
      "provider_name": "Novita",
      "accuracy": 0.61499975,
      "accuracy_std_dev": 0.03662873416714669,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.091530915,
      "avg_duration_per_task_ms": 150799.645,
      "avg_output_tokens_per_task": 3905.64,
      "last_run_timestamp": "2026-08-01T17:38:24.442Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Cerebras",
      "accuracy": 0.61458325,
      "accuracy_std_dev": 0.028281053484400142,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.065036325,
      "avg_duration_per_task_ms": 29159.9975,
      "avg_output_tokens_per_task": 20314.995,
      "last_run_timestamp": "2026-10-03T16:18:39.939Z"
    },
    {
      "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": 0.6133331666666667,
      "accuracy_std_dev": 0.031972140650972335,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.04558941666666667,
      "avg_duration_per_task_ms": 201450.47,
      "avg_output_tokens_per_task": 3388.8166666666666,
      "last_run_timestamp": "2026-10-01T16:01:02.677Z"
    },
    {
      "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": 0.613333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.01616616,
      "avg_duration_per_task_ms": 108573.58,
      "avg_output_tokens_per_task": 21305.1,
      "last_run_timestamp": "2026-10-03T16:13:54.745Z"
    },
    {
      "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": 0.608,
      "accuracy_std_dev": 0.11488351975457575,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.058990216,
      "avg_duration_per_task_ms": 582295.388,
      "avg_output_tokens_per_task": 35784.112,
      "last_run_timestamp": "2026-08-15T14:08:35.086Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "SambaNova",
      "accuracy": 0.6070889547511311,
      "accuracy_std_dev": 0.0286460576026238,
      "run_count": 9,
      "total_questions": 442,
      "avg_cost_per_task_usd": 0.041861986425339365,
      "avg_duration_per_task_ms": 87046.60407239819,
      "avg_output_tokens_per_task": 24951.95701357466,
      "last_run_timestamp": "2026-10-03T16:12:29.878Z"
    },
    {
      "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": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "GMICloud",
      "accuracy": 0.6066666666666667,
      "accuracy_std_dev": 0.15241273398753646,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.017056396666666668,
      "avg_duration_per_task_ms": 233868.72666666665,
      "avg_output_tokens_per_task": 8308.71,
      "last_run_timestamp": "2026-08-29T16:33:39.721Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-2.5-pro",
      "model_slug": "google/gemini-2.5-pro",
      "model_name": "Google: Gemini 2.5 Pro",
      "provider_name": "Google",
      "accuracy": 0.6060269955357143,
      "accuracy_std_dev": 0.032348691330028045,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.19772456473214287,
      "avg_duration_per_task_ms": 188670.52455357142,
      "avg_output_tokens_per_task": 13047.183035714286,
      "last_run_timestamp": "2026-10-01T19:04:55.725Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Chutes",
      "accuracy": 0.603333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.030717460000000002,
      "avg_duration_per_task_ms": 422658.18,
      "avg_output_tokens_per_task": 11680.98,
      "last_run_timestamp": "2026-07-18T16:00:09.161Z"
    },
    {
      "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": 0.6029999,
      "accuracy_std_dev": 0.026010556404852253,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.04766987600000001,
      "avg_duration_per_task_ms": 92389.166,
      "avg_output_tokens_per_task": 4752.396,
      "last_run_timestamp": "2026-10-03T16:14:40.781Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-next-2025-02-03",
      "model_slug": "qwen/qwen3-coder-next",
      "model_name": "Qwen: Qwen3 Coder Next",
      "provider_name": "Parasail",
      "accuracy": 0.6019345223214286,
      "accuracy_std_dev": 0.023580042960096588,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.019949879464285714,
      "avg_duration_per_task_ms": 99253.61830357143,
      "avg_output_tokens_per_task": 4312.604910714285,
      "last_run_timestamp": "2026-10-03T16:14:27.398Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-235b-a22b-thinking-2507",
      "model_slug": "qwen/qwen3-235b-a22b-thinking-2507",
      "model_name": "Qwen: Qwen3 235B A22B Thinking 2507",
      "provider_name": "Novita",
      "accuracy": 0.599552610738255,
      "accuracy_std_dev": 0.021214776019452947,
      "run_count": 6,
      "total_questions": 298,
      "avg_cost_per_task_usd": 0.1599610402684564,
      "avg_duration_per_task_ms": 609782.6946308725,
      "avg_output_tokens_per_task": 29203.84899328859,
      "last_run_timestamp": "2026-08-22T16:20:42.106Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-2.5-pro",
      "model_slug": "google/gemini-2.5-pro",
      "model_name": "Google: Gemini 2.5 Pro",
      "provider_name": "auto-routing",
      "accuracy": 0.5933333333333334,
      "accuracy_std_dev": 0.02880349773127486,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.21649743333333332,
      "avg_duration_per_task_ms": 191120.76666666666,
      "avg_output_tokens_per_task": 14122.313333333334,
      "last_run_timestamp": "2026-10-01T19:04:42.461Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "BaseTen",
      "accuracy": 0.593333,
      "accuracy_std_dev": 0.020000000000000018,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.04552234,
      "avg_duration_per_task_ms": 32685.39,
      "avg_output_tokens_per_task": 3125.98,
      "last_run_timestamp": "2026-09-26T08:22:32.737Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.6-20251208",
      "model_slug": "z-ai/glm-4.6v",
      "model_name": "Z.ai: GLM 4.6V",
      "provider_name": "Z.AI",
      "accuracy": 0.593333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.01896666,
      "avg_duration_per_task_ms": 306972.98,
      "avg_output_tokens_per_task": 7436.64,
      "last_run_timestamp": "2026-10-01T08:02:59.889Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2-0905",
      "model_slug": "moonshotai/kimi-k2-0905",
      "model_name": "MoonshotAI: Kimi K2 0905",
      "provider_name": "auto-routing",
      "accuracy": 0.5909091212121212,
      "accuracy_std_dev": 0.04025476018668477,
      "run_count": 4,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0993339494949495,
      "avg_duration_per_task_ms": 115416.30808080808,
      "avg_output_tokens_per_task": 3700.722222222222,
      "last_run_timestamp": "2026-08-01T17:38:23.537Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-2.5-flash",
      "model_slug": "google/gemini-2.5-flash",
      "model_name": "Google: Gemini 2.5 Flash",
      "provider_name": "Google",
      "accuracy": 0.5881481111111111,
      "accuracy_std_dev": 0.007554077566224961,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.029889999999999996,
      "avg_duration_per_task_ms": 78610.16,
      "avg_output_tokens_per_task": 8108.122222222222,
      "last_run_timestamp": "2026-10-01T19:05:46.198Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "openai/gpt-5.2-chat-20251211",
      "model_slug": "openai/gpt-5.2-chat",
      "model_name": "OpenAI: GPT-5.2 Chat",
      "provider_name": "OpenAI",
      "accuracy": 0.58,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.10846708,
      "avg_duration_per_task_ms": 79571.2,
      "avg_output_tokens_per_task": 4199.82,
      "last_run_timestamp": "2026-08-01T08:05:35.385Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Novita",
      "accuracy": 0.5799998333333334,
      "accuracy_std_dev": 0.042339037236797064,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.01084836,
      "avg_duration_per_task_ms": 298654.7033333333,
      "avg_output_tokens_per_task": 18007.093333333334,
      "last_run_timestamp": "2026-08-29T16:33:38.124Z"
    },
    {
      "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": 0.5791477009132421,
      "accuracy_std_dev": 0.037940164496542333,
      "run_count": 9,
      "total_questions": 438,
      "avg_cost_per_task_usd": 0.07539671461187215,
      "avg_duration_per_task_ms": 91646.01598173517,
      "avg_output_tokens_per_task": 5547.684931506849,
      "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": 0.571111,
      "accuracy_std_dev": 0.008314764498569157,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.028662713333333336,
      "avg_duration_per_task_ms": 78035.85333333333,
      "avg_output_tokens_per_task": 8001.593333333333,
      "last_run_timestamp": "2026-10-01T19:04:40.340Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Cerebras",
      "accuracy": 0.5648772953020134,
      "accuracy_std_dev": 0.0424873867714486,
      "run_count": 3,
      "total_questions": 149,
      "avg_cost_per_task_usd": 0.1486826308724832,
      "avg_duration_per_task_ms": 18002.88590604027,
      "avg_output_tokens_per_task": 2546.3288590604025,
      "last_run_timestamp": "2026-08-29T17:07:32.456Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-2.5-pro",
      "model_slug": "google/gemini-2.5-pro",
      "model_name": "Google: Gemini 2.5 Pro",
      "provider_name": "Google AI Studio",
      "accuracy": 0.5622223333333333,
      "accuracy_std_dev": 0.03456997912768952,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.23936231333333335,
      "avg_duration_per_task_ms": 165980.19333333333,
      "avg_output_tokens_per_task": 13303.673333333334,
      "last_run_timestamp": "2026-10-01T19:04:52.649Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-r1-0528",
      "model_slug": "deepseek/deepseek-r1-0528",
      "model_name": "DeepSeek: R1 0528",
      "provider_name": "Novita",
      "accuracy": 0.5616569826589596,
      "accuracy_std_dev": 0.040448501937396776,
      "run_count": 7,
      "total_questions": 346,
      "avg_cost_per_task_usd": 0.177089710982659,
      "avg_duration_per_task_ms": 1805453.9132947978,
      "avg_output_tokens_per_task": 38595.37283236994,
      "last_run_timestamp": "2026-10-03T16:17:09.625Z"
    },
    {
      "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": 0.5604491347150259,
      "accuracy_std_dev": 0.10678459157041047,
      "run_count": 8,
      "total_questions": 386,
      "avg_cost_per_task_usd": 0.07120025388601035,
      "avg_duration_per_task_ms": 510444.22279792745,
      "avg_output_tokens_per_task": 29300.067357512955,
      "last_run_timestamp": "2026-08-29T08:10:31.240Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Mancer 2",
      "accuracy": 0.5572390808080808,
      "accuracy_std_dev": 0.08974967405616299,
      "run_count": 4,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.04370982828282828,
      "avg_duration_per_task_ms": 3081206.1666666665,
      "avg_output_tokens_per_task": 93737.79292929293,
      "last_run_timestamp": "2026-10-03T16:33:53.219Z"
    },
    {
      "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": 0.5521540748299321,
      "accuracy_std_dev": 0.03338584001572458,
      "run_count": 3,
      "total_questions": 147,
      "avg_cost_per_task_usd": 0.0026003809523809525,
      "avg_duration_per_task_ms": 390785.49659863947,
      "avg_output_tokens_per_task": 6412.292517006803,
      "last_run_timestamp": "2026-10-03T16:14:16.049Z"
    },
    {
      "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": 0.5516073604060914,
      "accuracy_std_dev": 0.07298347728499496,
      "run_count": 8,
      "total_questions": 394,
      "avg_cost_per_task_usd": 0.051452035532994926,
      "avg_duration_per_task_ms": 435341.5304568528,
      "avg_output_tokens_per_task": 9213.700507614214,
      "last_run_timestamp": "2026-10-04T00:00:28.213Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-2.5-flash",
      "model_slug": "google/gemini-2.5-flash",
      "model_name": "Google: Gemini 2.5 Flash",
      "provider_name": "Google AI Studio",
      "accuracy": 0.5511113333333334,
      "accuracy_std_dev": 0.006285236475706844,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.035753513333333334,
      "avg_duration_per_task_ms": 61316.40666666667,
      "avg_output_tokens_per_task": 6659.566666666667,
      "last_run_timestamp": "2026-10-01T19:04:42.483Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-next-2025-02-03",
      "model_slug": "qwen/qwen3-coder-next",
      "model_name": "Qwen: Qwen3 Coder Next",
      "provider_name": "auto-routing",
      "accuracy": 0.5496295555555556,
      "accuracy_std_dev": 0.09428684395928454,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.04100789111111111,
      "avg_duration_per_task_ms": 99006.21111111112,
      "avg_output_tokens_per_task": 4319.34,
      "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/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "SiliconFlow",
      "accuracy": 0.5375,
      "accuracy_std_dev": 0.09219167231642998,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.029294462499999997,
      "avg_duration_per_task_ms": 409054.695,
      "avg_output_tokens_per_task": 11072.1825,
      "last_run_timestamp": "2026-10-03T16:13:14.992Z"
    },
    {
      "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": 0.5355556666666667,
      "accuracy_std_dev": 0.02199873172606905,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.09782582666666667,
      "avg_duration_per_task_ms": 375218.6066666667,
      "avg_output_tokens_per_task": 16827.4,
      "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": 0.533914726744186,
      "accuracy_std_dev": 0.015990278830002442,
      "run_count": 7,
      "total_questions": 344,
      "avg_cost_per_task_usd": 0.02433947674418605,
      "avg_duration_per_task_ms": 1793934.9767441861,
      "avg_output_tokens_per_task": 124117.57848837209,
      "last_run_timestamp": "2026-10-03T16:10:46.805Z"
    },
    {
      "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": "Fireworks",
      "accuracy": 0.526667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.03416076,
      "avg_duration_per_task_ms": 1845834.04,
      "avg_output_tokens_per_task": 105339.92,
      "last_run_timestamp": "2026-07-18T16:00:08.888Z"
    },
    {
      "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": 0.523035243902439,
      "accuracy_std_dev": 0.26305710961614404,
      "run_count": 5,
      "total_questions": 246,
      "avg_cost_per_task_usd": 0.01812669918699187,
      "avg_duration_per_task_ms": 110026.72764227643,
      "avg_output_tokens_per_task": 18887.439024390245,
      "last_run_timestamp": "2026-08-29T17:06:47.726Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4.1-2025-04-14",
      "model_slug": "openai/gpt-4.1",
      "model_name": "OpenAI: GPT-4.1",
      "provider_name": "Azure",
      "accuracy": 0.5221089591836734,
      "accuracy_std_dev": 0.023889008586586408,
      "run_count": 4,
      "total_questions": 196,
      "avg_cost_per_task_usd": 0.13644156632653062,
      "avg_duration_per_task_ms": 73125.92346938775,
      "avg_output_tokens_per_task": 2553.7397959183672,
      "last_run_timestamp": "2026-10-01T19:05:18.003Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3-nano-30b-a3b",
      "model_slug": "nvidia/nemotron-3-nano-30b-a3b",
      "model_name": "NVIDIA: Nemotron 3 Nano 30B A3B",
      "provider_name": "auto-routing",
      "accuracy": 0.517500125,
      "accuracy_std_dev": 0.024930426803594342,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0282700725,
      "avg_duration_per_task_ms": 444815,
      "avg_output_tokens_per_task": 58343.125,
      "last_run_timestamp": "2026-10-03T16:09:19.689Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-2.5-flash-lite",
      "model_slug": "google/gemini-2.5-flash-lite",
      "model_name": "Google: Gemini 2.5 Flash Lite",
      "provider_name": "Google AI Studio",
      "accuracy": 0.5156599127516779,
      "accuracy_std_dev": 0.03449671802746843,
      "run_count": 3,
      "total_questions": 149,
      "avg_cost_per_task_usd": 0.024214489932885906,
      "avg_duration_per_task_ms": 190286.47651006712,
      "avg_output_tokens_per_task": 48937.10738255033,
      "last_run_timestamp": "2026-10-01T19:05:00.102Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.6-20251208",
      "model_slug": "z-ai/glm-4.6v",
      "model_name": "Z.ai: GLM 4.6V",
      "provider_name": "Novita",
      "accuracy": 0.513333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.03238348,
      "avg_duration_per_task_ms": 271942.88,
      "avg_output_tokens_per_task": 7084.96,
      "last_run_timestamp": "2026-10-01T08:05:50.069Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-r1-0528",
      "model_slug": "deepseek/deepseek-r1-0528",
      "model_name": "DeepSeek: R1 0528",
      "provider_name": "auto-routing",
      "accuracy": 0.5125241502890173,
      "accuracy_std_dev": 0.04228836158315612,
      "run_count": 7,
      "total_questions": 346,
      "avg_cost_per_task_usd": 0.14881240173410404,
      "avg_duration_per_task_ms": 1730003.4537572255,
      "avg_output_tokens_per_task": 33598.51734104046,
      "last_run_timestamp": "2026-10-03T16:05:51.297Z"
    },
    {
      "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": 0.5119049598214286,
      "accuracy_std_dev": 0.02807387800381059,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.06533608482142857,
      "avg_duration_per_task_ms": 299963.359375,
      "avg_output_tokens_per_task": 4007.223214285714,
      "last_run_timestamp": "2026-10-04T00:00:34.892Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.1-terminus",
      "model_slug": "deepseek/deepseek-v3.1-terminus",
      "model_name": "DeepSeek: DeepSeek V3.1 Terminus",
      "provider_name": "DeepInfra",
      "accuracy": 0.5104164999999999,
      "accuracy_std_dev": 0.003472500000000045,
      "run_count": 2,
      "total_questions": 96,
      "avg_cost_per_task_usd": 0.024820281250000003,
      "avg_duration_per_task_ms": 480070.4583333333,
      "avg_output_tokens_per_task": 9578.8125,
      "last_run_timestamp": "2026-08-08T16:03:43.735Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "WandB",
      "accuracy": 0.51,
      "accuracy_std_dev": 0.016666999999999987,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.022937250000000003,
      "avg_duration_per_task_ms": 2318885.96,
      "avg_output_tokens_per_task": 152067.3,
      "last_run_timestamp": "2026-07-18T16:00:10.167Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4.1-2025-04-14",
      "model_slug": "openai/gpt-4.1",
      "model_name": "OpenAI: GPT-4.1",
      "provider_name": "OpenAI",
      "accuracy": 0.5083332500000001,
      "accuracy_std_dev": 0.026404802805313633,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.12635733,
      "avg_duration_per_task_ms": 53878.245,
      "avg_output_tokens_per_task": 2800.47,
      "last_run_timestamp": "2026-10-02T23:07:04.603Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "DeepInfra",
      "accuracy": 0.5033784729729729,
      "accuracy_std_dev": 0.012185696720153325,
      "run_count": 6,
      "total_questions": 296,
      "avg_cost_per_task_usd": 0.021273824324324325,
      "avg_duration_per_task_ms": 1274174.0878378379,
      "avg_output_tokens_per_task": 117319.30743243243,
      "last_run_timestamp": "2026-08-29T17:16:15.676Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-2.5-flash-lite",
      "model_slug": "google/gemini-2.5-flash-lite",
      "model_name": "Google: Gemini 2.5 Flash Lite",
      "provider_name": "auto-routing",
      "accuracy": 0.5033670303030303,
      "accuracy_std_dev": 0.030340499999999992,
      "run_count": 2,
      "total_questions": 99,
      "avg_cost_per_task_usd": 0.01769851515151515,
      "avg_duration_per_task_ms": 135563.53535353535,
      "avg_output_tokens_per_task": 30988.575757575756,
      "last_run_timestamp": "2026-10-01T19:04:42.466Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3-nano-30b-a3b",
      "model_slug": "nvidia/nemotron-3-nano-30b-a3b",
      "model_name": "NVIDIA: Nemotron 3 Nano 30B A3B",
      "provider_name": "DeepInfra",
      "accuracy": 0.5033669747474748,
      "accuracy_std_dev": 0.027292446564651823,
      "run_count": 8,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.025276146464646466,
      "avg_duration_per_task_ms": 577739.3712121212,
      "avg_output_tokens_per_task": 53598.469696969696,
      "last_run_timestamp": "2026-09-26T16:00:25.689Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Together",
      "accuracy": 0.49740733333333337,
      "accuracy_std_dev": 0.030943206527802244,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.02629459333333333,
      "avg_duration_per_task_ms": 283072.18,
      "avg_output_tokens_per_task": 19100.475555555557,
      "last_run_timestamp": "2026-10-03T16:15:51.386Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "CoreWeave",
      "accuracy": 0.49666650000000007,
      "accuracy_std_dev": 0.023333500000000007,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.23109727999999996,
      "avg_duration_per_task_ms": 80168.37,
      "avg_output_tokens_per_task": 2746.03,
      "last_run_timestamp": "2026-08-01T17:41:27.708Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.1-terminus",
      "model_slug": "deepseek/deepseek-v3.1-terminus",
      "model_name": "DeepSeek: DeepSeek V3.1 Terminus",
      "provider_name": "SiliconFlow",
      "accuracy": 0.49523814285714285,
      "accuracy_std_dev": 0.023289352137395136,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.06688620571428572,
      "avg_duration_per_task_ms": 308997.38285714283,
      "avg_output_tokens_per_task": 4074.4114285714286,
      "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": "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": "qwen/qwen3-32b-04-28",
      "model_slug": "qwen/qwen3-32b",
      "model_name": "Qwen: Qwen3 32B",
      "provider_name": "SiliconFlow",
      "accuracy": 0.49319713265306125,
      "accuracy_std_dev": 0.03082350135347216,
      "run_count": 4,
      "total_questions": 196,
      "avg_cost_per_task_usd": 0.05686764285714286,
      "avg_duration_per_task_ms": 2675498.8775510206,
      "avg_output_tokens_per_task": 42954.67857142857,
      "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": "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": 0.489899005050505,
      "accuracy_std_dev": 0.009452416119701875,
      "run_count": 4,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.014275666666666666,
      "avg_duration_per_task_ms": 117584.4797979798,
      "avg_output_tokens_per_task": 3141.5757575757575,
      "last_run_timestamp": "2026-10-03T16:05:51.297Z"
    },
    {
      "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": 0.4894028092485549,
      "accuracy_std_dev": 0.025397205898062335,
      "run_count": 7,
      "total_questions": 346,
      "avg_cost_per_task_usd": 0.023809176300578036,
      "avg_duration_per_task_ms": 845136.3179190751,
      "avg_output_tokens_per_task": 22279.682080924857,
      "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": "nvidia/nemotron-3-nano-30b-a3b",
      "model_slug": "nvidia/nemotron-3-nano-30b-a3b",
      "model_name": "NVIDIA: Nemotron 3 Nano 30B A3B",
      "provider_name": "Nebius",
      "accuracy": 0.48833333333333334,
      "accuracy_std_dev": 0.018831396165859726,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.03323320666666667,
      "avg_duration_per_task_ms": 665027.6466666666,
      "avg_output_tokens_per_task": 61694.416666666664,
      "last_run_timestamp": "2026-10-03T16:04:32.703Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3-nano-30b-a3b",
      "model_slug": "nvidia/nemotron-3-nano-30b-a3b",
      "model_name": "NVIDIA: Nemotron 3 Nano 30B A3B",
      "provider_name": "Crusoe",
      "accuracy": 0.4852150080645161,
      "accuracy_std_dev": 0.023420064714684302,
      "run_count": 10,
      "total_questions": 496,
      "avg_cost_per_task_usd": 0.028619552419354842,
      "avg_duration_per_task_ms": 386610.2439516129,
      "avg_output_tokens_per_task": 66355.3125,
      "last_run_timestamp": "2026-10-03T16:14:03.053Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-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": 0.4825,
      "accuracy_std_dev": 0.02355551362738668,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0347383225,
      "avg_duration_per_task_ms": 43524,
      "avg_output_tokens_per_task": 3653.295,
      "last_run_timestamp": "2026-10-03T16:17:11.217Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-235b-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": "qwen/qwen3-32b-04-28",
      "model_slug": "qwen/qwen3-32b",
      "model_name": "Qwen: Qwen3 32B",
      "provider_name": "auto-routing",
      "accuracy": 0.47833325000000004,
      "accuracy_std_dev": 0.028234971421758143,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.022111440000000003,
      "avg_duration_per_task_ms": 722757.885,
      "avg_output_tokens_per_task": 21962.2775,
      "last_run_timestamp": "2026-08-29T16:16:50.456Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4.1-2025-04-14",
      "model_slug": "openai/gpt-4.1",
      "model_name": "OpenAI: GPT-4.1",
      "provider_name": "auto-routing",
      "accuracy": 0.4777776666666667,
      "accuracy_std_dev": 0.013698571612479242,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.12510532,
      "avg_duration_per_task_ms": 53067.566666666666,
      "avg_output_tokens_per_task": 2709.2933333333335,
      "last_run_timestamp": "2026-10-01T19:04:26.703Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-r1-0528",
      "model_slug": "deepseek/deepseek-r1-0528",
      "model_name": "DeepSeek: R1 0528",
      "provider_name": "SiliconFlow",
      "accuracy": 0.47701135057471267,
      "accuracy_std_dev": 0.009988678401274195,
      "run_count": 7,
      "total_questions": 348,
      "avg_cost_per_task_usd": 0.16442305459770115,
      "avg_duration_per_task_ms": 2471430.7183908047,
      "avg_output_tokens_per_task": 44221.49137931035,
      "last_run_timestamp": "2026-10-03T16:12:22.579Z"
    },
    {
      "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": 0.4762964444444444,
      "accuracy_std_dev": 0.03444249058651928,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.023876493333333335,
      "avg_duration_per_task_ms": 58280.74444444444,
      "avg_output_tokens_per_task": 3755.9466666666667,
      "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-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.475168393939394,
      "accuracy_std_dev": 0.08821191621175678,
      "run_count": 16,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.04045829419191919,
      "avg_duration_per_task_ms": 704580.8383838384,
      "avg_output_tokens_per_task": 194395.8547979798,
      "last_run_timestamp": "2026-10-03T16:19:24.321Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.3-chat-20260303",
      "model_slug": "openai/gpt-5.3-chat",
      "model_name": "OpenAI: GPT-5.3 Chat",
      "provider_name": "OpenAI",
      "accuracy": 0.473333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.09545477999999999,
      "avg_duration_per_task_ms": 59370.42,
      "avg_output_tokens_per_task": 2482.38,
      "last_run_timestamp": "2026-08-01T08:06:06.635Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-32b-04-28",
      "model_slug": "qwen/qwen3-32b",
      "model_name": "Qwen: Qwen3 32B",
      "provider_name": "Alibaba",
      "accuracy": 0.473333,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.03316041,
      "avg_duration_per_task_ms": 478208.83,
      "avg_output_tokens_per_task": 36953.92,
      "last_run_timestamp": "2026-07-18T16:00:08.521Z"
    },
    {
      "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": 0.4727888571428572,
      "accuracy_std_dev": 0.053216013468691904,
      "run_count": 4,
      "total_questions": 196,
      "avg_cost_per_task_usd": 0.026510076530612242,
      "avg_duration_per_task_ms": 2225913.1275510206,
      "avg_output_tokens_per_task": 135590.98469387754,
      "last_run_timestamp": "2026-10-03T16:09:05.687Z"
    },
    {
      "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": 0.47083349999999996,
      "accuracy_std_dev": 0.03247871952910089,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0190439575,
      "avg_duration_per_task_ms": 88315.37,
      "avg_output_tokens_per_task": 3656.9575,
      "last_run_timestamp": "2026-10-03T16:12:20.062Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-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": 0.4696295555555556,
      "accuracy_std_dev": 0.026314631063814885,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.024747560000000002,
      "avg_duration_per_task_ms": 125553.2,
      "avg_output_tokens_per_task": 3799.4177777777777,
      "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": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Groq",
      "accuracy": 0.4682815259515572,
      "accuracy_std_dev": 0.039807100596362034,
      "run_count": 6,
      "total_questions": 289,
      "avg_cost_per_task_usd": 0.04939322491349481,
      "avg_duration_per_task_ms": 229004.3321799308,
      "avg_output_tokens_per_task": 98239.9169550173,
      "last_run_timestamp": "2026-10-03T16:14:31.002Z"
    },
    {
      "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": 0.4666665555555556,
      "accuracy_std_dev": 0.031426983767001204,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.10525973333333334,
      "avg_duration_per_task_ms": 99178.76444444444,
      "avg_output_tokens_per_task": 2760.7244444444445,
      "last_run_timestamp": "2026-10-03T16:07:50.332Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5-nano-2025-08-07",
      "model_slug": "openai/gpt-5-nano",
      "model_name": "OpenAI: GPT-5 Nano",
      "provider_name": "auto-routing",
      "accuracy": 0.4662164729729729,
      "accuracy_std_dev": 0.03892660093903227,
      "run_count": 3,
      "total_questions": 148,
      "avg_cost_per_task_usd": 0.04219703378378378,
      "avg_duration_per_task_ms": 711257.6756756756,
      "avg_output_tokens_per_task": 98782.22297297297,
      "last_run_timestamp": "2026-10-01T19:04:24.509Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "StreamLake",
      "accuracy": 0.4653333,
      "accuracy_std_dev": 0.0479628643390905,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.054455583999999994,
      "avg_duration_per_task_ms": 126748.042,
      "avg_output_tokens_per_task": 3614.196,
      "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": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "WandB",
      "accuracy": 0.4652775,
      "accuracy_std_dev": 0.006944499999999992,
      "run_count": 2,
      "total_questions": 96,
      "avg_cost_per_task_usd": 0.027522666666666668,
      "avg_duration_per_task_ms": 29612.604166666668,
      "avg_output_tokens_per_task": 1399.6145833333333,
      "last_run_timestamp": "2026-07-18T16:00:08.906Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Coder 30B A3B Instruct",
      "provider_name": "DigitalOcean",
      "accuracy": 0.46444433333333335,
      "accuracy_std_dev": 0.008314853589945066,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.033371,
      "avg_duration_per_task_ms": 305705.87333333335,
      "avg_output_tokens_per_task": 3594.2266666666665,
      "last_run_timestamp": "2026-08-15T08:15:42.180Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-next-80b-a3b-instruct-2509",
      "model_slug": "qwen/qwen3-next-80b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Next 80B A3B Instruct",
      "provider_name": "DeepInfra",
      "accuracy": 0.4640741111111111,
      "accuracy_std_dev": 0.03249991284242824,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.02241910888888889,
      "avg_duration_per_task_ms": 213262.18444444446,
      "avg_output_tokens_per_task": 3634.297777777778,
      "last_run_timestamp": "2026-10-03T16:04:32.682Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4o",
      "model_slug": "openai/gpt-4o",
      "model_name": "OpenAI: GPT-4o",
      "provider_name": "Azure",
      "accuracy": 0.463333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.26064434,
      "avg_duration_per_task_ms": 31467.4,
      "avg_output_tokens_per_task": 1819.56,
      "last_run_timestamp": "2026-08-01T08:06:06.230Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Google",
      "accuracy": 0.4626308744394619,
      "accuracy_std_dev": 0.044603677582420745,
      "run_count": 9,
      "total_questions": 446,
      "avg_cost_per_task_usd": 0.054874065022421516,
      "avg_duration_per_task_ms": 71821.13004484306,
      "avg_output_tokens_per_task": 3451.358744394619,
      "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-5-nano-2025-08-07",
      "model_slug": "openai/gpt-5-nano",
      "model_name": "OpenAI: GPT-5 Nano",
      "provider_name": "Azure",
      "accuracy": 0.4608501677852349,
      "accuracy_std_dev": 0.034994868451198195,
      "run_count": 6,
      "total_questions": 298,
      "avg_cost_per_task_usd": 0.04671855033557047,
      "avg_duration_per_task_ms": 767600.1241610738,
      "avg_output_tokens_per_task": 101728.0033557047,
      "last_run_timestamp": "2026-10-01T19:06:01.717Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-2.5-flash-lite",
      "model_slug": "google/gemini-2.5-flash-lite",
      "model_name": "Google: Gemini 2.5 Flash Lite",
      "provider_name": "Google",
      "accuracy": 0.46068351282051284,
      "accuracy_std_dev": 0.04225414673955088,
      "run_count": 4,
      "total_questions": 195,
      "avg_cost_per_task_usd": 0.01900871794871795,
      "avg_duration_per_task_ms": 170098.1794871795,
      "avg_output_tokens_per_task": 35854.89743589744,
      "last_run_timestamp": "2026-10-01T19:06:10.500Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4.1-2025-04-14",
      "model_slug": "openai/gpt-4.1",
      "model_name": "OpenAI: GPT-4.1",
      "provider_name": "FakeProvider",
      "accuracy": 0.46,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": null,
      "avg_duration_per_task_ms": 158124.84,
      "avg_output_tokens_per_task": 4515,
      "last_run_timestamp": "2026-10-01T19:06:51.244Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.3-chat-20260303",
      "model_slug": "openai/gpt-5.3-chat",
      "model_name": "OpenAI: GPT-5.3 Chat",
      "provider_name": "auto-routing",
      "accuracy": 0.46,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.09873096,
      "avg_duration_per_task_ms": 65244.1,
      "avg_output_tokens_per_task": 2748.04,
      "last_run_timestamp": "2026-08-01T08:05:59.441Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Nebius",
      "accuracy": 0.45962955555555557,
      "accuracy_std_dev": 0.03844019636934208,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.053124257777777774,
      "avg_duration_per_task_ms": 153327.5977777778,
      "avg_output_tokens_per_task": 3785.157777777778,
      "last_run_timestamp": "2026-10-03T16:12:33.495Z"
    },
    {
      "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": 0.45920115624999996,
      "accuracy_std_dev": 0.020597873615982785,
      "run_count": 4,
      "total_questions": 192,
      "avg_cost_per_task_usd": 0.08208188020833333,
      "avg_duration_per_task_ms": 47993.442708333336,
      "avg_output_tokens_per_task": 3348.3072916666665,
      "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-coder-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Coder 30B A3B Instruct",
      "provider_name": "Novita",
      "accuracy": 0.45777799999999996,
      "accuracy_std_dev": 0.008314764498569207,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.00622334,
      "avg_duration_per_task_ms": 39170.17333333333,
      "avg_output_tokens_per_task": 1174.36,
      "last_run_timestamp": "2026-08-15T08:16:13.926Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.45542626162790695,
      "accuracy_std_dev": 0.018912652487751393,
      "run_count": 7,
      "total_questions": 344,
      "avg_cost_per_task_usd": 0.046210372093023255,
      "avg_duration_per_task_ms": 174189.27034883722,
      "avg_output_tokens_per_task": 1898.7906976744187,
      "last_run_timestamp": "2026-08-29T16:23:01.234Z"
    },
    {
      "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": "WandB",
      "accuracy": 0.45333349999999994,
      "accuracy_std_dev": 0.013333499999999998,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.2293187,
      "avg_duration_per_task_ms": 78949.14,
      "avg_output_tokens_per_task": 2726.37,
      "last_run_timestamp": "2026-07-18T08:04:08.872Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "Novita",
      "accuracy": 0.451852,
      "accuracy_std_dev": 0.06252139759936415,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.15496766666666664,
      "avg_duration_per_task_ms": 120175.53777777778,
      "avg_output_tokens_per_task": 3250.7844444444445,
      "last_run_timestamp": "2026-10-03T16:19:14.095Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5-nano-2025-08-07",
      "model_slug": "openai/gpt-5-nano",
      "model_name": "OpenAI: GPT-5 Nano",
      "provider_name": "OpenAI",
      "accuracy": 0.4511785656565656,
      "accuracy_std_dev": 0.03515039216708685,
      "run_count": 4,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.04521708585858586,
      "avg_duration_per_task_ms": 779859.6717171717,
      "avg_output_tokens_per_task": 105433.07575757576,
      "last_run_timestamp": "2026-10-02T23:22:06.591Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-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": 0.44962944444444447,
      "accuracy_std_dev": 0.027054818100832535,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.032987291111111115,
      "avg_duration_per_task_ms": 107424.01111111112,
      "avg_output_tokens_per_task": 3035.311111111111,
      "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": "Venice",
      "accuracy": 0.44925922222222225,
      "accuracy_std_dev": 0.030622329285008784,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.04105088444444445,
      "avg_duration_per_task_ms": 296366.4222222222,
      "avg_output_tokens_per_task": 3639.817777777778,
      "last_run_timestamp": "2026-10-03T16:17:30.535Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Coder 30B A3B Instruct",
      "provider_name": "Alibaba",
      "accuracy": 0.448889,
      "accuracy_std_dev": 0.0031423825355930196,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.08242190666666667,
      "avg_duration_per_task_ms": 81216.82666666666,
      "avg_output_tokens_per_task": 4095.7466666666664,
      "last_run_timestamp": "2026-08-15T08:11:06.981Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-32b-04-28",
      "model_slug": "qwen/qwen3-32b",
      "model_name": "Qwen: Qwen3 32B",
      "provider_name": "DeepInfra",
      "accuracy": 0.44857142857142857,
      "accuracy_std_dev": 0.025125762588894082,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.021147157142857143,
      "avg_duration_per_task_ms": 505078.46285714285,
      "avg_output_tokens_per_task": 21834.525714285715,
      "last_run_timestamp": "2026-08-22T16:08:38.678Z"
    },
    {
      "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": 0.44388611055276384,
      "accuracy_std_dev": 0.025511183797197155,
      "run_count": 8,
      "total_questions": 398,
      "avg_cost_per_task_usd": 0.01678875879396985,
      "avg_duration_per_task_ms": 204045.45477386934,
      "avg_output_tokens_per_task": 2857.608040201005,
      "last_run_timestamp": "2026-10-03T16:14:29.926Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-14b-04-28",
      "model_slug": "qwen/qwen3-14b",
      "model_name": "Qwen: Qwen3 14B",
      "provider_name": "DeepInfra",
      "accuracy": 0.44333333333333336,
      "accuracy_std_dev": 0.025962822424039775,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.028878013333333334,
      "avg_duration_per_task_ms": 681765.0866666667,
      "avg_output_tokens_per_task": 19566.68,
      "last_run_timestamp": "2026-10-01T16:01:00.465Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4.1-mini-2025-04-14",
      "model_slug": "openai/gpt-4.1-mini",
      "model_name": "OpenAI: GPT-4.1 Mini",
      "provider_name": "OpenAI",
      "accuracy": 0.44333325,
      "accuracy_std_dev": 0.01914841159698371,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.025700495,
      "avg_duration_per_task_ms": 61023.305,
      "avg_output_tokens_per_task": 2441.375,
      "last_run_timestamp": "2026-10-02T23:07:02.872Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-chat-v3-0324",
      "model_slug": "deepseek/deepseek-chat-v3-0324",
      "model_name": "DeepSeek: DeepSeek V3 0324",
      "provider_name": "SiliconFlow",
      "accuracy": 0.44133320000000004,
      "accuracy_std_dev": 0.016000066667361108,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.101740272,
      "avg_duration_per_task_ms": 938564.588,
      "avg_output_tokens_per_task": 21680.24,
      "last_run_timestamp": "2026-08-29T16:10:09.206Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4o-2024-05-13",
      "model_slug": "openai/gpt-4o-2024-05-13",
      "model_name": "OpenAI: GPT-4o (2024-05-13)",
      "provider_name": "OpenAI",
      "accuracy": 0.44,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.7291186999999999,
      "avg_duration_per_task_ms": 35507.54,
      "avg_output_tokens_per_task": 3272.36,
      "last_run_timestamp": "2026-08-01T08:05:57.702Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-30b-a3b-04-28",
      "model_slug": "qwen/qwen3-30b-a3b",
      "model_name": "Qwen: Qwen3 30B A3B",
      "provider_name": "Alibaba",
      "accuracy": 0.44,
      "accuracy_std_dev": 0.02788886675718467,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.03487082,
      "avg_duration_per_task_ms": 281485.155,
      "avg_output_tokens_per_task": 21754.5,
      "last_run_timestamp": "2026-08-01T17:38:29.532Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "Google",
      "accuracy": 0.43999974999999997,
      "accuracy_std_dev": 0.047081383440671115,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.050863820000000004,
      "avg_duration_per_task_ms": 37018.795,
      "avg_output_tokens_per_task": 2487.295,
      "last_run_timestamp": "2026-10-03T16:12:29.903Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta-llama/llama-4-maverick-17b-128e-instruct",
      "model_slug": "meta-llama/llama-4-maverick",
      "model_name": "Meta: Llama 4 Maverick",
      "provider_name": "auto-routing",
      "accuracy": 0.43965518965517236,
      "accuracy_std_dev": 0.020525836824522862,
      "run_count": 7,
      "total_questions": 348,
      "avg_cost_per_task_usd": 0.045893882183908045,
      "avg_duration_per_task_ms": 128881.75862068965,
      "avg_output_tokens_per_task": 2035.2413793103449,
      "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": "deepseek/deepseek-chat-v3-0324",
      "model_slug": "deepseek/deepseek-chat-v3-0324",
      "model_name": "DeepSeek: DeepSeek V3 0324",
      "provider_name": "DeepInfra",
      "accuracy": 0.4387170412371134,
      "accuracy_std_dev": 0.022793011768351766,
      "run_count": 6,
      "total_questions": 291,
      "avg_cost_per_task_usd": 0.031083333333333338,
      "avg_duration_per_task_ms": 487106.5326460481,
      "avg_output_tokens_per_task": 4378.945017182131,
      "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/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "Io Net",
      "accuracy": 0.4383965367088608,
      "accuracy_std_dev": 0.03960836597472182,
      "run_count": 8,
      "total_questions": 395,
      "avg_cost_per_task_usd": 0.05240046329113924,
      "avg_duration_per_task_ms": 422067.2658227848,
      "avg_output_tokens_per_task": 20533.427848101266,
      "last_run_timestamp": "2026-09-19T16:00:30.864Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4o",
      "model_slug": "openai/gpt-4o",
      "model_name": "OpenAI: GPT-4o",
      "provider_name": "auto-routing",
      "accuracy": 0.43777799999999994,
      "accuracy_std_dev": 0.03861798340151904,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.19450383999999998,
      "avg_duration_per_task_ms": 50798.87333333334,
      "avg_output_tokens_per_task": 2047.5333333333333,
      "last_run_timestamp": "2026-10-01T16:01:13.623Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4.1-mini-2025-04-14",
      "model_slug": "openai/gpt-4.1-mini",
      "model_name": "OpenAI: GPT-4.1 Mini",
      "provider_name": "auto-routing",
      "accuracy": 0.4377776666666667,
      "accuracy_std_dev": 0.011330936570096718,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.02595394666666667,
      "avg_duration_per_task_ms": 66273.3,
      "avg_output_tokens_per_task": 2407.2933333333335,
      "last_run_timestamp": "2026-10-01T19:04:38.295Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4o-2024-08-06",
      "model_slug": "openai/gpt-4o-2024-08-06",
      "model_name": "OpenAI: GPT-4o (2024-08-06)",
      "provider_name": "auto-routing",
      "accuracy": 0.4377776666666667,
      "accuracy_std_dev": 0.016629618082071397,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.21387673999999998,
      "avg_duration_per_task_ms": 47267.2,
      "avg_output_tokens_per_task": 2224.56,
      "last_run_timestamp": "2026-10-01T16:00:49.880Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-chat-v3-0324",
      "model_slug": "deepseek/deepseek-chat-v3-0324",
      "model_name": "DeepSeek: DeepSeek V3 0324",
      "provider_name": "auto-routing",
      "accuracy": 0.4373334,
      "accuracy_std_dev": 0.019136942750606727,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.077255652,
      "avg_duration_per_task_ms": 788193.684,
      "avg_output_tokens_per_task": 19596.964,
      "last_run_timestamp": "2026-08-29T16:11:46.632Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "DeepInfra",
      "accuracy": 0.4372223333333333,
      "accuracy_std_dev": 0.025270150254576823,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.027693089999999997,
      "avg_duration_per_task_ms": 78315.11,
      "avg_output_tokens_per_task": 2793.7166666666667,
      "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": "Alibaba",
      "accuracy": 0.4370371111111111,
      "accuracy_std_dev": 0.04062616143828847,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.38184548444444444,
      "avg_duration_per_task_ms": 105579.94,
      "avg_output_tokens_per_task": 3104.1155555555556,
      "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": "mistralai/mistral-small-2603",
      "model_slug": "mistralai/mistral-small-2603",
      "model_name": "Mistral: Mistral Small 4",
      "provider_name": "auto-routing",
      "accuracy": 0.43666650000000007,
      "accuracy_std_dev": 0.023333500000000007,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.014376880000000002,
      "avg_duration_per_task_ms": 93606.58,
      "avg_output_tokens_per_task": 10159.34,
      "last_run_timestamp": "2026-09-01T13:02:23.335Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.4353740544217687,
      "accuracy_std_dev": 0.0333418611528944,
      "run_count": 6,
      "total_questions": 294,
      "avg_cost_per_task_usd": 0.03948419727891156,
      "avg_duration_per_task_ms": 53655.58163265306,
      "avg_output_tokens_per_task": 2073.0680272108843,
      "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": "openai/gpt-4o",
      "model_slug": "openai/gpt-4o",
      "model_name": "OpenAI: GPT-4o",
      "provider_name": "OpenAI",
      "accuracy": 0.43416662500000003,
      "accuracy_std_dev": 0.01613729505630899,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.2170375225,
      "avg_duration_per_task_ms": 46329.7025,
      "avg_output_tokens_per_task": 2265.4175,
      "last_run_timestamp": "2026-10-02T23:10:05.271Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-30b-a3b-04-28",
      "model_slug": "qwen/qwen3-30b-a3b",
      "model_name": "Qwen: Qwen3 30B A3B",
      "provider_name": "auto-routing",
      "accuracy": 0.4333336666666667,
      "accuracy_std_dev": 0.0339934634239519,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.03350491333333334,
      "avg_duration_per_task_ms": 439900.87333333335,
      "avg_output_tokens_per_task": 22781.906666666666,
      "last_run_timestamp": "2026-08-01T17:38:32.120Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-30b-a3b-04-28",
      "model_slug": "qwen/qwen3-30b-a3b",
      "model_name": "Qwen: Qwen3 30B A3B",
      "provider_name": "DeepInfra",
      "accuracy": 0.43333333333333335,
      "accuracy_std_dev": 0.014401440259764139,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.03429177333333333,
      "avg_duration_per_task_ms": 515166.3933333333,
      "avg_output_tokens_per_task": 24569.606666666667,
      "last_run_timestamp": "2026-08-01T17:38:35.770Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Coder 30B A3B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.4311113333333333,
      "accuracy_std_dev": 0.006285236475706817,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.023601506666666664,
      "avg_duration_per_task_ms": 216917.58666666667,
      "avg_output_tokens_per_task": 3146.213333333333,
      "last_run_timestamp": "2026-08-15T08:02:48.904Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-14b-04-28",
      "model_slug": "qwen/qwen3-14b",
      "model_name": "Qwen: Qwen3 14B",
      "provider_name": "auto-routing",
      "accuracy": 0.431111,
      "accuracy_std_dev": 0.013698950130089036,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.029769013333333334,
      "avg_duration_per_task_ms": 630152.1266666667,
      "avg_output_tokens_per_task": 18760.18666666667,
      "last_run_timestamp": "2026-10-01T16:00:49.878Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4o-2024-08-06",
      "model_slug": "openai/gpt-4o-2024-08-06",
      "model_name": "OpenAI: GPT-4o (2024-08-06)",
      "provider_name": "OpenAI",
      "accuracy": 0.43,
      "accuracy_std_dev": 0.0033330000000000026,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.24811575000000002,
      "avg_duration_per_task_ms": 46287.18,
      "avg_output_tokens_per_task": 2565.8,
      "last_run_timestamp": "2026-10-02T23:09:02.629Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-chat-v3",
      "model_slug": "deepseek/deepseek-chat",
      "model_name": "DeepSeek: DeepSeek V3",
      "provider_name": "DeepInfra",
      "accuracy": 0.42888899999999996,
      "accuracy_std_dev": 0.022825013355235225,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.07427926,
      "avg_duration_per_task_ms": 326411.7266666667,
      "avg_output_tokens_per_task": 3665.133333333333,
      "last_run_timestamp": "2026-10-01T19:04:28.526Z"
    },
    {
      "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": 0.4266666666666667,
      "accuracy_std_dev": 0.013052515015207695,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.024687224444444444,
      "avg_duration_per_task_ms": 229766.70888888888,
      "avg_output_tokens_per_task": 3591.1422222222222,
      "last_run_timestamp": "2026-10-03T16:13:22.758Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "DeepInfra",
      "accuracy": 0.4256352959641255,
      "accuracy_std_dev": 0.029267857939983613,
      "run_count": 9,
      "total_questions": 446,
      "avg_cost_per_task_usd": 0.024396058295964126,
      "avg_duration_per_task_ms": 309280.5,
      "avg_output_tokens_per_task": 3539.0829596412555,
      "last_run_timestamp": "2026-10-03T16:12:07.465Z"
    },
    {
      "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": 0.4225002499999999,
      "accuracy_std_dev": 0.03843280391394179,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0546106075,
      "avg_duration_per_task_ms": 139402.745,
      "avg_output_tokens_per_task": 3558.66,
      "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-vl-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 VL 30B A3B Instruct",
      "provider_name": "Novita",
      "accuracy": 0.41759616356877327,
      "accuracy_std_dev": 0.026805628250132423,
      "run_count": 11,
      "total_questions": 538,
      "avg_cost_per_task_usd": 0.0537584814126394,
      "avg_duration_per_task_ms": 212875.83271375464,
      "avg_output_tokens_per_task": 4238.7007434944235,
      "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": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "CoreWeave",
      "accuracy": 0.4166667027027027,
      "accuracy_std_dev": 0.06529892179218752,
      "run_count": 6,
      "total_questions": 296,
      "avg_cost_per_task_usd": 0.23566864189189185,
      "avg_duration_per_task_ms": 251017.3918918919,
      "avg_output_tokens_per_task": 9861.692567567568,
      "last_run_timestamp": "2026-10-04T00:00:34.897Z"
    },
    {
      "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": 0.4162963333333333,
      "accuracy_std_dev": 0.12078393478715067,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.02470130444444444,
      "avg_duration_per_task_ms": 179469.95555555556,
      "avg_output_tokens_per_task": 3348.8955555555553,
      "last_run_timestamp": "2026-10-03T16:09:19.691Z"
    },
    {
      "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": 0.4136055102040816,
      "accuracy_std_dev": 0.02936042507338748,
      "run_count": 10,
      "total_questions": 490,
      "avg_cost_per_task_usd": 0.011379914285714287,
      "avg_duration_per_task_ms": 120011.6918367347,
      "avg_output_tokens_per_task": 3351.665306122449,
      "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/qwen3-14b-04-28",
      "model_slug": "qwen/qwen3-14b",
      "model_name": "Qwen: Qwen3 14B",
      "provider_name": "Alibaba",
      "accuracy": 0.413333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.06003468,
      "avg_duration_per_task_ms": 392895.94,
      "avg_output_tokens_per_task": 17058.96,
      "last_run_timestamp": "2026-08-01T08:06:29.769Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "DeepInfra",
      "accuracy": 0.4124148571428572,
      "accuracy_std_dev": 0.023430017383889733,
      "run_count": 8,
      "total_questions": 392,
      "avg_cost_per_task_usd": 0.03184153571428572,
      "avg_duration_per_task_ms": 707919.5331632653,
      "avg_output_tokens_per_task": 7547.298469387755,
      "last_run_timestamp": "2026-10-04T00:00:37.084Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "mistralai/mistral-small-2603",
      "model_slug": "mistralai/mistral-small-2603",
      "model_name": "Mistral: Mistral Small 4",
      "provider_name": "Mistral",
      "accuracy": 0.41142871428571426,
      "accuracy_std_dev": 0.022737753675420126,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.014496525714285714,
      "avg_duration_per_task_ms": 78866.04571428572,
      "avg_output_tokens_per_task": 9061.385714285714,
      "last_run_timestamp": "2026-09-17T02:10:06.482Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_slug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_name": "Qwen: Qwen3 VL 235B A22B Instruct",
      "provider_name": "Alibaba",
      "accuracy": 0.4092593333333333,
      "accuracy_std_dev": 0.022758969050073914,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.06364212444444445,
      "avg_duration_per_task_ms": 99265.78666666667,
      "avg_output_tokens_per_task": 2928.4333333333334,
      "last_run_timestamp": "2026-10-03T16:12:33.500Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_slug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_name": "Qwen: Qwen3 VL 235B A22B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.4088887777777778,
      "accuracy_std_dev": 0.036514668117462,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.05922039999999999,
      "avg_duration_per_task_ms": 309197.2711111111,
      "avg_output_tokens_per_task": 7366.042222222222,
      "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": "openai/gpt-4.1-mini-2025-04-14",
      "model_slug": "openai/gpt-4.1-mini",
      "model_name": "OpenAI: GPT-4.1 Mini",
      "provider_name": "Azure",
      "accuracy": 0.40777783333333334,
      "accuracy_std_dev": 0.01781244190087991,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.025491406666666664,
      "avg_duration_per_task_ms": 73232.91333333333,
      "avg_output_tokens_per_task": 2860.7966666666666,
      "last_run_timestamp": "2026-10-01T19:05:00.133Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-chat-v3",
      "model_slug": "deepseek/deepseek-chat",
      "model_name": "DeepSeek: DeepSeek V3",
      "provider_name": "StreamLake",
      "accuracy": 0.4053599698492462,
      "accuracy_std_dev": 0.008101744842933528,
      "run_count": 4,
      "total_questions": 199,
      "avg_cost_per_task_usd": 0.06672347236180905,
      "avg_duration_per_task_ms": 212067.50753768845,
      "avg_output_tokens_per_task": 8080.718592964824,
      "last_run_timestamp": "2026-10-02T15:44:48.231Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "SiliconFlow",
      "accuracy": 0.4037037777777778,
      "accuracy_std_dev": 0.02436562151327963,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.026673933333333334,
      "avg_duration_per_task_ms": 283419.1955555556,
      "avg_output_tokens_per_task": 4179.831111111111,
      "last_run_timestamp": "2026-10-03T16:13:56.972Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4o-2024-05-13",
      "model_slug": "openai/gpt-4o-2024-05-13",
      "model_name": "OpenAI: GPT-4o (2024-05-13)",
      "provider_name": "auto-routing",
      "accuracy": 0.4,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 1.1315643,
      "avg_duration_per_task_ms": 71996.84,
      "avg_output_tokens_per_task": 3249.82,
      "last_run_timestamp": "2026-08-01T08:06:02.002Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Coder 30B A3B Instruct",
      "provider_name": "SiliconFlow",
      "accuracy": 0.4,
      "accuracy_std_dev": 0.019626323972325204,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.020620846666666668,
      "avg_duration_per_task_ms": 194089.08666666667,
      "avg_output_tokens_per_task": 3830.9066666666668,
      "last_run_timestamp": "2026-08-15T08:02:44.458Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_slug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_name": "Qwen: Qwen3 VL 235B A22B Instruct",
      "provider_name": "Novita",
      "accuracy": 0.39999977777777784,
      "accuracy_std_dev": 0.02330672557094853,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.07755904666666667,
      "avg_duration_per_task_ms": 122311.57111111112,
      "avg_output_tokens_per_task": 3122.713333333333,
      "last_run_timestamp": "2026-10-03T16:12:07.495Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Nebius",
      "accuracy": 0.3994974723618091,
      "accuracy_std_dev": 0.15640288967474153,
      "run_count": 8,
      "total_questions": 398,
      "avg_cost_per_task_usd": 0.1472113216080402,
      "avg_duration_per_task_ms": 361190.4924623116,
      "avg_output_tokens_per_task": 52739.01005025126,
      "last_run_timestamp": "2026-10-03T16:22:23.573Z"
    },
    {
      "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": 0.3992593333333333,
      "accuracy_std_dev": 0.02801133907307301,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.06248461111111111,
      "avg_duration_per_task_ms": 285179.9666666667,
      "avg_output_tokens_per_task": 10620.682222222222,
      "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/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "GMICloud",
      "accuracy": 0.39866660000000004,
      "accuracy_std_dev": 0.17082280826119212,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.016985011999999997,
      "avg_duration_per_task_ms": 130076.872,
      "avg_output_tokens_per_task": 3074.228,
      "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": "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": 0.39851855555555554,
      "accuracy_std_dev": 0.04784532061900936,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.047124519999999996,
      "avg_duration_per_task_ms": 86880.75777777778,
      "avg_output_tokens_per_task": 3704.46,
      "last_run_timestamp": "2026-10-03T16:05:51.296Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_slug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_name": "Qwen: Qwen3 VL 235B A22B Instruct",
      "provider_name": "Parasail",
      "accuracy": 0.39703688888888894,
      "accuracy_std_dev": 0.020393928467095,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.03641048222222223,
      "avg_duration_per_task_ms": 189988.64666666667,
      "avg_output_tokens_per_task": 3741.2244444444445,
      "last_run_timestamp": "2026-10-03T16:08:50.061Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "WandB",
      "accuracy": 0.39666650000000003,
      "accuracy_std_dev": 0.02333349999999998,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.0319114,
      "avg_duration_per_task_ms": 79638.86,
      "avg_output_tokens_per_task": 5369.64,
      "last_run_timestamp": "2026-07-18T08:03:14.950Z"
    },
    {
      "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": 0.3948895644599303,
      "accuracy_std_dev": 0.014575970313613202,
      "run_count": 6,
      "total_questions": 287,
      "avg_cost_per_task_usd": 0.010113132404181184,
      "avg_duration_per_task_ms": 61320.738675958186,
      "avg_output_tokens_per_task": 1041.9094076655051,
      "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": "auto-routing",
      "accuracy": 0.3944446666666666,
      "accuracy_std_dev": 0.015947545189011245,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.07351260000000001,
      "avg_duration_per_task_ms": 297595.3333333333,
      "avg_output_tokens_per_task": 7950.953333333333,
      "last_run_timestamp": "2026-10-01T19:04:42.463Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_slug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_name": "Qwen: Qwen3 VL 235B A22B Instruct",
      "provider_name": "DeepInfra",
      "accuracy": 0.390508134529148,
      "accuracy_std_dev": 0.028796290448254745,
      "run_count": 9,
      "total_questions": 446,
      "avg_cost_per_task_usd": 0.060798977578475334,
      "avg_duration_per_task_ms": 369562.0852017937,
      "avg_output_tokens_per_task": 3902.7331838565024,
      "last_run_timestamp": "2026-10-03T16:12:24.720Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "CoreWeave",
      "accuracy": 0.39,
      "accuracy_std_dev": 0.014782396941407474,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.028387520000000006,
      "avg_duration_per_task_ms": 78569.82333333333,
      "avg_output_tokens_per_task": 4258.516666666666,
      "last_run_timestamp": "2026-08-29T16:13:12.633Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.38814811111111114,
      "accuracy_std_dev": 0.0223944854097037,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.02571107333333333,
      "avg_duration_per_task_ms": 215286.6288888889,
      "avg_output_tokens_per_task": 4475.344444444445,
      "last_run_timestamp": "2026-10-03T16:03:16.197Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "Nebius",
      "accuracy": 0.385833375,
      "accuracy_std_dev": 0.008457413743832989,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.014628382499999999,
      "avg_duration_per_task_ms": 83726.5775,
      "avg_output_tokens_per_task": 599.935,
      "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-8b-instruct",
      "model_slug": "qwen/qwen3-vl-8b-instruct",
      "model_name": "Qwen: Qwen3 VL 8B Instruct",
      "provider_name": "Parasail",
      "accuracy": 0.38215508080808086,
      "accuracy_std_dev": 0.014660000000000006,
      "run_count": 2,
      "total_questions": 99,
      "avg_cost_per_task_usd": 0.05195866666666667,
      "avg_duration_per_task_ms": 134392.0202020202,
      "avg_output_tokens_per_task": 8684.282828282829,
      "last_run_timestamp": "2026-07-25T08:07:19.515Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.37978144444444445,
      "accuracy_std_dev": 0.018078801359923458,
      "run_count": 11,
      "total_questions": 549,
      "avg_cost_per_task_usd": 0.012327355191256831,
      "avg_duration_per_task_ms": 34864.229508196724,
      "avg_output_tokens_per_task": 609.8688524590164,
      "last_run_timestamp": "2026-10-03T16:12:33.526Z"
    },
    {
      "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": "WandB",
      "accuracy": 0.3766665,
      "accuracy_std_dev": 0.0033334999999999892,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.07082534,
      "avg_duration_per_task_ms": 16764.74,
      "avg_output_tokens_per_task": 860.73,
      "last_run_timestamp": "2026-07-18T16:00:09.280Z"
    },
    {
      "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": 0.374074,
      "accuracy_std_dev": 0.03118154021211909,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.05026599333333333,
      "avg_duration_per_task_ms": 71333.42666666667,
      "avg_output_tokens_per_task": 4374.031111111111,
      "last_run_timestamp": "2026-10-03T16:05:34.166Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "Nebius",
      "accuracy": 0.37166650000000007,
      "accuracy_std_dev": 0.01658317420610421,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.032970215000000004,
      "avg_duration_per_task_ms": 150852.9325,
      "avg_output_tokens_per_task": 4449.52,
      "last_run_timestamp": "2026-10-03T16:07:02.084Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4o-2024-05-13",
      "model_slug": "openai/gpt-4o-2024-05-13",
      "model_name": "OpenAI: GPT-4o (2024-05-13)",
      "provider_name": "Azure",
      "accuracy": 0.36666699999999997,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.80181,
      "avg_duration_per_task_ms": 63430.68,
      "avg_output_tokens_per_task": 3064.6,
      "last_run_timestamp": "2026-08-01T08:06:07.633Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.36476185714285714,
      "accuracy_std_dev": 0.029753169004943562,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.07516260571428572,
      "avg_duration_per_task_ms": 18915.674285714285,
      "avg_output_tokens_per_task": 875.96,
      "last_run_timestamp": "2026-10-03T16:05:34.168Z"
    },
    {
      "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": 0.3614163516483516,
      "accuracy_std_dev": 0.028501851452693475,
      "run_count": 11,
      "total_questions": 546,
      "avg_cost_per_task_usd": 0.070551663003663,
      "avg_duration_per_task_ms": 367421.45604395604,
      "avg_output_tokens_per_task": 7535.331501831502,
      "last_run_timestamp": "2026-10-03T08:32:03.332Z"
    },
    {
      "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": 0.3582909492385786,
      "accuracy_std_dev": 0.07485838274869422,
      "run_count": 8,
      "total_questions": 394,
      "avg_cost_per_task_usd": 0.0038857563451776644,
      "avg_duration_per_task_ms": 276001.0329949239,
      "avg_output_tokens_per_task": 4008.0659898477156,
      "last_run_timestamp": "2026-10-03T16:11:33.007Z"
    },
    {
      "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": 0.3564477189781022,
      "accuracy_std_dev": 0.023446226503206637,
      "run_count": 11,
      "total_questions": 548,
      "avg_cost_per_task_usd": 0.07946453284671531,
      "avg_duration_per_task_ms": 237669.46532846717,
      "avg_output_tokens_per_task": 7645.419708029197,
      "last_run_timestamp": "2026-10-03T08:12:20.019Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-vl-8b-instruct",
      "model_slug": "qwen/qwen3-vl-8b-instruct",
      "model_name": "Qwen: Qwen3 VL 8B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.3541665,
      "accuracy_std_dev": 0.015343212399950682,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.041681384999999994,
      "avg_duration_per_task_ms": 103610.79,
      "avg_output_tokens_per_task": 4507.515,
      "last_run_timestamp": "2026-08-01T17:38:36.457Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Parasail",
      "accuracy": 0.35259277777777775,
      "accuracy_std_dev": 0.07615423091445964,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.005360093333333335,
      "avg_duration_per_task_ms": 130268.62,
      "avg_output_tokens_per_task": 3481.0555555555557,
      "last_run_timestamp": "2026-10-03T16:07:05.240Z"
    },
    {
      "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": "Ionstream",
      "accuracy": 0.34353760714285714,
      "accuracy_std_dev": 0.24874446504633524,
      "run_count": 4,
      "total_questions": 196,
      "avg_cost_per_task_usd": 0.009734260204081633,
      "avg_duration_per_task_ms": 70782.52551020408,
      "avg_output_tokens_per_task": 2162.954081632653,
      "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": "deepseek/deepseek-chat-v3",
      "model_slug": "deepseek/deepseek-chat",
      "model_name": "DeepSeek: DeepSeek V3",
      "provider_name": "Novita",
      "accuracy": 0.34058,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 46,
      "avg_cost_per_task_usd": 0.11353834782608695,
      "avg_duration_per_task_ms": 397172.5,
      "avg_output_tokens_per_task": 14067.434782608696,
      "last_run_timestamp": "2026-08-01T08:05:57.857Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.34005986995515697,
      "accuracy_std_dev": 0.05113342105622878,
      "run_count": 9,
      "total_questions": 446,
      "avg_cost_per_task_usd": 0.02689973766816143,
      "avg_duration_per_task_ms": 99831.18609865471,
      "avg_output_tokens_per_task": 1162.4708520179372,
      "last_run_timestamp": "2026-10-03T16:14:36.877Z"
    },
    {
      "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": 0.3367119324324324,
      "accuracy_std_dev": 0.23078482961913535,
      "run_count": 6,
      "total_questions": 296,
      "avg_cost_per_task_usd": 0.014007192567567568,
      "avg_duration_per_task_ms": 1521898.9797297297,
      "avg_output_tokens_per_task": 94037.88851351352,
      "last_run_timestamp": "2026-10-03T16:12:24.696Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-vl-8b-instruct",
      "model_slug": "qwen/qwen3-vl-8b-instruct",
      "model_name": "Qwen: Qwen3 VL 8B Instruct",
      "provider_name": "Alibaba",
      "accuracy": 0.3341665,
      "accuracy_std_dev": 0.0059509803184013355,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.03100886,
      "avg_duration_per_task_ms": 84710.3,
      "avg_output_tokens_per_task": 3999.695,
      "last_run_timestamp": "2026-08-01T17:38:37.779Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 VL 30B A3B Instruct",
      "provider_name": "Alibaba",
      "accuracy": 0.328889,
      "accuracy_std_dev": 0.022986915227291085,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.08611619166666667,
      "avg_duration_per_task_ms": 152663.81666666668,
      "avg_output_tokens_per_task": 8236.506666666666,
      "last_run_timestamp": "2026-10-03T08:23:09.856Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "Venice",
      "accuracy": 0.3206666,
      "accuracy_std_dev": 0.3296116389532384,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.019227021999999996,
      "avg_duration_per_task_ms": 58892.988,
      "avg_output_tokens_per_task": 10442.698,
      "last_run_timestamp": "2026-10-03T16:12:07.499Z"
    },
    {
      "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": 0.31282059999999995,
      "accuracy_std_dev": 0.030993226130882217,
      "run_count": 8,
      "total_questions": 390,
      "avg_cost_per_task_usd": 0.015985989743589744,
      "avg_duration_per_task_ms": 332807.73846153845,
      "avg_output_tokens_per_task": 4072.7410256410258,
      "last_run_timestamp": "2026-08-29T16:43:42.449Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "Together",
      "accuracy": 0.30370366666666665,
      "accuracy_std_dev": 0.035847625249101234,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.16472546666666668,
      "avg_duration_per_task_ms": 120261.91777777778,
      "avg_output_tokens_per_task": 1852.1177777777777,
      "last_run_timestamp": "2026-10-03T16:16:35.584Z"
    },
    {
      "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": 0.2965195220883534,
      "accuracy_std_dev": 0.18696214545570447,
      "run_count": 10,
      "total_questions": 498,
      "avg_cost_per_task_usd": 0.14611226305220884,
      "avg_duration_per_task_ms": 854783.9919678714,
      "avg_output_tokens_per_task": 5611.467871485944,
      "last_run_timestamp": "2026-10-03T08:09:23.857Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4o-mini",
      "model_slug": "openai/gpt-4o-mini",
      "model_name": "OpenAI: GPT-4o-mini",
      "provider_name": "auto-routing",
      "accuracy": 0.2844443333333333,
      "accuracy_std_dev": 0.031894788459698065,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.02081186,
      "avg_duration_per_task_ms": 62184.12666666666,
      "avg_output_tokens_per_task": 3804.7533333333336,
      "last_run_timestamp": "2026-10-01T19:04:58.075Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4o-mini",
      "model_slug": "openai/gpt-4o-mini",
      "model_name": "OpenAI: GPT-4o-mini",
      "provider_name": "OpenAI",
      "accuracy": 0.28000025,
      "accuracy_std_dev": 0.027080189561513406,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.02535222,
      "avg_duration_per_task_ms": 67904.845,
      "avg_output_tokens_per_task": 4023.05,
      "last_run_timestamp": "2026-10-02T23:13:47.090Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.266667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.00574078,
      "avg_duration_per_task_ms": 359216.92,
      "avg_output_tokens_per_task": 2727.92,
      "last_run_timestamp": "2026-07-11T08:03:53.497Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen-2.5-72b-instruct",
      "model_slug": "qwen/qwen-2.5-72b-instruct",
      "model_name": "Qwen2.5 72B Instruct",
      "provider_name": "Novita",
      "accuracy": 0.2622223333333333,
      "accuracy_std_dev": 0.01911642045177101,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.08337879999999999,
      "avg_duration_per_task_ms": 576772.3666666667,
      "avg_output_tokens_per_task": 5220.566666666667,
      "last_run_timestamp": "2026-10-01T16:00:45.063Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen-2.5-72b-instruct",
      "model_slug": "qwen/qwen-2.5-72b-instruct",
      "model_name": "Qwen2.5 72B Instruct",
      "provider_name": "DeepInfra",
      "accuracy": 0.25,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.0713485,
      "avg_duration_per_task_ms": 660516.14,
      "avg_output_tokens_per_task": 6662.1,
      "last_run_timestamp": "2026-09-01T12:59:14.290Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen-2.5-72b-instruct",
      "model_slug": "qwen/qwen-2.5-72b-instruct",
      "model_name": "Qwen2.5 72B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.24200919178082192,
      "accuracy_std_dev": 0.027381797668443095,
      "run_count": 3,
      "total_questions": 146,
      "avg_cost_per_task_usd": 0.07592458219178082,
      "avg_duration_per_task_ms": 576526.0342465753,
      "avg_output_tokens_per_task": 5262.068493150685,
      "last_run_timestamp": "2026-10-01T16:00:49.856Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-next-2025-02-03",
      "model_slug": "qwen/qwen3-coder-next",
      "model_name": "Qwen: Qwen3 Coder Next",
      "provider_name": "Novita",
      "accuracy": 0.24,
      "accuracy_std_dev": 0.06798683968566596,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.014107426666666664,
      "avg_duration_per_task_ms": 23182.855555555554,
      "avg_output_tokens_per_task": 1167.951111111111,
      "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": "mistralai/mistral-nemo",
      "model_slug": "mistralai/mistral-nemo",
      "model_name": "Mistral: Mistral Nemo",
      "provider_name": "auto-routing",
      "accuracy": 0.23299353061224487,
      "accuracy_std_dev": 0.02687500000000001,
      "run_count": 2,
      "total_questions": 98,
      "avg_cost_per_task_usd": 0.028852255102040817,
      "avg_duration_per_task_ms": 377353.28571428574,
      "avg_output_tokens_per_task": 5486.489795918367,
      "last_run_timestamp": "2026-07-18T08:04:09.749Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "mistralai/mistral-nemo",
      "model_slug": "mistralai/mistral-nemo",
      "model_name": "Mistral: Mistral Nemo",
      "provider_name": "Mistral",
      "accuracy": 0.2316665,
      "accuracy_std_dev": 0.018333500000000003,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.02960246,
      "avg_duration_per_task_ms": 418677.18,
      "avg_output_tokens_per_task": 5444.6,
      "last_run_timestamp": "2026-07-18T08:04:25.389Z"
    },
    {
      "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": 0.21466659999999996,
      "accuracy_std_dev": 0.04182530015959241,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.005141112,
      "avg_duration_per_task_ms": 40939.9,
      "avg_output_tokens_per_task": 1789.784,
      "last_run_timestamp": "2026-10-03T16:07:44.178Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "amazon/nova-micro-v1",
      "model_slug": "amazon/nova-micro-v1",
      "model_name": "Amazon: Nova Micro 1.0",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.18766967073170732,
      "accuracy_std_dev": 0.018396157170452745,
      "run_count": 5,
      "total_questions": 246,
      "avg_cost_per_task_usd": 0.011055426829268291,
      "avg_duration_per_task_ms": 138140.34552845528,
      "avg_output_tokens_per_task": 5390.59756097561,
      "last_run_timestamp": "2026-10-01T19:04:52.650Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.1-terminus",
      "model_slug": "deepseek/deepseek-v3.1-terminus",
      "model_name": "DeepSeek: DeepSeek V3.1 Terminus",
      "provider_name": "AtlasCloud",
      "accuracy": 0.17333325,
      "accuracy_std_dev": 0.11522926663368771,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0127158975,
      "avg_duration_per_task_ms": 62680.3375,
      "avg_output_tokens_per_task": 1280.3475,
      "last_run_timestamp": "2026-08-29T16:17:08.780Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "AtlasCloud",
      "accuracy": 0.16858231034482757,
      "accuracy_std_dev": 0.10965908309383035,
      "run_count": 7,
      "total_questions": 348,
      "avg_cost_per_task_usd": 0.19030922126436778,
      "avg_duration_per_task_ms": 565575.5,
      "avg_output_tokens_per_task": 9876.801724137931,
      "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/qwen-2.5-7b-instruct",
      "model_slug": "qwen/qwen-2.5-7b-instruct",
      "model_name": "Qwen: Qwen2.5 7B Instruct",
      "provider_name": "Phala",
      "accuracy": 0.166667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.01877496,
      "avg_duration_per_task_ms": 63137.36,
      "avg_output_tokens_per_task": 2797.32,
      "last_run_timestamp": "2026-08-01T08:06:16.826Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen-2.5-7b-instruct",
      "model_slug": "qwen/qwen-2.5-7b-instruct",
      "model_name": "Qwen: Qwen2.5 7B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.166667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.016978299999999998,
      "avg_duration_per_task_ms": 57790.52,
      "avg_output_tokens_per_task": 2492.08,
      "last_run_timestamp": "2026-08-01T08:06:07.057Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4.1-nano-2025-04-14",
      "model_slug": "openai/gpt-4.1-nano",
      "model_name": "OpenAI: GPT-4.1 Nano",
      "provider_name": "Azure",
      "accuracy": 0.13333333333333333,
      "accuracy_std_dev": 0.021430473285072877,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.007961373333333334,
      "avg_duration_per_task_ms": 80728.30333333333,
      "avg_output_tokens_per_task": 2469.17,
      "last_run_timestamp": "2026-10-01T19:04:58.093Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4.1-nano-2025-04-14",
      "model_slug": "openai/gpt-4.1-nano",
      "model_name": "OpenAI: GPT-4.1 Nano",
      "provider_name": "OpenAI",
      "accuracy": 0.11833325,
      "accuracy_std_dev": 0.018484159885899604,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.007158584999999999,
      "avg_duration_per_task_ms": 55526.67,
      "avg_output_tokens_per_task": 2353.445,
      "last_run_timestamp": "2026-10-02T23:07:59.112Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4.1-nano-2025-04-14",
      "model_slug": "openai/gpt-4.1-nano",
      "model_name": "OpenAI: GPT-4.1 Nano",
      "provider_name": "auto-routing",
      "accuracy": 0.108889,
      "accuracy_std_dev": 0.008314764498569193,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.006775626666666668,
      "avg_duration_per_task_ms": 60620.74,
      "avg_output_tokens_per_task": 2327.0066666666667,
      "last_run_timestamp": "2026-10-01T19:04:36.194Z"
    }
  ]
}
