{
  "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": "2026-10-11",
  "retrieved_at": "2026-10-11T01:46:09.977Z",
  "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": 2173,
  "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": "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": "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": 0.9309763333333333,
      "accuracy_std_dev": 0.00880159775899302,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.026392959034792367,
      "avg_duration_per_task_ms": 18809.703142536477,
      "avg_output_tokens_per_task": 2248.6481481481483,
      "last_run_timestamp": "2026-10-10T08:25:06.484Z"
    },
    {
      "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": 0.930415,
      "accuracy_std_dev": 0.0034590473640392213,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.02488252693602694,
      "avg_duration_per_task_ms": 19612.508417508416,
      "avg_output_tokens_per_task": 2247.9511784511783,
      "last_run_timestamp": "2026-10-10T08:14:45.743Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.6-flash-20260721",
      "model_slug": "google/gemini-3.6-flash",
      "model_name": "Google: Gemini 3.6 Flash",
      "provider_name": "Google",
      "accuracy": 0.930415,
      "accuracy_std_dev": 0.004827415699246697,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.06216489141414142,
      "avg_duration_per_task_ms": 64761.851851851854,
      "avg_output_tokens_per_task": 11730.429292929293,
      "last_run_timestamp": "2026-10-01T19:12:42.153Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Parasail",
      "accuracy": 0.9297306163895488,
      "accuracy_std_dev": 0.01120962874243434,
      "run_count": 9,
      "total_questions": 1684,
      "avg_cost_per_task_usd": 0.1273893414489311,
      "avg_duration_per_task_ms": 147661.65023752968,
      "avg_output_tokens_per_task": 8445.026128266034,
      "last_run_timestamp": "2026-10-10T08:22:26.915Z"
    },
    {
      "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": "Relace",
      "accuracy": 0.9277397342419081,
      "accuracy_std_dev": 0.007149525570895533,
      "run_count": 6,
      "total_questions": 1174,
      "avg_cost_per_task_usd": 0.1081421950596252,
      "avg_duration_per_task_ms": 192835.26320272574,
      "avg_output_tokens_per_task": 9335.29471890971,
      "last_run_timestamp": "2026-10-10T08:22:51.532Z"
    },
    {
      "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": 0.9275660041194645,
      "accuracy_std_dev": 0.01457727458237649,
      "run_count": 6,
      "total_questions": 971,
      "avg_cost_per_task_usd": 0.10551570648815653,
      "avg_duration_per_task_ms": 239156.22245108135,
      "avg_output_tokens_per_task": 8138.333676622039,
      "last_run_timestamp": "2026-10-10T08:25:06.968Z"
    },
    {
      "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": 0.927048,
      "accuracy_std_dev": 0.0020998715833752102,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.025530393939393943,
      "avg_duration_per_task_ms": 19102.368686868685,
      "avg_output_tokens_per_task": 2313.7003367003367,
      "last_run_timestamp": "2026-10-10T08:14:48.725Z"
    },
    {
      "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": 0.9269767035321367,
      "accuracy_std_dev": 0.007983506267308428,
      "run_count": 28,
      "total_questions": 5181,
      "avg_cost_per_task_usd": 0.16678585485427527,
      "avg_duration_per_task_ms": 152403.25902335456,
      "avg_output_tokens_per_task": 9054.98861223702,
      "last_run_timestamp": "2026-10-10T08:23:04.185Z"
    },
    {
      "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": "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": 0.926487,
      "accuracy_std_dev": 0.0028613344206273234,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.02518170033670034,
      "avg_duration_per_task_ms": 20924.996632996634,
      "avg_output_tokens_per_task": 2280.7272727272725,
      "last_run_timestamp": "2026-10-10T08:45:10.390Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.5-flash-20260519",
      "model_slug": "google/gemini-3.5-flash",
      "model_name": "Google: Gemini 3.5 Flash",
      "provider_name": "Google AI Studio",
      "accuracy": 0.925926,
      "accuracy_std_dev": 0.005498287975967322,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.1441699393939394,
      "avg_duration_per_task_ms": 61658.986531986535,
      "avg_output_tokens_per_task": 15894.8367003367,
      "last_run_timestamp": "2026-10-01T19:12:44.398Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.5-flash-20260519",
      "model_slug": "google/gemini-3.5-flash",
      "model_name": "Google: Gemini 3.5 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.9259256666666668,
      "accuracy_std_dev": 0.0036366213562713575,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.1413088872053872,
      "avg_duration_per_task_ms": 82468.60606060606,
      "avg_output_tokens_per_task": 15584.752525252525,
      "last_run_timestamp": "2026-10-01T19:18:40.706Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.5-20260423",
      "model_slug": "openai/gpt-5.5",
      "model_name": "OpenAI: GPT-5.5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.9255440309278352,
      "accuracy_std_dev": 0.0047577793839843644,
      "run_count": 3,
      "total_questions": 582,
      "avg_cost_per_task_usd": 0.3407643092783505,
      "avg_duration_per_task_ms": 132309.46391752578,
      "avg_output_tokens_per_task": 10197.671821305841,
      "last_run_timestamp": "2026-10-01T19:13:06.559Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-sol-20260709",
      "model_slug": "openai/gpt-5.6-sol",
      "model_name": "OpenAI: GPT-5.6 Sol",
      "provider_name": "Azure",
      "accuracy": 0.9248735,
      "accuracy_std_dev": 0.00862277742377712,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.10364464204545455,
      "avg_duration_per_task_ms": 55123.33648989899,
      "avg_output_tokens_per_task": 3517.8642676767677,
      "last_run_timestamp": "2026-10-01T19:42:48.407Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "x-ai/grok-4.6-20260810",
      "model_slug": "x-ai/grok-4.6",
      "model_name": "SpaceXAI: Grok 4.6",
      "provider_name": "xAI",
      "accuracy": 0.9246375695652174,
      "accuracy_std_dev": 0.006863146440316182,
      "run_count": 7,
      "total_questions": 1380,
      "avg_cost_per_task_usd": 0.12292555942028985,
      "avg_duration_per_task_ms": 340480.1579710145,
      "avg_output_tokens_per_task": 19662.684057971015,
      "last_run_timestamp": "2026-10-01T16:09:31.184Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-sol-20260922",
      "model_slug": "openai/gpt-6-sol",
      "model_name": "OpenAI: GPT-6 Sol",
      "provider_name": "Azure",
      "accuracy": 0.924599700084246,
      "accuracy_std_dev": 0.006819818440072705,
      "run_count": 6,
      "total_questions": 1187,
      "avg_cost_per_task_usd": 0.02751449789385004,
      "avg_duration_per_task_ms": 24957.139848357205,
      "avg_output_tokens_per_task": 2420.7127211457455,
      "last_run_timestamp": "2026-10-01T08:24:52.927Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.5-flash-20260519",
      "model_slug": "google/gemini-3.5-flash",
      "model_name": "Google: Gemini 3.5 Flash",
      "provider_name": "Google",
      "accuracy": 0.9242421666666667,
      "accuracy_std_dev": 0.0048599613653014136,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.14878865656565657,
      "avg_duration_per_task_ms": 85040.50420875421,
      "avg_output_tokens_per_task": 15641.649831649831,
      "last_run_timestamp": "2026-10-01T19:13:26.156Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "unbiased/pareto",
      "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": "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": 0.9242419999999999,
      "accuracy_std_dev": 0.003367000000000009,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.025158247474747476,
      "avg_duration_per_task_ms": 20743.68181818182,
      "avg_output_tokens_per_task": 2280.7853535353534,
      "last_run_timestamp": "2026-10-10T08:14:39.188Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-sol-20260709",
      "model_slug": "openai/gpt-5.6-sol",
      "model_name": "OpenAI: GPT-5.6 Sol",
      "provider_name": "OpenAI",
      "accuracy": 0.9234005000000001,
      "accuracy_std_dev": 0.008625346413333228,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.05533751388888889,
      "avg_duration_per_task_ms": 61705.84722222222,
      "avg_output_tokens_per_task": 3564.286616161616,
      "last_run_timestamp": "2026-10-02T23:20:36.509Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Decart",
      "accuracy": 0.9228396666666667,
      "accuracy_std_dev": 0.0024135722257452555,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.12293840404040404,
      "avg_duration_per_task_ms": 199289.85858585857,
      "avg_output_tokens_per_task": 9078.558922558923,
      "last_run_timestamp": "2026-10-10T08:23:18.238Z"
    },
    {
      "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": "Amazon Bedrock",
      "accuracy": 0.9225589999999999,
      "accuracy_std_dev": 0.003367000000000009,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.15024601262626264,
      "avg_duration_per_task_ms": 83108.62878787878,
      "avg_output_tokens_per_task": 9763.2601010101,
      "last_run_timestamp": "2026-10-10T08:24:29.592Z"
    },
    {
      "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": 0.9211143939126598,
      "accuracy_std_dev": 0.007124160458822897,
      "run_count": 12,
      "total_questions": 2267,
      "avg_cost_per_task_usd": 0.15088417026907808,
      "avg_duration_per_task_ms": 133150.59550066167,
      "avg_output_tokens_per_task": 9907.07454786061,
      "last_run_timestamp": "2026-10-10T08:23:05.103Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-sol-20260709",
      "model_slug": "openai/gpt-5.6-sol",
      "model_name": "OpenAI: GPT-5.6 Sol",
      "provider_name": "auto-routing",
      "accuracy": 0.9208750000000001,
      "accuracy_std_dev": 0.004761657064510223,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.06043574579124578,
      "avg_duration_per_task_ms": 62971.38047138047,
      "avg_output_tokens_per_task": 3586.8703703703704,
      "last_run_timestamp": "2026-10-01T19:12:28.573Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-sol-20260922",
      "model_slug": "openai/gpt-6-sol",
      "model_name": "OpenAI: GPT-6 Sol",
      "provider_name": "OpenAI",
      "accuracy": 0.9208749999999999,
      "accuracy_std_dev": 0.002749143987983661,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.025590255892255893,
      "avg_duration_per_task_ms": 33502.22222222222,
      "avg_output_tokens_per_task": 2400.506734006734,
      "last_run_timestamp": "2026-10-02T23:17:42.236Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Alibaba",
      "accuracy": 0.9206347142857143,
      "accuracy_std_dev": 0.006819412411099165,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.17132878354978356,
      "avg_duration_per_task_ms": 313924.4264069264,
      "avg_output_tokens_per_task": 9747.513708513708,
      "last_run_timestamp": "2026-10-10T08:21:47.068Z"
    },
    {
      "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": 0.9205385999999999,
      "accuracy_std_dev": 0.008066750289924707,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.1561665070707071,
      "avg_duration_per_task_ms": 221042.0797979798,
      "avg_output_tokens_per_task": 12387.791919191919,
      "last_run_timestamp": "2026-10-10T08:21:47.460Z"
    },
    {
      "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": 0.9201543251770765,
      "accuracy_std_dev": 0.015248606655609568,
      "run_count": 8,
      "total_questions": 1553,
      "avg_cost_per_task_usd": 0.25049747198969735,
      "avg_duration_per_task_ms": 407837.91500321956,
      "avg_output_tokens_per_task": 36515.1790083709,
      "last_run_timestamp": "2026-10-10T08:27:33.606Z"
    },
    {
      "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": 0.9197903617021278,
      "accuracy_std_dev": 0.011919658050883864,
      "run_count": 10,
      "total_questions": 1974,
      "avg_cost_per_task_usd": 0.13309346149949342,
      "avg_duration_per_task_ms": 301635.15045592707,
      "avg_output_tokens_per_task": 10335.872340425532,
      "last_run_timestamp": "2026-10-10T08:24:29.954Z"
    },
    {
      "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": "BaseTen",
      "accuracy": 0.918596129750983,
      "accuracy_std_dev": 0.013046134717055308,
      "run_count": 12,
      "total_questions": 2289,
      "avg_cost_per_task_usd": 0.12497408082131936,
      "avg_duration_per_task_ms": 123328.45827872433,
      "avg_output_tokens_per_task": 8354.565749235475,
      "last_run_timestamp": "2026-10-10T08:22:13.673Z"
    },
    {
      "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": 0.91835,
      "accuracy_std_dev": 0.0070592047474674425,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.09014416835016834,
      "avg_duration_per_task_ms": 315664.6085858586,
      "avg_output_tokens_per_task": 25003.872053872055,
      "last_run_timestamp": "2026-10-10T08:22:27.653Z"
    },
    {
      "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": 0.9172897562692477,
      "accuracy_std_dev": 0.03762026976984024,
      "run_count": 12,
      "total_questions": 2273,
      "avg_cost_per_task_usd": 0.12029984205895294,
      "avg_duration_per_task_ms": 200092.88737351517,
      "avg_output_tokens_per_task": 9308.563572371315,
      "last_run_timestamp": "2026-10-10T08:25:06.885Z"
    },
    {
      "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": 0.9158840305164319,
      "accuracy_std_dev": 0.007323011048053776,
      "run_count": 11,
      "total_questions": 2130,
      "avg_cost_per_task_usd": 0.14128096619718308,
      "avg_duration_per_task_ms": 291601.6774647887,
      "avg_output_tokens_per_task": 9258.633333333333,
      "last_run_timestamp": "2026-10-10T08:23:05.358Z"
    },
    {
      "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": 0.9155198394495413,
      "accuracy_std_dev": 0.012423863218230725,
      "run_count": 7,
      "total_questions": 872,
      "avg_cost_per_task_usd": 0.10930780733944953,
      "avg_duration_per_task_ms": 186343.69036697247,
      "avg_output_tokens_per_task": 7334.917431192661,
      "last_run_timestamp": "2026-10-10T08:22:26.808Z"
    },
    {
      "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": 0.9152635,
      "accuracy_std_dev": 0.003022191630257765,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.02597409595959596,
      "avg_duration_per_task_ms": 19482.909090909092,
      "avg_output_tokens_per_task": 2237.35101010101,
      "last_run_timestamp": "2026-10-10T08:24:21.100Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Minimax",
      "accuracy": 0.9145621666666666,
      "accuracy_std_dev": 0.012656594788444839,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.033476913720538716,
      "avg_duration_per_task_ms": 229462.40782828283,
      "avg_output_tokens_per_task": 27719.715488215486,
      "last_run_timestamp": "2026-10-10T16:22:04.085Z"
    },
    {
      "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": 0.9141412857142857,
      "accuracy_std_dev": 0.006612511252256941,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.11183320346320347,
      "avg_duration_per_task_ms": 640230.5497835497,
      "avg_output_tokens_per_task": 40187.943001443,
      "last_run_timestamp": "2026-10-10T08:14:42.245Z"
    },
    {
      "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": 0.9131312,
      "accuracy_std_dev": 0.00943941759644102,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.15941067777777776,
      "avg_duration_per_task_ms": 524125.72525252524,
      "avg_output_tokens_per_task": 39938.53333333333,
      "last_run_timestamp": "2026-10-10T08:21:33.193Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "DeepInfra",
      "accuracy": 0.9128786136363636,
      "accuracy_std_dev": 0.01857399999999998,
      "run_count": 2,
      "total_questions": 352,
      "avg_cost_per_task_usd": 0.13041307102272726,
      "avg_duration_per_task_ms": 1907977.3181818181,
      "avg_output_tokens_per_task": 51917.37784090909,
      "last_run_timestamp": "2026-10-10T08:22:15.544Z"
    },
    {
      "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": 0.912456052026013,
      "accuracy_std_dev": 0.03101228282566373,
      "run_count": 21,
      "total_questions": 3998,
      "avg_cost_per_task_usd": 0.14550272386193094,
      "avg_duration_per_task_ms": 428036.3906953477,
      "avg_output_tokens_per_task": 8925.227113556779,
      "last_run_timestamp": "2026-10-10T08:15:00.283Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "GMICloud",
      "accuracy": 0.9117844,
      "accuracy_std_dev": 0.006079420781620585,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.03790721616161616,
      "avg_duration_per_task_ms": 129456.97070707071,
      "avg_output_tokens_per_task": 28484.87575757576,
      "last_run_timestamp": "2026-10-03T09:28:48.247Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Chutes",
      "accuracy": 0.9106839879573877,
      "accuracy_std_dev": 0.01509499450450079,
      "run_count": 11,
      "total_questions": 2159,
      "avg_cost_per_task_usd": 0.18564463547938861,
      "avg_duration_per_task_ms": 504241.5956461325,
      "avg_output_tokens_per_task": 12306.873089393237,
      "last_run_timestamp": "2026-10-10T08:15:01.297Z"
    },
    {
      "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": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "CoreWeave",
      "accuracy": 0.9093359666329627,
      "accuracy_std_dev": 0.004225510176298254,
      "run_count": 10,
      "total_questions": 1978,
      "avg_cost_per_task_usd": 0.03006448432760364,
      "avg_duration_per_task_ms": 457008.77856420627,
      "avg_output_tokens_per_task": 31093.431243680487,
      "last_run_timestamp": "2026-10-10T16:21:14.191Z"
    },
    {
      "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": "Alibaba",
      "accuracy": 0.9080194545454546,
      "accuracy_std_dev": 0.016023878405586615,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.13567432598714418,
      "avg_duration_per_task_ms": 551540.5821854912,
      "avg_output_tokens_per_task": 39959.97474747475,
      "last_run_timestamp": "2026-10-10T08:34:36.569Z"
    },
    {
      "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": "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": "GMICloud",
      "accuracy": 0.9073368212927757,
      "accuracy_std_dev": 0.025244605272986335,
      "run_count": 12,
      "total_questions": 2367,
      "avg_cost_per_task_usd": 0.03730512885509083,
      "avg_duration_per_task_ms": 249194.64807773553,
      "avg_output_tokens_per_task": 30447.69877482045,
      "last_run_timestamp": "2026-10-10T16:20:06.472Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/switchyard",
      "model_slug": "nvidia/switchyard",
      "model_name": "NVIDIA: Switchyard",
      "provider_name": "auto-routing",
      "accuracy": 0.9071381724573594,
      "accuracy_std_dev": 0.03371213456235009,
      "run_count": 8,
      "total_questions": 1583,
      "avg_cost_per_task_usd": 0.015826173089071385,
      "avg_duration_per_task_ms": 64047.12571067593,
      "avg_output_tokens_per_task": 3604.423246999368,
      "last_run_timestamp": "2026-09-30T03:11:21.812Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Together",
      "accuracy": 0.9068924337714864,
      "accuracy_std_dev": 0.016448572625003058,
      "run_count": 10,
      "total_questions": 1978,
      "avg_cost_per_task_usd": 0.035087561172901925,
      "avg_duration_per_task_ms": 328169.32204246713,
      "avg_output_tokens_per_task": 29201.10465116279,
      "last_run_timestamp": "2026-10-10T16:22:15.257Z"
    },
    {
      "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": "Novita",
      "accuracy": 0.9061828181818181,
      "accuracy_std_dev": 0.02164377168365568,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.03816088016528926,
      "avg_duration_per_task_ms": 270123.5055096419,
      "avg_output_tokens_per_task": 31994.34618916437,
      "last_run_timestamp": "2026-10-10T16:23:09.690Z"
    },
    {
      "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": 0.9056252185030218,
      "accuracy_std_dev": 0.011533357236316561,
      "run_count": 11,
      "total_questions": 2151,
      "avg_cost_per_task_usd": 0.028380999070199905,
      "avg_duration_per_task_ms": 289429.54486285447,
      "avg_output_tokens_per_task": 23515.902370990236,
      "last_run_timestamp": "2026-10-10T16:20:06.463Z"
    },
    {
      "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": 0.9055708181818182,
      "accuracy_std_dev": 0.027871027429338564,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.042199853076216716,
      "avg_duration_per_task_ms": 291574.9090909091,
      "avg_output_tokens_per_task": 32546.853076216714,
      "last_run_timestamp": "2026-10-10T16:22:42.647Z"
    },
    {
      "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": 0.9055076503284487,
      "accuracy_std_dev": 0.05364913641103649,
      "run_count": 10,
      "total_questions": 1979,
      "avg_cost_per_task_usd": 0.13999162253663464,
      "avg_duration_per_task_ms": 293935.19656392117,
      "avg_output_tokens_per_task": 10089.026275896918,
      "last_run_timestamp": "2026-10-10T08:21:49.772Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "InferenceNet",
      "accuracy": 0.905376,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 155,
      "avg_cost_per_task_usd": 0.09219292258064517,
      "avg_duration_per_task_ms": 892432.9419354838,
      "avg_output_tokens_per_task": 36790.02580645162,
      "last_run_timestamp": "2026-10-02T06:44:44.722Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "auto-routing",
      "accuracy": 0.9050427041719343,
      "accuracy_std_dev": 0.01882757447203436,
      "run_count": 12,
      "total_questions": 2373,
      "avg_cost_per_task_usd": 0.03372518246944795,
      "avg_duration_per_task_ms": 328089.9827222925,
      "avg_output_tokens_per_task": 29343.05309734513,
      "last_run_timestamp": "2026-10-10T16:21:42.544Z"
    },
    {
      "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": 0.9046013333333333,
      "accuracy_std_dev": 0.0028614651647170097,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.03432480808080808,
      "avg_duration_per_task_ms": 135428.62121212122,
      "avg_output_tokens_per_task": 28424.8265993266,
      "last_run_timestamp": "2026-10-10T08:22:10.433Z"
    },
    {
      "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": 0.904377,
      "accuracy_std_dev": 0.009122177678602841,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.033627452525252524,
      "avg_duration_per_task_ms": 247053.7898989899,
      "avg_output_tokens_per_task": 27843.98181818182,
      "last_run_timestamp": "2026-10-10T08:11:33.638Z"
    },
    {
      "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": 0.9038719,
      "accuracy_std_dev": 0.022667207355340432,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.15332785707070706,
      "avg_duration_per_task_ms": 585989.541919192,
      "avg_output_tokens_per_task": 38728.24494949495,
      "last_run_timestamp": "2026-10-10T08:21:19.694Z"
    },
    {
      "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": 0.9036284693670886,
      "accuracy_std_dev": 0.014048473697879075,
      "run_count": 10,
      "total_questions": 1975,
      "avg_cost_per_task_usd": 0.02646089265822785,
      "avg_duration_per_task_ms": 194010.8364556962,
      "avg_output_tokens_per_task": 25873.973164556963,
      "last_run_timestamp": "2026-10-10T08:25:15.295Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "x-ai/grok-4.6-20260810",
      "model_slug": "x-ai/grok-4.6",
      "model_name": "SpaceXAI: Grok 4.6",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.903508802631579,
      "accuracy_std_dev": 0.0984941779678147,
      "run_count": 3,
      "total_questions": 456,
      "avg_cost_per_task_usd": 0.11285829166666667,
      "avg_duration_per_task_ms": 423954.7916666667,
      "avg_output_tokens_per_task": 16846.184210526317,
      "last_run_timestamp": "2026-10-01T16:04:59.818Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "StreamLake",
      "accuracy": 0.9034791666666666,
      "accuracy_std_dev": 0.007995417675491095,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.022680801346801345,
      "avg_duration_per_task_ms": 208477.4335016835,
      "avg_output_tokens_per_task": 29237.281986531987,
      "last_run_timestamp": "2026-10-10T08:29:50.459Z"
    },
    {
      "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": 0.9032354285714285,
      "accuracy_std_dev": 0.022634937934174912,
      "run_count": 9,
      "total_questions": 1638,
      "avg_cost_per_task_usd": 0.1526553308913309,
      "avg_duration_per_task_ms": 567262.6489621489,
      "avg_output_tokens_per_task": 37775.336996337,
      "last_run_timestamp": "2026-10-10T08:21:36.296Z"
    },
    {
      "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": 0.9031220000000001,
      "accuracy_std_dev": 0.04051469049167676,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.14198908172635447,
      "avg_duration_per_task_ms": 209617.58539944905,
      "avg_output_tokens_per_task": 9568.357208448118,
      "last_run_timestamp": "2026-10-10T08:25:16.321Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.4-20260305",
      "model_slug": "openai/gpt-5.4",
      "model_name": "OpenAI: GPT-5.4",
      "provider_name": "auto-routing",
      "accuracy": 0.902918,
      "accuracy_std_dev": 0.0007933738084913016,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.14672305387205387,
      "avg_duration_per_task_ms": 133620.78956228957,
      "avg_output_tokens_per_task": 9652.131313131313,
      "last_run_timestamp": "2026-10-01T19:14:23.097Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Wafer",
      "accuracy": 0.9027134284064666,
      "accuracy_std_dev": 0.015276438772029145,
      "run_count": 9,
      "total_questions": 1732,
      "avg_cost_per_task_usd": 0.014871293879907621,
      "avg_duration_per_task_ms": 274837.21131639724,
      "avg_output_tokens_per_task": 18283.435912240184,
      "last_run_timestamp": "2026-10-10T08:14:58.091Z"
    },
    {
      "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": 0.9024151322662172,
      "accuracy_std_dev": 0.007540900139682681,
      "run_count": 6,
      "total_questions": 1187,
      "avg_cost_per_task_usd": 0.0312832594776748,
      "avg_duration_per_task_ms": 159175.5240101095,
      "avg_output_tokens_per_task": 29122.362257792756,
      "last_run_timestamp": "2026-10-10T08:15:01.919Z"
    },
    {
      "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": 0.9023917412834764,
      "accuracy_std_dev": 0.031521333558242756,
      "run_count": 10,
      "total_questions": 1979,
      "avg_cost_per_task_usd": 0.10391216371905002,
      "avg_duration_per_task_ms": 566571.5058110156,
      "avg_output_tokens_per_task": 39085.76503284487,
      "last_run_timestamp": "2026-10-10T08:36:06.052Z"
    },
    {
      "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": 0.902357,
      "accuracy_std_dev": 0.002749143987983661,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0541827138047138,
      "avg_duration_per_task_ms": 33856.301346801345,
      "avg_output_tokens_per_task": 2498.9175084175085,
      "last_run_timestamp": "2026-10-10T08:22:13.045Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "SiliconFlow",
      "accuracy": 0.9023568333333333,
      "accuracy_std_dev": 0.00927208580303744,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.03514278535353535,
      "avg_duration_per_task_ms": 197697.9335016835,
      "avg_output_tokens_per_task": 29114.67845117845,
      "last_run_timestamp": "2026-10-10T08:27:28.674Z"
    },
    {
      "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": 0.9017953333333333,
      "accuracy_std_dev": 0.0041993868070892585,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.03234963973063973,
      "avg_duration_per_task_ms": 866618.7946127946,
      "avg_output_tokens_per_task": 36837.547138047135,
      "last_run_timestamp": "2026-10-10T08:37:36.958Z"
    },
    {
      "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": 0.900939869848156,
      "accuracy_std_dev": 0.007380306298604699,
      "run_count": 7,
      "total_questions": 1383,
      "avg_cost_per_task_usd": 0.01601901446131598,
      "avg_duration_per_task_ms": 443238.66449746926,
      "avg_output_tokens_per_task": 28948.347071583514,
      "last_run_timestamp": "2026-10-10T08:23:04.424Z"
    },
    {
      "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": 0.90067325,
      "accuracy_std_dev": 0.005350183771376443,
      "run_count": 20,
      "total_questions": 3960,
      "avg_cost_per_task_usd": 0.20474679772727275,
      "avg_duration_per_task_ms": 98144.71237373738,
      "avg_output_tokens_per_task": 7846.206818181819,
      "last_run_timestamp": "2026-10-10T08:22:41.542Z"
    },
    {
      "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": 0.90067325,
      "accuracy_std_dev": 0.0031494064817835163,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.053791821969696965,
      "avg_duration_per_task_ms": 27301.564393939392,
      "avg_output_tokens_per_task": 2474.441919191919,
      "last_run_timestamp": "2026-10-10T08:23:49.959Z"
    },
    {
      "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": 0.900619,
      "accuracy_std_dev": 0.005948071674514441,
      "run_count": 31,
      "total_questions": 6138,
      "avg_cost_per_task_usd": 0.21661038025415447,
      "avg_duration_per_task_ms": 94471.30303030302,
      "avg_output_tokens_per_task": 7896.7298794395565,
      "last_run_timestamp": "2026-10-10T08:22:25.959Z"
    },
    {
      "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": 0.9005056897133222,
      "accuracy_std_dev": 0.017065352280115537,
      "run_count": 9,
      "total_questions": 1779,
      "avg_cost_per_task_usd": 0.1164296958965711,
      "avg_duration_per_task_ms": 597873.8173130973,
      "avg_output_tokens_per_task": 36394.80157391793,
      "last_run_timestamp": "2026-10-10T08:14:22.301Z"
    },
    {
      "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": 0.9003566096256683,
      "accuracy_std_dev": 0.009294020279657108,
      "run_count": 11,
      "total_questions": 1870,
      "avg_cost_per_task_usd": 0.023890174331550804,
      "avg_duration_per_task_ms": 191466.34438502675,
      "avg_output_tokens_per_task": 19764.336898395723,
      "last_run_timestamp": "2026-10-10T16:21:33.796Z"
    },
    {
      "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": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "Tencent",
      "accuracy": 0.8998315,
      "accuracy_std_dev": 0.007575499999999957,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.11587846212121213,
      "avg_duration_per_task_ms": 577782.9368686868,
      "avg_output_tokens_per_task": 46174.61616161616,
      "last_run_timestamp": "2026-10-10T08:11:30.726Z"
    },
    {
      "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": "DeepSeek",
      "accuracy": 0.8996633999999999,
      "accuracy_std_dev": 0.005792805834826492,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.007343192929292929,
      "avg_duration_per_task_ms": 219337.0404040404,
      "avg_output_tokens_per_task": 25809.762626262625,
      "last_run_timestamp": "2026-08-08T09:30:52.456Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.4-20260305",
      "model_slug": "openai/gpt-5.4",
      "model_name": "OpenAI: GPT-5.4",
      "provider_name": "OpenAI",
      "accuracy": 0.899551,
      "accuracy_std_dev": 0.006198077174952426,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.14109027272727273,
      "avg_duration_per_task_ms": 131303,
      "avg_output_tokens_per_task": 9278.422558922559,
      "last_run_timestamp": "2026-10-01T19:14:55.028Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-luna-pro-20260709",
      "model_slug": "openai/gpt-5.6-luna-pro",
      "model_name": "OpenAI: GPT-5.6 Luna Pro",
      "provider_name": "auto-routing",
      "accuracy": 0.899551,
      "accuracy_std_dev": 0.006198077174952426,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0412280202020202,
      "avg_duration_per_task_ms": 141835.40572390571,
      "avg_output_tokens_per_task": 27211.210437710437,
      "last_run_timestamp": "2026-10-01T16:12:34.647Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.7-opus-20260416",
      "model_slug": "anthropic/claude-opus-4.7",
      "model_name": "Anthropic: Claude Opus 4.7",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.8995507777777776,
      "accuracy_std_dev": 0.004558884848483037,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.2283120224466891,
      "avg_duration_per_task_ms": 110024.31481481482,
      "avg_output_tokens_per_task": 8923.117845117846,
      "last_run_timestamp": "2026-10-10T08:24:16.114Z"
    },
    {
      "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": 0.8994632500000004,
      "accuracy_std_dev": 0.0063860840847501914,
      "run_count": 32,
      "total_questions": 6336,
      "avg_cost_per_task_usd": 0.22312757212752524,
      "avg_duration_per_task_ms": 94690.671875,
      "avg_output_tokens_per_task": 8125.733585858586,
      "last_run_timestamp": "2026-10-10T08:26:06.025Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "Novita",
      "accuracy": 0.8994372877637131,
      "accuracy_std_dev": 0.01977995405637288,
      "run_count": 6,
      "total_questions": 1185,
      "avg_cost_per_task_usd": 0.03544181097046414,
      "avg_duration_per_task_ms": 152645.21687763714,
      "avg_output_tokens_per_task": 26597.288607594935,
      "last_run_timestamp": "2026-10-03T09:28:39.969Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Baidu",
      "accuracy": 0.8994013674033149,
      "accuracy_std_dev": 0.0057226206055478665,
      "run_count": 11,
      "total_questions": 2172,
      "avg_cost_per_task_usd": 0.01418124585635359,
      "avg_duration_per_task_ms": 183996.83977900553,
      "avg_output_tokens_per_task": 26261.59944751381,
      "last_run_timestamp": "2026-10-10T08:22:32.411Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "NextBit",
      "accuracy": 0.8993638888888887,
      "accuracy_std_dev": 0.004029367042778795,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.030411617283950615,
      "avg_duration_per_task_ms": 237553.6509539843,
      "avg_output_tokens_per_task": 26633.1290684624,
      "last_run_timestamp": "2026-10-03T10:19:34.320Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-fable-5.1-20260831",
      "model_slug": "anthropic/claude-fable-5.1",
      "model_name": "Anthropic: Claude Fable 5.1",
      "provider_name": "auto-routing",
      "accuracy": 0.8993264,
      "accuracy_std_dev": 0.006511664598242117,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.13052546969696968,
      "avg_duration_per_task_ms": 30874.333333333332,
      "avg_output_tokens_per_task": 2403.130303030303,
      "last_run_timestamp": "2026-10-01T08:24:30.121Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-opus-5.5-20260921",
      "model_slug": "anthropic/claude-opus-5.5",
      "model_name": "Anthropic: Claude Opus 5.5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.8992703333333333,
      "accuracy_std_dev": 0.004703538142245193,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.05869316456228956,
      "avg_duration_per_task_ms": 43102.90572390572,
      "avg_output_tokens_per_task": 2536.3127104377104,
      "last_run_timestamp": "2026-10-10T08:37:23.710Z"
    },
    {
      "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": 0.898999033256351,
      "accuracy_std_dev": 0.018913767762537594,
      "run_count": 11,
      "total_questions": 2165,
      "avg_cost_per_task_usd": 0.1172629778290993,
      "avg_duration_per_task_ms": 375593.448960739,
      "avg_output_tokens_per_task": 8022.391224018475,
      "last_run_timestamp": "2026-10-10T08:28:02.228Z"
    },
    {
      "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": "anthropic/claude-opus-5.5-20260921",
      "model_slug": "anthropic/claude-opus-5.5",
      "model_name": "Anthropic: Claude Opus 5.5",
      "provider_name": "Azure",
      "accuracy": 0.898989625,
      "accuracy_std_dev": 0.00564651885982635,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.057617888257575745,
      "avg_duration_per_task_ms": 25880.185606060608,
      "avg_output_tokens_per_task": 2525.5625,
      "last_run_timestamp": "2026-10-10T08:22:11.579Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-luna",
      "model_slug": "openai/gpt-6-luna",
      "model_name": "GPT-6-Luna",
      "provider_name": "auto-routing",
      "accuracy": 0.898477,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 197,
      "avg_cost_per_task_usd": 0.003442197969543147,
      "avg_duration_per_task_ms": 73312.30456852792,
      "avg_output_tokens_per_task": 6831.9390862944165,
      "last_run_timestamp": "2026-10-08T13:29:57.437Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "Xiaomi",
      "accuracy": 0.8984286666666665,
      "accuracy_std_dev": 0.002099515235053612,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.03384256734006734,
      "avg_duration_per_task_ms": 896786.5488215488,
      "avg_output_tokens_per_task": 38532.71885521885,
      "last_run_timestamp": "2026-10-10T08:14:49.069Z"
    },
    {
      "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": 0.8982883333333332,
      "accuracy_std_dev": 0.003854920542487059,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.0574033244949495,
      "avg_duration_per_task_ms": 25094.159511784514,
      "avg_output_tokens_per_task": 2482.843434343434,
      "last_run_timestamp": "2026-10-10T08:25:06.801Z"
    },
    {
      "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": "Wafer",
      "accuracy": 0.8982242269380685,
      "accuracy_std_dev": 0.025718218408254386,
      "run_count": 13,
      "total_questions": 2309,
      "avg_cost_per_task_usd": 0.2508324629709831,
      "avg_duration_per_task_ms": 414948.73408401903,
      "avg_output_tokens_per_task": 46979.851017756606,
      "last_run_timestamp": "2026-10-10T08:40:32.331Z"
    },
    {
      "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": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "ModelRun",
      "accuracy": 0.8978673333333335,
      "accuracy_std_dev": 0.021639482631420633,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.08683191161616162,
      "avg_duration_per_task_ms": 208079.4877946128,
      "avg_output_tokens_per_task": 29133.21127946128,
      "last_run_timestamp": "2026-10-10T16:33:12.083Z"
    },
    {
      "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": 0.8978673333333335,
      "accuracy_std_dev": 0.0103171593420325,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.12474947474747476,
      "avg_duration_per_task_ms": 901534.3451178451,
      "avg_output_tokens_per_task": 49639.851851851854,
      "last_run_timestamp": "2026-10-10T08:29:32.160Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "auto-routing",
      "accuracy": 0.897306,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.12030833838383839,
      "avg_duration_per_task_ms": 846579.9494949495,
      "avg_output_tokens_per_task": 47921.80303030303,
      "last_run_timestamp": "2026-10-10T08:14:41.263Z"
    },
    {
      "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": 0.8971503581183612,
      "accuracy_std_dev": 0.02073357692150583,
      "run_count": 10,
      "total_questions": 1977,
      "avg_cost_per_task_usd": 0.12949166211431462,
      "avg_duration_per_task_ms": 522272.583712696,
      "avg_output_tokens_per_task": 38320.582701062216,
      "last_run_timestamp": "2026-10-10T08:21:10.337Z"
    },
    {
      "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": 0.8969694999999998,
      "accuracy_std_dev": 0.01185165437607762,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.2005278484848485,
      "avg_duration_per_task_ms": 96563.55505050505,
      "avg_output_tokens_per_task": 7804.548484848485,
      "last_run_timestamp": "2026-10-10T08:14:40.993Z"
    },
    {
      "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": 0.8969322222222222,
      "accuracy_std_dev": 0.0084278559785429,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.1972121043771044,
      "avg_duration_per_task_ms": 95091.35016835017,
      "avg_output_tokens_per_task": 7682.2547699214365,
      "last_run_timestamp": "2026-10-10T08:22:27.812Z"
    },
    {
      "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": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "DigitalOcean",
      "accuracy": 0.8967648379746836,
      "accuracy_std_dev": 0.007137443646401371,
      "run_count": 6,
      "total_questions": 1185,
      "avg_cost_per_task_usd": 0.018252328270042195,
      "avg_duration_per_task_ms": 503009.7375527426,
      "avg_output_tokens_per_task": 19743.828691983123,
      "last_run_timestamp": "2026-10-10T08:22:09.106Z"
    },
    {
      "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": 0.8967118381024095,
      "accuracy_std_dev": 0.027051445539009556,
      "run_count": 8,
      "total_questions": 1328,
      "avg_cost_per_task_usd": 0.026119229668674696,
      "avg_duration_per_task_ms": 760619.1935240964,
      "avg_output_tokens_per_task": 23698.94503012048,
      "last_run_timestamp": "2026-10-10T16:24:01.594Z"
    },
    {
      "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": 0.8963888977035491,
      "accuracy_std_dev": 0.007117042546266783,
      "run_count": 22,
      "total_questions": 4311,
      "avg_cost_per_task_usd": 0.12574247158431917,
      "avg_duration_per_task_ms": 57691.72442588727,
      "avg_output_tokens_per_task": 4730.365576432382,
      "last_run_timestamp": "2026-10-10T08:28:02.111Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-flash-20260826",
      "model_slug": "qwen/qwen3.8-flash",
      "model_name": "Qwen: Qwen3.8 Flash",
      "provider_name": "Alibaba",
      "accuracy": 0.8962022556962025,
      "accuracy_std_dev": 0.0022689999999999655,
      "run_count": 2,
      "total_questions": 395,
      "avg_cost_per_task_usd": 0.020123096202531644,
      "avg_duration_per_task_ms": 653583.2962025317,
      "avg_output_tokens_per_task": 42659.516455696205,
      "last_run_timestamp": "2026-10-10T00:00:04.637Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Novita",
      "accuracy": 0.8957578329718004,
      "accuracy_std_dev": 0.016809562938053055,
      "run_count": 14,
      "total_questions": 2766,
      "avg_cost_per_task_usd": 0.023020330802603038,
      "avg_duration_per_task_ms": 236991.16630513375,
      "avg_output_tokens_per_task": 26771.403470715835,
      "last_run_timestamp": "2026-10-10T08:26:17.965Z"
    },
    {
      "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": 0.8956230000000002,
      "accuracy_std_dev": 0.007273551310513056,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.00771163468013468,
      "avg_duration_per_task_ms": 254636.3367003367,
      "avg_output_tokens_per_task": 27268.55387205387,
      "last_run_timestamp": "2026-08-08T10:06:09.471Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-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": "anthropic/claude-opus-5-20260723",
      "model_slug": "anthropic/claude-opus-5",
      "model_name": "Anthropic: Claude Opus 5",
      "provider_name": "Anthropic",
      "accuracy": 0.8950104545454546,
      "accuracy_std_dev": 0.007224079964870452,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.12169120569329661,
      "avg_duration_per_task_ms": 58147.77502295684,
      "avg_output_tokens_per_task": 4649.89898989899,
      "last_run_timestamp": "2026-10-10T08:30:55.092Z"
    },
    {
      "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": 0.8949752307692306,
      "accuracy_std_dev": 0.005999147723492375,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.20542784071484071,
      "avg_duration_per_task_ms": 100020.3480963481,
      "avg_output_tokens_per_task": 8004.257187257187,
      "last_run_timestamp": "2026-10-10T08:24:37.404Z"
    },
    {
      "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": 0.894781,
      "accuracy_std_dev": 0.004835703206359931,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.054542488636363635,
      "avg_duration_per_task_ms": 32603.121212121212,
      "avg_output_tokens_per_task": 2523.126262626263,
      "last_run_timestamp": "2026-10-10T08:22:14.032Z"
    },
    {
      "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": 0.894653484848485,
      "accuracy_std_dev": 0.00729743750841102,
      "run_count": 33,
      "total_questions": 6534,
      "avg_cost_per_task_usd": 0.20388223569023572,
      "avg_duration_per_task_ms": 70408.00857055403,
      "avg_output_tokens_per_task": 7423.640189776554,
      "last_run_timestamp": "2026-10-10T08:29:34.477Z"
    },
    {
      "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": 0.894286846153846,
      "accuracy_std_dev": 0.010109118939518566,
      "run_count": 15,
      "total_questions": 2964,
      "avg_cost_per_task_usd": 0.013039793522267208,
      "avg_duration_per_task_ms": 301749.21187584347,
      "avg_output_tokens_per_task": 27435.895074224023,
      "last_run_timestamp": "2026-10-10T08:29:49.095Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-terra-20260709",
      "model_slug": "openai/gpt-5.6-terra",
      "model_name": "OpenAI: GPT-5.6 Terra",
      "provider_name": "Azure",
      "accuracy": 0.8941264444444444,
      "accuracy_std_dev": 0.009636502052221474,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.05178651795735129,
      "avg_duration_per_task_ms": 45445.343434343435,
      "avg_output_tokens_per_task": 3671.5897867564536,
      "last_run_timestamp": "2026-10-01T19:13:06.514Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.7-opus-20260416",
      "model_slug": "anthropic/claude-opus-4.7",
      "model_name": "Anthropic: Claude Opus 4.7",
      "provider_name": "auto-routing",
      "accuracy": 0.8939462206896551,
      "accuracy_std_dev": 0.00852872826887441,
      "run_count": 11,
      "total_questions": 2175,
      "avg_cost_per_task_usd": 0.22604011034482757,
      "avg_duration_per_task_ms": 108518.86942528735,
      "avg_output_tokens_per_task": 8818.531954022988,
      "last_run_timestamp": "2026-10-10T08:14:40.266Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "GMICloud",
      "accuracy": 0.8936586666666667,
      "accuracy_std_dev": 0.02110358604334552,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.026894239057239058,
      "avg_duration_per_task_ms": 149425.4292929293,
      "avg_output_tokens_per_task": 27645.62373737374,
      "last_run_timestamp": "2026-10-10T08:22:18.155Z"
    },
    {
      "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": 0.8933884681122448,
      "accuracy_std_dev": 0.01364679809916505,
      "run_count": 16,
      "total_questions": 3136,
      "avg_cost_per_task_usd": 0.0038534390943877547,
      "avg_duration_per_task_ms": 324360.03539540817,
      "avg_output_tokens_per_task": 21118.27838010204,
      "last_run_timestamp": "2026-10-10T08:21:47.369Z"
    },
    {
      "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": 0.8933779999999999,
      "accuracy_std_dev": 0.008285503967774051,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.11363387037037037,
      "avg_duration_per_task_ms": 879366.8148148148,
      "avg_output_tokens_per_task": 45257.41750841751,
      "last_run_timestamp": "2026-10-10T08:22:57.323Z"
    },
    {
      "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-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Parasail",
      "accuracy": 0.8913205555555556,
      "accuracy_std_dev": 0.01785381330641292,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.14202826936026933,
      "avg_duration_per_task_ms": 616787.0375982042,
      "avg_output_tokens_per_task": 35812.23737373737,
      "last_run_timestamp": "2026-10-10T08:21:47.253Z"
    },
    {
      "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": 0.8912990824295012,
      "accuracy_std_dev": 0.011335972984040578,
      "run_count": 14,
      "total_questions": 2766,
      "avg_cost_per_task_usd": 0.20838120318148953,
      "avg_duration_per_task_ms": 100499.53289949386,
      "avg_output_tokens_per_task": 8107.096167751265,
      "last_run_timestamp": "2026-10-10T08:25:06.399Z"
    },
    {
      "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": 0.8910347038626609,
      "accuracy_std_dev": 0.006181130251620641,
      "run_count": 4,
      "total_questions": 699,
      "avg_cost_per_task_usd": 0.01261281974248927,
      "avg_duration_per_task_ms": 330150.28755364806,
      "avg_output_tokens_per_task": 23768.37339055794,
      "last_run_timestamp": "2026-10-10T08:33:55.559Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "GMICloud",
      "accuracy": 0.8910211333333333,
      "accuracy_std_dev": 0.013724636800861273,
      "run_count": 15,
      "total_questions": 2970,
      "avg_cost_per_task_usd": 0.01418473804713805,
      "avg_duration_per_task_ms": 211329.1047138047,
      "avg_output_tokens_per_task": 24782.838047138048,
      "last_run_timestamp": "2026-10-10T08:22:09.792Z"
    },
    {
      "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": 0.891002888888889,
      "accuracy_std_dev": 0.0356098077018116,
      "run_count": 10,
      "total_questions": 1971,
      "avg_cost_per_task_usd": 0.14153233739218668,
      "avg_duration_per_task_ms": 796678.4495180112,
      "avg_output_tokens_per_task": 35436.016235413495,
      "last_run_timestamp": "2026-10-10T08:24:22.693Z"
    },
    {
      "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": 0.8907346275957544,
      "accuracy_std_dev": 0.009380068570120645,
      "run_count": 33,
      "total_questions": 6501,
      "avg_cost_per_task_usd": 0.13349859406245196,
      "avg_duration_per_task_ms": 59227.62467312721,
      "avg_output_tokens_per_task": 4775.171665897554,
      "last_run_timestamp": "2026-10-10T08:22:14.249Z"
    },
    {
      "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": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "DeepInfra",
      "accuracy": 0.8902381885714286,
      "accuracy_std_dev": 0.0030799415984722714,
      "run_count": 4,
      "total_questions": 700,
      "avg_cost_per_task_usd": 0.03244278,
      "avg_duration_per_task_ms": 834708.2842857143,
      "avg_output_tokens_per_task": 37089.54,
      "last_run_timestamp": "2026-10-10T08:11:33.657Z"
    },
    {
      "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": 0.8901512500000001,
      "accuracy_std_dev": 0.006979218773437337,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.008867842171717173,
      "avg_duration_per_task_ms": 175715.7159090909,
      "avg_output_tokens_per_task": 31193.151515151516,
      "last_run_timestamp": "2026-10-10T08:26:16.926Z"
    },
    {
      "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": 0.890118616572311,
      "accuracy_std_dev": 0.05543710090986,
      "run_count": 10,
      "total_questions": 1943,
      "avg_cost_per_task_usd": 0.011268564590838909,
      "avg_duration_per_task_ms": 493758.26093669585,
      "avg_output_tokens_per_task": 21520.772516726713,
      "last_run_timestamp": "2026-10-10T08:48:02.661Z"
    },
    {
      "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": 0.8900365454545454,
      "accuracy_std_dev": 0.008514741728465947,
      "run_count": 22,
      "total_questions": 4356,
      "avg_cost_per_task_usd": 0.24217010583103765,
      "avg_duration_per_task_ms": 85203.68801652892,
      "avg_output_tokens_per_task": 9007.777777777777,
      "last_run_timestamp": "2026-10-10T08:24:21.894Z"
    },
    {
      "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": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8898569891067537,
      "accuracy_std_dev": 0.020559295245207023,
      "run_count": 7,
      "total_questions": 1377,
      "avg_cost_per_task_usd": 0.19048945533769066,
      "avg_duration_per_task_ms": 1402381.5344952797,
      "avg_output_tokens_per_task": 55920.59404502542,
      "last_run_timestamp": "2026-10-10T16:21:04.066Z"
    },
    {
      "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": "Azure",
      "accuracy": 0.8893492596685083,
      "accuracy_std_dev": 0.01020692558839056,
      "run_count": 11,
      "total_questions": 2172,
      "avg_cost_per_task_usd": 0.20542008609576426,
      "avg_duration_per_task_ms": 100076.91850828729,
      "avg_output_tokens_per_task": 7990.411602209945,
      "last_run_timestamp": "2026-10-10T08:21:58.851Z"
    },
    {
      "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": 0.8888376274149035,
      "accuracy_std_dev": 0.008182609561727207,
      "run_count": 11,
      "total_questions": 2174,
      "avg_cost_per_task_usd": 0.015841472401103954,
      "avg_duration_per_task_ms": 163351.5689972401,
      "avg_output_tokens_per_task": 26304.647194112236,
      "last_run_timestamp": "2026-10-10T08:22:23.779Z"
    },
    {
      "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": 0.8884680000000001,
      "accuracy_std_dev": 0.008615114711946651,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.02470814393939394,
      "avg_duration_per_task_ms": 276879.46338383836,
      "avg_output_tokens_per_task": 29963.959595959597,
      "last_run_timestamp": "2026-10-10T08:22:50.656Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Sail Research",
      "accuracy": 0.8878433783257624,
      "accuracy_std_dev": 0.020080093052184153,
      "run_count": 8,
      "total_questions": 1541,
      "avg_cost_per_task_usd": 0.155311282284231,
      "avg_duration_per_task_ms": 605967.6262167423,
      "avg_output_tokens_per_task": 42247.91953277093,
      "last_run_timestamp": "2026-10-10T08:22:51.630Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-terra-20260709",
      "model_slug": "openai/gpt-5.6-terra",
      "model_name": "OpenAI: GPT-5.6 Terra",
      "provider_name": "auto-routing",
      "accuracy": 0.8877663333333334,
      "accuracy_std_dev": 0.005555502277522277,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.03440870875420875,
      "avg_duration_per_task_ms": 50195.98148148148,
      "avg_output_tokens_per_task": 3309.2491582491584,
      "last_run_timestamp": "2026-10-01T19:13:07.806Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "NextBit",
      "accuracy": 0.887537924012158,
      "accuracy_std_dev": 0.020250680116973837,
      "run_count": 5,
      "total_questions": 987,
      "avg_cost_per_task_usd": 0.12033997061803446,
      "avg_duration_per_task_ms": 580755.4113475177,
      "avg_output_tokens_per_task": 34357.5663627153,
      "last_run_timestamp": "2026-10-03T16:44:56.199Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-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": 0.8873838777089784,
      "accuracy_std_dev": 0.02655499827504712,
      "run_count": 7,
      "total_questions": 1292,
      "avg_cost_per_task_usd": 0.009403260061919503,
      "avg_duration_per_task_ms": 395325.9891640867,
      "avg_output_tokens_per_task": 19091.013157894737,
      "last_run_timestamp": "2026-10-10T08:23:40.493Z"
    },
    {
      "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": 0.8872052500000001,
      "accuracy_std_dev": 0.0031496737430248356,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.012731213383838385,
      "avg_duration_per_task_ms": 417944.53661616164,
      "avg_output_tokens_per_task": 23604.305555555555,
      "last_run_timestamp": "2026-10-10T08:23:04.653Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-fable-5.1-20260831",
      "model_slug": "anthropic/claude-fable-5.1",
      "model_name": "Anthropic: Claude Fable 5.1",
      "provider_name": "Azure",
      "accuracy": 0.8872051666666666,
      "accuracy_std_dev": 0.024761493125527695,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.1434989882154882,
      "avg_duration_per_task_ms": 42121.643939393936,
      "avg_output_tokens_per_task": 2660.465488215488,
      "last_run_timestamp": "2026-10-01T08:24:34.535Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Wafer",
      "accuracy": 0.887205,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.007206747474747474,
      "avg_duration_per_task_ms": 979612.8939393939,
      "avg_output_tokens_per_task": 42001.40909090909,
      "last_run_timestamp": "2026-10-10T16:19:50.069Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8868686,
      "accuracy_std_dev": 0.01990232228258803,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.029374592929292932,
      "avg_duration_per_task_ms": 212707.40101010102,
      "avg_output_tokens_per_task": 27438.87777777778,
      "last_run_timestamp": "2026-10-10T08:11:30.591Z"
    },
    {
      "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": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Inceptron",
      "accuracy": 0.8865611646002564,
      "accuracy_std_dev": 0.008235396939431653,
      "run_count": 12,
      "total_questions": 2339,
      "avg_cost_per_task_usd": 0.09613296280461736,
      "avg_duration_per_task_ms": 530353.1603249252,
      "avg_output_tokens_per_task": 29119.90594271056,
      "last_run_timestamp": "2026-10-10T08:48:04.149Z"
    },
    {
      "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": 0.8860826666666668,
      "accuracy_std_dev": 0.009654676391377534,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.006178414141414142,
      "avg_duration_per_task_ms": 550604.5437710438,
      "avg_output_tokens_per_task": 33932.367003367006,
      "last_run_timestamp": "2026-10-10T08:22:13.710Z"
    },
    {
      "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": 0.8855269617834396,
      "accuracy_std_dev": 0.01525692535211471,
      "run_count": 5,
      "total_questions": 942,
      "avg_cost_per_task_usd": 0.03748663375796179,
      "avg_duration_per_task_ms": 257130.8949044586,
      "avg_output_tokens_per_task": 34302.59554140127,
      "last_run_timestamp": "2026-10-10T08:22:50.466Z"
    },
    {
      "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": "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": 0.8853288347386172,
      "accuracy_std_dev": 0.0037367403204159573,
      "run_count": 3,
      "total_questions": 593,
      "avg_cost_per_task_usd": 0.032569026981450255,
      "avg_duration_per_task_ms": 806245.0269814503,
      "avg_output_tokens_per_task": 37423.73524451939,
      "last_run_timestamp": "2026-10-10T08:21:50.114Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-luna-20260709",
      "model_slug": "openai/gpt-5.6-luna",
      "model_name": "OpenAI: GPT-5.6 Luna",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.8853190481927712,
      "accuracy_std_dev": 0.005795847495189989,
      "run_count": 4,
      "total_questions": 747,
      "avg_cost_per_task_usd": 0.0097919718875502,
      "avg_duration_per_task_ms": 50542.31994645247,
      "avg_output_tokens_per_task": 7272.5394912985275,
      "last_run_timestamp": "2026-10-01T19:09:12.892Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "DekaLLM",
      "accuracy": 0.8850689656218401,
      "accuracy_std_dev": 0.014411932672615443,
      "run_count": 5,
      "total_questions": 989,
      "avg_cost_per_task_usd": 0.026149528816986855,
      "avg_duration_per_task_ms": 355797.47320525785,
      "avg_output_tokens_per_task": 26995.41152679474,
      "last_run_timestamp": "2026-10-10T08:22:45.418Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "GMICloud",
      "accuracy": 0.8850408571428571,
      "accuracy_std_dev": 0.015130548440900905,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.00522727344877345,
      "avg_duration_per_task_ms": 245222.45887445888,
      "avg_output_tokens_per_task": 28404.085137085138,
      "last_run_timestamp": "2026-10-10T16:19:44.342Z"
    },
    {
      "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": 0.8848482000000001,
      "accuracy_std_dev": 0.003626295707743655,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.029178751515151518,
      "avg_duration_per_task_ms": 89030.6888888889,
      "avg_output_tokens_per_task": 24171.247474747473,
      "last_run_timestamp": "2026-10-10T08:26:07.846Z"
    },
    {
      "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": 0.8848054784010315,
      "accuracy_std_dev": 0.03136720764189888,
      "run_count": 8,
      "total_questions": 1551,
      "avg_cost_per_task_usd": 0.009929504835589941,
      "avg_duration_per_task_ms": 329283.4983881367,
      "avg_output_tokens_per_task": 25727.783365570598,
      "last_run_timestamp": "2026-10-10T08:21:36.060Z"
    },
    {
      "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": 0.8846618235507244,
      "accuracy_std_dev": 0.01231654174229908,
      "run_count": 14,
      "total_questions": 2760,
      "avg_cost_per_task_usd": 0.007433793115942029,
      "avg_duration_per_task_ms": 452053.6221014493,
      "avg_output_tokens_per_task": 26553.594202898552,
      "last_run_timestamp": "2026-10-10T08:23:40.407Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Parasail",
      "accuracy": 0.8845116000000001,
      "accuracy_std_dev": 0.005694030895595855,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.028268917171717173,
      "avg_duration_per_task_ms": 139965.10808080807,
      "avg_output_tokens_per_task": 23399.145454545454,
      "last_run_timestamp": "2026-10-10T08:22:26.722Z"
    },
    {
      "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": 0.8842973636363637,
      "accuracy_std_dev": 0.009633322699432706,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.02382341873278237,
      "avg_duration_per_task_ms": 351638.5702479339,
      "avg_output_tokens_per_task": 26165.873278236915,
      "last_run_timestamp": "2026-10-10T08:28:12.999Z"
    },
    {
      "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": "z-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": 0.883922059602649,
      "accuracy_std_dev": 0.013047095505854499,
      "run_count": 6,
      "total_questions": 906,
      "avg_cost_per_task_usd": 0.007205040838852097,
      "avg_duration_per_task_ms": 953103.2119205298,
      "avg_output_tokens_per_task": 28596.360927152316,
      "last_run_timestamp": "2026-10-10T08:21:47.633Z"
    },
    {
      "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": 0.8838929412284579,
      "accuracy_std_dev": 0.013075471353610644,
      "run_count": 23,
      "total_questions": 4526,
      "avg_cost_per_task_usd": 0.19182242841361027,
      "avg_duration_per_task_ms": 375443.09456473705,
      "avg_output_tokens_per_task": 31243.90676093681,
      "last_run_timestamp": "2026-10-10T08:22:13.945Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3-flash-preview-20251217",
      "model_slug": "google/gemini-3-flash-preview",
      "model_name": "Google: Gemini 3 Flash Preview",
      "provider_name": "auto-routing",
      "accuracy": 0.8838383333333335,
      "accuracy_std_dev": 0.001374572014200141,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.13180945791245793,
      "avg_duration_per_task_ms": 248970.3754208754,
      "avg_output_tokens_per_task": 43817.097643097644,
      "last_run_timestamp": "2026-10-01T19:15:31.426Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Ionstream",
      "accuracy": 0.8838382499999999,
      "accuracy_std_dev": 0.01653796711169484,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.08040183459595959,
      "avg_duration_per_task_ms": 318120.2462121212,
      "avg_output_tokens_per_task": 28808.45580808081,
      "last_run_timestamp": "2026-10-10T08:22:11.610Z"
    },
    {
      "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": 0.8838381818181819,
      "accuracy_std_dev": 0.019315258494108086,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.025690312672176308,
      "avg_duration_per_task_ms": 215061.82185491276,
      "avg_output_tokens_per_task": 25361.899449035813,
      "last_run_timestamp": "2026-10-10T08:21:47.719Z"
    },
    {
      "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": 0.8838381111111112,
      "accuracy_std_dev": 0.013444611638078868,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.18439962401795731,
      "avg_duration_per_task_ms": 304152.4012345679,
      "avg_output_tokens_per_task": 46284.17845117845,
      "last_run_timestamp": "2026-10-10T08:14:46.816Z"
    },
    {
      "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": "Novita",
      "accuracy": 0.8836850909090909,
      "accuracy_std_dev": 0.01588134704301956,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.11592730165289256,
      "avg_duration_per_task_ms": 601761.9191919192,
      "avg_output_tokens_per_task": 30287.78328741965,
      "last_run_timestamp": "2026-10-10T08:14:58.268Z"
    },
    {
      "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": 0.883660005882353,
      "accuracy_std_dev": 0.0111840129195372,
      "run_count": 8,
      "total_questions": 1530,
      "avg_cost_per_task_usd": 0.06956762679738562,
      "avg_duration_per_task_ms": 311666.74901960784,
      "avg_output_tokens_per_task": 19224.162745098038,
      "last_run_timestamp": "2026-10-10T16:19:40.670Z"
    },
    {
      "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": 0.8835341342512908,
      "accuracy_std_dev": 0.0025738249530394865,
      "run_count": 3,
      "total_questions": 581,
      "avg_cost_per_task_usd": 0.12498561962134253,
      "avg_duration_per_task_ms": 454716.72461273667,
      "avg_output_tokens_per_task": 36016.14457831325,
      "last_run_timestamp": "2026-10-10T16:20:08.973Z"
    },
    {
      "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": 0.8834197726683938,
      "accuracy_std_dev": 0.04493897289102186,
      "run_count": 8,
      "total_questions": 1544,
      "avg_cost_per_task_usd": 0.05442232836787564,
      "avg_duration_per_task_ms": 582124.3406735752,
      "avg_output_tokens_per_task": 18408.36658031088,
      "last_run_timestamp": "2026-10-10T08:21:48.903Z"
    },
    {
      "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": 0.8833552031872511,
      "accuracy_std_dev": 0.02275586572183091,
      "run_count": 4,
      "total_questions": 753,
      "avg_cost_per_task_usd": 0.015250491367861886,
      "avg_duration_per_task_ms": 298505.31341301464,
      "avg_output_tokens_per_task": 26478.50996015936,
      "last_run_timestamp": "2026-10-10T08:14:42.775Z"
    },
    {
      "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": 0.8828329620853078,
      "accuracy_std_dev": 0.07234827378689748,
      "run_count": 14,
      "total_questions": 2532,
      "avg_cost_per_task_usd": 0.009765114533965244,
      "avg_duration_per_task_ms": 152136.68799368088,
      "avg_output_tokens_per_task": 23748.714060031594,
      "last_run_timestamp": "2026-10-10T08:25:04.507Z"
    },
    {
      "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": 0.8824462686258934,
      "accuracy_std_dev": 0.010300883361604722,
      "run_count": 33,
      "total_questions": 6295,
      "avg_cost_per_task_usd": 0.13718419205718824,
      "avg_duration_per_task_ms": 62608.661318506754,
      "avg_output_tokens_per_task": 4975.615250198571,
      "last_run_timestamp": "2026-10-10T08:22:10.164Z"
    },
    {
      "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": 0.8823811431670283,
      "accuracy_std_dev": 0.01015848669631376,
      "run_count": 7,
      "total_questions": 1383,
      "avg_cost_per_task_usd": 0.00803116413593637,
      "avg_duration_per_task_ms": 247598.04772234274,
      "avg_output_tokens_per_task": 28350.216919739698,
      "last_run_timestamp": "2026-10-03T16:57:09.267Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "StreamLake",
      "accuracy": 0.8820143333333333,
      "accuracy_std_dev": 0.012187637016893623,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.00562553324915825,
      "avg_duration_per_task_ms": 391236.59764309763,
      "avg_output_tokens_per_task": 29487.287037037036,
      "last_run_timestamp": "2026-10-10T08:22:27.077Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-luna-20260922",
      "model_slug": "openai/gpt-6-luna",
      "model_name": "OpenAI: GPT-6 Luna",
      "provider_name": "OpenAI",
      "accuracy": 0.8815936666666665,
      "accuracy_std_dev": 0.00975199857579064,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.002981575757575758,
      "avg_duration_per_task_ms": 52993.44276094276,
      "avg_output_tokens_per_task": 5804.632996632997,
      "last_run_timestamp": "2026-10-02T23:20:31.396Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "DeepInfra",
      "accuracy": 0.8811927142857143,
      "accuracy_std_dev": 0.007888829783113527,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.015891973304473304,
      "avg_duration_per_task_ms": 279407.10173160175,
      "avg_output_tokens_per_task": 24276.554834054834,
      "last_run_timestamp": "2026-10-08T22:18:51.642Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Novita",
      "accuracy": 0.8811927142857143,
      "accuracy_std_dev": 0.02691867172330274,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.09661549711399711,
      "avg_duration_per_task_ms": 473415.0324675325,
      "avg_output_tokens_per_task": 35715.57359307359,
      "last_run_timestamp": "2026-10-10T16:26:22.667Z"
    },
    {
      "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": 0.8811446,
      "accuracy_std_dev": 0.0062631755715451785,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.028506177777777777,
      "avg_duration_per_task_ms": 118218.53838383839,
      "avg_output_tokens_per_task": 23561.454545454544,
      "last_run_timestamp": "2026-10-10T08:30:31.475Z"
    },
    {
      "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": 0.8810132745849297,
      "accuracy_std_dev": 0.029062612893329138,
      "run_count": 4,
      "total_questions": 783,
      "avg_cost_per_task_usd": 0.03828848659003832,
      "avg_duration_per_task_ms": 742925.4495530013,
      "avg_output_tokens_per_task": 25587.997445721583,
      "last_run_timestamp": "2026-08-22T10:03:34.994Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2-thinking-20251106",
      "model_slug": "moonshotai/kimi-k2-thinking",
      "model_name": "MoonshotAI: Kimi K2 Thinking",
      "provider_name": "Novita",
      "accuracy": 0.8808888960000001,
      "accuracy_std_dev": 0.009359000000000006,
      "run_count": 2,
      "total_questions": 375,
      "avg_cost_per_task_usd": 0.10911519200000001,
      "avg_duration_per_task_ms": 1097247.664,
      "avg_output_tokens_per_task": 43639.90933333334,
      "last_run_timestamp": "2026-08-01T17:46:48.312Z"
    },
    {
      "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": 0.8808079825757577,
      "accuracy_std_dev": 0.02428919623862362,
      "run_count": 7,
      "total_questions": 1320,
      "avg_cost_per_task_usd": 0.10256016742424243,
      "avg_duration_per_task_ms": 656305.7287878788,
      "avg_output_tokens_per_task": 29271.418181818182,
      "last_run_timestamp": "2026-10-10T08:22:26.435Z"
    },
    {
      "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": 0.8805554416666667,
      "accuracy_std_dev": 0.010188565454298696,
      "run_count": 11,
      "total_questions": 2160,
      "avg_cost_per_task_usd": 0.005995147685185185,
      "avg_duration_per_task_ms": 391554.91064814816,
      "avg_output_tokens_per_task": 18833.464351851853,
      "last_run_timestamp": "2026-10-10T08:14:58.180Z"
    },
    {
      "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": 0.880471,
      "accuracy_std_dev": 0.004761657064510223,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.042269898989898996,
      "avg_duration_per_task_ms": 255330.39562289562,
      "avg_output_tokens_per_task": 37185.93771043771,
      "last_run_timestamp": "2026-10-10T08:21:41.951Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-luna-20260922",
      "model_slug": "openai/gpt-6-luna",
      "model_name": "OpenAI: GPT-6 Luna",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.880471,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0032475000000000004,
      "avg_duration_per_task_ms": 58949.292929292926,
      "avg_output_tokens_per_task": 5746.025252525253,
      "last_run_timestamp": "2026-09-23T06:09:34.299Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "auto-routing",
      "accuracy": 0.8801906666666667,
      "accuracy_std_dev": 0.011226983784119793,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.0227881835016835,
      "avg_duration_per_task_ms": 234700.68434343435,
      "avg_output_tokens_per_task": 25264.190235690236,
      "last_run_timestamp": "2026-10-03T09:30:18.263Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Decart",
      "accuracy": 0.8798644568527919,
      "accuracy_std_dev": 0.004527248931623837,
      "run_count": 3,
      "total_questions": 591,
      "avg_cost_per_task_usd": 0.0029938477157360403,
      "avg_duration_per_task_ms": 316818.90016920475,
      "avg_output_tokens_per_task": 18321.88663282572,
      "last_run_timestamp": "2026-08-22T10:20:26.941Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openrouter/auto",
      "model_slug": "openrouter/auto",
      "model_name": "Auto Router",
      "provider_name": "auto-routing",
      "accuracy": 0.8795957999999999,
      "accuracy_std_dev": 0.026351833738091142,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.08917041111111111,
      "avg_duration_per_task_ms": 196788.44343434344,
      "avg_output_tokens_per_task": 37445.42121212121,
      "last_run_timestamp": "2026-09-29T14:28:17.204Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-opus-5-20260723",
      "model_slug": "anthropic/claude-opus-5",
      "model_name": "Anthropic: Claude Opus 5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.8794386469072164,
      "accuracy_std_dev": 0.018595844993461137,
      "run_count": 12,
      "total_questions": 2328,
      "avg_cost_per_task_usd": 0.12945956872852238,
      "avg_duration_per_task_ms": 63011.24012027492,
      "avg_output_tokens_per_task": 4952.514604810996,
      "last_run_timestamp": "2026-10-10T08:14:49.614Z"
    },
    {
      "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": 0.8793878407750632,
      "accuracy_std_dev": 0.007247192184257008,
      "run_count": 6,
      "total_questions": 1187,
      "avg_cost_per_task_usd": 0.015630656276326874,
      "avg_duration_per_task_ms": 322819.4338668913,
      "avg_output_tokens_per_task": 28000.395956192082,
      "last_run_timestamp": "2026-10-10T08:36:28.938Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "OpenInference",
      "accuracy": 0.879386,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 76,
      "avg_cost_per_task_usd": 0.03120878947368421,
      "avg_duration_per_task_ms": 274583.63157894736,
      "avg_output_tokens_per_task": 9741.526315789473,
      "last_run_timestamp": "2026-10-09T00:00:06.206Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Reka",
      "accuracy": 0.879189,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 189,
      "avg_cost_per_task_usd": 0.020723571428571428,
      "avg_duration_per_task_ms": 141095.54497354498,
      "avg_output_tokens_per_task": 17734.21164021164,
      "last_run_timestamp": "2026-09-15T00:33:26.568Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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-v4.1-flash",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "Deepseek V4.1 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.878788,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.010643959595959596,
      "avg_duration_per_task_ms": 127627.76767676767,
      "avg_output_tokens_per_task": 14277.949494949495,
      "last_run_timestamp": "2026-10-08T13:29:57.446Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-pro",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Mimo V2.6 Pro",
      "provider_name": "auto-routing",
      "accuracy": 0.878788,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.01249009595959596,
      "avg_duration_per_task_ms": 315799.5808080808,
      "avg_output_tokens_per_task": 14226.873737373737,
      "last_run_timestamp": "2026-10-08T01:55:44.175Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.2-20251211",
      "model_slug": "openai/gpt-5.2",
      "model_name": "OpenAI: GPT-5.2",
      "provider_name": "OpenAI",
      "accuracy": 0.878788,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.13332395959595958,
      "avg_duration_per_task_ms": 138478.24242424243,
      "avg_output_tokens_per_task": 9426.934343434343,
      "last_run_timestamp": "2026-08-01T08:50:28.136Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.7-plus-20260602",
      "model_slug": "qwen/qwen3.7-plus",
      "model_name": "Qwen: Qwen3.7 Plus",
      "provider_name": "auto-routing",
      "accuracy": 0.878788,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.043995565656565656,
      "avg_duration_per_task_ms": 588519.7272727273,
      "avg_output_tokens_per_task": 34166.82828282828,
      "last_run_timestamp": "2026-10-01T08:47:14.818Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-luna-20260709",
      "model_slug": "openai/gpt-5.6-luna",
      "model_name": "OpenAI: GPT-5.6 Luna",
      "provider_name": "auto-routing",
      "accuracy": 0.8787876666666666,
      "accuracy_std_dev": 0.005991714074916775,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.007907055555555554,
      "avg_duration_per_task_ms": 79561.398989899,
      "avg_output_tokens_per_task": 7834.653198653199,
      "last_run_timestamp": "2026-10-01T19:14:22.294Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Fireworks",
      "accuracy": 0.8787801424227066,
      "accuracy_std_dev": 0.016401176693323544,
      "run_count": 11,
      "total_questions": 1973,
      "avg_cost_per_task_usd": 0.011341015205271159,
      "avg_duration_per_task_ms": 261740.6791687785,
      "avg_output_tokens_per_task": 24571.48302078054,
      "last_run_timestamp": "2026-09-19T08:54:58.652Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "GMICloud",
      "accuracy": 0.8779461000000001,
      "accuracy_std_dev": 0.03428938607922281,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.10926873686868688,
      "avg_duration_per_task_ms": 517475.14797979797,
      "avg_output_tokens_per_task": 35212.96313131313,
      "last_run_timestamp": "2026-10-10T08:22:15.909Z"
    },
    {
      "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": 0.8772414110344827,
      "accuracy_std_dev": 0.0072833211636968214,
      "run_count": 11,
      "total_questions": 2175,
      "avg_cost_per_task_usd": 0.11771683494252876,
      "avg_duration_per_task_ms": 751137.611954023,
      "avg_output_tokens_per_task": 30859.994942528734,
      "last_run_timestamp": "2026-10-10T08:22:03.480Z"
    },
    {
      "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": 0.8772163628528975,
      "accuracy_std_dev": 0.0321266435991701,
      "run_count": 17,
      "total_questions": 3365,
      "avg_cost_per_task_usd": 0.14701722585438334,
      "avg_duration_per_task_ms": 392848.22288261517,
      "avg_output_tokens_per_task": 11120.41337295691,
      "last_run_timestamp": "2026-10-10T08:22:13.763Z"
    },
    {
      "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": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Wafer",
      "accuracy": 0.8771433591105211,
      "accuracy_std_dev": 0.1534252530590106,
      "run_count": 17,
      "total_questions": 3013,
      "avg_cost_per_task_usd": 0.14607643743776966,
      "avg_duration_per_task_ms": 264965.2558911384,
      "avg_output_tokens_per_task": 10925.349817457683,
      "last_run_timestamp": "2026-10-10T08:37:55.990Z"
    },
    {
      "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-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8771043333333333,
      "accuracy_std_dev": 0.01914541124017855,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.008207621212121211,
      "avg_duration_per_task_ms": 369369.9638047138,
      "avg_output_tokens_per_task": 28948.207912457914,
      "last_run_timestamp": "2026-10-10T16:21:38.272Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "GMICloud",
      "accuracy": 0.877104,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.07509321212121212,
      "avg_duration_per_task_ms": 263175.32828282827,
      "avg_output_tokens_per_task": 20738.459595959597,
      "last_run_timestamp": "2026-08-29T19:12:43.653Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "AkashML",
      "accuracy": 0.8765966933797908,
      "accuracy_std_dev": 0.020189124617118887,
      "run_count": 6,
      "total_questions": 1148,
      "avg_cost_per_task_usd": 0.004125730836236933,
      "avg_duration_per_task_ms": 453615.28397212544,
      "avg_output_tokens_per_task": 16054.564459930314,
      "last_run_timestamp": "2026-09-05T08:24:32.148Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-fable-5.1-20260831",
      "model_slug": "anthropic/claude-fable-5.1",
      "model_name": "Anthropic: Claude Fable 5.1",
      "provider_name": "Google",
      "accuracy": 0.8765431666666668,
      "accuracy_std_dev": 0.03217293038884639,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.11818556060606061,
      "avg_duration_per_task_ms": 28260.9031986532,
      "avg_output_tokens_per_task": 2161.995791245791,
      "last_run_timestamp": "2026-10-01T08:25:56.890Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Parasail",
      "accuracy": 0.8761782480719793,
      "accuracy_std_dev": 0.00788648157846706,
      "run_count": 4,
      "total_questions": 778,
      "avg_cost_per_task_usd": 0.0930879910025707,
      "avg_duration_per_task_ms": 616690.5848329049,
      "avg_output_tokens_per_task": 27406.845758354757,
      "last_run_timestamp": "2026-08-22T09:51:24.721Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Parasail",
      "accuracy": 0.8760934928027095,
      "accuracy_std_dev": 0.008565775498394134,
      "run_count": 6,
      "total_questions": 1181,
      "avg_cost_per_task_usd": 0.11545865114309908,
      "avg_duration_per_task_ms": 580224.5131244708,
      "avg_output_tokens_per_task": 32981.541913632514,
      "last_run_timestamp": "2026-10-10T16:22:35.001Z"
    },
    {
      "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": 0.876032909090909,
      "accuracy_std_dev": 0.03183332512108125,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.1551542842056933,
      "avg_duration_per_task_ms": 356025.97061524336,
      "avg_output_tokens_per_task": 35808.172635445364,
      "last_run_timestamp": "2026-10-10T08:22:25.674Z"
    },
    {
      "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": 0.8757575000000001,
      "accuracy_std_dev": 0.017781528276557126,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.13141335,
      "avg_duration_per_task_ms": 63646.402525252524,
      "avg_output_tokens_per_task": 5036.028787878788,
      "last_run_timestamp": "2026-10-10T08:11:16.872Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Together",
      "accuracy": 0.8756554578651687,
      "accuracy_std_dev": 0.014258761571593646,
      "run_count": 9,
      "total_questions": 1780,
      "avg_cost_per_task_usd": 0.0903369808988764,
      "avg_duration_per_task_ms": 218770.61797752808,
      "avg_output_tokens_per_task": 22903.363483146066,
      "last_run_timestamp": "2026-10-10T08:21:17.581Z"
    },
    {
      "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": 0.8755794209860935,
      "accuracy_std_dev": 0.004386999116708366,
      "run_count": 8,
      "total_questions": 1582,
      "avg_cost_per_task_usd": 0.12302590328697852,
      "avg_duration_per_task_ms": 915411.7926675095,
      "avg_output_tokens_per_task": 43199.48293299621,
      "last_run_timestamp": "2026-10-10T16:19:34.448Z"
    },
    {
      "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": "WandB",
      "accuracy": 0.875421,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.1538974696969697,
      "avg_duration_per_task_ms": 363407.6313131313,
      "avg_output_tokens_per_task": 34764.520202020205,
      "last_run_timestamp": "2026-07-18T08:26:40.271Z"
    },
    {
      "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": 0.8754207500000001,
      "accuracy_std_dev": 0.014884023314306964,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.11076118223905723,
      "avg_duration_per_task_ms": 245485.56776094277,
      "avg_output_tokens_per_task": 31501.30387205387,
      "last_run_timestamp": "2026-10-10T08:22:18.403Z"
    },
    {
      "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": 0.8754207142857143,
      "accuracy_std_dev": 0.015996248780924358,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.22405389971139966,
      "avg_duration_per_task_ms": 1033226.1435786436,
      "avg_output_tokens_per_task": 56216.72438672439,
      "last_run_timestamp": "2026-10-10T16:29:33.708Z"
    },
    {
      "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": 0.8749732275334607,
      "accuracy_std_dev": 0.01165560971968007,
      "run_count": 8,
      "total_questions": 1569,
      "avg_cost_per_task_usd": 0.042417240917782026,
      "avg_duration_per_task_ms": 238116.32058636073,
      "avg_output_tokens_per_task": 32745.878903760356,
      "last_run_timestamp": "2026-10-10T08:28:02.892Z"
    },
    {
      "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": 0.8747894999999999,
      "accuracy_std_dev": 0.005115959905042276,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.0784861994949495,
      "avg_duration_per_task_ms": 339416.61994949495,
      "avg_output_tokens_per_task": 22314.78724747475,
      "last_run_timestamp": "2026-10-10T16:21:08.442Z"
    },
    {
      "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-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Novita",
      "accuracy": 0.8744587142857143,
      "accuracy_std_dev": 0.028125314066636473,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.1966599458874459,
      "avg_duration_per_task_ms": 1101198.378787879,
      "avg_output_tokens_per_task": 58277.85137085137,
      "last_run_timestamp": "2026-10-10T16:22:34.553Z"
    },
    {
      "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": 0.8743913028237584,
      "accuracy_std_dev": 0.02063518118543075,
      "run_count": 6,
      "total_questions": 1027,
      "avg_cost_per_task_usd": 0.1379305988315482,
      "avg_duration_per_task_ms": 596483.0662122688,
      "avg_output_tokens_per_task": 50109.39727361246,
      "last_run_timestamp": "2026-10-03T17:01:24.374Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Phala",
      "accuracy": 0.8741992451960031,
      "accuracy_std_dev": 0.025720260433664145,
      "run_count": 7,
      "total_questions": 1301,
      "avg_cost_per_task_usd": 0.009777912375096079,
      "avg_duration_per_task_ms": 392901.29208301305,
      "avg_output_tokens_per_task": 21503.2475019216,
      "last_run_timestamp": "2026-10-10T08:23:18.139Z"
    },
    {
      "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": 0.8741581250000001,
      "accuracy_std_dev": 0.007610645331335265,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.049696884469696966,
      "avg_duration_per_task_ms": 277247.4457070707,
      "avg_output_tokens_per_task": 21110.815025252527,
      "last_run_timestamp": "2026-10-10T16:11:14.296Z"
    },
    {
      "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": 0.8741300360531311,
      "accuracy_std_dev": 0.0120411700506014,
      "run_count": 8,
      "total_questions": 1581,
      "avg_cost_per_task_usd": 0.008937657811511701,
      "avg_duration_per_task_ms": 436322.61416824796,
      "avg_output_tokens_per_task": 21945.78494623656,
      "last_run_timestamp": "2026-10-10T08:31:34.941Z"
    },
    {
      "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": 0.127235829004329,
      "avg_duration_per_task_ms": 1398882.1861471862,
      "avg_output_tokens_per_task": 55498.13997113997,
      "last_run_timestamp": "2026-10-10T16:19:40.668Z"
    },
    {
      "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": 0.8738600668693008,
      "accuracy_std_dev": 0.0037938444143111693,
      "run_count": 5,
      "total_questions": 987,
      "avg_cost_per_task_usd": 0.044807096251266464,
      "avg_duration_per_task_ms": 186049.48733535968,
      "avg_output_tokens_per_task": 38929.8591691996,
      "last_run_timestamp": "2026-10-10T08:16:44.988Z"
    },
    {
      "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": 0.873737,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.008290222222222223,
      "avg_duration_per_task_ms": 529447,
      "avg_output_tokens_per_task": 29241.171717171717,
      "last_run_timestamp": "2026-07-18T08:31:23.912Z"
    },
    {
      "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": "StreamLake",
      "accuracy": 0.8733163749999999,
      "accuracy_std_dev": 0.015051805696805129,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.025130481060606065,
      "avg_duration_per_task_ms": 723244.2184343435,
      "avg_output_tokens_per_task": 35149.60164141414,
      "last_run_timestamp": "2026-10-10T16:32:56.553Z"
    },
    {
      "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": 0.873140968503937,
      "accuracy_std_dev": 0.008141500000000024,
      "run_count": 2,
      "total_questions": 381,
      "avg_cost_per_task_usd": 0.10504513910761155,
      "avg_duration_per_task_ms": 1003143.2047244095,
      "avg_output_tokens_per_task": 42008.77427821522,
      "last_run_timestamp": "2026-08-01T17:53:10.275Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "ModelRun",
      "accuracy": 0.8727272475524476,
      "accuracy_std_dev": 0.029233686851601524,
      "run_count": 8,
      "total_questions": 1430,
      "avg_cost_per_task_usd": 0.09700304615384615,
      "avg_duration_per_task_ms": 194874.37272727274,
      "avg_output_tokens_per_task": 26397.99300699301,
      "last_run_timestamp": "2026-09-26T09:19:13.012Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "DeepInfra",
      "accuracy": 0.8727271395348837,
      "accuracy_std_dev": 0.05132840041585165,
      "run_count": 12,
      "total_questions": 2365,
      "avg_cost_per_task_usd": 0.07667611374207188,
      "avg_duration_per_task_ms": 546814.6951374207,
      "avg_output_tokens_per_task": 38662.21902748414,
      "last_run_timestamp": "2026-10-10T08:21:47.281Z"
    },
    {
      "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": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Venice",
      "accuracy": 0.8723906000000001,
      "accuracy_std_dev": 0.017443444896006063,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.032869219191919194,
      "avg_duration_per_task_ms": 141993.59494949496,
      "avg_output_tokens_per_task": 24250.128282828282,
      "last_run_timestamp": "2026-10-10T08:24:43.682Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.8723343333333334,
      "accuracy_std_dev": 0.0042828368194716806,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.0043181245791245795,
      "avg_duration_per_task_ms": 40010.954545454544,
      "avg_output_tokens_per_task": 7867.228956228956,
      "last_run_timestamp": "2026-10-10T08:23:23.511Z"
    },
    {
      "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": 0.8721631051412692,
      "accuracy_std_dev": 0.00936942620732698,
      "run_count": 11,
      "total_questions": 2159,
      "avg_cost_per_task_usd": 0.0081905757295044,
      "avg_duration_per_task_ms": 336770.76748494676,
      "avg_output_tokens_per_task": 28956.01343214451,
      "last_run_timestamp": "2026-10-10T08:22:41.871Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Crusoe",
      "accuracy": 0.8720540000000001,
      "accuracy_std_dev": 0.003367000000000009,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.04668046464646465,
      "avg_duration_per_task_ms": 217379.2803030303,
      "avg_output_tokens_per_task": 38760.13888888889,
      "last_run_timestamp": "2026-10-10T08:22:24.805Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Morph",
      "accuracy": 0.871875,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 160,
      "avg_cost_per_task_usd": 0.04507655,
      "avg_duration_per_task_ms": 1504372.61875,
      "avg_output_tokens_per_task": 18693.5375,
      "last_run_timestamp": "2026-08-27T15:00:05.074Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Inceptron",
      "accuracy": 0.8717762017353579,
      "accuracy_std_dev": 0.02072198720771574,
      "run_count": 7,
      "total_questions": 1383,
      "avg_cost_per_task_usd": 0.1588113485177151,
      "avg_duration_per_task_ms": 951795.9934924078,
      "avg_output_tokens_per_task": 53171.428054953,
      "last_run_timestamp": "2026-10-10T16:20:06.463Z"
    },
    {
      "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": 0.8716425648854962,
      "accuracy_std_dev": 0.013438421323454144,
      "run_count": 6,
      "total_questions": 1179,
      "avg_cost_per_task_usd": 0.1836925326547922,
      "avg_duration_per_task_ms": 775815.9338422392,
      "avg_output_tokens_per_task": 53092.24766751484,
      "last_run_timestamp": "2026-10-10T16:26:47.810Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "DeepInfra",
      "accuracy": 0.8714859264426126,
      "accuracy_std_dev": 0.00752103815972237,
      "run_count": 8,
      "total_questions": 1577,
      "avg_cost_per_task_usd": 0.06945589727330374,
      "avg_duration_per_task_ms": 276858.71528218134,
      "avg_output_tokens_per_task": 11308.629676601142,
      "last_run_timestamp": "2026-10-10T08:22:14.215Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "Chutes",
      "accuracy": 0.8714225,
      "accuracy_std_dev": 0.004885840587862051,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.06337124747474747,
      "avg_duration_per_task_ms": 284598.4861111111,
      "avg_output_tokens_per_task": 21034.015151515152,
      "last_run_timestamp": "2026-08-15T15:55:02.752Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8714174896551725,
      "accuracy_std_dev": 0.09994490478357737,
      "run_count": 11,
      "total_questions": 2175,
      "avg_cost_per_task_usd": 0.0273483167816092,
      "avg_duration_per_task_ms": 210323.05977011495,
      "avg_output_tokens_per_task": 23588.830344827587,
      "last_run_timestamp": "2026-10-10T16:22:50.823Z"
    },
    {
      "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": 0.8713650909090909,
      "accuracy_std_dev": 0.010760647289203607,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.006425433425160699,
      "avg_duration_per_task_ms": 272145.9591368228,
      "avg_output_tokens_per_task": 18480.175849403124,
      "last_run_timestamp": "2026-10-10T08:22:19.761Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "Anthropic",
      "accuracy": 0.871212,
      "accuracy_std_dev": 0.007576000000000027,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.00393519191919192,
      "avg_duration_per_task_ms": 37885.9494949495,
      "avg_output_tokens_per_task": 7630.535353535353,
      "last_run_timestamp": "2026-10-10T08:33:36.051Z"
    },
    {
      "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": 0.8712119166666668,
      "accuracy_std_dev": 0.01635831591504423,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.008297640993265992,
      "avg_duration_per_task_ms": 229029.07449494948,
      "avg_output_tokens_per_task": 29392.178030303032,
      "last_run_timestamp": "2026-10-10T08:22:41.239Z"
    },
    {
      "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": 0.8711735553145336,
      "accuracy_std_dev": 0.01949786803730092,
      "run_count": 7,
      "total_questions": 1383,
      "avg_cost_per_task_usd": 0.05907025379609545,
      "avg_duration_per_task_ms": 512258.996384671,
      "avg_output_tokens_per_task": 33994.65726681128,
      "last_run_timestamp": "2026-10-10T16:29:14.779Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "thinkingmachines/inkling-small-20260730",
      "model_slug": "thinkingmachines/inkling-small",
      "model_name": "Thinking Machines: Inkling Small",
      "provider_name": "Together",
      "accuracy": 0.8710116425661915,
      "accuracy_std_dev": 0.01025225770257459,
      "run_count": 5,
      "total_questions": 982,
      "avg_cost_per_task_usd": 0.025059682281059068,
      "avg_duration_per_task_ms": 328330.6405295316,
      "avg_output_tokens_per_task": 20642.19348268839,
      "last_run_timestamp": "2026-09-12T08:54:29.103Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Parasail",
      "accuracy": 0.8706108571428572,
      "accuracy_std_dev": 0.005578448284720969,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.008177433621933622,
      "avg_duration_per_task_ms": 408394.5281385281,
      "avg_output_tokens_per_task": 28879.512265512265,
      "last_run_timestamp": "2026-10-10T16:19:50.069Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.8703705,
      "accuracy_std_dev": 0.0016835000000000044,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.004056603535353536,
      "avg_duration_per_task_ms": 39122.805555555555,
      "avg_output_tokens_per_task": 7875.414141414141,
      "last_run_timestamp": "2026-10-10T08:24:29.911Z"
    },
    {
      "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": "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": 0.8701297142857144,
      "accuracy_std_dev": 0.009751370859148623,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.0770424466089466,
      "avg_duration_per_task_ms": 106865.99062049061,
      "avg_output_tokens_per_task": 12502.755411255412,
      "last_run_timestamp": "2026-10-10T08:24:11.799Z"
    },
    {
      "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": "Alibaba",
      "accuracy": 0.8698495694200352,
      "accuracy_std_dev": 0.013406448590229597,
      "run_count": 9,
      "total_questions": 1707,
      "avg_cost_per_task_usd": 0.00814071236086702,
      "avg_duration_per_task_ms": 243885.67662565905,
      "avg_output_tokens_per_task": 29621.29173989455,
      "last_run_timestamp": "2026-10-10T16:20:12.054Z"
    },
    {
      "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": "auto-routing",
      "accuracy": 0.8694081428571427,
      "accuracy_std_dev": 0.012000957395481546,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.0817209256854257,
      "avg_duration_per_task_ms": 202473.6810966811,
      "avg_output_tokens_per_task": 13277.033189033189,
      "last_run_timestamp": "2026-10-10T08:13:06.615Z"
    },
    {
      "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": 0.8693955131578947,
      "accuracy_std_dev": 0.019648291273268125,
      "run_count": 7,
      "total_questions": 1368,
      "avg_cost_per_task_usd": 0.07663164254385965,
      "avg_duration_per_task_ms": 433978.8881578947,
      "avg_output_tokens_per_task": 23705.32821637427,
      "last_run_timestamp": "2026-10-08T23:14:17.356Z"
    },
    {
      "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": 0.8693361387478848,
      "accuracy_std_dev": 0.024099081962433728,
      "run_count": 9,
      "total_questions": 1773,
      "avg_cost_per_task_usd": 0.08116604963338973,
      "avg_duration_per_task_ms": 411609.69035532995,
      "avg_output_tokens_per_task": 32356.587704455724,
      "last_run_timestamp": "2026-10-10T16:22:22.711Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Sail Research",
      "accuracy": 0.8692809647058825,
      "accuracy_std_dev": 0.009104176033557385,
      "run_count": 4,
      "total_questions": 765,
      "avg_cost_per_task_usd": 0.0047775359477124175,
      "avg_duration_per_task_ms": 482494.045751634,
      "avg_output_tokens_per_task": 26353.58431372549,
      "last_run_timestamp": "2026-08-22T16:48:48.690Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.1-20251113",
      "model_slug": "openai/gpt-5.1",
      "model_name": "OpenAI: GPT-5.1",
      "provider_name": "OpenAI",
      "accuracy": 0.8692479999999999,
      "accuracy_std_dev": 0.004827531943619518,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.22171807575757574,
      "avg_duration_per_task_ms": 260359.50673400675,
      "avg_output_tokens_per_task": 22076.11111111111,
      "last_run_timestamp": "2026-10-01T16:12:49.000Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Baidu",
      "accuracy": 0.8691678571428573,
      "accuracy_std_dev": 0.022635118532483796,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.15808130735930737,
      "avg_duration_per_task_ms": 1043145.4696969697,
      "avg_output_tokens_per_task": 60299.43795093795,
      "last_run_timestamp": "2026-10-10T16:27:14.209Z"
    },
    {
      "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": 0.8687731366353544,
      "accuracy_std_dev": 0.028735792378518143,
      "run_count": 13,
      "total_questions": 2342,
      "avg_cost_per_task_usd": 0.011254341588385995,
      "avg_duration_per_task_ms": 345256.6985482494,
      "avg_output_tokens_per_task": 20267.622117847994,
      "last_run_timestamp": "2026-10-10T08:22:27.733Z"
    },
    {
      "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": "openai/gpt-5.1-20251113",
      "model_slug": "openai/gpt-5.1",
      "model_name": "OpenAI: GPT-5.1",
      "provider_name": "Azure",
      "accuracy": 0.8686867500000001,
      "accuracy_std_dev": 0.009143975622643609,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.2541581641414141,
      "avg_duration_per_task_ms": 251149.7840909091,
      "avg_output_tokens_per_task": 24116.972222222223,
      "last_run_timestamp": "2026-10-01T16:14:35.237Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "Modal",
      "accuracy": 0.8682058571428571,
      "accuracy_std_dev": 0.007673379706651383,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.07688844300144301,
      "avg_duration_per_task_ms": 83582.13492063493,
      "avg_output_tokens_per_task": 12574.14357864358,
      "last_run_timestamp": "2026-10-10T08:21:46.324Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "Google",
      "accuracy": 0.8681256666666667,
      "accuracy_std_dev": 0.00955634355923971,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.004278478114478115,
      "avg_duration_per_task_ms": 35528.6632996633,
      "avg_output_tokens_per_task": 7802.548821548821,
      "last_run_timestamp": "2026-10-10T08:40:38.947Z"
    },
    {
      "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": 0.8680553750000001,
      "accuracy_std_dev": 0.004605583891796467,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.07070611994949497,
      "avg_duration_per_task_ms": 232289.42803030304,
      "avg_output_tokens_per_task": 11517.214015151516,
      "last_run_timestamp": "2026-10-10T08:22:11.510Z"
    },
    {
      "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": 0.8678449999999999,
      "accuracy_std_dev": 0.007760578667212899,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.0038241401515151513,
      "avg_duration_per_task_ms": 408271.3705808081,
      "avg_output_tokens_per_task": 29802.972222222223,
      "last_run_timestamp": "2026-10-10T16:19:50.076Z"
    },
    {
      "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": 0.8676699264214046,
      "accuracy_std_dev": 0.009490809964264255,
      "run_count": 9,
      "total_questions": 1495,
      "avg_cost_per_task_usd": 0.05216587558528428,
      "avg_duration_per_task_ms": 421713.7277591973,
      "avg_output_tokens_per_task": 17294.008695652174,
      "last_run_timestamp": "2026-10-10T16:20:24.763Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "thinkingmachines/inkling-small-20260730",
      "model_slug": "thinkingmachines/inkling-small",
      "model_name": "Thinking Machines: Inkling Small",
      "provider_name": "DeepInfra",
      "accuracy": 0.867634625,
      "accuracy_std_dev": 0.007431114989311851,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.025694838383838382,
      "avg_duration_per_task_ms": 248942.36553030304,
      "avg_output_tokens_per_task": 21167.008207070707,
      "last_run_timestamp": "2026-10-08T22:14:16.087Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "WandB",
      "accuracy": 0.867003,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.11555696464646466,
      "avg_duration_per_task_ms": 292702.6111111111,
      "avg_output_tokens_per_task": 29334.075757575756,
      "last_run_timestamp": "2026-07-18T08:27:05.304Z"
    },
    {
      "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": 0.8666407869362363,
      "accuracy_std_dev": 0.0883070185402466,
      "run_count": 13,
      "total_questions": 2572,
      "avg_cost_per_task_usd": 0.26275894012441675,
      "avg_duration_per_task_ms": 494855.53265940904,
      "avg_output_tokens_per_task": 63072.154743390354,
      "last_run_timestamp": "2026-10-10T08:36:25.212Z"
    },
    {
      "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": 0.8665821595946803,
      "accuracy_std_dev": 0.017909527324426373,
      "run_count": 8,
      "total_questions": 1579,
      "avg_cost_per_task_usd": 0.006400538315389487,
      "avg_duration_per_task_ms": 337664.5319822673,
      "avg_output_tokens_per_task": 21313.250158328057,
      "last_run_timestamp": "2026-10-10T08:29:42.266Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "CoreWeave",
      "accuracy": 0.8662817142857142,
      "accuracy_std_dev": 0.018765141853327668,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.19692205699855703,
      "avg_duration_per_task_ms": 428251.06204906205,
      "avg_output_tokens_per_task": 57606.430014430014,
      "last_run_timestamp": "2026-10-10T16:23:16.833Z"
    },
    {
      "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": "DeepSeek",
      "accuracy": 0.8661615,
      "accuracy_std_dev": 0.012626499999999985,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.030201459595959598,
      "avg_duration_per_task_ms": 496509.8611111111,
      "avg_output_tokens_per_task": 34831.292929292926,
      "last_run_timestamp": "2026-08-08T09:24:32.269Z"
    },
    {
      "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": 0.8659881251908396,
      "accuracy_std_dev": 0.03841903799472341,
      "run_count": 10,
      "total_questions": 1965,
      "avg_cost_per_task_usd": 0.09208694758269721,
      "avg_duration_per_task_ms": 534219.7852417303,
      "avg_output_tokens_per_task": 23722.84885496183,
      "last_run_timestamp": "2026-10-10T08:26:20.990Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "Azure",
      "accuracy": 0.86574075,
      "accuracy_std_dev": 0.004187816696979481,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.00415591287878788,
      "avg_duration_per_task_ms": 35585.20075757576,
      "avg_output_tokens_per_task": 7676.222222222223,
      "last_run_timestamp": "2026-10-10T08:24:11.975Z"
    },
    {
      "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": 0.865689794871795,
      "accuracy_std_dev": 0.004877364201651907,
      "run_count": 3,
      "total_questions": 546,
      "avg_cost_per_task_usd": 0.005803672161172161,
      "avg_duration_per_task_ms": 153675.5,
      "avg_output_tokens_per_task": 20639.26923076923,
      "last_run_timestamp": "2026-08-08T09:25:45.716Z"
    },
    {
      "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": "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": 0.865530125,
      "accuracy_std_dev": 0.009082931319754368,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.07008002651515152,
      "avg_duration_per_task_ms": 228487.37373737374,
      "avg_output_tokens_per_task": 11410.660353535353,
      "last_run_timestamp": "2026-10-10T08:31:34.860Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "NextBit",
      "accuracy": 0.8653195,
      "accuracy_std_dev": 0.0016835000000000044,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.008587179292929294,
      "avg_duration_per_task_ms": 240899.30303030304,
      "avg_output_tokens_per_task": 30387.573232323233,
      "last_run_timestamp": "2026-08-29T18:38:15.907Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Inceptron",
      "accuracy": 0.8651996980273142,
      "accuracy_std_dev": 0.03307372620434535,
      "run_count": 10,
      "total_questions": 1977,
      "avg_cost_per_task_usd": 0.13402964896307534,
      "avg_duration_per_task_ms": 611597.9382903388,
      "avg_output_tokens_per_task": 40351.92817400101,
      "last_run_timestamp": "2026-10-10T08:14:39.277Z"
    },
    {
      "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": 0.8650791428571427,
      "accuracy_std_dev": 0.02076397686536241,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.11897097835497836,
      "avg_duration_per_task_ms": 585047.1832611832,
      "avg_output_tokens_per_task": 35346.81024531025,
      "last_run_timestamp": "2026-10-10T16:21:22.563Z"
    },
    {
      "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": 0.8649788658227848,
      "accuracy_std_dev": 0.00034200000000000896,
      "run_count": 2,
      "total_questions": 395,
      "avg_cost_per_task_usd": 0.5347964379746836,
      "avg_duration_per_task_ms": 1600008.7797468354,
      "avg_output_tokens_per_task": 61634.33417721519,
      "last_run_timestamp": "2026-10-10T16:21:45.788Z"
    },
    {
      "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": 0.8648158710912572,
      "accuracy_std_dev": 0.014610716473170469,
      "run_count": 8,
      "total_questions": 1567,
      "avg_cost_per_task_usd": 0.07521307466496491,
      "avg_duration_per_task_ms": 273268.63114231016,
      "avg_output_tokens_per_task": 20767.77855775367,
      "last_run_timestamp": "2026-10-10T16:19:34.427Z"
    },
    {
      "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": 0.8646463,
      "accuracy_std_dev": 0.01622113306184252,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.08817965,
      "avg_duration_per_task_ms": 193592.76363636364,
      "avg_output_tokens_per_task": 24362.06313131313,
      "last_run_timestamp": "2026-10-10T16:11:14.298Z"
    },
    {
      "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": 0.8645964220532318,
      "accuracy_std_dev": 0.028631942149459588,
      "run_count": 8,
      "total_questions": 1578,
      "avg_cost_per_task_usd": 0.10253569771863118,
      "avg_duration_per_task_ms": 364033.2845373891,
      "avg_output_tokens_per_task": 23065.25855513308,
      "last_run_timestamp": "2026-10-10T08:22:13.815Z"
    },
    {
      "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": 0.8643883344182262,
      "accuracy_std_dev": 0.01774344793127301,
      "run_count": 7,
      "total_questions": 1229,
      "avg_cost_per_task_usd": 0.0051325646867371845,
      "avg_duration_per_task_ms": 651023.4035801465,
      "avg_output_tokens_per_task": 28234.647681041497,
      "last_run_timestamp": "2026-10-08T23:29:09.758Z"
    },
    {
      "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": 0.8643575714285714,
      "accuracy_std_dev": 0.014782409815695945,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.042783132756132755,
      "avg_duration_per_task_ms": 500704.0137085137,
      "avg_output_tokens_per_task": 34254.27994227994,
      "last_run_timestamp": "2026-10-10T16:29:43.875Z"
    },
    {
      "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": 0.8642674999999999,
      "accuracy_std_dev": 0.01013372239110584,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.008602727904040404,
      "avg_duration_per_task_ms": 273218.547979798,
      "avg_output_tokens_per_task": 30406.29861111111,
      "last_run_timestamp": "2026-10-10T16:10:48.867Z"
    },
    {
      "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": 0.86405725,
      "accuracy_std_dev": 0.017018053235828706,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.07367723863636363,
      "avg_duration_per_task_ms": 277870.3768939394,
      "avg_output_tokens_per_task": 20967.248106060608,
      "last_run_timestamp": "2026-10-10T16:19:34.397Z"
    },
    {
      "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": "Venice",
      "accuracy": 0.863779637554585,
      "accuracy_std_dev": 0.014583686901214414,
      "run_count": 7,
      "total_questions": 1374,
      "avg_cost_per_task_usd": 0.13604484788937413,
      "avg_duration_per_task_ms": 292893.29330422124,
      "avg_output_tokens_per_task": 27835.631004366813,
      "last_run_timestamp": "2026-10-10T08:22:50.053Z"
    },
    {
      "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": 0.86363625,
      "accuracy_std_dev": 0.015612150218259494,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.049225773358585856,
      "avg_duration_per_task_ms": 723327.8787878788,
      "avg_output_tokens_per_task": 33961.57638888889,
      "last_run_timestamp": "2026-10-10T16:21:33.430Z"
    },
    {
      "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": 0.863636,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.008496712121212122,
      "avg_duration_per_task_ms": 739364.4242424242,
      "avg_output_tokens_per_task": 30053.868686868685,
      "last_run_timestamp": "2026-07-18T08:30:06.336Z"
    },
    {
      "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": "anthropic/claude-5-fable-20260609",
      "model_slug": "anthropic/claude-fable-5",
      "model_name": "Anthropic: Claude Fable 5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.8632620000000001,
      "accuracy_std_dev": 0.0411666386369675,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.17732408585858583,
      "avg_duration_per_task_ms": 45341.72558922559,
      "avg_output_tokens_per_task": 3345.6531986531986,
      "last_run_timestamp": "2026-10-10T08:22:26.757Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "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": "z-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": 0.8628614010378057,
      "accuracy_std_dev": 0.04100774563321383,
      "run_count": 7,
      "total_questions": 1349,
      "avg_cost_per_task_usd": 0.015537737583395106,
      "avg_duration_per_task_ms": 567965.6375092661,
      "avg_output_tokens_per_task": 38495.872498146775,
      "last_run_timestamp": "2026-10-10T08:29:25.989Z"
    },
    {
      "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": 0.8625952061068702,
      "accuracy_std_dev": 0.005736999999999992,
      "run_count": 2,
      "total_questions": 393,
      "avg_cost_per_task_usd": 0.043527526717557254,
      "avg_duration_per_task_ms": 365946.2213740458,
      "avg_output_tokens_per_task": 24956.239185750637,
      "last_run_timestamp": "2026-08-01T17:57:47.536Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "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": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Inceptron",
      "accuracy": 0.8624318766716196,
      "accuracy_std_dev": 0.015635397712647187,
      "run_count": 7,
      "total_questions": 1346,
      "avg_cost_per_task_usd": 0.08403349479940564,
      "avg_duration_per_task_ms": 364844.5950965825,
      "avg_output_tokens_per_task": 25975.783060921247,
      "last_run_timestamp": "2026-10-10T08:23:49.783Z"
    },
    {
      "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": 0.862290694980695,
      "accuracy_std_dev": 0.013680701809483308,
      "run_count": 4,
      "total_questions": 777,
      "avg_cost_per_task_usd": 0.02531409137709138,
      "avg_duration_per_task_ms": 145459.53539253538,
      "avg_output_tokens_per_task": 38737.666666666664,
      "last_run_timestamp": "2026-10-10T08:21:21.538Z"
    },
    {
      "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": 0.8622895,
      "accuracy_std_dev": 0.009153046839714083,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.8048450656565656,
      "avg_duration_per_task_ms": 460341.97323232325,
      "avg_output_tokens_per_task": 32021.969191919194,
      "last_run_timestamp": "2026-10-10T08:14:45.483Z"
    },
    {
      "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": 0.8621211000000002,
      "accuracy_std_dev": 0.014333635361972893,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.7483498358585858,
      "avg_duration_per_task_ms": 493892.65252525255,
      "avg_output_tokens_per_task": 29772.12373737374,
      "last_run_timestamp": "2026-10-10T08:21:50.157Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "auto-routing",
      "accuracy": 0.8619530000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.004047368686868687,
      "avg_duration_per_task_ms": 38582.818181818184,
      "avg_output_tokens_per_task": 7859.823232323232,
      "last_run_timestamp": "2026-10-10T08:13:08.565Z"
    },
    {
      "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": 0.8619530000000001,
      "accuracy_std_dev": 0.009912179612308596,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.07266585185185186,
      "avg_duration_per_task_ms": 469919.06565656565,
      "avg_output_tokens_per_task": 56748.079124579126,
      "last_run_timestamp": "2026-10-10T16:19:34.424Z"
    },
    {
      "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": 0.8614478,
      "accuracy_std_dev": 0.008749266401247596,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.8440670252525252,
      "avg_duration_per_task_ms": 414934.17070707073,
      "avg_output_tokens_per_task": 33571.59191919192,
      "last_run_timestamp": "2026-10-10T08:22:13.907Z"
    },
    {
      "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": 0.8614377764705882,
      "accuracy_std_dev": 0.012431326538225896,
      "run_count": 4,
      "total_questions": 765,
      "avg_cost_per_task_usd": 0.051462403921568624,
      "avg_duration_per_task_ms": 413319.64183006535,
      "avg_output_tokens_per_task": 26795.732026143793,
      "last_run_timestamp": "2026-08-01T17:44:01.464Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Ambient",
      "accuracy": 0.8613915000000001,
      "accuracy_std_dev": 0.04603966955336814,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.004525512626262627,
      "avg_duration_per_task_ms": 565963.2584175084,
      "avg_output_tokens_per_task": 18978.082491582492,
      "last_run_timestamp": "2026-08-29T10:57:26.955Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Morph",
      "accuracy": 0.8612792,
      "accuracy_std_dev": 0.03136197207702347,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.005120404040404041,
      "avg_duration_per_task_ms": 530889.794949495,
      "avg_output_tokens_per_task": 17542.636363636364,
      "last_run_timestamp": "2026-09-05T09:22:54.724Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Baidu",
      "accuracy": 0.8612312857142858,
      "accuracy_std_dev": 0.008191248094578664,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.005079496392496392,
      "avg_duration_per_task_ms": 242062.32467532466,
      "avg_output_tokens_per_task": 28953.08080808081,
      "last_run_timestamp": "2026-10-10T16:20:24.760Z"
    },
    {
      "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": 0.861142954,
      "accuracy_std_dev": 0.038697731064696234,
      "run_count": 18,
      "total_questions": 3500,
      "avg_cost_per_task_usd": 0.004615875999999999,
      "avg_duration_per_task_ms": 95064.51371428571,
      "avg_output_tokens_per_task": 17377.474,
      "last_run_timestamp": "2026-10-10T08:28:15.461Z"
    },
    {
      "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": 0.8609939949452401,
      "accuracy_std_dev": 0.034291003424934796,
      "run_count": 6,
      "total_questions": 1187,
      "avg_cost_per_task_usd": 0.12772796882898063,
      "avg_duration_per_task_ms": 665372.3294018534,
      "avg_output_tokens_per_task": 42817.63689974726,
      "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-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Together",
      "accuracy": 0.8609694285714284,
      "accuracy_std_dev": 0.023293052022008633,
      "run_count": 4,
      "total_questions": 784,
      "avg_cost_per_task_usd": 0.10234140178571428,
      "avg_duration_per_task_ms": 463624.993622449,
      "avg_output_tokens_per_task": 30066.539540816328,
      "last_run_timestamp": "2026-08-22T16:49:15.579Z"
    },
    {
      "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": 0.86077435,
      "accuracy_std_dev": 0.007771512898239295,
      "run_count": 20,
      "total_questions": 3960,
      "avg_cost_per_task_usd": 0.8830627252525252,
      "avg_duration_per_task_ms": 416252.63737373735,
      "avg_output_tokens_per_task": 33448.936111111114,
      "last_run_timestamp": "2026-10-10T08:21:47.547Z"
    },
    {
      "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": 0.8606576691394658,
      "accuracy_std_dev": 0.011001380598888357,
      "run_count": 24,
      "total_questions": 4044,
      "avg_cost_per_task_usd": 0.16706140232443123,
      "avg_duration_per_task_ms": 253403.07789317507,
      "avg_output_tokens_per_task": 30580.46612265084,
      "last_run_timestamp": "2026-09-26T08:54:22.466Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Inceptron",
      "accuracy": 0.8605299818688982,
      "accuracy_std_dev": 0.04545858647428449,
      "run_count": 5,
      "total_questions": 717,
      "avg_cost_per_task_usd": 0.016286825662482563,
      "avg_duration_per_task_ms": 469492.70013947,
      "avg_output_tokens_per_task": 32621.504881450488,
      "last_run_timestamp": "2026-10-10T08:14:43.213Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "auto-routing",
      "accuracy": 0.8604376000000001,
      "accuracy_std_dev": 0.018325598255991542,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.0985075101010101,
      "avg_duration_per_task_ms": 504799.2101010101,
      "avg_output_tokens_per_task": 26512.831818181818,
      "last_run_timestamp": "2026-10-10T08:14:42.420Z"
    },
    {
      "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": 0.8604375000000001,
      "accuracy_std_dev": 0.010417175387311071,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.8523096944444445,
      "avg_duration_per_task_ms": 482090.7853535354,
      "avg_output_tokens_per_task": 33901.62272727273,
      "last_run_timestamp": "2026-10-10T08:22:14.416Z"
    },
    {
      "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": 0.8604060246555475,
      "accuracy_std_dev": 0.014343180030514657,
      "run_count": 7,
      "total_questions": 1379,
      "avg_cost_per_task_usd": 0.07309712690355329,
      "avg_duration_per_task_ms": 193422.33792603336,
      "avg_output_tokens_per_task": 24090.86439448876,
      "last_run_timestamp": "2026-10-10T08:24:32.509Z"
    },
    {
      "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": 0.8603742847113885,
      "accuracy_std_dev": 0.024463459108454412,
      "run_count": 7,
      "total_questions": 1282,
      "avg_cost_per_task_usd": 0.02198177067082683,
      "avg_duration_per_task_ms": 379545.007800312,
      "avg_output_tokens_per_task": 23938.183307332292,
      "last_run_timestamp": "2026-10-10T16:11:11.319Z"
    },
    {
      "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": 0.8602841643713905,
      "accuracy_std_dev": 0.03073868173578963,
      "run_count": 12,
      "total_questions": 2251,
      "avg_cost_per_task_usd": 0.025296473567303423,
      "avg_duration_per_task_ms": 476437.5064415815,
      "avg_output_tokens_per_task": 24104.674811195026,
      "last_run_timestamp": "2026-10-10T08:22:28.072Z"
    },
    {
      "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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "BaseTen",
      "accuracy": 0.859932032938258,
      "accuracy_std_dev": 0.020959658153797075,
      "run_count": 37,
      "total_questions": 7256,
      "avg_cost_per_task_usd": 0.26751970507166484,
      "avg_duration_per_task_ms": 334415.0935777288,
      "avg_output_tokens_per_task": 48371.618936052924,
      "last_run_timestamp": "2026-10-10T08:27:36.984Z"
    },
    {
      "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": "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": 0.8598485,
      "accuracy_std_dev": 0.007374290067525129,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.06667829040404041,
      "avg_duration_per_task_ms": 218736.0157828283,
      "avg_output_tokens_per_task": 10845.020202020201,
      "last_run_timestamp": "2026-10-10T08:14:46.373Z"
    },
    {
      "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": 0.8597080000000001,
      "accuracy_std_dev": 0.011568646146661513,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.843727255892256,
      "avg_duration_per_task_ms": 453475.51683501684,
      "avg_output_tokens_per_task": 33571.67003367003,
      "last_run_timestamp": "2026-10-10T08:30:17.849Z"
    },
    {
      "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": 0.8587461904761905,
      "accuracy_std_dev": 0.026229751534903395,
      "run_count": 21,
      "total_questions": 4158,
      "avg_cost_per_task_usd": 0.15894341366041365,
      "avg_duration_per_task_ms": 153104.68903318903,
      "avg_output_tokens_per_task": 15468.237133237133,
      "last_run_timestamp": "2026-10-10T08:22:11.707Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.858586,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.08651329797979797,
      "avg_duration_per_task_ms": 499241.5707070707,
      "avg_output_tokens_per_task": 35858.23737373737,
      "last_run_timestamp": "2026-10-04T03:06:10.317Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "WandB",
      "accuracy": 0.858586,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.2511099898989899,
      "avg_duration_per_task_ms": 290522.3080808081,
      "avg_output_tokens_per_task": 62590.06565656565,
      "last_run_timestamp": "2026-07-18T08:26:40.483Z"
    },
    {
      "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": 0.8585856363636363,
      "accuracy_std_dev": 0.036178356787833384,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.18056699632690543,
      "avg_duration_per_task_ms": 44952.12350780533,
      "avg_output_tokens_per_task": 3404.9237832874196,
      "last_run_timestamp": "2026-10-10T08:14:48.641Z"
    },
    {
      "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": 0.8579736363636363,
      "accuracy_std_dev": 0.04127002151082713,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.1750287943067034,
      "avg_duration_per_task_ms": 43575.79752066116,
      "avg_output_tokens_per_task": 3294.965564738292,
      "last_run_timestamp": "2026-10-10T08:14:22.551Z"
    },
    {
      "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": 0.8577798183779121,
      "accuracy_std_dev": 0.020787162919696366,
      "run_count": 12,
      "total_questions": 2318,
      "avg_cost_per_task_usd": 0.03407388869715272,
      "avg_duration_per_task_ms": 74054.28817946506,
      "avg_output_tokens_per_task": 23021.116048317515,
      "last_run_timestamp": "2026-10-10T08:38:05.766Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3-ultra-550b-a55b-20260604",
      "model_slug": "nvidia/nemotron-3-ultra-550b-a55b",
      "model_name": "NVIDIA: Nemotron 3 Ultra",
      "provider_name": "BaseTen",
      "accuracy": 0.857418,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 173,
      "avg_cost_per_task_usd": 0.08778269942196533,
      "avg_duration_per_task_ms": 317630.5202312139,
      "avg_output_tokens_per_task": 36575.57225433526,
      "last_run_timestamp": "2026-08-04T03:12:38.954Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Together",
      "accuracy": 0.8573280311284046,
      "accuracy_std_dev": 0.008776662705721345,
      "run_count": 4,
      "total_questions": 771,
      "avg_cost_per_task_usd": 0.11767329182879378,
      "avg_duration_per_task_ms": 355412.9636835279,
      "avg_output_tokens_per_task": 29743.00907911803,
      "last_run_timestamp": "2026-08-22T10:19:36.962Z"
    },
    {
      "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": 0.8569023750000001,
      "accuracy_std_dev": 0.057848586717692385,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.10924076767676767,
      "avg_duration_per_task_ms": 469852.4330808081,
      "avg_output_tokens_per_task": 27711.996843434343,
      "last_run_timestamp": "2026-10-10T08:22:14.167Z"
    },
    {
      "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": 0.8569019999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.006055888888888888,
      "avg_duration_per_task_ms": 490894.14646464644,
      "avg_output_tokens_per_task": 17030.60606060606,
      "last_run_timestamp": "2026-07-18T08:04:02.357Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5-2025-08-07",
      "model_slug": "openai/gpt-5",
      "model_name": "OpenAI: GPT-5",
      "provider_name": "auto-routing",
      "accuracy": 0.8569019999999999,
      "accuracy_std_dev": 0.0033669999999999534,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.20989276262626264,
      "avg_duration_per_task_ms": 263850.72222222225,
      "avg_output_tokens_per_task": 20893.070707070707,
      "last_run_timestamp": "2026-10-01T16:31:36.833Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "Minimax",
      "accuracy": 0.8568846555023923,
      "accuracy_std_dev": 0.008223840494828117,
      "run_count": 16,
      "total_questions": 3135,
      "avg_cost_per_task_usd": 0.06289737799043063,
      "avg_duration_per_task_ms": 600548.7228070175,
      "avg_output_tokens_per_task": 34655.311323763955,
      "last_run_timestamp": "2026-10-10T16:26:11.494Z"
    },
    {
      "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": 0.8565061764705884,
      "accuracy_std_dev": 0.012525463463235917,
      "run_count": 17,
      "total_questions": 3366,
      "avg_cost_per_task_usd": 0.13928274509803923,
      "avg_duration_per_task_ms": 437654.1420083185,
      "avg_output_tokens_per_task": 42837.04515745692,
      "last_run_timestamp": "2026-10-10T08:21:41.870Z"
    },
    {
      "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": 0.8563971000000001,
      "accuracy_std_dev": 0.01879277008559409,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.7556142424242424,
      "avg_duration_per_task_ms": 501990.4429292929,
      "avg_output_tokens_per_task": 30104.112626262628,
      "last_run_timestamp": "2026-10-10T08:30:49.776Z"
    },
    {
      "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": 0.8560605,
      "accuracy_std_dev": 0.012626499999999985,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.02547615909090909,
      "avg_duration_per_task_ms": 266390.93434343435,
      "avg_output_tokens_per_task": 21055.325757575756,
      "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-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "AkashML",
      "accuracy": 0.8557689653846154,
      "accuracy_std_dev": 0.016989351495046,
      "run_count": 3,
      "total_questions": 520,
      "avg_cost_per_task_usd": 0.06363424807692308,
      "avg_duration_per_task_ms": 459711.01153846155,
      "avg_output_tokens_per_task": 23017.521153846155,
      "last_run_timestamp": "2026-08-15T08:47:51.369Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.5-opus-20251124",
      "model_slug": "anthropic/claude-opus-4.5",
      "model_name": "Anthropic: Claude Opus 4.5",
      "provider_name": "Azure",
      "accuracy": 0.8557238,
      "accuracy_std_dev": 0.018226201572461534,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.8383144191919192,
      "avg_duration_per_task_ms": 453248.6974747475,
      "avg_output_tokens_per_task": 33360.03333333333,
      "last_run_timestamp": "2026-10-10T08:22:13.690Z"
    },
    {
      "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": 0.8555989032258063,
      "accuracy_std_dev": 0.02602705082664132,
      "run_count": 31,
      "total_questions": 6138,
      "avg_cost_per_task_usd": 0.17111695454545456,
      "avg_duration_per_task_ms": 157950.86656891496,
      "avg_output_tokens_per_task": 15997.16568914956,
      "last_run_timestamp": "2026-10-10T08:40:15.478Z"
    },
    {
      "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": 0.8554993333333333,
      "accuracy_std_dev": 0.02849418848740135,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.12659392845117845,
      "avg_duration_per_task_ms": 524253.6026936027,
      "avg_output_tokens_per_task": 33349.10774410774,
      "last_run_timestamp": "2026-10-10T16:20:22.055Z"
    },
    {
      "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": 0.8554712499999998,
      "accuracy_std_dev": 0.018079248695327472,
      "run_count": 20,
      "total_questions": 3960,
      "avg_cost_per_task_usd": 0.8025045977272727,
      "avg_duration_per_task_ms": 479096.1358585859,
      "avg_output_tokens_per_task": 30416.972222222223,
      "last_run_timestamp": "2026-10-10T08:24:12.220Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Cloudflare",
      "accuracy": 0.8552682396551725,
      "accuracy_std_dev": 0.0376233031656357,
      "run_count": 30,
      "total_questions": 5220,
      "avg_cost_per_task_usd": 0.013701367241379304,
      "avg_duration_per_task_ms": 406055.5988505747,
      "avg_output_tokens_per_task": 16990.452490421456,
      "last_run_timestamp": "2026-10-10T08:22:25.039Z"
    },
    {
      "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": 0.855219,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.19856715151515153,
      "avg_duration_per_task_ms": 638184.691919192,
      "avg_output_tokens_per_task": 58287.33838383838,
      "last_run_timestamp": "2026-07-18T08:26:48.377Z"
    },
    {
      "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": 0.8550479847715735,
      "accuracy_std_dev": 0.00436888279642389,
      "run_count": 3,
      "total_questions": 591,
      "avg_cost_per_task_usd": 0.24340484940778345,
      "avg_duration_per_task_ms": 425015.06429780036,
      "avg_output_tokens_per_task": 54319.2165820643,
      "last_run_timestamp": "2026-08-15T08:57:45.830Z"
    },
    {
      "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": 0.8550316666666666,
      "accuracy_std_dev": 0.029012228253311092,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.8331611588103255,
      "avg_duration_per_task_ms": 427437.11616161617,
      "avg_output_tokens_per_task": 33239.09371492705,
      "last_run_timestamp": "2026-10-10T08:21:42.444Z"
    },
    {
      "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": 0.854800172815534,
      "accuracy_std_dev": 0.012641847777693532,
      "run_count": 8,
      "total_questions": 1545,
      "avg_cost_per_task_usd": 0.1002615708737864,
      "avg_duration_per_task_ms": 668327.793527508,
      "avg_output_tokens_per_task": 36663.679611650485,
      "last_run_timestamp": "2026-09-19T08:55:10.654Z"
    },
    {
      "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": 0.854592,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 196,
      "avg_cost_per_task_usd": 0.08009020408163266,
      "avg_duration_per_task_ms": 1565820.9795918367,
      "avg_output_tokens_per_task": 40916.48469387755,
      "last_run_timestamp": "2026-07-18T16:22:39.277Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "auto-routing",
      "accuracy": 0.8545453999999999,
      "accuracy_std_dev": 0.008010288199559339,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.006016572727272727,
      "avg_duration_per_task_ms": 589898.7232323232,
      "avg_output_tokens_per_task": 28997.2,
      "last_run_timestamp": "2026-10-03T17:45:03.327Z"
    },
    {
      "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": 0.8543769166666666,
      "accuracy_std_dev": 0.015710205910269136,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.7564824368686868,
      "avg_duration_per_task_ms": 502571.60101010103,
      "avg_output_tokens_per_task": 30160.646464646463,
      "last_run_timestamp": "2026-10-10T08:22:14.712Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8543189326500733,
      "accuracy_std_dev": 0.02429014124777243,
      "run_count": 7,
      "total_questions": 1366,
      "avg_cost_per_task_usd": 0.13498622767203514,
      "avg_duration_per_task_ms": 368358.08931185945,
      "avg_output_tokens_per_task": 30688.598828696926,
      "last_run_timestamp": "2026-10-10T08:25:24.926Z"
    },
    {
      "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": 0.8542019060913706,
      "accuracy_std_dev": 0.006301576321753829,
      "run_count": 12,
      "total_questions": 2364,
      "avg_cost_per_task_usd": 0.05433487690355329,
      "avg_duration_per_task_ms": 423724.30245346867,
      "avg_output_tokens_per_task": 27804.93189509306,
      "last_run_timestamp": "2026-08-29T19:17:08.288Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Baidu",
      "accuracy": 0.8540081848101264,
      "accuracy_std_dev": 0.01478060943056438,
      "run_count": 6,
      "total_questions": 1185,
      "avg_cost_per_task_usd": 0.10410900675105485,
      "avg_duration_per_task_ms": 306238.13670886075,
      "avg_output_tokens_per_task": 29563.783966244726,
      "last_run_timestamp": "2026-10-10T08:34:20.699Z"
    },
    {
      "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": 0.8539943636363636,
      "accuracy_std_dev": 0.039917099808734806,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.17991296786042243,
      "avg_duration_per_task_ms": 45861.22451790634,
      "avg_output_tokens_per_task": 3391.032598714417,
      "last_run_timestamp": "2026-10-10T08:28:02.617Z"
    },
    {
      "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": 0.8539559999999999,
      "accuracy_std_dev": 0.062180043611274514,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.07175601767676767,
      "avg_duration_per_task_ms": 649147.8876262626,
      "avg_output_tokens_per_task": 27046.439393939392,
      "last_run_timestamp": "2026-10-10T16:30:18.009Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Near AI",
      "accuracy": 0.8537527991169976,
      "accuracy_std_dev": 0.017478527754711547,
      "run_count": 6,
      "total_questions": 906,
      "avg_cost_per_task_usd": 0.0038477715231788083,
      "avg_duration_per_task_ms": 380553.4061810154,
      "avg_output_tokens_per_task": 8747.8233995585,
      "last_run_timestamp": "2026-10-10T08:22:09.526Z"
    },
    {
      "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": 0.853535090909091,
      "accuracy_std_dev": 0.02817953127377952,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.15646832736455463,
      "avg_duration_per_task_ms": 152222.10376492195,
      "avg_output_tokens_per_task": 15431.129476584021,
      "last_run_timestamp": "2026-10-10T08:22:41.710Z"
    },
    {
      "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": 0.853478567104149,
      "accuracy_std_dev": 0.016412097348675454,
      "run_count": 25,
      "total_questions": 4724,
      "avg_cost_per_task_usd": 0.16086846613039799,
      "avg_duration_per_task_ms": 197709.6486028789,
      "avg_output_tokens_per_task": 28618.75550381033,
      "last_run_timestamp": "2026-10-10T08:24:01.170Z"
    },
    {
      "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": 0.0424615430210325,
      "avg_duration_per_task_ms": 655970.3365200765,
      "avg_output_tokens_per_task": 35467.44550669216,
      "last_run_timestamp": "2026-10-10T16:20:00.481Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "DekaLLM",
      "accuracy": 0.8530642646239555,
      "accuracy_std_dev": 0.022329477337983097,
      "run_count": 4,
      "total_questions": 718,
      "avg_cost_per_task_usd": 0.02860370194986072,
      "avg_duration_per_task_ms": 460751.2409470752,
      "avg_output_tokens_per_task": 34791.85933147632,
      "last_run_timestamp": "2026-10-10T08:22:41.628Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8529741666666667,
      "accuracy_std_dev": 0.015541427019600089,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.11713632154882155,
      "avg_duration_per_task_ms": 720822.8728956229,
      "avg_output_tokens_per_task": 37315.70791245791,
      "last_run_timestamp": "2026-10-10T16:20:03.022Z"
    },
    {
      "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": 0.8523328571428571,
      "accuracy_std_dev": 0.014361701465341489,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.10223797113997113,
      "avg_duration_per_task_ms": 552415.2619047619,
      "avg_output_tokens_per_task": 29613.10101010101,
      "last_run_timestamp": "2026-10-10T08:22:27.730Z"
    },
    {
      "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": 0.8523109090909091,
      "accuracy_std_dev": 0.012141784887917529,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.0287087217630854,
      "avg_duration_per_task_ms": 734317.4224058769,
      "avg_output_tokens_per_task": 32736.72819100092,
      "last_run_timestamp": "2026-10-10T08:34:24.655Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Sail Research",
      "accuracy": 0.8522881577868852,
      "accuracy_std_dev": 0.009664339658766143,
      "run_count": 5,
      "total_questions": 976,
      "avg_cost_per_task_usd": 0.09954490881147542,
      "avg_duration_per_task_ms": 487265.8411885246,
      "avg_output_tokens_per_task": 30959.861680327867,
      "last_run_timestamp": "2026-08-29T11:19:17.684Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "BaseTen",
      "accuracy": 0.851852,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 144,
      "avg_cost_per_task_usd": 0.15792222916666668,
      "avg_duration_per_task_ms": 798429.2847222222,
      "avg_output_tokens_per_task": 42935.319444444445,
      "last_run_timestamp": "2026-07-18T08:29:37.313Z"
    },
    {
      "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": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "DigitalOcean",
      "accuracy": 0.8513466999999999,
      "accuracy_std_dev": 0.07725069644611626,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.007058363131313131,
      "avg_duration_per_task_ms": 665246.3,
      "avg_output_tokens_per_task": 29067.647474747475,
      "last_run_timestamp": "2026-10-10T08:21:25.466Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "DigitalOcean",
      "accuracy": 0.8512543790322581,
      "accuracy_std_dev": 0.05785076520439789,
      "run_count": 8,
      "total_questions": 1488,
      "avg_cost_per_task_usd": 0.010551217069892472,
      "avg_duration_per_task_ms": 536964.1794354839,
      "avg_output_tokens_per_task": 20899.05376344086,
      "last_run_timestamp": "2026-10-10T08:22:13.781Z"
    },
    {
      "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": 0.8510100909090909,
      "accuracy_std_dev": 0.009650973153517579,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.03987833700642792,
      "avg_duration_per_task_ms": 393524.6414141414,
      "avg_output_tokens_per_task": 40088.10973370064,
      "last_run_timestamp": "2026-10-10T16:21:08.437Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Nebius",
      "accuracy": 0.8510099999999999,
      "accuracy_std_dev": 0.0025249999999999995,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.26341453282828287,
      "avg_duration_per_task_ms": 421660.03282828286,
      "avg_output_tokens_per_task": 59612.60101010101,
      "last_run_timestamp": "2026-10-10T08:15:00.041Z"
    },
    {
      "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": 0.8509659760869565,
      "accuracy_std_dev": 0.01639823188577327,
      "run_count": 7,
      "total_questions": 1380,
      "avg_cost_per_task_usd": 0.07163253550724637,
      "avg_duration_per_task_ms": 259347.84275362318,
      "avg_output_tokens_per_task": 18767.492753623188,
      "last_run_timestamp": "2026-10-10T08:35:07.242Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Mara",
      "accuracy": 0.850575,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 145,
      "avg_cost_per_task_usd": 0.015425489655172412,
      "avg_duration_per_task_ms": 62324.744827586204,
      "avg_output_tokens_per_task": 15913.096551724138,
      "last_run_timestamp": "2026-08-01T18:26:19.887Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Chutes",
      "accuracy": 0.850427,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 156,
      "avg_cost_per_task_usd": 0.014483525641025641,
      "avg_duration_per_task_ms": 276126.60256410256,
      "avg_output_tokens_per_task": 12035.96794871795,
      "last_run_timestamp": "2026-07-18T16:53:38.018Z"
    },
    {
      "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": 0.8504244789808917,
      "accuracy_std_dev": 0.0585455720762414,
      "run_count": 8,
      "total_questions": 1570,
      "avg_cost_per_task_usd": 0.0790043025477707,
      "avg_duration_per_task_ms": 541595.1828025478,
      "avg_output_tokens_per_task": 21736.728025477707,
      "last_run_timestamp": "2026-10-10T08:21:42.123Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5-2025-08-07",
      "model_slug": "openai/gpt-5",
      "model_name": "OpenAI: GPT-5",
      "provider_name": "OpenAI",
      "accuracy": 0.8501685000000001,
      "accuracy_std_dev": 0.0016835000000000044,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.20853091414141414,
      "avg_duration_per_task_ms": 280266.345959596,
      "avg_output_tokens_per_task": 20755.358585858587,
      "last_run_timestamp": "2026-10-01T16:20:44.419Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Decart",
      "accuracy": 0.8500727423580785,
      "accuracy_std_dev": 0.017073044892999246,
      "run_count": 7,
      "total_questions": 1374,
      "avg_cost_per_task_usd": 0.10693941775836974,
      "avg_duration_per_task_ms": 193416.326783115,
      "avg_output_tokens_per_task": 29417.200145560408,
      "last_run_timestamp": "2026-10-10T08:26:05.435Z"
    },
    {
      "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": 0.8496872857142856,
      "accuracy_std_dev": 0.02328767377645021,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.014274932178932179,
      "avg_duration_per_task_ms": 598242.8015873015,
      "avg_output_tokens_per_task": 35895.869408369406,
      "last_run_timestamp": "2026-10-10T08:11:18.084Z"
    },
    {
      "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": "Together",
      "accuracy": 0.8495985437262357,
      "accuracy_std_dev": 0.01598636257095404,
      "run_count": 8,
      "total_questions": 1578,
      "avg_cost_per_task_usd": 0.01962292712294043,
      "avg_duration_per_task_ms": 346902.4087452472,
      "avg_output_tokens_per_task": 39034.471482889734,
      "last_run_timestamp": "2026-10-10T08:22:48.052Z"
    },
    {
      "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": 0.8494668333333334,
      "accuracy_std_dev": 0.08246985866245651,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.09358024705387205,
      "avg_duration_per_task_ms": 746941.3282828283,
      "avg_output_tokens_per_task": 41492.88888888889,
      "last_run_timestamp": "2026-10-10T08:22:24.596Z"
    },
    {
      "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": 0.8493871292775664,
      "accuracy_std_dev": 0.028830055092028094,
      "run_count": 8,
      "total_questions": 1578,
      "avg_cost_per_task_usd": 0.012782455006337136,
      "avg_duration_per_task_ms": 567459.4005069708,
      "avg_output_tokens_per_task": 36588.9752851711,
      "last_run_timestamp": "2026-10-10T08:24:22.181Z"
    },
    {
      "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": 0.849375142857143,
      "accuracy_std_dev": 0.05373653913902533,
      "run_count": 5,
      "total_questions": 987,
      "avg_cost_per_task_usd": 0.09692596960486322,
      "avg_duration_per_task_ms": 658494.2613981763,
      "avg_output_tokens_per_task": 27042.529888551166,
      "last_run_timestamp": "2026-08-29T11:31:24.094Z"
    },
    {
      "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": 0.8493666190476191,
      "accuracy_std_dev": 0.05184374805167303,
      "run_count": 21,
      "total_questions": 4158,
      "avg_cost_per_task_usd": 0.1855372905242905,
      "avg_duration_per_task_ms": 40993.36219336219,
      "avg_output_tokens_per_task": 3341.2967772967772,
      "last_run_timestamp": "2026-10-10T08:22:13.823Z"
    },
    {
      "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": 0.8493151205479452,
      "accuracy_std_dev": 0.01972059190006447,
      "run_count": 3,
      "total_questions": 365,
      "avg_cost_per_task_usd": 0.06587333698630138,
      "avg_duration_per_task_ms": 279697.01369863015,
      "avg_output_tokens_per_task": 16490.610958904108,
      "last_run_timestamp": "2026-08-15T08:47:41.523Z"
    },
    {
      "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": 0.8490420331034483,
      "accuracy_std_dev": 0.010617213932381244,
      "run_count": 11,
      "total_questions": 2175,
      "avg_cost_per_task_usd": 0.030508747126436777,
      "avg_duration_per_task_ms": 708403.5742528735,
      "avg_output_tokens_per_task": 31823.873563218393,
      "last_run_timestamp": "2026-10-10T08:22:26.996Z"
    },
    {
      "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": 0.848969343283582,
      "accuracy_std_dev": 0.009530391676106496,
      "run_count": 12,
      "total_questions": 2345,
      "avg_cost_per_task_usd": 0.07047954541577824,
      "avg_duration_per_task_ms": 213416.39104477613,
      "avg_output_tokens_per_task": 21924.39616204691,
      "last_run_timestamp": "2026-10-10T08:14:48.985Z"
    },
    {
      "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": 0.8487653333333334,
      "accuracy_std_dev": 0.03509082742762781,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.015535906565656565,
      "avg_duration_per_task_ms": 365861.50505050505,
      "avg_output_tokens_per_task": 27936.9234006734,
      "last_run_timestamp": "2026-10-10T16:10:44.131Z"
    },
    {
      "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": 0.8486160259927799,
      "accuracy_std_dev": 0.03690580233015742,
      "run_count": 7,
      "total_questions": 1385,
      "avg_cost_per_task_usd": 0.16769674440433213,
      "avg_duration_per_task_ms": 1001771.3018050542,
      "avg_output_tokens_per_task": 54149.948736462095,
      "last_run_timestamp": "2026-10-10T16:21:33.712Z"
    },
    {
      "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": "Parasail",
      "accuracy": 0.8485233052899935,
      "accuracy_std_dev": 0.011509616348949264,
      "run_count": 8,
      "total_questions": 1569,
      "avg_cost_per_task_usd": 0.006999760994263861,
      "avg_duration_per_task_ms": 550728.0395156151,
      "avg_output_tokens_per_task": 17320.901848311027,
      "last_run_timestamp": "2026-10-10T16:19:34.420Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Fireworks",
      "accuracy": 0.84850889,
      "accuracy_std_dev": 0.013594201563902184,
      "run_count": 10,
      "total_questions": 1900,
      "avg_cost_per_task_usd": 0.01998638105263158,
      "avg_duration_per_task_ms": 280445.96315789473,
      "avg_output_tokens_per_task": 36076.418947368424,
      "last_run_timestamp": "2026-10-10T08:23:11.162Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "inclusionai/ling-3.0-flash-vl-20260910",
      "model_slug": "inclusionai/ling-3.0-flash-vl",
      "model_name": "inclusionAI: Ling 3.0 Flash VL",
      "provider_name": "auto-routing",
      "accuracy": 0.848485,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0012582575757575758,
      "avg_duration_per_task_ms": 88400.0101010101,
      "avg_output_tokens_per_task": 17501.79797979798,
      "last_run_timestamp": "2026-10-01T08:22:54.644Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Fireworks",
      "accuracy": 0.8484668985163207,
      "accuracy_std_dev": 0.020724910315372824,
      "run_count": 9,
      "total_questions": 1685,
      "avg_cost_per_task_usd": 0.13222807833827893,
      "avg_duration_per_task_ms": 291436.95014836796,
      "avg_output_tokens_per_task": 26711.21543026706,
      "last_run_timestamp": "2026-10-10T08:28:24.294Z"
    },
    {
      "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": 0.8484593299406276,
      "accuracy_std_dev": 0.0127989667296598,
      "run_count": 7,
      "total_questions": 1179,
      "avg_cost_per_task_usd": 0.006712759966072943,
      "avg_duration_per_task_ms": 475640.88295165397,
      "avg_output_tokens_per_task": 16540.466497031382,
      "last_run_timestamp": "2026-10-10T16:22:19.237Z"
    },
    {
      "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": 0.8484146666666667,
      "accuracy_std_dev": 0.01925144376115435,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.11280877988215487,
      "avg_duration_per_task_ms": 573583.9625420875,
      "avg_output_tokens_per_task": 45058.854797979795,
      "last_run_timestamp": "2026-10-10T08:11:17.024Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Decart",
      "accuracy": 0.8482337114256125,
      "accuracy_std_dev": 0.02225972291072427,
      "run_count": 15,
      "total_questions": 2897,
      "avg_cost_per_task_usd": 0.178282096306524,
      "avg_duration_per_task_ms": 293597.77562996204,
      "avg_output_tokens_per_task": 43073.98826372109,
      "last_run_timestamp": "2026-10-10T08:24:31.407Z"
    },
    {
      "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": 0.8481106666666667,
      "accuracy_std_dev": 0.036070316075379444,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.7001586167227833,
      "avg_duration_per_task_ms": 475392.94332211,
      "avg_output_tokens_per_task": 28044.415263748597,
      "last_run_timestamp": "2026-10-10T08:22:31.415Z"
    },
    {
      "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": 0.8477956427969672,
      "accuracy_std_dev": 0.03319167551259403,
      "run_count": 6,
      "total_questions": 1187,
      "avg_cost_per_task_usd": 0.09577104128053918,
      "avg_duration_per_task_ms": 1415253.807919124,
      "avg_output_tokens_per_task": 42907.37742207245,
      "last_run_timestamp": "2026-08-29T18:36:10.818Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8477632857142857,
      "accuracy_std_dev": 0.008091757370740825,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.010911548340548341,
      "avg_duration_per_task_ms": 434100.7842712843,
      "avg_output_tokens_per_task": 18528.194083694085,
      "last_run_timestamp": "2026-10-10T16:21:25.762Z"
    },
    {
      "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": 0.8476972604562738,
      "accuracy_std_dev": 0.016172357759461055,
      "run_count": 8,
      "total_questions": 1578,
      "avg_cost_per_task_usd": 0.0376128732572877,
      "avg_duration_per_task_ms": 613711.562103929,
      "avg_output_tokens_per_task": 35115.68567807351,
      "last_run_timestamp": "2026-10-10T16:19:34.452Z"
    },
    {
      "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": 0.8474194046372475,
      "accuracy_std_dev": 0.028897507729553278,
      "run_count": 7,
      "total_questions": 1337,
      "avg_cost_per_task_usd": 0.1101656335078534,
      "avg_duration_per_task_ms": 533472.9222139118,
      "avg_output_tokens_per_task": 33787.17053103964,
      "last_run_timestamp": "2026-10-10T08:35:43.320Z"
    },
    {
      "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": 0.8472713558033451,
      "accuracy_std_dev": 0.026110591357723027,
      "run_count": 10,
      "total_questions": 1973,
      "avg_cost_per_task_usd": 0.00887132336543335,
      "avg_duration_per_task_ms": 219350.43081601622,
      "avg_output_tokens_per_task": 13273.467308667005,
      "last_run_timestamp": "2026-10-10T08:23:49.870Z"
    },
    {
      "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": 0.8471839090909092,
      "accuracy_std_dev": 0.009929703716668814,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.09975296602387511,
      "avg_duration_per_task_ms": 580036.968778696,
      "avg_output_tokens_per_task": 31307.20523415978,
      "last_run_timestamp": "2026-10-10T16:20:22.049Z"
    },
    {
      "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": 0.846689,
      "accuracy_std_dev": 0.010530862908612949,
      "run_count": 15,
      "total_questions": 2970,
      "avg_cost_per_task_usd": 0.14702261952861953,
      "avg_duration_per_task_ms": 151760.5090909091,
      "avg_output_tokens_per_task": 32387.174074074075,
      "last_run_timestamp": "2026-10-10T08:29:41.483Z"
    },
    {
      "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": 0.8461861403286979,
      "accuracy_std_dev": 0.016702053748258686,
      "run_count": 12,
      "total_questions": 2373,
      "avg_cost_per_task_usd": 0.05117073999157185,
      "avg_duration_per_task_ms": 277147.41508638853,
      "avg_output_tokens_per_task": 45853.79603876949,
      "last_run_timestamp": "2026-10-10T16:20:22.056Z"
    },
    {
      "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": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Venice",
      "accuracy": 0.846084,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 183,
      "avg_cost_per_task_usd": 0.0059167213114754095,
      "avg_duration_per_task_ms": 533061.174863388,
      "avg_output_tokens_per_task": 16599.48087431694,
      "last_run_timestamp": "2026-10-10T08:22:47.742Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "auto-routing",
      "accuracy": 0.8459596249999999,
      "accuracy_std_dev": 0.02667178620254698,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.07075826704545456,
      "avg_duration_per_task_ms": 553373.509469697,
      "avg_output_tokens_per_task": 22211.534090909092,
      "last_run_timestamp": "2026-10-10T16:10:30.028Z"
    },
    {
      "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": 0.845959375,
      "accuracy_std_dev": 0.012876192798509,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.10514613573232325,
      "avg_duration_per_task_ms": 332982.2455808081,
      "avg_output_tokens_per_task": 29278.01388888889,
      "last_run_timestamp": "2026-10-10T08:14:49.956Z"
    },
    {
      "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": 0.8456079554432706,
      "accuracy_std_dev": 0.03356297487340206,
      "run_count": 33,
      "total_questions": 6531,
      "avg_cost_per_task_usd": 0.16561795467769108,
      "avg_duration_per_task_ms": 150136.5763282805,
      "avg_output_tokens_per_task": 15309.536211912418,
      "last_run_timestamp": "2026-10-10T08:28:02.729Z"
    },
    {
      "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": 0.8455446459920203,
      "accuracy_std_dev": 0.010937911610913816,
      "run_count": 14,
      "total_questions": 2757,
      "avg_cost_per_task_usd": 0.12664673558215453,
      "avg_duration_per_task_ms": 332593.6597751179,
      "avg_output_tokens_per_task": 33927.599927457384,
      "last_run_timestamp": "2026-10-10T08:26:05.194Z"
    },
    {
      "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": 0.8450010520231215,
      "accuracy_std_dev": 0.013556052895639447,
      "run_count": 8,
      "total_questions": 1557,
      "avg_cost_per_task_usd": 0.0843938554913295,
      "avg_duration_per_task_ms": 983145.6628131021,
      "avg_output_tokens_per_task": 42748.55427103404,
      "last_run_timestamp": "2026-10-10T16:20:24.762Z"
    },
    {
      "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": 0.8449914928670043,
      "accuracy_std_dev": 0.014523064670429856,
      "run_count": 11,
      "total_questions": 2173,
      "avg_cost_per_task_usd": 0.0884558996778647,
      "avg_duration_per_task_ms": 407622.08145421074,
      "avg_output_tokens_per_task": 21997.700414173953,
      "last_run_timestamp": "2026-10-10T08:27:34.920Z"
    },
    {
      "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": 0.8448371666666666,
      "accuracy_std_dev": 0.007861279696857409,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.1419990378787879,
      "avg_duration_per_task_ms": 155301.35437710438,
      "avg_output_tokens_per_task": 32100.082491582492,
      "last_run_timestamp": "2026-10-10T08:14:47.788Z"
    },
    {
      "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": 0.8447187666034155,
      "accuracy_std_dev": 0.033801436872417136,
      "run_count": 5,
      "total_questions": 527,
      "avg_cost_per_task_usd": 0.10369237191650856,
      "avg_duration_per_task_ms": 295443.4819734345,
      "avg_output_tokens_per_task": 23792.745730550283,
      "last_run_timestamp": "2026-08-22T17:29:49.714Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-122b-a10b-20260224",
      "model_slug": "qwen/qwen3.5-122b-a10b",
      "model_name": "Qwen: Qwen3.5-122B-A10B",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8445006460176991,
      "accuracy_std_dev": 0.03377447178960267,
      "run_count": 12,
      "total_questions": 2373,
      "avg_cost_per_task_usd": 0.05367159629161399,
      "avg_duration_per_task_ms": 217527.86177833966,
      "avg_output_tokens_per_task": 22261.738727349348,
      "last_run_timestamp": "2026-10-10T08:22:49.962Z"
    },
    {
      "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": "AtlasCloud",
      "accuracy": 0.8444865,
      "accuracy_std_dev": 0.01261755690496381,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.019410517676767675,
      "avg_duration_per_task_ms": 420835.0101010101,
      "avg_output_tokens_per_task": 38594.204545454544,
      "last_run_timestamp": "2026-10-10T08:22:11.854Z"
    },
    {
      "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": 0.8444072953657173,
      "accuracy_std_dev": 0.011477740217046216,
      "run_count": 10,
      "total_questions": 1791,
      "avg_cost_per_task_usd": 0.015475027359017308,
      "avg_duration_per_task_ms": 603617.405360134,
      "avg_output_tokens_per_task": 16554.757118927973,
      "last_run_timestamp": "2026-08-29T18:29:51.235Z"
    },
    {
      "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": 0.8436994003795065,
      "accuracy_std_dev": 0.021313050412206686,
      "run_count": 24,
      "total_questions": 4743,
      "avg_cost_per_task_usd": 0.2397977305502846,
      "avg_duration_per_task_ms": 521390.0229812355,
      "avg_output_tokens_per_task": 46489.35800126502,
      "last_run_timestamp": "2026-10-10T08:26:02.713Z"
    },
    {
      "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": 0.8436971598984772,
      "accuracy_std_dev": 0.01736385407066934,
      "run_count": 8,
      "total_questions": 1576,
      "avg_cost_per_task_usd": 0.00971490228426396,
      "avg_duration_per_task_ms": 459655.7252538071,
      "avg_output_tokens_per_task": 30262.047588832487,
      "last_run_timestamp": "2026-10-10T08:38:35.477Z"
    },
    {
      "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": 0.8436747142857144,
      "accuracy_std_dev": 0.038362541267358635,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.1561614963924964,
      "avg_duration_per_task_ms": 1365496.6435786437,
      "avg_output_tokens_per_task": 51685.106782106785,
      "last_run_timestamp": "2026-10-10T16:24:14.682Z"
    },
    {
      "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": 0.8435435936936937,
      "accuracy_std_dev": 0.023817522835089276,
      "run_count": 8,
      "total_questions": 1110,
      "avg_cost_per_task_usd": 0.015176712612612612,
      "avg_duration_per_task_ms": 82041.15225225226,
      "avg_output_tokens_per_task": 13011.012612612612,
      "last_run_timestamp": "2026-10-10T16:24:45.975Z"
    },
    {
      "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": 0.8433861695238094,
      "accuracy_std_dev": 0.02133976042178954,
      "run_count": 8,
      "total_questions": 1575,
      "avg_cost_per_task_usd": 0.04683724698412699,
      "avg_duration_per_task_ms": 634662.0095238095,
      "avg_output_tokens_per_task": 37604.758095238096,
      "last_run_timestamp": "2026-10-10T16:20:00.384Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8432812727272728,
      "accuracy_std_dev": 0.012668660387607238,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.1344134775022957,
      "avg_duration_per_task_ms": 424945.45224977046,
      "avg_output_tokens_per_task": 37266.27272727273,
      "last_run_timestamp": "2026-10-10T08:22:18.320Z"
    },
    {
      "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": 0.843273301664355,
      "accuracy_std_dev": 0.026075377685423836,
      "run_count": 8,
      "total_questions": 1442,
      "avg_cost_per_task_usd": 0.010784043689320387,
      "avg_duration_per_task_ms": 367303.58252427186,
      "avg_output_tokens_per_task": 21354.485436893203,
      "last_run_timestamp": "2026-10-10T08:28:19.345Z"
    },
    {
      "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": 0.8431938571428572,
      "accuracy_std_dev": 0.04476285754771171,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.11224030519480518,
      "avg_duration_per_task_ms": 1310504.1515151516,
      "avg_output_tokens_per_task": 44998.9531024531,
      "last_run_timestamp": "2026-10-10T16:11:11.295Z"
    },
    {
      "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": 0.8430194788376499,
      "accuracy_std_dev": 0.012894642985921526,
      "run_count": 8,
      "total_questions": 1583,
      "avg_cost_per_task_usd": 0.1371234333543904,
      "avg_duration_per_task_ms": 198575.00252684776,
      "avg_output_tokens_per_task": 30938.104232469992,
      "last_run_timestamp": "2026-10-10T08:24:22.033Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.8426664382222222,
      "accuracy_std_dev": 0.00961840561483182,
      "run_count": 7,
      "total_questions": 1125,
      "avg_cost_per_task_usd": 0.1371568737777778,
      "avg_duration_per_task_ms": 125791.95822222222,
      "avg_output_tokens_per_task": 30936.946666666667,
      "last_run_timestamp": "2026-10-10T08:24:22.575Z"
    },
    {
      "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": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Inceptron",
      "accuracy": 0.8419940950570343,
      "accuracy_std_dev": 0.005052124553096465,
      "run_count": 4,
      "total_questions": 789,
      "avg_cost_per_task_usd": 0.02034613561470215,
      "avg_duration_per_task_ms": 621368.3954372624,
      "avg_output_tokens_per_task": 23198.26235741445,
      "last_run_timestamp": "2026-08-15T14:57:08.773Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Together",
      "accuracy": 0.8414709276830954,
      "accuracy_std_dev": 0.01569670882911814,
      "run_count": 11,
      "total_questions": 2171,
      "avg_cost_per_task_usd": 0.21642636665131276,
      "avg_duration_per_task_ms": 299666.9097190235,
      "avg_output_tokens_per_task": 49086.817134960846,
      "last_run_timestamp": "2026-10-10T08:18:24.473Z"
    },
    {
      "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/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Novita",
      "accuracy": 0.8412499316666666,
      "accuracy_std_dev": 0.012659388006293123,
      "run_count": 7,
      "total_questions": 1200,
      "avg_cost_per_task_usd": 0.0068970816666666665,
      "avg_duration_per_task_ms": 636725.1641666667,
      "avg_output_tokens_per_task": 17029.4825,
      "last_run_timestamp": "2026-10-10T16:23:21.036Z"
    },
    {
      "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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Novita",
      "accuracy": 0.8408532168508287,
      "accuracy_std_dev": 0.05657006571975314,
      "run_count": 11,
      "total_questions": 2172,
      "avg_cost_per_task_usd": 0.07583489180478822,
      "avg_duration_per_task_ms": 651639.9650092081,
      "avg_output_tokens_per_task": 45206.226979742176,
      "last_run_timestamp": "2026-10-10T08:22:51.094Z"
    },
    {
      "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": 0.8407408,
      "accuracy_std_dev": 0.1041355491182526,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.11180628585858585,
      "avg_duration_per_task_ms": 1526581.008080808,
      "avg_output_tokens_per_task": 27815.140404040405,
      "last_run_timestamp": "2026-10-10T08:36:30.960Z"
    },
    {
      "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": 0.8407406999999999,
      "accuracy_std_dev": 0.01043760000239519,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.07050250505050504,
      "avg_duration_per_task_ms": 363872.355050505,
      "avg_output_tokens_per_task": 22481.775757575757,
      "last_run_timestamp": "2026-10-10T08:46:09.054Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.4-mini-20260317",
      "model_slug": "openai/gpt-5.4-mini",
      "model_name": "OpenAI: GPT-5.4 Mini",
      "provider_name": "OpenAI",
      "accuracy": 0.84048825,
      "accuracy_std_dev": 0.004667830993887862,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.06197197222222223,
      "avg_duration_per_task_ms": 84839.70707070707,
      "avg_output_tokens_per_task": 13643.420454545454,
      "last_run_timestamp": "2026-10-02T23:24:30.140Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.5-flash-lite-20260721",
      "model_slug": "google/gemini-3.5-flash-lite",
      "model_name": "Google: Gemini 3.5 Flash Lite",
      "provider_name": "Google",
      "accuracy": 0.8402775531250001,
      "accuracy_std_dev": 0.007466846475587956,
      "run_count": 10,
      "total_questions": 1920,
      "avg_cost_per_task_usd": 0.039408469791666664,
      "avg_duration_per_task_ms": 51791.722395833334,
      "avg_output_tokens_per_task": 14757.431770833333,
      "last_run_timestamp": "2026-10-01T19:13:57.987Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "thinkingmachines/inkling-20260715",
      "model_slug": "thinkingmachines/inkling",
      "model_name": "Thinking Machines: Inkling",
      "provider_name": "DeepInfra",
      "accuracy": 0.8402425498181818,
      "accuracy_std_dev": 0.00976753775827742,
      "run_count": 8,
      "total_questions": 1375,
      "avg_cost_per_task_usd": 0.0796483170909091,
      "avg_duration_per_task_ms": 509723.63709090906,
      "avg_output_tokens_per_task": 19683.845818181817,
      "last_run_timestamp": "2026-10-08T22:17:54.321Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "GMICloud",
      "accuracy": 0.8401708461538463,
      "accuracy_std_dev": 0.01634594423702712,
      "run_count": 8,
      "total_questions": 1560,
      "avg_cost_per_task_usd": 0.03290970448717949,
      "avg_duration_per_task_ms": 545901.551923077,
      "avg_output_tokens_per_task": 32901.243589743586,
      "last_run_timestamp": "2026-10-10T16:25:08.026Z"
    },
    {
      "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": 0.8399772366412214,
      "accuracy_std_dev": 0.010642150299634022,
      "run_count": 6,
      "total_questions": 1179,
      "avg_cost_per_task_usd": 0.02758528159457167,
      "avg_duration_per_task_ms": 524755.5767599661,
      "avg_output_tokens_per_task": 22800.932994062765,
      "last_run_timestamp": "2026-08-29T18:16:39.840Z"
    },
    {
      "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": 0.8398466400000001,
      "accuracy_std_dev": 0.04372123622631255,
      "run_count": 11,
      "total_questions": 2175,
      "avg_cost_per_task_usd": 0.16602242022988506,
      "avg_duration_per_task_ms": 162702.748045977,
      "avg_output_tokens_per_task": 16381.855172413792,
      "last_run_timestamp": "2026-10-10T08:14:40.728Z"
    },
    {
      "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": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Phala",
      "accuracy": 0.8394795097613883,
      "accuracy_std_dev": 0.06253985350713477,
      "run_count": 7,
      "total_questions": 1383,
      "avg_cost_per_task_usd": 0.2298788691250904,
      "avg_duration_per_task_ms": 1437430.4526391902,
      "avg_output_tokens_per_task": 53441.45336225597,
      "last_run_timestamp": "2026-10-10T16:20:12.051Z"
    },
    {
      "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": 0.8393612296564196,
      "accuracy_std_dev": 0.01918273370044586,
      "run_count": 6,
      "total_questions": 1106,
      "avg_cost_per_task_usd": 0.1468148047016275,
      "avg_duration_per_task_ms": 272284.95840867993,
      "avg_output_tokens_per_task": 34433.43580470163,
      "last_run_timestamp": "2026-10-10T08:21:50.326Z"
    },
    {
      "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": 0.8391883298001213,
      "accuracy_std_dev": 0.103275715403127,
      "run_count": 17,
      "total_questions": 3302,
      "avg_cost_per_task_usd": 0.008241699273167776,
      "avg_duration_per_task_ms": 633419.09812235,
      "avg_output_tokens_per_task": 27504.60357359176,
      "last_run_timestamp": "2026-10-10T08:22:05.598Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Friendli",
      "accuracy": 0.8391142192408377,
      "accuracy_std_dev": 0.016722181240421387,
      "run_count": 8,
      "total_questions": 1528,
      "avg_cost_per_task_usd": 0.019121354712041884,
      "avg_duration_per_task_ms": 242204.53206806284,
      "avg_output_tokens_per_task": 38024.85013089005,
      "last_run_timestamp": "2026-10-10T08:22:13.945Z"
    },
    {
      "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": 0.8388045000000002,
      "accuracy_std_dev": 0.02728541628782673,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.14381466666666667,
      "avg_duration_per_task_ms": 923603.7752525252,
      "avg_output_tokens_per_task": 54078.98484848485,
      "last_run_timestamp": "2026-10-10T16:23:01.146Z"
    },
    {
      "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": 0.838739,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 185,
      "avg_cost_per_task_usd": 0.0289266,
      "avg_duration_per_task_ms": 182176.34594594594,
      "avg_output_tokens_per_task": 24823.56216216216,
      "last_run_timestamp": "2026-07-18T08:31:24.271Z"
    },
    {
      "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": 0.8386610687977762,
      "accuracy_std_dev": 0.022495213007090526,
      "run_count": 8,
      "total_questions": 1439,
      "avg_cost_per_task_usd": 0.07294077206393328,
      "avg_duration_per_task_ms": 384771.50799166085,
      "avg_output_tokens_per_task": 24034.94023627519,
      "last_run_timestamp": "2026-10-10T08:21:47.164Z"
    },
    {
      "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": 0.8384554097714735,
      "accuracy_std_dev": 0.027250356683679502,
      "run_count": 7,
      "total_questions": 1269,
      "avg_cost_per_task_usd": 0.0903084074074074,
      "avg_duration_per_task_ms": 1604100.3364854215,
      "avg_output_tokens_per_task": 41888.97872340425,
      "last_run_timestamp": "2026-10-10T16:21:58.069Z"
    },
    {
      "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": 0.8383456936090224,
      "accuracy_std_dev": 0.05167297206793249,
      "run_count": 6,
      "total_questions": 1064,
      "avg_cost_per_task_usd": 0.17379162406015036,
      "avg_duration_per_task_ms": 1701245.0780075188,
      "avg_output_tokens_per_task": 51051.67105263158,
      "last_run_timestamp": "2026-10-10T16:20:06.404Z"
    },
    {
      "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": 0.8379778156484459,
      "accuracy_std_dev": 0.014079857290552125,
      "run_count": 11,
      "total_questions": 1866,
      "avg_cost_per_task_usd": 0.09540400214362274,
      "avg_duration_per_task_ms": 623450.2524115755,
      "avg_output_tokens_per_task": 29739.112004287246,
      "last_run_timestamp": "2026-10-10T16:21:21.249Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "InferenceNet",
      "accuracy": 0.8378226666666666,
      "accuracy_std_dev": 0.00975199857579064,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.15828378956228956,
      "avg_duration_per_task_ms": 446608.7861952862,
      "avg_output_tokens_per_task": 56298.93097643097,
      "last_run_timestamp": "2026-10-10T08:30:28.864Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Parasail",
      "accuracy": 0.8377989469160769,
      "accuracy_std_dev": 0.022618776770639035,
      "run_count": 10,
      "total_questions": 1978,
      "avg_cost_per_task_usd": 0.023259668857431744,
      "avg_duration_per_task_ms": 408040.1359959555,
      "avg_output_tokens_per_task": 43341.32406471183,
      "last_run_timestamp": "2026-10-10T08:39:09.098Z"
    },
    {
      "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": 0.837571,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 118,
      "avg_cost_per_task_usd": 0.01632071186440678,
      "avg_duration_per_task_ms": 269506.5254237288,
      "avg_output_tokens_per_task": 11788.906779661016,
      "last_run_timestamp": "2026-07-18T16:47:35.300Z"
    },
    {
      "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": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Novita",
      "accuracy": 0.836910625,
      "accuracy_std_dev": 0.02484845377954887,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.11375521527777778,
      "avg_duration_per_task_ms": 402363.48232323234,
      "avg_output_tokens_per_task": 35795.318181818184,
      "last_run_timestamp": "2026-10-10T08:22:05.690Z"
    },
    {
      "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": 0.8368532727272727,
      "accuracy_std_dev": 0.07122412090410091,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.195708673553719,
      "avg_duration_per_task_ms": 540689.1230486685,
      "avg_output_tokens_per_task": 44503.561983471074,
      "last_run_timestamp": "2026-10-10T08:23:29.599Z"
    },
    {
      "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": 0.8367697340206185,
      "accuracy_std_dev": 0.008932455147382513,
      "run_count": 5,
      "total_questions": 970,
      "avg_cost_per_task_usd": 0.027574227835051545,
      "avg_duration_per_task_ms": 610547.7175257732,
      "avg_output_tokens_per_task": 14641.665979381443,
      "last_run_timestamp": "2026-10-10T08:23:40.323Z"
    },
    {
      "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": 0.8367169999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 197,
      "avg_cost_per_task_usd": 0.10272846700507614,
      "avg_duration_per_task_ms": 406781.3959390863,
      "avg_output_tokens_per_task": 31764.908629441623,
      "last_run_timestamp": "2026-07-18T08:30:34.727Z"
    },
    {
      "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": 0.8367000000000001,
      "accuracy_std_dev": 0.006734000000000018,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.02250940404040404,
      "avg_duration_per_task_ms": 13191.318181818182,
      "avg_output_tokens_per_task": 14558.808080808081,
      "last_run_timestamp": "2026-10-10T08:14:48.050Z"
    },
    {
      "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": 0.8365800000000001,
      "accuracy_std_dev": 0.012840289382586837,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.006623383116883117,
      "avg_duration_per_task_ms": 371056.68975468975,
      "avg_output_tokens_per_task": 26074.506493506495,
      "last_run_timestamp": "2026-10-10T16:19:34.423Z"
    },
    {
      "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": 0.8364249544688027,
      "accuracy_std_dev": 0.025478198310237742,
      "run_count": 9,
      "total_questions": 1779,
      "avg_cost_per_task_usd": 0.15890397301854978,
      "avg_duration_per_task_ms": 937364.9814502529,
      "avg_output_tokens_per_task": 52191.70320404722,
      "last_run_timestamp": "2026-10-10T16:25:51.190Z"
    },
    {
      "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": 0.8363261111111111,
      "accuracy_std_dev": 0.010093985221621908,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.24586164983164982,
      "avg_duration_per_task_ms": 359650.97138047137,
      "avg_output_tokens_per_task": 40134.10662177329,
      "last_run_timestamp": "2026-10-10T08:34:27.150Z"
    },
    {
      "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": 0.8362795,
      "accuracy_std_dev": 0.015277669703961197,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.07403895328282828,
      "avg_duration_per_task_ms": 689404.1414141414,
      "avg_output_tokens_per_task": 53986.75673400673,
      "last_run_timestamp": "2026-10-10T08:28:33.489Z"
    },
    {
      "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": "inclusionai/ling-3.0-flash-vl-20260910",
      "model_slug": "inclusionai/ling-3.0-flash-vl",
      "model_name": "inclusionAI: Ling 3.0 Flash VL",
      "provider_name": "Novita",
      "accuracy": 0.8358585,
      "accuracy_std_dev": 0.01430950000000003,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.0011451691919191918,
      "avg_duration_per_task_ms": 116299.33333333333,
      "avg_output_tokens_per_task": 18355.219696969696,
      "last_run_timestamp": "2026-10-01T08:23:48.135Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Cloudflare",
      "accuracy": 0.8353882517845557,
      "accuracy_std_dev": 0.03954502959917974,
      "run_count": 8,
      "total_questions": 1541,
      "avg_cost_per_task_usd": 0.02963127903958469,
      "avg_duration_per_task_ms": 279716.5554834523,
      "avg_output_tokens_per_task": 9152.79299156392,
      "last_run_timestamp": "2026-10-10T08:50:00.756Z"
    },
    {
      "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": 0.8353515086042066,
      "accuracy_std_dev": 0.02210739585947427,
      "run_count": 8,
      "total_questions": 1569,
      "avg_cost_per_task_usd": 0.09059525812619502,
      "avg_duration_per_task_ms": 1011217.5761631612,
      "avg_output_tokens_per_task": 41548.43977055449,
      "last_run_timestamp": "2026-10-10T16:11:14.320Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "inclusionai/ling-3.0-flash-vl-20260910",
      "model_slug": "inclusionai/ling-3.0-flash-vl",
      "model_name": "inclusionAI: Ling 3.0 Flash VL",
      "provider_name": "DeepInfra",
      "accuracy": 0.8351849999999998,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 180,
      "avg_cost_per_task_usd": 0.0030483999999999997,
      "avg_duration_per_task_ms": 288802.22222222225,
      "avg_output_tokens_per_task": 16703.5,
      "last_run_timestamp": "2026-10-01T08:43:11.415Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Friendli",
      "accuracy": 0.8351813548597521,
      "accuracy_std_dev": 0.013967532009539117,
      "run_count": 8,
      "total_questions": 1533,
      "avg_cost_per_task_usd": 0.22283928962818,
      "avg_duration_per_task_ms": 428460.4801043705,
      "avg_output_tokens_per_task": 50510.284409654276,
      "last_run_timestamp": "2026-10-10T16:19:37.318Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Mistral",
      "accuracy": 0.8351770476190475,
      "accuracy_std_dev": 0.01163942668558729,
      "run_count": 21,
      "total_questions": 4158,
      "avg_cost_per_task_usd": 0.2585508424723425,
      "avg_duration_per_task_ms": 272269.33116883115,
      "avg_output_tokens_per_task": 56376.68253968254,
      "last_run_timestamp": "2026-10-10T08:26:21.636Z"
    },
    {
      "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": 0.8350167777777778,
      "accuracy_std_dev": 0.012294439174573386,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.03943477890011223,
      "avg_duration_per_task_ms": 224321.7744107744,
      "avg_output_tokens_per_task": 14566.167227833894,
      "last_run_timestamp": "2026-10-10T08:14:46.285Z"
    },
    {
      "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": 0.8350167272727272,
      "accuracy_std_dev": 0.011463201140177447,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.08175606887052343,
      "avg_duration_per_task_ms": 770659.5785123967,
      "avg_output_tokens_per_task": 48084.065197428834,
      "last_run_timestamp": "2026-10-10T16:30:58.078Z"
    },
    {
      "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": 0.8350166666666667,
      "accuracy_std_dev": 0.019584493003961645,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.025601432659932657,
      "avg_duration_per_task_ms": 248785.20707070708,
      "avg_output_tokens_per_task": 21309.069865319867,
      "last_run_timestamp": "2026-08-29T18:12:43.943Z"
    },
    {
      "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-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Relace",
      "accuracy": 0.8349160504017531,
      "accuracy_std_dev": 0.018287978469196532,
      "run_count": 7,
      "total_questions": 1369,
      "avg_cost_per_task_usd": 0.015230915997078159,
      "avg_duration_per_task_ms": 440198.80131482834,
      "avg_output_tokens_per_task": 39488.13586559532,
      "last_run_timestamp": "2026-10-10T08:48:34.734Z"
    },
    {
      "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": 0.8348179217065166,
      "accuracy_std_dev": 0.020088025050917314,
      "run_count": 11,
      "total_questions": 2133,
      "avg_cost_per_task_usd": 0.21873191092358182,
      "avg_duration_per_task_ms": 294912.9784341303,
      "avg_output_tokens_per_task": 49538.53023909986,
      "last_run_timestamp": "2026-10-10T08:39:15.395Z"
    },
    {
      "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": 0.8347633260247855,
      "accuracy_std_dev": 0.010241443659085482,
      "run_count": 6,
      "total_questions": 1049,
      "avg_cost_per_task_usd": 0.047060826501429925,
      "avg_duration_per_task_ms": 90755.15919923737,
      "avg_output_tokens_per_task": 15002.908484270734,
      "last_run_timestamp": "2026-10-10T08:22:14.331Z"
    },
    {
      "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": 0.8346427777777778,
      "accuracy_std_dev": 0.050646967751239176,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.1631596778900112,
      "avg_duration_per_task_ms": 161457.49102132436,
      "avg_output_tokens_per_task": 16154.749719416386,
      "last_run_timestamp": "2026-10-10T08:25:10.471Z"
    },
    {
      "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": 0.8344556666666666,
      "accuracy_std_dev": 0.016627808360962375,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.1539143468013468,
      "avg_duration_per_task_ms": 212926.7558922559,
      "avg_output_tokens_per_task": 34726.19528619529,
      "last_run_timestamp": "2026-10-10T08:23:16.645Z"
    },
    {
      "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": 0.8342810781990522,
      "accuracy_std_dev": 0.009652908469324662,
      "run_count": 13,
      "total_questions": 2110,
      "avg_cost_per_task_usd": 0.018507007109004738,
      "avg_duration_per_task_ms": 188704.51184834124,
      "avg_output_tokens_per_task": 36823.764454976306,
      "last_run_timestamp": "2026-10-10T08:36:35.871Z"
    },
    {
      "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": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8341792639593909,
      "accuracy_std_dev": 0.020282949394476377,
      "run_count": 6,
      "total_questions": 1182,
      "avg_cost_per_task_usd": 0.07842123350253806,
      "avg_duration_per_task_ms": 728101.9196277496,
      "avg_output_tokens_per_task": 42337.8130287648,
      "last_run_timestamp": "2026-10-04T00:07:08.217Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "thinkingmachines/inkling-20260715",
      "model_slug": "thinkingmachines/inkling",
      "model_name": "Thinking Machines: Inkling",
      "provider_name": "auto-routing",
      "accuracy": 0.8340364430379746,
      "accuracy_std_dev": 0.01069546220626094,
      "run_count": 6,
      "total_questions": 1185,
      "avg_cost_per_task_usd": 0.0954248565400844,
      "avg_duration_per_task_ms": 293770.05400843883,
      "avg_output_tokens_per_task": 23403.932489451476,
      "last_run_timestamp": "2026-10-01T16:11:05.832Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8338943333333333,
      "accuracy_std_dev": 0.011690806112307073,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.03773407239057239,
      "avg_duration_per_task_ms": 276396.1228956229,
      "avg_output_tokens_per_task": 28290.345117845118,
      "last_run_timestamp": "2026-09-12T09:56:44.384Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Modal",
      "accuracy": 0.833648875,
      "accuracy_std_dev": 0.013388255986474674,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.13926744002525251,
      "avg_duration_per_task_ms": 187514.4393939394,
      "avg_output_tokens_per_task": 31415.290404040403,
      "last_run_timestamp": "2026-10-10T08:23:04.853Z"
    },
    {
      "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": 0.8334175866261399,
      "accuracy_std_dev": 0.031437998183885686,
      "run_count": 10,
      "total_questions": 1974,
      "avg_cost_per_task_usd": 0.01607202178318136,
      "avg_duration_per_task_ms": 132425.65501519758,
      "avg_output_tokens_per_task": 17865.155015197568,
      "last_run_timestamp": "2026-10-10T16:25:38.295Z"
    },
    {
      "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": 0.8334175326251898,
      "accuracy_std_dev": 0.0340950993933146,
      "run_count": 10,
      "total_questions": 1977,
      "avg_cost_per_task_usd": 0.10936984522003032,
      "avg_duration_per_task_ms": 1005533.760748609,
      "avg_output_tokens_per_task": 34172.51643904906,
      "last_run_timestamp": "2026-10-10T16:20:24.740Z"
    },
    {
      "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": 0.8333335829015545,
      "accuracy_std_dev": 0.019006499999999982,
      "run_count": 2,
      "total_questions": 386,
      "avg_cost_per_task_usd": 0.023828624352331605,
      "avg_duration_per_task_ms": 290054.69430051814,
      "avg_output_tokens_per_task": 19802.816062176167,
      "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": "Crusoe",
      "accuracy": 0.8333334266055047,
      "accuracy_std_dev": 0.015888679779606903,
      "run_count": 6,
      "total_questions": 1090,
      "avg_cost_per_task_usd": 0.02116503211009174,
      "avg_duration_per_task_ms": 295187.03211009176,
      "avg_output_tokens_per_task": 42108.07706422018,
      "last_run_timestamp": "2026-10-10T08:23:44.530Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-flash-20260826",
      "model_slug": "qwen/qwen3.8-flash",
      "model_name": "Qwen: Qwen3.8 Flash",
      "provider_name": "Makora",
      "accuracy": 0.8333329999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 156,
      "avg_cost_per_task_usd": 0.003982064102564103,
      "avg_duration_per_task_ms": 34178.44871794872,
      "avg_output_tokens_per_task": 8341.692307692309,
      "last_run_timestamp": "2026-09-09T15:03:02.824Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Cloudflare",
      "accuracy": 0.8332223213453214,
      "accuracy_std_dev": 0.019068183116368872,
      "run_count": 16,
      "total_questions": 3003,
      "avg_cost_per_task_usd": 0.19684398967698968,
      "avg_duration_per_task_ms": 375126.0233100233,
      "avg_output_tokens_per_task": 39605.68764568765,
      "last_run_timestamp": "2026-10-10T08:22:27.406Z"
    },
    {
      "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": 0.054762767255892256,
      "avg_duration_per_task_ms": 267512.9158249158,
      "avg_output_tokens_per_task": 23368.800084175084,
      "last_run_timestamp": "2026-10-10T08:22:25.200Z"
    },
    {
      "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": 0.8331589885057472,
      "accuracy_std_dev": 0.013081913080821427,
      "run_count": 12,
      "total_questions": 1914,
      "avg_cost_per_task_usd": 0.11308331609195402,
      "avg_duration_per_task_ms": 223576.03500522466,
      "avg_output_tokens_per_task": 27212.787356321838,
      "last_run_timestamp": "2026-10-10T08:14:48.303Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "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": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "Phala",
      "accuracy": 0.8325081287128712,
      "accuracy_std_dev": 0.06854471399887452,
      "run_count": 4,
      "total_questions": 606,
      "avg_cost_per_task_usd": 0.13443122607260727,
      "avg_duration_per_task_ms": 1513362.3762376239,
      "avg_output_tokens_per_task": 46533.88613861386,
      "last_run_timestamp": "2026-08-29T18:25:52.753Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-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": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "Venice",
      "accuracy": 0.8320707500000001,
      "accuracy_std_dev": 0.05130359405721493,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.09154002398989898,
      "avg_duration_per_task_ms": 446836.672979798,
      "avg_output_tokens_per_task": 20695.150883838385,
      "last_run_timestamp": "2026-10-10T16:11:11.317Z"
    },
    {
      "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": 0.8314910184229919,
      "accuracy_std_dev": 0.014032999291029713,
      "run_count": 7,
      "total_questions": 1357,
      "avg_cost_per_task_usd": 0.15096693072955047,
      "avg_duration_per_task_ms": 266519.0810611643,
      "avg_output_tokens_per_task": 34117.86366985999,
      "last_run_timestamp": "2026-10-10T08:23:25.246Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Io Net",
      "accuracy": 0.8312629347826086,
      "accuracy_std_dev": 0.044536486139877485,
      "run_count": 4,
      "total_questions": 644,
      "avg_cost_per_task_usd": 0.005694647515527951,
      "avg_duration_per_task_ms": 270388.4906832298,
      "avg_output_tokens_per_task": 13521.194099378881,
      "last_run_timestamp": "2026-08-22T09:56:05.550Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.4-mini-20260317",
      "model_slug": "openai/gpt-5.4-mini",
      "model_name": "OpenAI: GPT-5.4 Mini",
      "provider_name": "Azure",
      "accuracy": 0.8310885000000001,
      "accuracy_std_dev": 0.012955632388399007,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.06418621717171719,
      "avg_duration_per_task_ms": 117116.24831649831,
      "avg_output_tokens_per_task": 13483.090067340067,
      "last_run_timestamp": "2026-10-01T19:14:05.661Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "CoreWeave",
      "accuracy": 0.8309619512195122,
      "accuracy_std_dev": 0.016029553150759313,
      "run_count": 13,
      "total_questions": 1968,
      "avg_cost_per_task_usd": 0.036442337906504065,
      "avg_duration_per_task_ms": 251079.88109756098,
      "avg_output_tokens_per_task": 29050.197662601626,
      "last_run_timestamp": "2026-10-10T16:21:24.224Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8307314545454545,
      "accuracy_std_dev": 0.026853345034104435,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.13328792148760332,
      "avg_duration_per_task_ms": 548052.8255280073,
      "avg_output_tokens_per_task": 41331.14049586777,
      "last_run_timestamp": "2026-10-10T08:14:46.985Z"
    },
    {
      "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": 0.830415961308153,
      "accuracy_std_dev": 0.034463455713538205,
      "run_count": 11,
      "total_questions": 2171,
      "avg_cost_per_task_usd": 0.06569626301243668,
      "avg_duration_per_task_ms": 1046585.6020267158,
      "avg_output_tokens_per_task": 33076.09580838323,
      "last_run_timestamp": "2026-10-10T16:21:31.573Z"
    },
    {
      "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": 0.8303493580562661,
      "accuracy_std_dev": 0.021497934689561845,
      "run_count": 6,
      "total_questions": 1173,
      "avg_cost_per_task_usd": 0.0706463444160273,
      "avg_duration_per_task_ms": 871151.2676896845,
      "avg_output_tokens_per_task": 40306.14833759591,
      "last_run_timestamp": "2026-08-30T00:37:59.094Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Friendli",
      "accuracy": 0.8301192727272728,
      "accuracy_std_dev": 0.022568976244920873,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.23170999265381084,
      "avg_duration_per_task_ms": 339625.2497704316,
      "avg_output_tokens_per_task": 52536.71900826446,
      "last_run_timestamp": "2026-10-10T08:22:18.240Z"
    },
    {
      "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": 0.8296903524590165,
      "accuracy_std_dev": 0.03798493151237738,
      "run_count": 3,
      "total_questions": 366,
      "avg_cost_per_task_usd": 0.08753724316939891,
      "avg_duration_per_task_ms": 313974.91803278687,
      "avg_output_tokens_per_task": 36231.311475409835,
      "last_run_timestamp": "2026-10-10T16:20:11.998Z"
    },
    {
      "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": 0.8295325781071835,
      "accuracy_std_dev": 0.006460861466467855,
      "run_count": 9,
      "total_questions": 1754,
      "avg_cost_per_task_usd": 0.035354884834663626,
      "avg_duration_per_task_ms": 201763.31527936147,
      "avg_output_tokens_per_task": 14117.418472063853,
      "last_run_timestamp": "2026-10-10T08:15:01.141Z"
    },
    {
      "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": 0.8292925999999999,
      "accuracy_std_dev": 0.004466830267650664,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.057570920202020204,
      "avg_duration_per_task_ms": 214582.1888888889,
      "avg_output_tokens_per_task": 15222.584848484848,
      "last_run_timestamp": "2026-10-10T08:19:01.753Z"
    },
    {
      "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": 0.8290087877529839,
      "accuracy_std_dev": 0.02323794562004137,
      "run_count": 10,
      "total_questions": 1927,
      "avg_cost_per_task_usd": 0.02390980435910742,
      "avg_duration_per_task_ms": 846858.6066424494,
      "avg_output_tokens_per_task": 46042.94706798132,
      "last_run_timestamp": "2026-10-10T08:17:16.897Z"
    },
    {
      "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": 0.8288439999999999,
      "accuracy_std_dev": 0.005203901805376433,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.02148271043771044,
      "avg_duration_per_task_ms": 661167.3181818182,
      "avg_output_tokens_per_task": 55670.02693602694,
      "last_run_timestamp": "2026-10-10T08:22:14.536Z"
    },
    {
      "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": "deepseek/deepseek-v3.2-exp",
      "model_slug": "deepseek/deepseek-v3.2-exp",
      "model_name": "DeepSeek: DeepSeek V3.2 Exp",
      "provider_name": "AtlasCloud",
      "accuracy": 0.828283,
      "accuracy_std_dev": 0.003367000000000009,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.009031212121212122,
      "avg_duration_per_task_ms": 399484.5707070707,
      "avg_output_tokens_per_task": 21544.277777777777,
      "last_run_timestamp": "2026-08-01T20:03:34.768Z"
    },
    {
      "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": 0.8280361855845629,
      "accuracy_std_dev": 0.008498377450075449,
      "run_count": 9,
      "total_questions": 1762,
      "avg_cost_per_task_usd": 0.007634015323496027,
      "avg_duration_per_task_ms": 329436.5028376845,
      "avg_output_tokens_per_task": 19170.439841089672,
      "last_run_timestamp": "2026-10-10T16:21:04.886Z"
    },
    {
      "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": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "Phala",
      "accuracy": 0.8276705454545453,
      "accuracy_std_dev": 0.013998517625180167,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.06217915610651974,
      "avg_duration_per_task_ms": 529389.9205693296,
      "avg_output_tokens_per_task": 48974.31083562902,
      "last_run_timestamp": "2026-10-10T16:25:44.293Z"
    },
    {
      "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": 0.8275666632911391,
      "accuracy_std_dev": 0.031862730949862604,
      "run_count": 6,
      "total_questions": 1185,
      "avg_cost_per_task_usd": 0.025537988185654007,
      "avg_duration_per_task_ms": 492003.4506329114,
      "avg_output_tokens_per_task": 23813.98734177215,
      "last_run_timestamp": "2026-08-29T18:05:36.579Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-122b-a10b-20260224",
      "model_slug": "qwen/qwen3.5-122b-a10b",
      "model_name": "Qwen: Qwen3.5-122B-A10B",
      "provider_name": "SiliconFlow",
      "accuracy": 0.827288,
      "accuracy_std_dev": 0.05899557909168817,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.05165088016528926,
      "avg_duration_per_task_ms": 574742.4444444445,
      "avg_output_tokens_per_task": 24729.172635445364,
      "last_run_timestamp": "2026-10-10T08:28:19.656Z"
    },
    {
      "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": 0.8272306250000001,
      "accuracy_std_dev": 0.010511419182697223,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.04333868371212121,
      "avg_duration_per_task_ms": 227486.86805555556,
      "avg_output_tokens_per_task": 14329.91856060606,
      "last_run_timestamp": "2026-10-10T08:22:04.190Z"
    },
    {
      "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": 0.827217112385321,
      "accuracy_std_dev": 0.014186141135664356,
      "run_count": 9,
      "total_questions": 1744,
      "avg_cost_per_task_usd": 0.036017568233944953,
      "avg_duration_per_task_ms": 436504.3365825688,
      "avg_output_tokens_per_task": 47925.99369266055,
      "last_run_timestamp": "2026-10-10T08:14:40.551Z"
    },
    {
      "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": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8265992857142858,
      "accuracy_std_dev": 0.05408198806498543,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.09954717027417027,
      "avg_duration_per_task_ms": 788800.0598845598,
      "avg_output_tokens_per_task": 39778.419913419915,
      "last_run_timestamp": "2026-10-03T17:58:22.995Z"
    },
    {
      "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": 0.8265991666666667,
      "accuracy_std_dev": 0.013432985736817995,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.024518459595959594,
      "avg_duration_per_task_ms": 477551.41245791246,
      "avg_output_tokens_per_task": 22795.994107744107,
      "last_run_timestamp": "2026-08-29T18:09:03.379Z"
    },
    {
      "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-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "CoreWeave",
      "accuracy": 0.8263887499999999,
      "accuracy_std_dev": 0.018961582467913917,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.12611158901515152,
      "avg_duration_per_task_ms": 326364.59911616164,
      "avg_output_tokens_per_task": 52045.35858585859,
      "last_run_timestamp": "2026-10-10T08:22:18.071Z"
    },
    {
      "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": 0.8263887499999999,
      "accuracy_std_dev": 0.0068091963686987495,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.13881603345959595,
      "avg_duration_per_task_ms": 161822.7279040404,
      "avg_output_tokens_per_task": 34223.880681818184,
      "last_run_timestamp": "2026-10-10T08:21:09.697Z"
    },
    {
      "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": 0.8262248861473387,
      "accuracy_std_dev": 0.021314932165659498,
      "run_count": 15,
      "total_questions": 2837,
      "avg_cost_per_task_usd": 0.04278339231582658,
      "avg_duration_per_task_ms": 372895.77722946776,
      "avg_output_tokens_per_task": 46085.86993302785,
      "last_run_timestamp": "2026-10-10T16:22:35.588Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "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": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Ionstream",
      "accuracy": 0.8258778571428571,
      "accuracy_std_dev": 0.01206814403328714,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.03852732972582973,
      "avg_duration_per_task_ms": 268250.60461760464,
      "avg_output_tokens_per_task": 15536.08658008658,
      "last_run_timestamp": "2026-10-10T08:21:53.740Z"
    },
    {
      "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": 0.8258626249999999,
      "accuracy_std_dev": 0.012410001157710468,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.09867501325757576,
      "avg_duration_per_task_ms": 367182.26704545453,
      "avg_output_tokens_per_task": 45431.65340909091,
      "last_run_timestamp": "2026-10-10T16:11:14.320Z"
    },
    {
      "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": 0.8255893999999999,
      "accuracy_std_dev": 0.010301039221360171,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.14328596767676768,
      "avg_duration_per_task_ms": 412091.31313131313,
      "avg_output_tokens_per_task": 28667.273737373736,
      "last_run_timestamp": "2026-10-10T08:24:21.666Z"
    },
    {
      "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": 0.8254515454545454,
      "accuracy_std_dev": 0.024386361607048986,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.037847392102846655,
      "avg_duration_per_task_ms": 348957.9407713499,
      "avg_output_tokens_per_task": 39061.49908172635,
      "last_run_timestamp": "2026-10-10T16:22:04.171Z"
    },
    {
      "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": 0.8250688181818181,
      "accuracy_std_dev": 0.022841304837819883,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.010740173094582185,
      "avg_duration_per_task_ms": 682886.4674012855,
      "avg_output_tokens_per_task": 38122.29706152433,
      "last_run_timestamp": "2026-10-10T08:14:49.245Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.824916,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.022821656565656566,
      "avg_duration_per_task_ms": 516406.73232323234,
      "avg_output_tokens_per_task": 45403.40909090909,
      "last_run_timestamp": "2026-10-04T03:04:38.670Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "CoreWeave",
      "accuracy": 0.8248312595648912,
      "accuracy_std_dev": 0.01663705107798342,
      "run_count": 7,
      "total_questions": 1333,
      "avg_cost_per_task_usd": 0.02528466016504126,
      "avg_duration_per_task_ms": 296846.2640660165,
      "avg_output_tokens_per_task": 50351.12678169542,
      "last_run_timestamp": "2026-10-10T08:35:20.063Z"
    },
    {
      "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": 0.8243470488301118,
      "accuracy_std_dev": 0.01897362841419636,
      "run_count": 5,
      "total_questions": 983,
      "avg_cost_per_task_usd": 0.09452765412004069,
      "avg_duration_per_task_ms": 201477.20142421158,
      "avg_output_tokens_per_task": 27045.996948118005,
      "last_run_timestamp": "2026-08-29T18:58:14.477Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8242846250000001,
      "accuracy_std_dev": 0.007803239774886767,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.04825054608585859,
      "avg_duration_per_task_ms": 444851.5814393939,
      "avg_output_tokens_per_task": 26726.752525252527,
      "last_run_timestamp": "2026-10-10T16:10:48.867Z"
    },
    {
      "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": 0.8233164,
      "accuracy_std_dev": 0.059578276225819106,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.027962910606060605,
      "avg_duration_per_task_ms": 676433.6525252525,
      "avg_output_tokens_per_task": 29240.15101010101,
      "last_run_timestamp": "2026-10-10T08:21:58.760Z"
    },
    {
      "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": 0.8231511800643088,
      "accuracy_std_dev": 0.016091623665357716,
      "run_count": 8,
      "total_questions": 1555,
      "avg_cost_per_task_usd": 0.14175151639871383,
      "avg_duration_per_task_ms": 272966.68681672023,
      "avg_output_tokens_per_task": 31995.940836012862,
      "last_run_timestamp": "2026-10-10T08:25:06.567Z"
    },
    {
      "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": 0.823129,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 196,
      "avg_cost_per_task_usd": 0.032144397959183675,
      "avg_duration_per_task_ms": 345053.9591836735,
      "avg_output_tokens_per_task": 27884.80612244898,
      "last_run_timestamp": "2026-07-18T08:29:12.575Z"
    },
    {
      "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": 0.8229467130434782,
      "accuracy_std_dev": 0.020047430385493913,
      "run_count": 7,
      "total_questions": 1380,
      "avg_cost_per_task_usd": 0.006798507246376812,
      "avg_duration_per_task_ms": 630419.7739130434,
      "avg_output_tokens_per_task": 18294.628260869566,
      "last_run_timestamp": "2026-10-10T16:19:44.346Z"
    },
    {
      "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": 0.8224291477721376,
      "accuracy_std_dev": 0.010721893131391866,
      "run_count": 9,
      "total_questions": 1773,
      "avg_cost_per_task_usd": 0.025065345741680765,
      "avg_duration_per_task_ms": 452443.4557247603,
      "avg_output_tokens_per_task": 25021.49464184997,
      "last_run_timestamp": "2026-10-10T16:11:11.289Z"
    },
    {
      "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": 0.8223905000000001,
      "accuracy_std_dev": 0.007575500000000013,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.007053464646464647,
      "avg_duration_per_task_ms": 622803.648989899,
      "avg_output_tokens_per_task": 16887.560606060608,
      "last_run_timestamp": "2026-08-01T18:25:37.046Z"
    },
    {
      "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": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "StreamLake",
      "accuracy": 0.8223905,
      "accuracy_std_dev": 0.021663157431839608,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.10283124074074075,
      "avg_duration_per_task_ms": 597427.9688552188,
      "avg_output_tokens_per_task": 38154.76262626263,
      "last_run_timestamp": "2026-08-29T18:26:14.942Z"
    },
    {
      "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": 0.8222970000000001,
      "accuracy_std_dev": 0.02180780391918859,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.03627528900112234,
      "avg_duration_per_task_ms": 268542.3265993266,
      "avg_output_tokens_per_task": 14420.646464646465,
      "last_run_timestamp": "2026-10-10T08:43:05.187Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "Novita",
      "accuracy": 0.8217018181818182,
      "accuracy_std_dev": 0.0209343466485894,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.012272595959595959,
      "avg_duration_per_task_ms": 695475.8053259872,
      "avg_output_tokens_per_task": 36304.134527089074,
      "last_run_timestamp": "2026-10-10T08:14:48.813Z"
    },
    {
      "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": 0.8215404398976982,
      "accuracy_std_dev": 0.032891408388902345,
      "run_count": 6,
      "total_questions": 1173,
      "avg_cost_per_task_usd": 0.02587831713554987,
      "avg_duration_per_task_ms": 460054.41943734017,
      "avg_output_tokens_per_task": 21496.38874680307,
      "last_run_timestamp": "2026-08-29T19:19:06.625Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Makora",
      "accuracy": 0.8213506405228759,
      "accuracy_std_dev": 0.02248216924587128,
      "run_count": 3,
      "total_questions": 459,
      "avg_cost_per_task_usd": 0.08758356862745097,
      "avg_duration_per_task_ms": 304618.9760348584,
      "avg_output_tokens_per_task": 34720.03050108933,
      "last_run_timestamp": "2026-08-29T10:57:24.797Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "AkashML",
      "accuracy": 0.8212957816756088,
      "accuracy_std_dev": 0.017294239247104652,
      "run_count": 13,
      "total_questions": 2423,
      "avg_cost_per_task_usd": 0.031653728023111845,
      "avg_duration_per_task_ms": 257591.64630623194,
      "avg_output_tokens_per_task": 14017.31324803962,
      "last_run_timestamp": "2026-10-10T08:14:46.104Z"
    },
    {
      "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": 0.8212274003378379,
      "accuracy_std_dev": 0.0061029639361178676,
      "run_count": 6,
      "total_questions": 1184,
      "avg_cost_per_task_usd": 0.039089869087837834,
      "avg_duration_per_task_ms": 209017.77111486485,
      "avg_output_tokens_per_task": 14834.131756756757,
      "last_run_timestamp": "2026-10-10T08:21:53.663Z"
    },
    {
      "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": 0.821210220113852,
      "accuracy_std_dev": 0.012972061902411654,
      "run_count": 8,
      "total_questions": 1581,
      "avg_cost_per_task_usd": 0.19423610879190387,
      "avg_duration_per_task_ms": 987635.4212523719,
      "avg_output_tokens_per_task": 51696.76660341556,
      "last_run_timestamp": "2026-10-10T16:21:24.336Z"
    },
    {
      "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": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "Alibaba",
      "accuracy": 0.8211085620915033,
      "accuracy_std_dev": 0.010055056516550273,
      "run_count": 7,
      "total_questions": 1377,
      "avg_cost_per_task_usd": 0.022989802469135803,
      "avg_duration_per_task_ms": 339486.9368191721,
      "avg_output_tokens_per_task": 20491.059549745823,
      "last_run_timestamp": "2026-10-04T00:17:30.035Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "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": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8208700266159695,
      "accuracy_std_dev": 0.003273364894779542,
      "run_count": 8,
      "total_questions": 1578,
      "avg_cost_per_task_usd": 0.007889804816223067,
      "avg_duration_per_task_ms": 461257.1007604563,
      "avg_output_tokens_per_task": 20299.61216730038,
      "last_run_timestamp": "2026-10-10T16:10:51.722Z"
    },
    {
      "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": 0.8208440232558138,
      "accuracy_std_dev": 0.005892000000000008,
      "run_count": 2,
      "total_questions": 387,
      "avg_cost_per_task_usd": 0.007978992248062014,
      "avg_duration_per_task_ms": 646108.8320413437,
      "avg_output_tokens_per_task": 19074.485788113696,
      "last_run_timestamp": "2026-08-01T18:07:34.003Z"
    },
    {
      "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": 0.8205865714285715,
      "accuracy_std_dev": 0.030540716690332988,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.09537167965367965,
      "avg_duration_per_task_ms": 1474899.9105339106,
      "avg_output_tokens_per_task": 44220.89033189033,
      "last_run_timestamp": "2026-10-10T16:26:06.158Z"
    },
    {
      "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": 0.8205389659969088,
      "accuracy_std_dev": 0.05440642468688786,
      "run_count": 10,
      "total_questions": 1941,
      "avg_cost_per_task_usd": 0.08737539618753219,
      "avg_duration_per_task_ms": 555621.9278722308,
      "avg_output_tokens_per_task": 29961.489438433797,
      "last_run_timestamp": "2026-10-10T08:48:03.287Z"
    },
    {
      "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": 0.8202530101265821,
      "accuracy_std_dev": 0.008020329102765456,
      "run_count": 6,
      "total_questions": 1185,
      "avg_cost_per_task_usd": 0.018872242194092826,
      "avg_duration_per_task_ms": 516686.08607594937,
      "avg_output_tokens_per_task": 18743.98734177215,
      "last_run_timestamp": "2026-08-29T18:35:57.732Z"
    },
    {
      "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": 0.8198651666666666,
      "accuracy_std_dev": 0.10272117659862329,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.17633816414141412,
      "avg_duration_per_task_ms": 1166115.4705387205,
      "avg_output_tokens_per_task": 47461.85521885522,
      "last_run_timestamp": "2026-10-10T16:33:35.864Z"
    },
    {
      "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": 0.819865,
      "accuracy_std_dev": 0.006733999999999962,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.008803729797979798,
      "avg_duration_per_task_ms": 1022564.7878787878,
      "avg_output_tokens_per_task": 21115.48232323232,
      "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-5.3-20260816",
      "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": "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": 0.8195591818181818,
      "accuracy_std_dev": 0.014154784085686241,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.31459056198347113,
      "avg_duration_per_task_ms": 289293.8429752066,
      "avg_output_tokens_per_task": 20783.41092745638,
      "last_run_timestamp": "2026-10-10T08:22:14.123Z"
    },
    {
      "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": 0.819444375,
      "accuracy_std_dev": 0.015469708868765922,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.03466034659090909,
      "avg_duration_per_task_ms": 238239.72095959596,
      "avg_output_tokens_per_task": 13747.289772727272,
      "last_run_timestamp": "2026-10-10T08:21:42.364Z"
    },
    {
      "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": 0.8192329035532995,
      "accuracy_std_dev": 0.01830389154129678,
      "run_count": 6,
      "total_questions": 1182,
      "avg_cost_per_task_usd": 0.022038197123519458,
      "avg_duration_per_task_ms": 499412.81133671745,
      "avg_output_tokens_per_task": 22395.89847715736,
      "last_run_timestamp": "2026-08-29T18:16:26.161Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "CoreWeave",
      "accuracy": 0.8191919000000001,
      "accuracy_std_dev": 0.019432571659201477,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.04233793585858586,
      "avg_duration_per_task_ms": 245621.01414141414,
      "avg_output_tokens_per_task": 14127.980303030303,
      "last_run_timestamp": "2026-10-10T08:23:34.137Z"
    },
    {
      "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": 0.8190038027586207,
      "accuracy_std_dev": 0.04321943578037032,
      "run_count": 11,
      "total_questions": 2175,
      "avg_cost_per_task_usd": 0.0850050151724138,
      "avg_duration_per_task_ms": 798913.6754022988,
      "avg_output_tokens_per_task": 31455.114942528737,
      "last_run_timestamp": "2026-10-10T16:23:10.504Z"
    },
    {
      "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": 0.8187577756653992,
      "accuracy_std_dev": 0.007032873860263311,
      "run_count": 8,
      "total_questions": 1578,
      "avg_cost_per_task_usd": 0.008747721166032953,
      "avg_duration_per_task_ms": 704234.403041825,
      "avg_output_tokens_per_task": 18267.351077313055,
      "last_run_timestamp": "2026-10-10T16:19:34.424Z"
    },
    {
      "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": 0.8184820903010034,
      "accuracy_std_dev": 0.06228079588716609,
      "run_count": 22,
      "total_questions": 4186,
      "avg_cost_per_task_usd": 0.007251606784519827,
      "avg_duration_per_task_ms": 497071.1039178213,
      "avg_output_tokens_per_task": 17273.298853320593,
      "last_run_timestamp": "2026-10-10T16:27:03.857Z"
    },
    {
      "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": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "SambaNova",
      "accuracy": 0.8180962405063291,
      "accuracy_std_dev": 0.020585850087863757,
      "run_count": 5,
      "total_questions": 711,
      "avg_cost_per_task_usd": 0.10679723066104078,
      "avg_duration_per_task_ms": 387578.5597749648,
      "avg_output_tokens_per_task": 44272.65400843882,
      "last_run_timestamp": "2026-10-10T16:20:25.617Z"
    },
    {
      "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": "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": 0.817901,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 54,
      "avg_cost_per_task_usd": 0.008988981481481481,
      "avg_duration_per_task_ms": 114091.12962962964,
      "avg_output_tokens_per_task": 16251.87037037037,
      "last_run_timestamp": "2026-07-18T09:01:08.846Z"
    },
    {
      "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": "meta/muse-glimmer-30b-20260810",
      "model_slug": "meta/muse-glimmer-30b",
      "model_name": "Meta: Muse Glimmer 30B",
      "provider_name": "Phala",
      "accuracy": 0.816967559819413,
      "accuracy_std_dev": 0.022532849378432182,
      "run_count": 9,
      "total_questions": 1772,
      "avg_cost_per_task_usd": 0.021389000564334085,
      "avg_duration_per_task_ms": 297135.54796839727,
      "avg_output_tokens_per_task": 17946.336343115123,
      "last_run_timestamp": "2026-10-03T09:25:02.390Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "auto-routing",
      "accuracy": 0.8168565855513307,
      "accuracy_std_dev": 0.03334781637130645,
      "run_count": 8,
      "total_questions": 1578,
      "avg_cost_per_task_usd": 0.08050372496831433,
      "avg_duration_per_task_ms": 908697.1989860584,
      "avg_output_tokens_per_task": 40687.09125475285,
      "last_run_timestamp": "2026-10-10T16:10:34.759Z"
    },
    {
      "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": 0.8167003199999999,
      "accuracy_std_dev": 0.011578936088328655,
      "run_count": 25,
      "total_questions": 4950,
      "avg_cost_per_task_usd": 0.3337196343434343,
      "avg_duration_per_task_ms": 298714.2325252525,
      "avg_output_tokens_per_task": 21127.507474747475,
      "last_run_timestamp": "2026-10-10T08:23:34.485Z"
    },
    {
      "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": 0.8162174999999999,
      "accuracy_std_dev": 0.009642570572725913,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.019066546296296293,
      "avg_duration_per_task_ms": 525999.5555555555,
      "avg_output_tokens_per_task": 18868.49074074074,
      "last_run_timestamp": "2026-08-30T00:16:38.394Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.1-flash-lite-20260507",
      "model_slug": "google/gemini-3.1-flash-lite",
      "model_name": "Google: Gemini 3.1 Flash Lite",
      "provider_name": "auto-routing",
      "accuracy": 0.815937,
      "accuracy_std_dev": 0.021042110493009022,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.04137441245791246,
      "avg_duration_per_task_ms": 82874.20875420875,
      "avg_output_tokens_per_task": 27465.196969696968,
      "last_run_timestamp": "2026-10-01T19:23:01.327Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "WandB",
      "accuracy": 0.815838,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 181,
      "avg_cost_per_task_usd": 0.03376955248618784,
      "avg_duration_per_task_ms": 178211.5359116022,
      "avg_output_tokens_per_task": 26893.3591160221,
      "last_run_timestamp": "2026-07-18T08:04:57.082Z"
    },
    {
      "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": 0.8157000210325048,
      "accuracy_std_dev": 0.029811887613993424,
      "run_count": 8,
      "total_questions": 1569,
      "avg_cost_per_task_usd": 0.2467402931803696,
      "avg_duration_per_task_ms": 963446.5092415551,
      "avg_output_tokens_per_task": 56213.97705544933,
      "last_run_timestamp": "2026-10-10T16:11:11.321Z"
    },
    {
      "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": 0.8156999235181644,
      "accuracy_std_dev": 0.04951986238986445,
      "run_count": 9,
      "total_questions": 1569,
      "avg_cost_per_task_usd": 0.04647088910133843,
      "avg_duration_per_task_ms": 572516.0726577438,
      "avg_output_tokens_per_task": 26962.156787762906,
      "last_run_timestamp": "2026-09-26T09:56:00.145Z"
    },
    {
      "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": 0.8155362857142857,
      "accuracy_std_dev": 0.02021913729058464,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.0062872914862914875,
      "avg_duration_per_task_ms": 495775.87012987013,
      "avg_output_tokens_per_task": 18627.747474747473,
      "last_run_timestamp": "2026-10-10T16:21:33.333Z"
    },
    {
      "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": 0.8155034545454546,
      "accuracy_std_dev": 0.00991632568374574,
      "run_count": 22,
      "total_questions": 4356,
      "avg_cost_per_task_usd": 0.3643461124885216,
      "avg_duration_per_task_ms": 358943.8365472911,
      "avg_output_tokens_per_task": 22948.309228650138,
      "last_run_timestamp": "2026-10-10T08:22:14.499Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "auto-routing",
      "accuracy": 0.8152958571428572,
      "accuracy_std_dev": 0.016658861214968797,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.006618044011544012,
      "avg_duration_per_task_ms": 520273.7655122655,
      "avg_output_tokens_per_task": 17626.31096681097,
      "last_run_timestamp": "2026-10-10T16:21:38.183Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Darkbloom",
      "accuracy": 0.8152955714285716,
      "accuracy_std_dev": 0.00996837593674748,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.033789289321789324,
      "avg_duration_per_task_ms": 681304.6854256854,
      "avg_output_tokens_per_task": 15943.984848484848,
      "last_run_timestamp": "2026-10-10T08:14:49.700Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "auto-routing",
      "accuracy": 0.8149562466539196,
      "accuracy_std_dev": 0.025669785816390737,
      "run_count": 8,
      "total_questions": 1569,
      "avg_cost_per_task_usd": 0.20883088782664117,
      "avg_duration_per_task_ms": 930609.0956022944,
      "avg_output_tokens_per_task": 55931.00191204589,
      "last_run_timestamp": "2026-10-10T16:20:08.951Z"
    },
    {
      "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": 0.8148146250000001,
      "accuracy_std_dev": 0.021706775042239102,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.0411005606060606,
      "avg_duration_per_task_ms": 178924.70391414143,
      "avg_output_tokens_per_task": 22731.602272727272,
      "last_run_timestamp": "2026-10-10T16:10:48.868Z"
    },
    {
      "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": "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": 0.8142806714727086,
      "accuracy_std_dev": 0.02584273780697392,
      "run_count": 5,
      "total_questions": 971,
      "avg_cost_per_task_usd": 0.06495255406797117,
      "avg_duration_per_task_ms": 232305.07929969105,
      "avg_output_tokens_per_task": 18958.6982492276,
      "last_run_timestamp": "2026-10-10T08:46:13.460Z"
    },
    {
      "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": 0.8141833749999999,
      "accuracy_std_dev": 0.00914133656717522,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.02917629608585859,
      "avg_duration_per_task_ms": 198512.0467171717,
      "avg_output_tokens_per_task": 23747.22474747475,
      "last_run_timestamp": "2026-10-10T16:21:37.928Z"
    },
    {
      "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": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "StreamLake",
      "accuracy": 0.8139730000000001,
      "accuracy_std_dev": 0.007059363734312956,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.006141271885521885,
      "avg_duration_per_task_ms": 521938.39983164985,
      "avg_output_tokens_per_task": 18689.60521885522,
      "last_run_timestamp": "2026-08-29T19:09:30.030Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Morph",
      "accuracy": 0.8136272825651302,
      "accuracy_std_dev": 0.008773473238499465,
      "run_count": 3,
      "total_questions": 499,
      "avg_cost_per_task_usd": 0.0068851422845691386,
      "avg_duration_per_task_ms": 1090136,
      "avg_output_tokens_per_task": 17404.21643286573,
      "last_run_timestamp": "2026-08-15T10:05:32.709Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Z.AI",
      "accuracy": 0.8135448247619047,
      "accuracy_std_dev": 0.049998493601682646,
      "run_count": 8,
      "total_questions": 1575,
      "avg_cost_per_task_usd": 0.22882532634920633,
      "avg_duration_per_task_ms": 1050315.3606349207,
      "avg_output_tokens_per_task": 52425.60063492064,
      "last_run_timestamp": "2026-10-10T16:22:35.078Z"
    },
    {
      "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": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Mancer 2",
      "accuracy": 0.8131312857142856,
      "accuracy_std_dev": 0.08184342879584912,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.011368694805194807,
      "avg_duration_per_task_ms": 741530.3095238095,
      "avg_output_tokens_per_task": 25627.320346320346,
      "last_run_timestamp": "2026-10-10T16:19:34.384Z"
    },
    {
      "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": "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": 0.811973875,
      "accuracy_std_dev": 0.028210901188359352,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.01455338888888889,
      "avg_duration_per_task_ms": 917219.5435606061,
      "avg_output_tokens_per_task": 43021.98421717172,
      "last_run_timestamp": "2026-10-10T08:36:08.861Z"
    },
    {
      "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": 0.811448,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.004932363636363637,
      "avg_duration_per_task_ms": 1604879.5959595959,
      "avg_output_tokens_per_task": 24145.00505050505,
      "last_run_timestamp": "2026-07-18T08:42:46.001Z"
    },
    {
      "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": 0.8114476666666668,
      "accuracy_std_dev": 0.026594828950171668,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.11157255948372617,
      "avg_duration_per_task_ms": 774642.1705948373,
      "avg_output_tokens_per_task": 58026.93097643097,
      "last_run_timestamp": "2026-10-10T16:20:12.057Z"
    },
    {
      "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": 0.8110362107795959,
      "accuracy_std_dev": 0.03156884697855712,
      "run_count": 6,
      "total_questions": 1039,
      "avg_cost_per_task_usd": 0.07300517420596726,
      "avg_duration_per_task_ms": 1437301.4898941289,
      "avg_output_tokens_per_task": 41806.461982675646,
      "last_run_timestamp": "2026-10-10T16:10:51.723Z"
    },
    {
      "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": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "auto-routing",
      "accuracy": 0.810185,
      "accuracy_std_dev": 0.02559748825568634,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.008685453914141415,
      "avg_duration_per_task_ms": 555085.3888888889,
      "avg_output_tokens_per_task": 18946.522095959597,
      "last_run_timestamp": "2026-10-10T16:10:48.868Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Relace",
      "accuracy": 0.8091166153846153,
      "accuracy_std_dev": 0.03183613126622014,
      "run_count": 3,
      "total_questions": 585,
      "avg_cost_per_task_usd": 0.19830280854700855,
      "avg_duration_per_task_ms": 1060541.5008547008,
      "avg_output_tokens_per_task": 55592.3264957265,
      "last_run_timestamp": "2026-10-03T16:58:49.489Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.5-sonnet-20250929",
      "model_slug": "anthropic/claude-sonnet-4.5",
      "model_name": "Anthropic: Claude Sonnet 4.5",
      "provider_name": "Anthropic",
      "accuracy": 0.8089225,
      "accuracy_std_dev": 0.015608773342258514,
      "run_count": 14,
      "total_questions": 2772,
      "avg_cost_per_task_usd": 0.34001660497835495,
      "avg_duration_per_task_ms": 354284.26948051946,
      "avg_output_tokens_per_task": 22498.079725829724,
      "last_run_timestamp": "2026-10-10T08:22:14.582Z"
    },
    {
      "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": "openai/gpt-5-mini-2025-08-07",
      "model_slug": "openai/gpt-5-mini",
      "model_name": "OpenAI: GPT-5 Mini",
      "provider_name": "auto-routing",
      "accuracy": 0.808081,
      "accuracy_std_dev": 0.018027342695657245,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.044340436026936034,
      "avg_duration_per_task_ms": 217888.1734006734,
      "avg_output_tokens_per_task": 22075.030303030304,
      "last_run_timestamp": "2026-10-01T19:19:28.473Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.5-sonnet-20250929",
      "model_slug": "anthropic/claude-sonnet-4.5",
      "model_name": "Anthropic: Claude Sonnet 4.5",
      "provider_name": "Azure",
      "accuracy": 0.807927818181818,
      "accuracy_std_dev": 0.021837768516611848,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.3410933209366391,
      "avg_duration_per_task_ms": 355919.63177226816,
      "avg_output_tokens_per_task": 22581.024334251608,
      "last_run_timestamp": "2026-10-10T08:25:00.704Z"
    },
    {
      "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": 0.8075253162083936,
      "accuracy_std_dev": 0.05075034301643404,
      "run_count": 7,
      "total_questions": 1382,
      "avg_cost_per_task_usd": 0.19541867366136037,
      "avg_duration_per_task_ms": 620138.4428364689,
      "avg_output_tokens_per_task": 46512.787264833576,
      "last_run_timestamp": "2026-10-10T16:19:34.446Z"
    },
    {
      "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": 0.807420008605852,
      "accuracy_std_dev": 0.03225256635592289,
      "run_count": 9,
      "total_questions": 1743,
      "avg_cost_per_task_usd": 0.017075846815834766,
      "avg_duration_per_task_ms": 580474.0820424556,
      "avg_output_tokens_per_task": 18206.01204819277,
      "last_run_timestamp": "2026-10-10T16:11:14.298Z"
    },
    {
      "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": 0.8066155190839694,
      "accuracy_std_dev": 0.009958499999999981,
      "run_count": 2,
      "total_questions": 393,
      "avg_cost_per_task_usd": 0.07586704071246819,
      "avg_duration_per_task_ms": 385930.98473282444,
      "avg_output_tokens_per_task": 32444.25445292621,
      "last_run_timestamp": "2026-08-01T17:52:55.742Z"
    },
    {
      "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": 0.8063972857142858,
      "accuracy_std_dev": 0.04410277557094591,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.04038211976911977,
      "avg_duration_per_task_ms": 957664.380952381,
      "avg_output_tokens_per_task": 25231.96176046176,
      "last_run_timestamp": "2026-09-26T09:34:13.954Z"
    },
    {
      "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": "qwen/qwen3.5-122b-a10b-20260224",
      "model_slug": "qwen/qwen3.5-122b-a10b",
      "model_name": "Qwen: Qwen3.5-122B-A10B",
      "provider_name": "Alibaba",
      "accuracy": 0.8061798193604147,
      "accuracy_std_dev": 0.10627998128250551,
      "run_count": 12,
      "total_questions": 2314,
      "avg_cost_per_task_usd": 0.04518147709593777,
      "avg_duration_per_task_ms": 176487.23595505618,
      "avg_output_tokens_per_task": 21622.97363872083,
      "last_run_timestamp": "2026-10-10T08:21:29.013Z"
    },
    {
      "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-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8058417989690723,
      "accuracy_std_dev": 0.019860476718151096,
      "run_count": 6,
      "total_questions": 1164,
      "avg_cost_per_task_usd": 0.1695892783505155,
      "avg_duration_per_task_ms": 882585.4982817869,
      "avg_output_tokens_per_task": 53988.948453608245,
      "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-5.2-20260616",
      "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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "GMICloud",
      "accuracy": 0.8053260000000001,
      "accuracy_std_dev": 0.037748749347133695,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.11406567493112948,
      "avg_duration_per_task_ms": 444708.36363636365,
      "avg_output_tokens_per_task": 31880.327364554636,
      "last_run_timestamp": "2026-10-10T08:22:15.514Z"
    },
    {
      "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": 0.8051542427745665,
      "accuracy_std_dev": 0.11605686870967129,
      "run_count": 7,
      "total_questions": 1384,
      "avg_cost_per_task_usd": 0.12536348916184972,
      "avg_duration_per_task_ms": 697081.9812138728,
      "avg_output_tokens_per_task": 39514.423410404626,
      "last_run_timestamp": "2026-10-10T16:19:37.317Z"
    },
    {
      "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": 0.805068,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 171,
      "avg_cost_per_task_usd": 0.07262640935672515,
      "avg_duration_per_task_ms": 352658,
      "avg_output_tokens_per_task": 15572.333333333334,
      "last_run_timestamp": "2026-07-18T08:30:38.228Z"
    },
    {
      "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": "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": 0.8039654444444444,
      "accuracy_std_dev": 0.022307357058010714,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.3361001139169472,
      "avg_duration_per_task_ms": 353464.27497194166,
      "avg_output_tokens_per_task": 22284.79012345679,
      "last_run_timestamp": "2026-10-10T08:23:05.021Z"
    },
    {
      "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": 0.8032717017208413,
      "accuracy_std_dev": 0.028736543137614867,
      "run_count": 8,
      "total_questions": 1569,
      "avg_cost_per_task_usd": 0.258291845124283,
      "avg_duration_per_task_ms": 931434.8087954111,
      "avg_output_tokens_per_task": 65094.10834926705,
      "last_run_timestamp": "2026-10-10T16:19:34.447Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "ModelRun",
      "accuracy": 0.8024693333333334,
      "accuracy_std_dev": 0.005722799567422317,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.054554207070707066,
      "avg_duration_per_task_ms": 52570.52356902357,
      "avg_output_tokens_per_task": 15236.12962962963,
      "last_run_timestamp": "2026-10-10T08:22:47.631Z"
    },
    {
      "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": 0.8024180909090908,
      "accuracy_std_dev": 0.10219243254179791,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.09314315335169882,
      "avg_duration_per_task_ms": 521856.9338842975,
      "avg_output_tokens_per_task": 34359.95913682277,
      "last_run_timestamp": "2026-10-10T16:22:50.286Z"
    },
    {
      "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": 0.8016821288180611,
      "accuracy_std_dev": 0.026730709555864766,
      "run_count": 5,
      "total_questions": 753,
      "avg_cost_per_task_usd": 0.050311730411686593,
      "avg_duration_per_task_ms": 763057.3041168659,
      "avg_output_tokens_per_task": 21804.07835325365,
      "last_run_timestamp": "2026-08-22T16:06:14.894Z"
    },
    {
      "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": 0.8011870934718102,
      "accuracy_std_dev": 0.030695975976321205,
      "run_count": 6,
      "total_questions": 674,
      "avg_cost_per_task_usd": 0.030323706231454008,
      "avg_duration_per_task_ms": 103892.11572700297,
      "avg_output_tokens_per_task": 31484.501483679524,
      "last_run_timestamp": "2026-08-29T19:44:46.543Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "auto-routing",
      "accuracy": 0.8009257499999999,
      "accuracy_std_dev": 0.04236236185209107,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.026970955808080805,
      "avg_duration_per_task_ms": 357603.6609848485,
      "avg_output_tokens_per_task": 25554.109217171717,
      "last_run_timestamp": "2026-10-10T16:10:34.758Z"
    },
    {
      "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": 0.8008437316455697,
      "accuracy_std_dev": 0.022930653266364256,
      "run_count": 6,
      "total_questions": 1185,
      "avg_cost_per_task_usd": 0.028163265822784812,
      "avg_duration_per_task_ms": 479213.82784810127,
      "avg_output_tokens_per_task": 23504.80253164557,
      "last_run_timestamp": "2026-08-29T18:09:57.600Z"
    },
    {
      "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": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Phala",
      "accuracy": 0.8006493194805195,
      "accuracy_std_dev": 0.04527692613517397,
      "run_count": 4,
      "total_questions": 770,
      "avg_cost_per_task_usd": 0.00850762987012987,
      "avg_duration_per_task_ms": 622509.1441558441,
      "avg_output_tokens_per_task": 19295.080519480518,
      "last_run_timestamp": "2026-08-29T19:08:46.030Z"
    },
    {
      "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": 0.799032,
      "accuracy_std_dev": 0.02239639471544471,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.028761780303030306,
      "avg_duration_per_task_ms": 300265.7859848485,
      "avg_output_tokens_per_task": 19022.03345959596,
      "last_run_timestamp": "2026-10-10T16:11:11.316Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Near AI",
      "accuracy": 0.7988215,
      "accuracy_std_dev": 0.04292950000000001,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.02789190404040404,
      "avg_duration_per_task_ms": 202775.87121212122,
      "avg_output_tokens_per_task": 12888.320707070707,
      "last_run_timestamp": "2026-10-10T08:25:24.843Z"
    },
    {
      "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": "deepseek/deepseek-v3.1-terminus",
      "model_slug": "deepseek/deepseek-v3.1-terminus",
      "model_name": "DeepSeek: DeepSeek V3.1 Terminus",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7979796666666668,
      "accuracy_std_dev": 0.009816464683830371,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.01347027946127946,
      "avg_duration_per_task_ms": 287878.5387205387,
      "avg_output_tokens_per_task": 14030.314814814816,
      "last_run_timestamp": "2026-08-29T19:12:57.906Z"
    },
    {
      "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": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "WandB",
      "accuracy": 0.797481,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 172,
      "avg_cost_per_task_usd": 0.07670138953488372,
      "avg_duration_per_task_ms": 240811.48837209304,
      "avg_output_tokens_per_task": 21222.406976744187,
      "last_run_timestamp": "2026-07-18T08:37:00.147Z"
    },
    {
      "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": 0.7972860381679391,
      "accuracy_std_dev": 0.010877999999999999,
      "run_count": 2,
      "total_questions": 393,
      "avg_cost_per_task_usd": 0.01688720610687023,
      "avg_duration_per_task_ms": 167801.82442748093,
      "avg_output_tokens_per_task": 6563.4376590330785,
      "last_run_timestamp": "2026-08-01T17:44:11.144Z"
    },
    {
      "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": 0.7972582857142857,
      "accuracy_std_dev": 0.005891180301683102,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.014391279220779221,
      "avg_duration_per_task_ms": 318122.0873015873,
      "avg_output_tokens_per_task": 14823.277056277057,
      "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-r1-0528",
      "model_slug": "deepseek/deepseek-r1-0528",
      "model_name": "DeepSeek: R1 0528",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7971204358638745,
      "accuracy_std_dev": 0.006421655480433002,
      "run_count": 8,
      "total_questions": 1528,
      "avg_cost_per_task_usd": 0.05209169568062827,
      "avg_duration_per_task_ms": 1143446.0909685863,
      "avg_output_tokens_per_task": 23739.34227748691,
      "last_run_timestamp": "2026-10-10T16:19:37.314Z"
    },
    {
      "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-v3.1-terminus",
      "model_slug": "deepseek/deepseek-v3.1-terminus",
      "model_name": "DeepSeek: DeepSeek V3.1 Terminus",
      "provider_name": "StreamLake",
      "accuracy": 0.7968575,
      "accuracy_std_dev": 0.009506569460992069,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.015575840067340066,
      "avg_duration_per_task_ms": 306607.4048821549,
      "avg_output_tokens_per_task": 14907.355218855218,
      "last_run_timestamp": "2026-08-29T16:10:33.070Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "DekaLLM",
      "accuracy": 0.7966842408376964,
      "accuracy_std_dev": 0.016353751013826204,
      "run_count": 3,
      "total_questions": 573,
      "avg_cost_per_task_usd": 0.034378357766143106,
      "avg_duration_per_task_ms": 555130.9528795811,
      "avg_output_tokens_per_task": 34351.207678883075,
      "last_run_timestamp": "2026-10-10T16:22:09.823Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "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": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Phala",
      "accuracy": 0.7961511097178684,
      "accuracy_std_dev": 0.09902156060353723,
      "run_count": 10,
      "total_questions": 1914,
      "avg_cost_per_task_usd": 0.07970391797283176,
      "avg_duration_per_task_ms": 762521.6421107628,
      "avg_output_tokens_per_task": 29044.879310344826,
      "last_run_timestamp": "2026-10-10T16:29:40.552Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "Novita",
      "accuracy": 0.7956228,
      "accuracy_std_dev": 0.021522407276138952,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.006844225252525253,
      "avg_duration_per_task_ms": 363476.91515151516,
      "avg_output_tokens_per_task": 16652.98888888889,
      "last_run_timestamp": "2026-08-29T19:20:50.265Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-thinking-2507",
      "model_slug": "qwen/qwen3-235b-a22b-thinking-2507",
      "model_name": "Qwen: Qwen3 235B A22B Thinking 2507",
      "provider_name": "Alibaba",
      "accuracy": 0.7956228,
      "accuracy_std_dev": 0.009072301348610534,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.051206081818181814,
      "avg_duration_per_task_ms": 443200.41616161616,
      "avg_output_tokens_per_task": 24093.476767676766,
      "last_run_timestamp": "2026-08-22T17:18:01.115Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-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": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "SambaNova",
      "accuracy": 0.794007441011236,
      "accuracy_std_dev": 0.03030512423976348,
      "run_count": 4,
      "total_questions": 356,
      "avg_cost_per_task_usd": 0.07293162640449438,
      "avg_duration_per_task_ms": 115799.61235955056,
      "avg_output_tokens_per_task": 31686.629213483146,
      "last_run_timestamp": "2026-08-15T14:54:40.470Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-r1-0528",
      "model_slug": "deepseek/deepseek-r1-0528",
      "model_name": "DeepSeek: R1 0528",
      "provider_name": "DeepInfra",
      "accuracy": 0.7937475724725943,
      "accuracy_std_dev": 0.018702947078792137,
      "run_count": 9,
      "total_questions": 1642,
      "avg_cost_per_task_usd": 0.042084668696711326,
      "avg_duration_per_task_ms": 722711.6863580999,
      "avg_output_tokens_per_task": 19447.1790499391,
      "last_run_timestamp": "2026-10-10T16:21:30.038Z"
    },
    {
      "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": 0.7929292857142858,
      "accuracy_std_dev": 0.015164807649990954,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.017839756854256855,
      "avg_duration_per_task_ms": 60613.86363636364,
      "avg_output_tokens_per_task": 5792.939393939394,
      "last_run_timestamp": "2026-10-10T16:11:11.318Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "Io Net",
      "accuracy": 0.7925169999999999,
      "accuracy_std_dev": 0.001701000000000008,
      "run_count": 2,
      "total_questions": 392,
      "avg_cost_per_task_usd": 0.01095936224489796,
      "avg_duration_per_task_ms": 929468.150510204,
      "avg_output_tokens_per_task": 38601.85459183674,
      "last_run_timestamp": "2026-10-10T16:21:21.465Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta/muse-glimmer-30b-20260810",
      "model_slug": "meta/muse-glimmer-30b",
      "model_name": "Meta: Muse Glimmer 30B",
      "provider_name": "DeepInfra",
      "accuracy": 0.7920900579268293,
      "accuracy_std_dev": 0.011512971285033247,
      "run_count": 10,
      "total_questions": 1968,
      "avg_cost_per_task_usd": 0.01824498475609756,
      "avg_duration_per_task_ms": 181543.2469512195,
      "avg_output_tokens_per_task": 15051.83943089431,
      "last_run_timestamp": "2026-10-08T22:29:11.504Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Reka",
      "accuracy": 0.7906976325581395,
      "accuracy_std_dev": 0.021874000000000005,
      "run_count": 2,
      "total_questions": 215,
      "avg_cost_per_task_usd": 0.04791340930232558,
      "avg_duration_per_task_ms": 264094.9441860465,
      "avg_output_tokens_per_task": 20226.018604651163,
      "last_run_timestamp": "2026-10-10T16:22:35.230Z"
    },
    {
      "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": 0.7905518785249456,
      "accuracy_std_dev": 0.008184215100810029,
      "run_count": 7,
      "total_questions": 1383,
      "avg_cost_per_task_usd": 0.1894372812725958,
      "avg_duration_per_task_ms": 850759.7664497469,
      "avg_output_tokens_per_task": 47361.36731742589,
      "last_run_timestamp": "2026-10-10T16:22:21.156Z"
    },
    {
      "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": 0.7905404425675676,
      "accuracy_std_dev": 0.02562269386086286,
      "run_count": 4,
      "total_questions": 592,
      "avg_cost_per_task_usd": 0.010724150337837838,
      "avg_duration_per_task_ms": 1222418.4949324324,
      "avg_output_tokens_per_task": 38033.55743243243,
      "last_run_timestamp": "2026-10-10T08:14:41.347Z"
    },
    {
      "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": 0.7903146482300885,
      "accuracy_std_dev": 0.04010989916950373,
      "run_count": 7,
      "total_questions": 1356,
      "avg_cost_per_task_usd": 0.14060938569321535,
      "avg_duration_per_task_ms": 1540325.2824483777,
      "avg_output_tokens_per_task": 46576.46017699115,
      "last_run_timestamp": "2026-10-03T16:58:27.160Z"
    },
    {
      "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": 0.7902677562043796,
      "accuracy_std_dev": 0.015064519424931276,
      "run_count": 11,
      "total_questions": 2055,
      "avg_cost_per_task_usd": 0.0048938481751824815,
      "avg_duration_per_task_ms": 1278859.9883211679,
      "avg_output_tokens_per_task": 32183.068613138686,
      "last_run_timestamp": "2026-10-10T16:21:36.238Z"
    },
    {
      "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": 0.7899609875730994,
      "accuracy_std_dev": 0.0876067404547278,
      "run_count": 7,
      "total_questions": 1368,
      "avg_cost_per_task_usd": 0.12032361038011698,
      "avg_duration_per_task_ms": 899417.8004385965,
      "avg_output_tokens_per_task": 33979.93567251462,
      "last_run_timestamp": "2026-10-10T16:23:26.035Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "Venice",
      "accuracy": 0.7892815833333334,
      "accuracy_std_dev": 0.01859263003835273,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.004778486531986532,
      "avg_duration_per_task_ms": 256643.10227272726,
      "avg_output_tokens_per_task": 31311.578703703704,
      "last_run_timestamp": "2026-10-10T16:23:45.466Z"
    },
    {
      "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": 0.788807633295838,
      "accuracy_std_dev": 0.019769970874916862,
      "run_count": 9,
      "total_questions": 1778,
      "avg_cost_per_task_usd": 0.062063327896512935,
      "avg_duration_per_task_ms": 174797.196287964,
      "avg_output_tokens_per_task": 19478.599550056242,
      "last_run_timestamp": "2026-10-10T08:21:28.045Z"
    },
    {
      "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": 0.788428937579618,
      "accuracy_std_dev": 0.09448363391936922,
      "run_count": 8,
      "total_questions": 1570,
      "avg_cost_per_task_usd": 0.05568409044585987,
      "avg_duration_per_task_ms": 1688216.5210191084,
      "avg_output_tokens_per_task": 18918.31210191083,
      "last_run_timestamp": "2026-10-10T16:19:34.421Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Venice",
      "accuracy": 0.788165837398374,
      "accuracy_std_dev": 0.004809000000000008,
      "run_count": 2,
      "total_questions": 369,
      "avg_cost_per_task_usd": 0.0017273360433604334,
      "avg_duration_per_task_ms": 800398.4254742548,
      "avg_output_tokens_per_task": 11344.837398373984,
      "last_run_timestamp": "2026-10-10T16:23:10.694Z"
    },
    {
      "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": 0.78787875,
      "accuracy_std_dev": 0.011355865409007798,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.01991093813131313,
      "avg_duration_per_task_ms": 480055.4835858586,
      "avg_output_tokens_per_task": 26459.91287878788,
      "last_run_timestamp": "2026-10-10T16:20:24.760Z"
    },
    {
      "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": 0.7875786128564525,
      "accuracy_std_dev": 0.015275190978232898,
      "run_count": 11,
      "total_questions": 2069,
      "avg_cost_per_task_usd": 0.008747213629772837,
      "avg_duration_per_task_ms": 399214.60318994685,
      "avg_output_tokens_per_task": 34637.547124214594,
      "last_run_timestamp": "2026-10-10T16:21:37.697Z"
    },
    {
      "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": 0.7875725454545454,
      "accuracy_std_dev": 0.0347531688845875,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.04297859274563819,
      "avg_duration_per_task_ms": 486549.3893480257,
      "avg_output_tokens_per_task": 42831.01193755739,
      "last_run_timestamp": "2026-10-10T16:19:57.542Z"
    },
    {
      "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": 0.7853823536121673,
      "accuracy_std_dev": 0.0034548367906458246,
      "run_count": 4,
      "total_questions": 789,
      "avg_cost_per_task_usd": 0.013724529784537389,
      "avg_duration_per_task_ms": 462082.7122940431,
      "avg_output_tokens_per_task": 14299.747782002534,
      "last_run_timestamp": "2026-08-15T14:56:55.269Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7839506666666667,
      "accuracy_std_dev": 0.054971360848677896,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.0028603939393939395,
      "avg_duration_per_task_ms": 827726.2415824916,
      "avg_output_tokens_per_task": 17815.154882154882,
      "last_run_timestamp": "2026-10-10T16:11:11.297Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Phala",
      "accuracy": 0.7835498571428571,
      "accuracy_std_dev": 0.056925857417292156,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.17942975324675325,
      "avg_duration_per_task_ms": 1414227.1096681096,
      "avg_output_tokens_per_task": 45704.106782106785,
      "last_run_timestamp": "2026-10-03T17:00:45.414Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "CoreWeave",
      "accuracy": 0.783249125,
      "accuracy_std_dev": 0.011716709034937033,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.021132419191919192,
      "avg_duration_per_task_ms": 279205.4166666667,
      "avg_output_tokens_per_task": 12569.390782828283,
      "last_run_timestamp": "2026-10-10T16:11:11.285Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "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": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7828282857142858,
      "accuracy_std_dev": 0.12584004230395515,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.14693832467532467,
      "avg_duration_per_task_ms": 1871581.0685425685,
      "avg_output_tokens_per_task": 51842.28354978355,
      "last_run_timestamp": "2026-10-10T16:26:01.027Z"
    },
    {
      "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": 0.7822926363636363,
      "accuracy_std_dev": 0.04547173391594292,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.19412152387511478,
      "avg_duration_per_task_ms": 821398.3898071626,
      "avg_output_tokens_per_task": 64866.6712580349,
      "last_run_timestamp": "2026-10-10T08:27:35.179Z"
    },
    {
      "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": 0.7821018375634519,
      "accuracy_std_dev": 0.023657804732591572,
      "run_count": 9,
      "total_questions": 1773,
      "avg_cost_per_task_usd": 0.05520209024252679,
      "avg_duration_per_task_ms": 511206.14438804286,
      "avg_output_tokens_per_task": 23994.640721940214,
      "last_run_timestamp": "2026-10-10T16:19:34.450Z"
    },
    {
      "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": 0.7814065297191379,
      "accuracy_std_dev": 0.1322181089733929,
      "run_count": 8,
      "total_questions": 1531,
      "avg_cost_per_task_usd": 0.023490106466361852,
      "avg_duration_per_task_ms": 346686.11430437624,
      "avg_output_tokens_per_task": 23958.977139124756,
      "last_run_timestamp": "2026-10-10T16:20:24.759Z"
    },
    {
      "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": 0.781374562749004,
      "accuracy_std_dev": 0.09476171410543395,
      "run_count": 10,
      "total_questions": 1004,
      "avg_cost_per_task_usd": 0.004642227091633467,
      "avg_duration_per_task_ms": 236467.00996015937,
      "avg_output_tokens_per_task": 19258.703187250994,
      "last_run_timestamp": "2026-10-10T16:20:22.051Z"
    },
    {
      "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": 0.781145,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.020832949494949494,
      "avg_duration_per_task_ms": 266171.0505050505,
      "avg_output_tokens_per_task": 12380.515151515152,
      "last_run_timestamp": "2026-07-18T16:16:42.667Z"
    },
    {
      "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": 0.7790441154411766,
      "accuracy_std_dev": 0.014485910522101155,
      "run_count": 9,
      "total_questions": 1360,
      "avg_cost_per_task_usd": 0.05105861102941177,
      "avg_duration_per_task_ms": 1000104.9169117647,
      "avg_output_tokens_per_task": 26002.289705882355,
      "last_run_timestamp": "2026-10-10T16:22:34.468Z"
    },
    {
      "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": 0.7784707032418954,
      "accuracy_std_dev": 0.1210615547882977,
      "run_count": 5,
      "total_questions": 802,
      "avg_cost_per_task_usd": 0.01958750872817955,
      "avg_duration_per_task_ms": 560824.7643391521,
      "avg_output_tokens_per_task": 20557.403990024937,
      "last_run_timestamp": "2026-10-10T08:36:49.489Z"
    },
    {
      "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": 0.7784511999999999,
      "accuracy_std_dev": 0.0206074889047647,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.05598715454545455,
      "avg_duration_per_task_ms": 643404.0878787879,
      "avg_output_tokens_per_task": 23899.834343434344,
      "last_run_timestamp": "2026-08-22T17:09:41.940Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7777777142857143,
      "accuracy_std_dev": 0.011167032598735643,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.018922173881673884,
      "avg_duration_per_task_ms": 673737.0562770562,
      "avg_output_tokens_per_task": 47091.12914862915,
      "last_run_timestamp": "2026-10-10T16:25:17.654Z"
    },
    {
      "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": 0.7774715454545453,
      "accuracy_std_dev": 0.019178897604889684,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.005792984389348025,
      "avg_duration_per_task_ms": 704409.8140495868,
      "avg_output_tokens_per_task": 36399.0564738292,
      "last_run_timestamp": "2026-10-10T16:20:12.054Z"
    },
    {
      "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": "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": 0.777134539797395,
      "accuracy_std_dev": 0.02043543185504141,
      "run_count": 7,
      "total_questions": 1382,
      "avg_cost_per_task_usd": 0.022169380607814763,
      "avg_duration_per_task_ms": 860418.1266280753,
      "avg_output_tokens_per_task": 38276.162807525325,
      "last_run_timestamp": "2026-10-10T16:27:48.414Z"
    },
    {
      "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": "inclusionai/ling-3.0-flash-20260723",
      "model_slug": "inclusionai/ling-3.0-flash",
      "model_name": "inclusionAI: Ling 3.0 Flash",
      "provider_name": "DeepInfra",
      "accuracy": 0.7741139703903096,
      "accuracy_std_dev": 0.09718478875928833,
      "run_count": 4,
      "total_questions": 743,
      "avg_cost_per_task_usd": 0.004475604306864065,
      "avg_duration_per_task_ms": 312870.95827725437,
      "avg_output_tokens_per_task": 22926.07267833109,
      "last_run_timestamp": "2026-10-08T22:27:01.555Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "DeepInfra",
      "accuracy": 0.7739002377332412,
      "accuracy_std_dev": 0.03919556245291041,
      "run_count": 9,
      "total_questions": 1447,
      "avg_cost_per_task_usd": 0.05904209744298548,
      "avg_duration_per_task_ms": 844538.2218382861,
      "avg_output_tokens_per_task": 50153.10711817553,
      "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-r1-0528",
      "model_slug": "deepseek/deepseek-r1-0528",
      "model_name": "DeepSeek: R1 0528",
      "provider_name": "auto-routing",
      "accuracy": 0.7737795,
      "accuracy_std_dev": 0.0353227557921236,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.05213979987373737,
      "avg_duration_per_task_ms": 865410.4703282828,
      "avg_output_tokens_per_task": 23453.041035353537,
      "last_run_timestamp": "2026-10-10T16:19:57.542Z"
    },
    {
      "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": 0.7736572813299233,
      "accuracy_std_dev": 0.015466646781596444,
      "run_count": 3,
      "total_questions": 391,
      "avg_cost_per_task_usd": 0.0087913273657289,
      "avg_duration_per_task_ms": 647179.7749360614,
      "avg_output_tokens_per_task": 29178.061381074167,
      "last_run_timestamp": "2026-10-10T16:25:53.261Z"
    },
    {
      "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": 0.7730078333333333,
      "accuracy_std_dev": 0.02074231892706846,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.015851826599326598,
      "avg_duration_per_task_ms": 424890.48232323234,
      "avg_output_tokens_per_task": 15979.028619528619,
      "last_run_timestamp": "2026-08-29T18:21:38.666Z"
    },
    {
      "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": 0.7728757843137255,
      "accuracy_std_dev": 0.0424837728239854,
      "run_count": 5,
      "total_questions": 714,
      "avg_cost_per_task_usd": 0.13942839355742298,
      "avg_duration_per_task_ms": 1052327.7044817926,
      "avg_output_tokens_per_task": 40065.01820728291,
      "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": "BaseTen",
      "accuracy": 0.7722222999999999,
      "accuracy_std_dev": 0.010833447964983244,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.014700993434343435,
      "avg_duration_per_task_ms": 139170.2106060606,
      "avg_output_tokens_per_task": 29356.293434343435,
      "last_run_timestamp": "2026-10-10T16:22:35.299Z"
    },
    {
      "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": 0.7715388278781039,
      "accuracy_std_dev": 0.007999827803702316,
      "run_count": 9,
      "total_questions": 1772,
      "avg_cost_per_task_usd": 0.011440101015801354,
      "avg_duration_per_task_ms": 999816.4932279909,
      "avg_output_tokens_per_task": 11904.952031602708,
      "last_run_timestamp": "2026-10-10T16:20:24.761Z"
    },
    {
      "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": "Parasail",
      "accuracy": 0.7710935525846703,
      "accuracy_std_dev": 0.018076003519091888,
      "run_count": 6,
      "total_questions": 1122,
      "avg_cost_per_task_usd": 0.0213185614973262,
      "avg_duration_per_task_ms": 243272.5650623886,
      "avg_output_tokens_per_task": 28339.197860962566,
      "last_run_timestamp": "2026-10-10T16:21:03.979Z"
    },
    {
      "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": 0.7704125,
      "accuracy_std_dev": 0.011714588949254664,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.005121838383838384,
      "avg_duration_per_task_ms": 1054586.8964646466,
      "avg_output_tokens_per_task": 28369.739898989897,
      "last_run_timestamp": "2026-10-10T16:19:34.425Z"
    },
    {
      "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": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Reka",
      "accuracy": 0.7701042403314916,
      "accuracy_std_dev": 0.009818378139451328,
      "run_count": 11,
      "total_questions": 2172,
      "avg_cost_per_task_usd": 0.018828466850828732,
      "avg_duration_per_task_ms": 104392.93784530387,
      "avg_output_tokens_per_task": 5944.6809392265195,
      "last_run_timestamp": "2026-10-10T08:22:24.148Z"
    },
    {
      "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": 0.7693388734177216,
      "accuracy_std_dev": 0.018315649062573065,
      "run_count": 6,
      "total_questions": 1185,
      "avg_cost_per_task_usd": 0.058541679324894506,
      "avg_duration_per_task_ms": 602157.2911392405,
      "avg_output_tokens_per_task": 26894.77299578059,
      "last_run_timestamp": "2026-08-29T18:26:43.144Z"
    },
    {
      "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": 0.7692142947976879,
      "accuracy_std_dev": 0.049901425655830466,
      "run_count": 8,
      "total_questions": 1557,
      "avg_cost_per_task_usd": 0.2430246191393706,
      "avg_duration_per_task_ms": 900127.0802825948,
      "avg_output_tokens_per_task": 70448.42453436095,
      "last_run_timestamp": "2026-10-10T16:19:34.420Z"
    },
    {
      "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": 0.7691335028462997,
      "accuracy_std_dev": 0.0413125157602088,
      "run_count": 8,
      "total_questions": 1581,
      "avg_cost_per_task_usd": 0.14881955850727388,
      "avg_duration_per_task_ms": 1030577.8734977862,
      "avg_output_tokens_per_task": 67777.00316255534,
      "last_run_timestamp": "2026-10-10T16:11:14.272Z"
    },
    {
      "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": 0.768254,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 105,
      "avg_cost_per_task_usd": 0.04709281904761905,
      "avg_duration_per_task_ms": 1085275.6761904762,
      "avg_output_tokens_per_task": 21896.628571428573,
      "last_run_timestamp": "2026-07-18T16:43:10.588Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "AkashML",
      "accuracy": 0.7682437087198516,
      "accuracy_std_dev": 0.017109400315529023,
      "run_count": 11,
      "total_questions": 2156,
      "avg_cost_per_task_usd": 0.00780624628942486,
      "avg_duration_per_task_ms": 455549.38729128015,
      "avg_output_tokens_per_task": 27461.665584415583,
      "last_run_timestamp": "2026-10-10T16:19:40.682Z"
    },
    {
      "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": 0.7682173900573613,
      "accuracy_std_dev": 0.06513957389797216,
      "run_count": 8,
      "total_questions": 1569,
      "avg_cost_per_task_usd": 0.260235,
      "avg_duration_per_task_ms": 959583.7954110899,
      "avg_output_tokens_per_task": 74013.42192479286,
      "last_run_timestamp": "2026-10-10T16:11:11.315Z"
    },
    {
      "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": 0.7680976249999999,
      "accuracy_std_dev": 0.03807231169740517,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.1701674728535354,
      "avg_duration_per_task_ms": 546025.8289141414,
      "avg_output_tokens_per_task": 52982.59659090909,
      "last_run_timestamp": "2026-10-10T16:20:22.051Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4-sonnet-20250522",
      "model_slug": "anthropic/claude-sonnet-4",
      "model_name": "Anthropic: Claude Sonnet 4",
      "provider_name": "Google",
      "accuracy": 0.7672559999999999,
      "accuracy_std_dev": 0.008469894233892938,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.27037940151515155,
      "avg_duration_per_task_ms": 263755.6675084175,
      "avg_output_tokens_per_task": 17837.74579124579,
      "last_run_timestamp": "2026-09-01T13:23:02.059Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "GMICloud",
      "accuracy": 0.7671156666666666,
      "accuracy_std_dev": 0.03093018545405481,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.013063415824915825,
      "avg_duration_per_task_ms": 936092.1515151515,
      "avg_output_tokens_per_task": 46359.02693602694,
      "last_run_timestamp": "2026-10-10T08:15:34.059Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4-sonnet-20250522",
      "model_slug": "anthropic/claude-sonnet-4",
      "model_name": "Anthropic: Claude Sonnet 4",
      "provider_name": "auto-routing",
      "accuracy": 0.7668349999999999,
      "accuracy_std_dev": 0.01935999999999999,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.2670539015151515,
      "avg_duration_per_task_ms": 177095.6616161616,
      "avg_output_tokens_per_task": 17616.689393939392,
      "last_run_timestamp": "2026-09-01T13:19:59.945Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Mara",
      "accuracy": 0.7665065498199279,
      "accuracy_std_dev": 0.0370963104843463,
      "run_count": 7,
      "total_questions": 833,
      "avg_cost_per_task_usd": 0.012696330132052819,
      "avg_duration_per_task_ms": 53169.84753901561,
      "avg_output_tokens_per_task": 16839.417767106843,
      "last_run_timestamp": "2026-10-10T16:20:08.978Z"
    },
    {
      "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Crusoe",
      "accuracy": 0.7659934999999999,
      "accuracy_std_dev": 0.008417499999999967,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.007677666666666667,
      "avg_duration_per_task_ms": 154673.01515151514,
      "avg_output_tokens_per_task": 30515.89393939394,
      "last_run_timestamp": "2026-10-10T16:21:31.490Z"
    },
    {
      "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": "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": 0.7652717142857143,
      "accuracy_std_dev": 0.015187772344077012,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.019384612554112553,
      "avg_duration_per_task_ms": 1111853.9927849928,
      "avg_output_tokens_per_task": 48423.183982683986,
      "last_run_timestamp": "2026-10-10T16:21:25.172Z"
    },
    {
      "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": 0.7650936065232479,
      "accuracy_std_dev": 0.10998854890673562,
      "run_count": 8,
      "total_questions": 1441,
      "avg_cost_per_task_usd": 0.036296903539208884,
      "avg_duration_per_task_ms": 147941.55586398335,
      "avg_output_tokens_per_task": 29262.30326162387,
      "last_run_timestamp": "2026-10-10T16:22:02.247Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "SambaNova",
      "accuracy": 0.7638156348684211,
      "accuracy_std_dev": 0.006318496220175736,
      "run_count": 8,
      "total_questions": 1520,
      "avg_cost_per_task_usd": 0.024435126315789476,
      "avg_duration_per_task_ms": 70438.73486842106,
      "avg_output_tokens_per_task": 25588.723684210527,
      "last_run_timestamp": "2026-10-10T16:21:31.026Z"
    },
    {
      "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": 0.7630074302030456,
      "accuracy_std_dev": 0.05691748932171465,
      "run_count": 8,
      "total_questions": 1576,
      "avg_cost_per_task_usd": 0.007098315989847716,
      "avg_duration_per_task_ms": 964817.2912436548,
      "avg_output_tokens_per_task": 18426.44416243655,
      "last_run_timestamp": "2026-10-10T16:11:11.298Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-coder-next-2025-02-03",
      "model_slug": "qwen/qwen3-coder-next",
      "model_name": "Qwen: Qwen3 Coder Next",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.7626260000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.022027883838383836,
      "avg_duration_per_task_ms": 90904.42929292929,
      "avg_output_tokens_per_task": 14522.767676767677,
      "last_run_timestamp": "2026-10-04T03:02:47.748Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "InferenceNet",
      "accuracy": 0.7626259999999999,
      "accuracy_std_dev": 0.01346799999999998,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.01717197474747475,
      "avg_duration_per_task_ms": 622307.4974747475,
      "avg_output_tokens_per_task": 61038.34090909091,
      "last_run_timestamp": "2026-10-03T10:05:09.589Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "Novita",
      "accuracy": 0.7613855279025016,
      "accuracy_std_dev": 0.0736067708315614,
      "run_count": 8,
      "total_questions": 1559,
      "avg_cost_per_task_usd": 0.24665575304682488,
      "avg_duration_per_task_ms": 1221066.671584349,
      "avg_output_tokens_per_task": 78685.70686337395,
      "last_run_timestamp": "2026-10-10T16:19:34.422Z"
    },
    {
      "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": 0.7611530000000001,
      "accuracy_std_dev": 0.04936588290509955,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.01190072095959596,
      "avg_duration_per_task_ms": 817020.8775252525,
      "avg_output_tokens_per_task": 36142.90782828283,
      "last_run_timestamp": "2026-10-10T08:14:45.919Z"
    },
    {
      "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": "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": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7604232514285715,
      "accuracy_std_dev": 0.03409113144550026,
      "run_count": 8,
      "total_questions": 1575,
      "avg_cost_per_task_usd": 0.11787924444444446,
      "avg_duration_per_task_ms": 816938.3003174603,
      "avg_output_tokens_per_task": 46021.84,
      "last_run_timestamp": "2026-10-10T16:11:11.318Z"
    },
    {
      "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": 0.7596333333333334,
      "accuracy_std_dev": 0.09860502035562557,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.03538440628507295,
      "avg_duration_per_task_ms": 1290567.9764309765,
      "avg_output_tokens_per_task": 20956.203703703704,
      "last_run_timestamp": "2026-10-10T16:22:34.911Z"
    },
    {
      "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": "Groq",
      "accuracy": 0.75925925,
      "accuracy_std_dev": 0.006627946132664344,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.017700304292929292,
      "avg_duration_per_task_ms": 80251.25441919192,
      "avg_output_tokens_per_task": 29259.58712121212,
      "last_run_timestamp": "2026-10-10T16:22:35.205Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "mistralai/mistral-small-2603",
      "model_slug": "mistralai/mistral-small-2603",
      "model_name": "Mistral: Mistral Small 4",
      "provider_name": "Venice",
      "accuracy": 0.759259,
      "accuracy_std_dev": 0.003367000000000009,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.018210444444444442,
      "avg_duration_per_task_ms": 254627.46717171717,
      "avg_output_tokens_per_task": 24079.401515151516,
      "last_run_timestamp": "2026-09-01T12:59:55.618Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/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": "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": 0.7586981111111112,
      "accuracy_std_dev": 0.018373432640300367,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.016781338945005612,
      "avg_duration_per_task_ms": 1197308.150392817,
      "avg_output_tokens_per_task": 51632.826599326596,
      "last_run_timestamp": "2026-10-10T16:22:45.582Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-2.5-flash",
      "model_slug": "google/gemini-2.5-flash",
      "model_name": "Google: Gemini 2.5 Flash",
      "provider_name": "Google AI Studio",
      "accuracy": 0.758698,
      "accuracy_std_dev": 0.012009662443216293,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.056970961279461274,
      "avg_duration_per_task_ms": 99388.35016835017,
      "avg_output_tokens_per_task": 22699.07239057239,
      "last_run_timestamp": "2026-10-01T19:11:42.323Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7586279375000001,
      "accuracy_std_dev": 0.013868153578742699,
      "run_count": 16,
      "total_questions": 3168,
      "avg_cost_per_task_usd": 0.01862196054292929,
      "avg_duration_per_task_ms": 165357.07323232322,
      "avg_output_tokens_per_task": 30815.23611111111,
      "last_run_timestamp": "2026-10-10T16:25:17.748Z"
    },
    {
      "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": 0.7585079293193718,
      "accuracy_std_dev": 0.0038942041789690614,
      "run_count": 4,
      "total_questions": 764,
      "avg_cost_per_task_usd": 0.01819221335078534,
      "avg_duration_per_task_ms": 1032018.7041884817,
      "avg_output_tokens_per_task": 55015.78272251309,
      "last_run_timestamp": "2026-10-10T16:22:44.205Z"
    },
    {
      "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": "Nebius",
      "accuracy": 0.757365375,
      "accuracy_std_dev": 0.009939804285516652,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.018617486111111112,
      "avg_duration_per_task_ms": 159086.00441919192,
      "avg_output_tokens_per_task": 30836.404671717173,
      "last_run_timestamp": "2026-10-10T16:19:34.445Z"
    },
    {
      "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": 0.7569024000000001,
      "accuracy_std_dev": 0.0779853723171211,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.017876224242424243,
      "avg_duration_per_task_ms": 1933669.407070707,
      "avg_output_tokens_per_task": 17409.9303030303,
      "last_run_timestamp": "2026-10-10T16:21:15.010Z"
    },
    {
      "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": 0.7561492633587785,
      "accuracy_std_dev": 0.013604073603497425,
      "run_count": 8,
      "total_questions": 1572,
      "avg_cost_per_task_usd": 0.01678382569974555,
      "avg_duration_per_task_ms": 286656.0801526718,
      "avg_output_tokens_per_task": 28394.44720101781,
      "last_run_timestamp": "2026-10-10T16:19:40.649Z"
    },
    {
      "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": 0.7558386418966737,
      "accuracy_std_dev": 0.03258665299412283,
      "run_count": 8,
      "total_questions": 1413,
      "avg_cost_per_task_usd": 0.007369956829440905,
      "avg_duration_per_task_ms": 1050257.7034677991,
      "avg_output_tokens_per_task": 18996.81599433829,
      "last_run_timestamp": "2026-10-10T16:10:48.870Z"
    },
    {
      "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": 0.7556517142857143,
      "accuracy_std_dev": 0.007605047378725084,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.022500825396825398,
      "avg_duration_per_task_ms": 97229.08225108225,
      "avg_output_tokens_per_task": 14839.808080808081,
      "last_run_timestamp": "2026-10-03T16:56:57.458Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Io Net",
      "accuracy": 0.7553378326797385,
      "accuracy_std_dev": 0.026370845649125493,
      "run_count": 4,
      "total_questions": 765,
      "avg_cost_per_task_usd": 0.015925873202614377,
      "avg_duration_per_task_ms": 239486.98169934642,
      "avg_output_tokens_per_task": 49403.856209150326,
      "last_run_timestamp": "2026-10-10T08:36:35.786Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "DeepInfra",
      "accuracy": 0.7552047101754386,
      "accuracy_std_dev": 0.036034774046985296,
      "run_count": 9,
      "total_questions": 1425,
      "avg_cost_per_task_usd": 0.04509595859649123,
      "avg_duration_per_task_ms": 263424.8596491228,
      "avg_output_tokens_per_task": 16991.12350877193,
      "last_run_timestamp": "2026-10-10T08:22:13.853Z"
    },
    {
      "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": 0.75515575,
      "accuracy_std_dev": 0.06516829345385299,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.1108029071969697,
      "avg_duration_per_task_ms": 571979.9936868687,
      "avg_output_tokens_per_task": 33259.57323232323,
      "last_run_timestamp": "2026-10-10T16:21:36.325Z"
    },
    {
      "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": 0.7535393918918919,
      "accuracy_std_dev": 0.07402326310019032,
      "run_count": 8,
      "total_questions": 1554,
      "avg_cost_per_task_usd": 0.13893272265122267,
      "avg_duration_per_task_ms": 1010369.6351351351,
      "avg_output_tokens_per_task": 58604.81145431146,
      "last_run_timestamp": "2026-10-10T16:10:32.155Z"
    },
    {
      "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": 0.753367,
      "accuracy_std_dev": 0.004209000000000018,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.015499042929292928,
      "avg_duration_per_task_ms": 1045731.0252525252,
      "avg_output_tokens_per_task": 55053.381313131315,
      "last_run_timestamp": "2026-10-10T08:14:49.445Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "mistralai/mistral-small-2603",
      "model_slug": "mistralai/mistral-small-2603",
      "model_name": "Mistral: Mistral Small 4",
      "provider_name": "Mistral",
      "accuracy": 0.7528056666666667,
      "accuracy_std_dev": 0.004601995207395269,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.015185508417508417,
      "avg_duration_per_task_ms": 152622.4537037037,
      "avg_output_tokens_per_task": 24359.508417508416,
      "last_run_timestamp": "2026-09-17T02:17:49.284Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.7525250000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.008515843434343435,
      "avg_duration_per_task_ms": 110884.25757575757,
      "avg_output_tokens_per_task": 8312.732323232323,
      "last_run_timestamp": "2026-10-04T03:06:11.710Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "DeepInfra",
      "accuracy": 0.7524220592034444,
      "accuracy_std_dev": 0.028639192749568526,
      "run_count": 7,
      "total_questions": 929,
      "avg_cost_per_task_usd": 0.05453179763186222,
      "avg_duration_per_task_ms": 423386.32615715824,
      "avg_output_tokens_per_task": 15836.034445640473,
      "last_run_timestamp": "2026-09-19T08:53:38.031Z"
    },
    {
      "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": 0.7509663565217392,
      "accuracy_std_dev": 0.031143314111326263,
      "run_count": 7,
      "total_questions": 1380,
      "avg_cost_per_task_usd": 0.0046609500000000005,
      "avg_duration_per_task_ms": 782901.7971014492,
      "avg_output_tokens_per_task": 28031.711594202898,
      "last_run_timestamp": "2026-10-10T16:21:18.041Z"
    },
    {
      "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": 0.750842,
      "accuracy_std_dev": 0.003367000000000009,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.016549714646464646,
      "avg_duration_per_task_ms": 972340.9924242424,
      "avg_output_tokens_per_task": 58750.5101010101,
      "last_run_timestamp": "2026-10-10T08:22:25.593Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "GMICloud",
      "accuracy": 0.75076231097561,
      "accuracy_std_dev": 0.10261490128027215,
      "run_count": 10,
      "total_questions": 1968,
      "avg_cost_per_task_usd": 0.006507009654471546,
      "avg_duration_per_task_ms": 566183.218495935,
      "avg_output_tokens_per_task": 27810.678353658535,
      "last_run_timestamp": "2026-10-10T08:22:53.947Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-coder-next-2025-02-03",
      "model_slug": "qwen/qwen3-coder-next",
      "model_name": "Qwen: Qwen3 Coder Next",
      "provider_name": "StreamLake",
      "accuracy": 0.749368625,
      "accuracy_std_dev": 0.008643231874384451,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.013557867424242424,
      "avg_duration_per_task_ms": 96888.43560606061,
      "avg_output_tokens_per_task": 14901.809343434343,
      "last_run_timestamp": "2026-10-03T16:54:39.591Z"
    },
    {
      "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": 0.748737375,
      "accuracy_std_dev": 0.14136326117306566,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.02991949936868687,
      "avg_duration_per_task_ms": 136944.42866161617,
      "avg_output_tokens_per_task": 22918.396464646463,
      "last_run_timestamp": "2026-10-10T16:20:08.973Z"
    },
    {
      "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": 0.748737375,
      "accuracy_std_dev": 0.03686209625054409,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.052284164141414145,
      "avg_duration_per_task_ms": 944565.4507575758,
      "avg_output_tokens_per_task": 27129.21275252525,
      "last_run_timestamp": "2026-10-10T16:23:01.061Z"
    },
    {
      "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": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "Google",
      "accuracy": 0.7481442884615385,
      "accuracy_std_dev": 0.0227081467548543,
      "run_count": 5,
      "total_questions": 988,
      "avg_cost_per_task_usd": 0.020536520242914982,
      "avg_duration_per_task_ms": 151472.34615384616,
      "avg_output_tokens_per_task": 11878.324898785426,
      "last_run_timestamp": "2026-08-22T17:12:08.919Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "Wafer",
      "accuracy": 0.7481013645569621,
      "accuracy_std_dev": 0.004003499999999993,
      "run_count": 2,
      "total_questions": 395,
      "avg_cost_per_task_usd": 0.010130237974683544,
      "avg_duration_per_task_ms": 298779.6734177215,
      "avg_output_tokens_per_task": 60721.07341772152,
      "last_run_timestamp": "2026-10-10T08:39:12.933Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.747826043478261,
      "accuracy_std_dev": 0.06491713177676944,
      "run_count": 7,
      "total_questions": 1380,
      "avg_cost_per_task_usd": 0.010265418115942029,
      "avg_duration_per_task_ms": 524781.3072463768,
      "avg_output_tokens_per_task": 34823.62463768116,
      "last_run_timestamp": "2026-10-10T16:20:00.365Z"
    },
    {
      "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": 0.7478116,
      "accuracy_std_dev": 0.02151720496347049,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.08178134747474748,
      "avg_duration_per_task_ms": 885057.2202020203,
      "avg_output_tokens_per_task": 23371.92828282828,
      "last_run_timestamp": "2026-08-22T17:38:59.682Z"
    },
    {
      "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": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "Z.AI",
      "accuracy": 0.7471013608695652,
      "accuracy_std_dev": 0.06255498618570607,
      "run_count": 7,
      "total_questions": 1380,
      "avg_cost_per_task_usd": 0.2582489231884058,
      "avg_duration_per_task_ms": 1276224.2644927537,
      "avg_output_tokens_per_task": 82450.74855072463,
      "last_run_timestamp": "2026-10-10T16:20:06.465Z"
    },
    {
      "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": 0.7470487959527824,
      "accuracy_std_dev": 0.08033451694532995,
      "run_count": 9,
      "total_questions": 1779,
      "avg_cost_per_task_usd": 0.005070625070264193,
      "avg_duration_per_task_ms": 656330.5952782462,
      "avg_output_tokens_per_task": 33326.38673412029,
      "last_run_timestamp": "2026-09-26T18:04:35.621Z"
    },
    {
      "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": 0.746633,
      "accuracy_std_dev": 0.017677,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.017912724747474747,
      "avg_duration_per_task_ms": 70698.09595959596,
      "avg_output_tokens_per_task": 6119.75,
      "last_run_timestamp": "2026-08-01T17:45:27.258Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-35b-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": "qwen/qwen3-coder-next-2025-02-03",
      "model_slug": "qwen/qwen3-coder-next",
      "model_name": "Qwen: Qwen3 Coder Next",
      "provider_name": "auto-routing",
      "accuracy": 0.74642275,
      "accuracy_std_dev": 0.008945525344411051,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.01488374558080808,
      "avg_duration_per_task_ms": 107961.75883838384,
      "avg_output_tokens_per_task": 14400.517676767677,
      "last_run_timestamp": "2026-10-03T16:43:33.673Z"
    },
    {
      "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": 0.7461957648578811,
      "accuracy_std_dev": 0.05992671984854132,
      "run_count": 6,
      "total_questions": 1161,
      "avg_cost_per_task_usd": 0.1759588509905254,
      "avg_duration_per_task_ms": 1560060.6244616709,
      "avg_output_tokens_per_task": 42375.20930232558,
      "last_run_timestamp": "2026-10-10T16:23:46.181Z"
    },
    {
      "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": 0.7459782222222222,
      "accuracy_std_dev": 0.009438398095931553,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.006886203142536476,
      "avg_duration_per_task_ms": 163678.69079685747,
      "avg_output_tokens_per_task": 8003.4517396184065,
      "last_run_timestamp": "2026-10-03T16:58:32.590Z"
    },
    {
      "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": 0.7454796043395022,
      "accuracy_std_dev": 0.1116514794950676,
      "run_count": 8,
      "total_questions": 1567,
      "avg_cost_per_task_usd": 0.16252882514358646,
      "avg_duration_per_task_ms": 1066174.2463305679,
      "avg_output_tokens_per_task": 46648.48819400128,
      "last_run_timestamp": "2026-10-10T08:21:42.283Z"
    },
    {
      "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": 0.7434143503039515,
      "accuracy_std_dev": 0.004376274946591358,
      "run_count": 7,
      "total_questions": 1316,
      "avg_cost_per_task_usd": 0.009106649696048632,
      "avg_duration_per_task_ms": 140201.952887538,
      "avg_output_tokens_per_task": 11539.308510638299,
      "last_run_timestamp": "2026-10-03T17:03:40.665Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Nebius",
      "accuracy": 0.743266,
      "accuracy_std_dev": 0.006355106254029135,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.005372063762626263,
      "avg_duration_per_task_ms": 150638.75315656565,
      "avg_output_tokens_per_task": 8682.62563131313,
      "last_run_timestamp": "2026-10-10T16:21:31.113Z"
    },
    {
      "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": 0.743145857142857,
      "accuracy_std_dev": 0.011088924595914524,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.023026453102453104,
      "avg_duration_per_task_ms": 99585.75685425685,
      "avg_output_tokens_per_task": 15242.642857142857,
      "last_run_timestamp": "2026-10-03T16:58:08.462Z"
    },
    {
      "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/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Phala",
      "accuracy": 0.740951125,
      "accuracy_std_dev": 0.06026863513561074,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.01223913446969697,
      "avg_duration_per_task_ms": 98981.7702020202,
      "avg_output_tokens_per_task": 5413.777777777777,
      "last_run_timestamp": "2026-10-10T08:14:41.617Z"
    },
    {
      "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": 0.740741,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.02792859090909091,
      "avg_duration_per_task_ms": 682232.3585858586,
      "avg_output_tokens_per_task": 23364.934343434343,
      "last_run_timestamp": "2026-07-18T09:15:42.531Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Alibaba",
      "accuracy": 0.739899,
      "accuracy_std_dev": 0.010812416138403098,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.0048791994949494944,
      "avg_duration_per_task_ms": 108739.66098484848,
      "avg_output_tokens_per_task": 7956.729797979798,
      "last_run_timestamp": "2026-10-03T16:59:36.480Z"
    },
    {
      "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": 0.7392292244897959,
      "accuracy_std_dev": 0.110097,
      "run_count": 2,
      "total_questions": 294,
      "avg_cost_per_task_usd": 0.03822905782312925,
      "avg_duration_per_task_ms": 1403982.7517006802,
      "avg_output_tokens_per_task": 27800.656462585033,
      "last_run_timestamp": "2026-08-01T23:27:37.975Z"
    },
    {
      "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": 0.7390017868020304,
      "accuracy_std_dev": 0.059270287748631335,
      "run_count": 6,
      "total_questions": 1182,
      "avg_cost_per_task_usd": 0.01651792470389171,
      "avg_duration_per_task_ms": 1126017.0372250422,
      "avg_output_tokens_per_task": 41901.6192893401,
      "last_run_timestamp": "2026-10-10T16:29:23.083Z"
    },
    {
      "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": 0.738906892857143,
      "accuracy_std_dev": 0.016676086646452728,
      "run_count": 5,
      "total_questions": 616,
      "avg_cost_per_task_usd": 0.015153974025974027,
      "avg_duration_per_task_ms": 72126.25162337662,
      "avg_output_tokens_per_task": 8710.556818181818,
      "last_run_timestamp": "2026-10-10T16:10:51.752Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-2.5-flash",
      "model_slug": "google/gemini-2.5-flash",
      "model_name": "Google: Gemini 2.5 Flash",
      "provider_name": "Google",
      "accuracy": 0.738309,
      "accuracy_std_dev": 0.008136215650479735,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.059655349046015715,
      "avg_duration_per_task_ms": 125601.6632996633,
      "avg_output_tokens_per_task": 23853.966891133558,
      "last_run_timestamp": "2026-10-01T19:15:31.968Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "Makora",
      "accuracy": 0.7373739999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.020083924242424243,
      "avg_duration_per_task_ms": 186579.25757575757,
      "avg_output_tokens_per_task": 62586.09595959596,
      "last_run_timestamp": "2026-10-10T16:27:29.137Z"
    },
    {
      "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": 0.7356996188835285,
      "accuracy_std_dev": 0.006961170963817794,
      "run_count": 8,
      "total_questions": 1451,
      "avg_cost_per_task_usd": 0.00648628256374914,
      "avg_duration_per_task_ms": 228531.34045485873,
      "avg_output_tokens_per_task": 8024.518952446589,
      "last_run_timestamp": "2026-10-10T16:22:35.208Z"
    },
    {
      "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": 0.7348483749999999,
      "accuracy_std_dev": 0.07315972270644808,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.006649549242424242,
      "avg_duration_per_task_ms": 101923.77272727272,
      "avg_output_tokens_per_task": 16353.064393939394,
      "last_run_timestamp": "2026-10-10T16:19:40.666Z"
    },
    {
      "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": 0.7343453415559772,
      "accuracy_std_dev": 0.11879281780682018,
      "run_count": 8,
      "total_questions": 1581,
      "avg_cost_per_task_usd": 0.005205407969639469,
      "avg_duration_per_task_ms": 325171.4794433903,
      "avg_output_tokens_per_task": 14219.585072738773,
      "last_run_timestamp": "2026-10-10T16:11:11.296Z"
    },
    {
      "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": 0.73327025,
      "accuracy_std_dev": 0.13440860173157632,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.01872256628787879,
      "avg_duration_per_task_ms": 698828.0744949495,
      "avg_output_tokens_per_task": 35401.32323232323,
      "last_run_timestamp": "2026-10-10T08:33:01.735Z"
    },
    {
      "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": 0.731691875,
      "accuracy_std_dev": 0.013065716091526514,
      "run_count": 32,
      "total_questions": 6336,
      "avg_cost_per_task_usd": 0.2044065787563131,
      "avg_duration_per_task_ms": 243404.04545454544,
      "avg_output_tokens_per_task": 38177.99210858586,
      "last_run_timestamp": "2026-10-10T08:24:42.716Z"
    },
    {
      "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": 0.7310988181818183,
      "accuracy_std_dev": 0.0161010559157872,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.1857441469237833,
      "avg_duration_per_task_ms": 294606.93709825526,
      "avg_output_tokens_per_task": 36982.24701561065,
      "last_run_timestamp": "2026-10-10T08:24:31.744Z"
    },
    {
      "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": 0.7309332329663789,
      "accuracy_std_dev": 0.016291390037575088,
      "run_count": 17,
      "total_questions": 3361,
      "avg_cost_per_task_usd": 0.0067620446295745314,
      "avg_duration_per_task_ms": 112562.21005653079,
      "avg_output_tokens_per_task": 7053.808985421006,
      "last_run_timestamp": "2026-10-10T16:26:34.195Z"
    },
    {
      "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": 0.730219,
      "accuracy_std_dev": 0.02007444496742065,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.005502738005050506,
      "avg_duration_per_task_ms": 219251.19823232322,
      "avg_output_tokens_per_task": 7924.819444444444,
      "last_run_timestamp": "2026-10-10T16:20:00.369Z"
    },
    {
      "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": 0.7301589081632653,
      "accuracy_std_dev": 0.01053611207018773,
      "run_count": 3,
      "total_questions": 588,
      "avg_cost_per_task_usd": 0.01497738095238095,
      "avg_duration_per_task_ms": 995236.0867346938,
      "avg_output_tokens_per_task": 53138.4693877551,
      "last_run_timestamp": "2026-10-10T08:23:30.721Z"
    },
    {
      "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": 0.7284971818181819,
      "accuracy_std_dev": 0.009660041669571141,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.19039750918273646,
      "avg_duration_per_task_ms": 242398.78512396695,
      "avg_output_tokens_per_task": 37888.70890725436,
      "last_run_timestamp": "2026-10-10T08:14:40.906Z"
    },
    {
      "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": 0.7277537142857143,
      "accuracy_std_dev": 0.011584046612404796,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.006112832611832612,
      "avg_duration_per_task_ms": 377298.1096681097,
      "avg_output_tokens_per_task": 8091.412698412699,
      "last_run_timestamp": "2026-10-10T16:21:31.197Z"
    },
    {
      "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": 0.7274832499999999,
      "accuracy_std_dev": 0.02903040114151885,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.007206551136363636,
      "avg_duration_per_task_ms": 297296.3061868687,
      "avg_output_tokens_per_task": 17913.31376262626,
      "last_run_timestamp": "2026-10-10T16:19:34.422Z"
    },
    {
      "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": 0.7272725714285714,
      "accuracy_std_dev": 0.015455781912624552,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.004581955266955267,
      "avg_duration_per_task_ms": 160113.02813852814,
      "avg_output_tokens_per_task": 7774.143578643579,
      "last_run_timestamp": "2026-10-03T16:57:10.991Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "Phala",
      "accuracy": 0.7271808509532062,
      "accuracy_std_dev": 0.03540222199908109,
      "run_count": 6,
      "total_questions": 1154,
      "avg_cost_per_task_usd": 0.011152869150779895,
      "avg_duration_per_task_ms": 504501.5831889082,
      "avg_output_tokens_per_task": 55506.62998266898,
      "last_run_timestamp": "2026-10-10T08:24:27.616Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Tenstorrent",
      "accuracy": 0.7264309999999999,
      "accuracy_std_dev": 0.021044000000000007,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.1303771111111111,
      "avg_duration_per_task_ms": 318087.7601010101,
      "avg_output_tokens_per_task": 43267.416666666664,
      "last_run_timestamp": "2026-10-04T05:57:29.940Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-2.5-flash",
      "model_slug": "google/gemini-2.5-flash",
      "model_name": "Google: Gemini 2.5 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.7261503333333333,
      "accuracy_std_dev": 0.004418546090690373,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.059926168350168355,
      "avg_duration_per_task_ms": 123042.70033670033,
      "avg_output_tokens_per_task": 23940.483164983165,
      "last_run_timestamp": "2026-10-01T19:12:41.089Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Novita",
      "accuracy": 0.7258268816793894,
      "accuracy_std_dev": 0.021662280079307778,
      "run_count": 8,
      "total_questions": 1572,
      "avg_cost_per_task_usd": 0.006591062340966921,
      "avg_duration_per_task_ms": 353326.7900763359,
      "avg_output_tokens_per_task": 26171.935114503816,
      "last_run_timestamp": "2026-10-10T16:19:40.654Z"
    },
    {
      "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": 0.7243107387218045,
      "accuracy_std_dev": 0.009277815175446843,
      "run_count": 11,
      "total_questions": 2128,
      "avg_cost_per_task_usd": 0.18669253665413532,
      "avg_duration_per_task_ms": 347016.6790413534,
      "avg_output_tokens_per_task": 37148.297462406015,
      "last_run_timestamp": "2026-10-10T08:14:48.218Z"
    },
    {
      "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": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Makora",
      "accuracy": 0.723064,
      "accuracy_std_dev": 0.007575999999999972,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.01875409595959596,
      "avg_duration_per_task_ms": 273477.0126262626,
      "avg_output_tokens_per_task": 66626.44696969698,
      "last_run_timestamp": "2026-10-10T08:47:12.235Z"
    },
    {
      "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": 0.7229704444444446,
      "accuracy_std_dev": 0.011078144932063405,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.004416624017957351,
      "avg_duration_per_task_ms": 624923.5415263749,
      "avg_output_tokens_per_task": 7906.3456790123455,
      "last_run_timestamp": "2026-10-10T16:21:49.240Z"
    },
    {
      "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": 0.7219672121212123,
      "accuracy_std_dev": 0.013487941678127918,
      "run_count": 33,
      "total_questions": 6534,
      "avg_cost_per_task_usd": 0.19738882996633,
      "avg_duration_per_task_ms": 265847.20844811754,
      "avg_output_tokens_per_task": 36814.46908478727,
      "last_run_timestamp": "2026-10-10T08:31:35.247Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "ModelRun",
      "accuracy": 0.7219408075949366,
      "accuracy_std_dev": 0.025598999999999983,
      "run_count": 2,
      "total_questions": 395,
      "avg_cost_per_task_usd": 0.27939689620253166,
      "avg_duration_per_task_ms": 450249.90379746835,
      "avg_output_tokens_per_task": 45941.00253164557,
      "last_run_timestamp": "2026-08-15T08:16:15.061Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "Io Net",
      "accuracy": 0.7189313475609755,
      "accuracy_std_dev": 0.11211128771311958,
      "run_count": 7,
      "total_questions": 1148,
      "avg_cost_per_task_usd": 0.004322345818815331,
      "avg_duration_per_task_ms": 557688.2717770035,
      "avg_output_tokens_per_task": 24194.70644599303,
      "last_run_timestamp": "2026-10-10T08:48:56.419Z"
    },
    {
      "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": 0.7184945714285713,
      "accuracy_std_dev": 0.04791285835722166,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.015674489177489176,
      "avg_duration_per_task_ms": 909238.3787878788,
      "avg_output_tokens_per_task": 44831.91558441558,
      "last_run_timestamp": "2026-10-10T16:20:06.461Z"
    },
    {
      "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": "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": 0.7184345000000001,
      "accuracy_std_dev": 0.002104500000000009,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.005931222222222222,
      "avg_duration_per_task_ms": 277051.3181818182,
      "avg_output_tokens_per_task": 7213.616161616162,
      "last_run_timestamp": "2026-08-01T17:52:16.799Z"
    },
    {
      "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": 0.7171302234809476,
      "accuracy_std_dev": 0.05915089253318348,
      "run_count": 6,
      "total_questions": 971,
      "avg_cost_per_task_usd": 0.035917489186405756,
      "avg_duration_per_task_ms": 751381.8146240988,
      "avg_output_tokens_per_task": 38973.5623069001,
      "last_run_timestamp": "2026-08-29T18:51:07.939Z"
    },
    {
      "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": 0.715769,
      "accuracy_std_dev": 0.047798870551509894,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.00680639898989899,
      "avg_duration_per_task_ms": 127099.18602693603,
      "avg_output_tokens_per_task": 7554.406565656565,
      "last_run_timestamp": "2026-08-29T18:01:46.940Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "WandB",
      "accuracy": 0.715488,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0037085252525252527,
      "avg_duration_per_task_ms": 839996.601010101,
      "avg_output_tokens_per_task": 28561.196969696968,
      "last_run_timestamp": "2026-07-18T16:28:34.215Z"
    },
    {
      "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": 0.7153070167364017,
      "accuracy_std_dev": 0.05075792219782052,
      "run_count": 5,
      "total_questions": 956,
      "avg_cost_per_task_usd": 0.06210530125523012,
      "avg_duration_per_task_ms": 1717196.740585774,
      "avg_output_tokens_per_task": 33817.89644351464,
      "last_run_timestamp": "2026-10-10T16:11:11.295Z"
    },
    {
      "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": 0.7133716737044147,
      "accuracy_std_dev": 0.07918700388461619,
      "run_count": 3,
      "total_questions": 521,
      "avg_cost_per_task_usd": 0.07439642610364683,
      "avg_duration_per_task_ms": 652167.8637236084,
      "avg_output_tokens_per_task": 28734.940499040305,
      "last_run_timestamp": "2026-08-15T09:29:41.209Z"
    },
    {
      "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": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "CoreWeave",
      "accuracy": 0.7131104709511569,
      "accuracy_std_dev": 0.0367116985872351,
      "run_count": 10,
      "total_questions": 1945,
      "avg_cost_per_task_usd": 0.010703536246786632,
      "avg_duration_per_task_ms": 271949.64678663237,
      "avg_output_tokens_per_task": 47587.55115681234,
      "last_run_timestamp": "2026-10-10T08:14:46.902Z"
    },
    {
      "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": 0.7118807142857142,
      "accuracy_std_dev": 0.008414141170235586,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.011084775613275614,
      "avg_duration_per_task_ms": 94339.78354978355,
      "avg_output_tokens_per_task": 10485.185425685426,
      "last_run_timestamp": "2026-10-03T16:59:33.949Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-thinking-2507",
      "model_slug": "qwen/qwen3-235b-a22b-thinking-2507",
      "model_name": "Qwen: Qwen3 235B A22B Thinking 2507",
      "provider_name": "Novita",
      "accuracy": 0.71170025,
      "accuracy_std_dev": 0.018456257615982147,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.0649322058080808,
      "avg_duration_per_task_ms": 1066951.6603535353,
      "avg_output_tokens_per_task": 21770.531565656565,
      "last_run_timestamp": "2026-08-22T17:45:06.645Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-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": 0.7113995714285716,
      "accuracy_std_dev": 0.010488730194930244,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.007985493506493507,
      "avg_duration_per_task_ms": 77746.15223665224,
      "avg_output_tokens_per_task": 10141.235930735931,
      "last_run_timestamp": "2026-10-03T16:40:12.993Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_slug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_name": "Qwen: Qwen3 VL 235B A22B Instruct",
      "provider_name": "Novita",
      "accuracy": 0.7109188571428572,
      "accuracy_std_dev": 0.009120000930898651,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.008814419191919191,
      "avg_duration_per_task_ms": 134404.79509379508,
      "avg_output_tokens_per_task": 5713.798701298701,
      "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": "Novita",
      "accuracy": 0.7106782857142858,
      "accuracy_std_dev": 0.008509694097478088,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.015819831890331888,
      "avg_duration_per_task_ms": 94996.398989899,
      "avg_output_tokens_per_task": 10465.311688311689,
      "last_run_timestamp": "2026-10-03T16:46:22.205Z"
    },
    {
      "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": 0.7084164392026578,
      "accuracy_std_dev": 0.2593601641821839,
      "run_count": 8,
      "total_questions": 1505,
      "avg_cost_per_task_usd": 0.025249661129568106,
      "avg_duration_per_task_ms": 375367.3375415282,
      "avg_output_tokens_per_task": 15131.079734219269,
      "last_run_timestamp": "2026-10-10T16:19:34.421Z"
    },
    {
      "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": 0.7080899875730994,
      "accuracy_std_dev": 0.01572159476124274,
      "run_count": 7,
      "total_questions": 1368,
      "avg_cost_per_task_usd": 0.010683734649122806,
      "avg_duration_per_task_ms": 102665.67105263157,
      "avg_output_tokens_per_task": 9845.541666666666,
      "last_run_timestamp": "2026-10-03T16:43:17.105Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_slug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_name": "Qwen: Qwen3 VL 235B A22B Instruct",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.707071,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.008867828282828283,
      "avg_duration_per_task_ms": 116045.65151515152,
      "avg_output_tokens_per_task": 5749.39898989899,
      "last_run_timestamp": "2026-10-04T03:05:59.018Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_slug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_name": "Qwen: Qwen3 VL 235B A22B Instruct",
      "provider_name": "Alibaba",
      "accuracy": 0.7051465714285715,
      "accuracy_std_dev": 0.008927717479163138,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.006021121212121212,
      "avg_duration_per_task_ms": 109220.25974025975,
      "avg_output_tokens_per_task": 5588.1832611832615,
      "last_run_timestamp": "2026-10-03T17:02:24.058Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-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": "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": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "GMICloud",
      "accuracy": 0.7028622499999999,
      "accuracy_std_dev": 0.08200476941122087,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.0070032777777777775,
      "avg_duration_per_task_ms": 172461.12373737374,
      "avg_output_tokens_per_task": 7212.60101010101,
      "last_run_timestamp": "2026-10-03T16:56:36.098Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "DeepInfra",
      "accuracy": 0.7024286137142859,
      "accuracy_std_dev": 0.04779753153655171,
      "run_count": 18,
      "total_questions": 3500,
      "avg_cost_per_task_usd": 0.008325981428571428,
      "avg_duration_per_task_ms": 489576.588,
      "avg_output_tokens_per_task": 25239.30742857143,
      "last_run_timestamp": "2026-10-10T16:21:31.662Z"
    },
    {
      "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": 0.701832040429564,
      "accuracy_std_dev": 0.009254626092359155,
      "run_count": 8,
      "total_questions": 1583,
      "avg_cost_per_task_usd": 0.010471590018951359,
      "avg_duration_per_task_ms": 144460.25710675932,
      "avg_output_tokens_per_task": 9453.23373341756,
      "last_run_timestamp": "2026-10-08T22:40:37.311Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_slug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_name": "Qwen: Qwen3 VL 235B A22B Instruct",
      "provider_name": "Parasail",
      "accuracy": 0.7017798571428572,
      "accuracy_std_dev": 0.013719199460490942,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.011180141414141413,
      "avg_duration_per_task_ms": 199426.63852813852,
      "avg_output_tokens_per_task": 5820.556277056277,
      "last_run_timestamp": "2026-10-03T16:59:57.847Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-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": 0.7010584285714286,
      "accuracy_std_dev": 0.024870525026666448,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.011166472582972582,
      "avg_duration_per_task_ms": 61418.856421356424,
      "avg_output_tokens_per_task": 9244.735930735931,
      "last_run_timestamp": "2026-10-03T16:36:16.895Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "auto-routing",
      "accuracy": 0.6984750588235294,
      "accuracy_std_dev": 0.028411848992761817,
      "run_count": 8,
      "total_questions": 1530,
      "avg_cost_per_task_usd": 0.009441421568627451,
      "avg_duration_per_task_ms": 385390.5339869281,
      "avg_output_tokens_per_task": 44748.05424836601,
      "last_run_timestamp": "2026-10-10T08:22:04.276Z"
    },
    {
      "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": 0.69633825,
      "accuracy_std_dev": 0.11651475266972633,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.015558469065656567,
      "avg_duration_per_task_ms": 463991.38194444444,
      "avg_output_tokens_per_task": 14784.056186868687,
      "last_run_timestamp": "2026-10-10T16:10:48.866Z"
    },
    {
      "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": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "GMICloud",
      "accuracy": 0.6941227451737451,
      "accuracy_std_dev": 0.08714860354302875,
      "run_count": 8,
      "total_questions": 1554,
      "avg_cost_per_task_usd": 0.13150269498069497,
      "avg_duration_per_task_ms": 1023131.1537966538,
      "avg_output_tokens_per_task": 71194.34684684685,
      "last_run_timestamp": "2026-10-10T16:10:46.571Z"
    },
    {
      "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": 0.6928812857142858,
      "accuracy_std_dev": 0.0184605601038251,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.009158278499278501,
      "avg_duration_per_task_ms": 267700.48124098126,
      "avg_output_tokens_per_task": 6402.472582972583,
      "last_run_timestamp": "2026-10-03T17:02:48.284Z"
    },
    {
      "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": 0.692129625,
      "accuracy_std_dev": 0.010010686751385976,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.005500493055555555,
      "avg_duration_per_task_ms": 380445.4021464646,
      "avg_output_tokens_per_task": 6066.023358585859,
      "last_run_timestamp": "2026-10-08T23:09:15.796Z"
    },
    {
      "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": 0.691919,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.01248349494949495,
      "avg_duration_per_task_ms": 400806.7474747475,
      "avg_output_tokens_per_task": 31001.929292929293,
      "last_run_timestamp": "2026-07-18T08:04:42.524Z"
    },
    {
      "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": 0.691462222392638,
      "accuracy_std_dev": 0.028807852029534505,
      "run_count": 6,
      "total_questions": 652,
      "avg_cost_per_task_usd": 0.00933088036809816,
      "avg_duration_per_task_ms": 52547.14110429448,
      "avg_output_tokens_per_task": 6038.897239263803,
      "last_run_timestamp": "2026-10-10T16:11:11.293Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5-nano-2025-08-07",
      "model_slug": "openai/gpt-5-nano",
      "model_name": "OpenAI: GPT-5 Nano",
      "provider_name": "Azure",
      "accuracy": 0.689815,
      "accuracy_std_dev": 0.0034448118235979193,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.013852137626262627,
      "avg_duration_per_task_ms": 234308.42045454544,
      "avg_output_tokens_per_task": 32893.11994949495,
      "last_run_timestamp": "2026-10-01T19:27:10.142Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "DeepInfra",
      "accuracy": 0.6873114230769231,
      "accuracy_std_dev": 0.04923369976869075,
      "run_count": 9,
      "total_questions": 1768,
      "avg_cost_per_task_usd": 0.007741963800904979,
      "avg_duration_per_task_ms": 279050.26527149323,
      "avg_output_tokens_per_task": 40832.55995475113,
      "last_run_timestamp": "2026-10-10T08:15:00.521Z"
    },
    {
      "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": 0.6856376949064449,
      "accuracy_std_dev": 0.10684923261228746,
      "run_count": 11,
      "total_questions": 1924,
      "avg_cost_per_task_usd": 0.031270327962577955,
      "avg_duration_per_task_ms": 1092631.9272349272,
      "avg_output_tokens_per_task": 32831.02286902287,
      "last_run_timestamp": "2026-10-10T16:21:35.979Z"
    },
    {
      "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": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "Phala",
      "accuracy": 0.685185,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.06544819696969698,
      "avg_duration_per_task_ms": 945099.904040404,
      "avg_output_tokens_per_task": 33006.14646464647,
      "last_run_timestamp": "2026-07-18T16:15:33.587Z"
    },
    {
      "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": 0.6809765714285714,
      "accuracy_std_dev": 0.0445115533536989,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.014742932178932178,
      "avg_duration_per_task_ms": 268442.60533910536,
      "avg_output_tokens_per_task": 7991.235930735931,
      "last_run_timestamp": "2026-10-03T17:36:48.437Z"
    },
    {
      "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": 0.6782486254237288,
      "accuracy_std_dev": 0.021428719439237503,
      "run_count": 6,
      "total_questions": 590,
      "avg_cost_per_task_usd": 0.004103340677966102,
      "avg_duration_per_task_ms": 494419.3559322034,
      "avg_output_tokens_per_task": 27366.218644067798,
      "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-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "Parasail",
      "accuracy": 0.6779694919852034,
      "accuracy_std_dev": 0.0848101181666433,
      "run_count": 5,
      "total_questions": 811,
      "avg_cost_per_task_usd": 0.007035279901356351,
      "avg_duration_per_task_ms": 476700.88409371144,
      "avg_output_tokens_per_task": 34252.66584463625,
      "last_run_timestamp": "2026-08-29T12:30:09.617Z"
    },
    {
      "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": 0.6774892857142858,
      "accuracy_std_dev": 0.07499128841922582,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.00669294227994228,
      "avg_duration_per_task_ms": 714224.0743145744,
      "avg_output_tokens_per_task": 23217.833333333332,
      "last_run_timestamp": "2026-10-10T16:24:54.998Z"
    },
    {
      "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": 0.676784615514334,
      "accuracy_std_dev": 0.11674700703943836,
      "run_count": 3,
      "total_questions": 593,
      "avg_cost_per_task_usd": 0.05877558516020236,
      "avg_duration_per_task_ms": 524798.5919055649,
      "avg_output_tokens_per_task": 24486.29679595278,
      "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": "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-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "AkashML",
      "accuracy": 0.6744752254758419,
      "accuracy_std_dev": 0.018240929771167532,
      "run_count": 6,
      "total_questions": 683,
      "avg_cost_per_task_usd": 0.004128366032210834,
      "avg_duration_per_task_ms": 765372.7452415812,
      "avg_output_tokens_per_task": 41143.20790629576,
      "last_run_timestamp": "2026-10-10T16:22:35.317Z"
    },
    {
      "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": 0.6741221428571429,
      "accuracy_std_dev": 0.028652840424784446,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.005719461038961039,
      "avg_duration_per_task_ms": 425768.6572871573,
      "avg_output_tokens_per_task": 15674.55266955267,
      "last_run_timestamp": "2026-10-10T16:20:00.451Z"
    },
    {
      "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": 0.6725437394736843,
      "accuracy_std_dev": 0.2068414494859287,
      "run_count": 10,
      "total_questions": 1900,
      "avg_cost_per_task_usd": 0.013809886842105263,
      "avg_duration_per_task_ms": 353880.6878947368,
      "avg_output_tokens_per_task": 16644.034736842106,
      "last_run_timestamp": "2026-10-10T08:24:33.802Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-chat-v3",
      "model_slug": "deepseek/deepseek-chat",
      "model_name": "DeepSeek: DeepSeek V3",
      "provider_name": "StreamLake",
      "accuracy": 0.6689113333333334,
      "accuracy_std_dev": 0.0067804647005611455,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0026409175084175086,
      "avg_duration_per_task_ms": 47935.90572390572,
      "avg_output_tokens_per_task": 2383.6902356902356,
      "last_run_timestamp": "2026-10-02T15:53:31.869Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.6-20251208",
      "model_slug": "z-ai/glm-4.6v",
      "model_name": "Z.ai: GLM 4.6V",
      "provider_name": "auto-routing",
      "accuracy": 0.66835,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.013946772727272728,
      "avg_duration_per_task_ms": 324889.0353535354,
      "avg_output_tokens_per_task": 15315.242424242424,
      "last_run_timestamp": "2026-10-01T08:41:30.639Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "DeepInfra",
      "accuracy": 0.6678543283847981,
      "accuracy_std_dev": 0.01614207710791971,
      "run_count": 9,
      "total_questions": 1684,
      "avg_cost_per_task_usd": 0.007296722090261282,
      "avg_duration_per_task_ms": 505935.9150831354,
      "avg_output_tokens_per_task": 51932.642517814726,
      "last_run_timestamp": "2026-10-10T16:25:08.075Z"
    },
    {
      "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": 0.6678278397720075,
      "accuracy_std_dev": 0.09103865920988938,
      "run_count": 8,
      "total_questions": 1579,
      "avg_cost_per_task_usd": 0.01146532425585814,
      "avg_duration_per_task_ms": 1011964.7701076631,
      "avg_output_tokens_per_task": 33559.86700443319,
      "last_run_timestamp": "2026-10-10T16:24:41.108Z"
    },
    {
      "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": 0.666857044520548,
      "accuracy_std_dev": 0.030362698141707424,
      "run_count": 9,
      "total_questions": 1752,
      "avg_cost_per_task_usd": 0.0025397779680365294,
      "avg_duration_per_task_ms": 226314.88356164383,
      "avg_output_tokens_per_task": 5199.7414383561645,
      "last_run_timestamp": "2026-10-10T16:21:38.360Z"
    },
    {
      "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": 0.666667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.009070929292929292,
      "avg_duration_per_task_ms": 492518.0909090909,
      "avg_output_tokens_per_task": 69555.08585858585,
      "last_run_timestamp": "2026-07-18T17:04:25.343Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "DigitalOcean",
      "accuracy": 0.666667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.002231131313131313,
      "avg_duration_per_task_ms": 1386615.121212121,
      "avg_output_tokens_per_task": 14198.39898989899,
      "last_run_timestamp": "2026-07-18T09:58:41.490Z"
    },
    {
      "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": 0.6666666666666666,
      "accuracy_std_dev": 0.1312433939450753,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.6057920415263749,
      "avg_duration_per_task_ms": 626284.2744107745,
      "avg_output_tokens_per_task": 40322.78170594837,
      "last_run_timestamp": "2026-10-10T08:23:04.939Z"
    },
    {
      "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": 0.6655442787878788,
      "accuracy_std_dev": 0.1380673949197909,
      "run_count": 8,
      "total_questions": 1485,
      "avg_cost_per_task_usd": 0.09031919797979798,
      "avg_duration_per_task_ms": 682014.4538720539,
      "avg_output_tokens_per_task": 28128.345454545455,
      "last_run_timestamp": "2026-10-10T16:11:14.295Z"
    },
    {
      "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": "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": 0.6632997142857142,
      "accuracy_std_dev": 0.009351528808768375,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.0021804473304473304,
      "avg_duration_per_task_ms": 240867.02741702742,
      "avg_output_tokens_per_task": 3011.5793650793653,
      "last_run_timestamp": "2026-10-05T20:44:48.611Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-4-maverick-17b-128e-instruct",
      "model_slug": "meta-llama/llama-4-maverick",
      "model_name": "Meta: Llama 4 Maverick",
      "provider_name": "auto-routing",
      "accuracy": 0.663019,
      "accuracy_std_dev": 0.007921062218330754,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.0027949452861952863,
      "avg_duration_per_task_ms": 83787.81902356903,
      "avg_output_tokens_per_task": 2948.4612794612794,
      "last_run_timestamp": "2026-08-29T19:49:56.943Z"
    },
    {
      "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": 0.6622475,
      "accuracy_std_dev": 0.023522840681983983,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.02822281313131313,
      "avg_duration_per_task_ms": 831472.6881313132,
      "avg_output_tokens_per_task": 18950.592171717173,
      "last_run_timestamp": "2026-10-10T16:22:25.741Z"
    },
    {
      "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": 0.6616161818181817,
      "accuracy_std_dev": 0.013544295262167041,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.005160062442607897,
      "avg_duration_per_task_ms": 106935.65518824609,
      "avg_output_tokens_per_task": 9720.085399449035,
      "last_run_timestamp": "2026-10-03T09:23:08.256Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-chat-v3",
      "model_slug": "deepseek/deepseek-chat",
      "model_name": "DeepSeek: DeepSeek V3",
      "provider_name": "Novita",
      "accuracy": 0.661616,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.003362974747474748,
      "avg_duration_per_task_ms": 65871.15656565657,
      "avg_output_tokens_per_task": 2362.121212121212,
      "last_run_timestamp": "2026-08-01T09:30:53.884Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4.1-2025-04-14",
      "model_slug": "openai/gpt-4.1",
      "model_name": "OpenAI: GPT-4.1",
      "provider_name": "OpenAI",
      "accuracy": 0.6607744999999999,
      "accuracy_std_dev": 0.011572094721786573,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.015565636363636365,
      "avg_duration_per_task_ms": 14159.460858585859,
      "avg_output_tokens_per_task": 1751.3977272727273,
      "last_run_timestamp": "2026-10-02T23:17:48.072Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.6-20251208",
      "model_slug": "z-ai/glm-4.6v",
      "model_name": "Z.ai: GLM 4.6V",
      "provider_name": "Z.AI",
      "accuracy": 0.659933,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.014984520202020202,
      "avg_duration_per_task_ms": 348987.1767676768,
      "avg_output_tokens_per_task": 16461.863636363636,
      "last_run_timestamp": "2026-10-01T08:28:22.621Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-chat-v3-0324",
      "model_slug": "deepseek/deepseek-chat-v3-0324",
      "model_name": "DeepSeek: DeepSeek V3 0324",
      "provider_name": "SiliconFlow",
      "accuracy": 0.6595962,
      "accuracy_std_dev": 0.011187342962473291,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.0024944898989898985,
      "avg_duration_per_task_ms": 89627.26666666666,
      "avg_output_tokens_per_task": 2311.8535353535353,
      "last_run_timestamp": "2026-08-29T20:54:16.286Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 VL 30B A3B Instruct",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.658249,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0066776414141414145,
      "avg_duration_per_task_ms": 101359.9898989899,
      "avg_output_tokens_per_task": 9307.358585858587,
      "last_run_timestamp": "2026-10-04T03:07:35.168Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.6579336250000001,
      "accuracy_std_dev": 0.01612810650260517,
      "run_count": 16,
      "total_questions": 3168,
      "avg_cost_per_task_usd": 0.008984744633838386,
      "avg_duration_per_task_ms": 321798.5044191919,
      "avg_output_tokens_per_task": 59577.13005050505,
      "last_run_timestamp": "2026-10-10T16:21:42.797Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "ibm-granite/granite-4.2-8b-20260831",
      "model_slug": "ibm-granite/granite-4.2-8b",
      "model_name": "IBM: Granite 4.2 8B",
      "provider_name": "auto-routing",
      "accuracy": 0.65736,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 197,
      "avg_cost_per_task_usd": 0.014864365482233502,
      "avg_duration_per_task_ms": 1717906.416243655,
      "avg_output_tokens_per_task": 64415.350253807104,
      "last_run_timestamp": "2026-10-01T08:48:38.304Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.6-sonnet-20260217",
      "model_slug": "anthropic/claude-sonnet-4.6",
      "model_name": "Anthropic: Claude Sonnet 4.6",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.6564535333333333,
      "accuracy_std_dev": 0.13470588451282378,
      "run_count": 30,
      "total_questions": 5940,
      "avg_cost_per_task_usd": 0.6553498678451178,
      "avg_duration_per_task_ms": 477953.10454545455,
      "avg_output_tokens_per_task": 40779.67929292929,
      "last_run_timestamp": "2026-10-10T08:22:15.381Z"
    },
    {
      "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": 0.6558634490329324,
      "accuracy_std_dev": 0.13914333566507864,
      "run_count": 29,
      "total_questions": 5739,
      "avg_cost_per_task_usd": 0.6623900515769298,
      "avg_duration_per_task_ms": 598314.6978567694,
      "avg_output_tokens_per_task": 41294.878027530925,
      "last_run_timestamp": "2026-10-10T08:22:42.388Z"
    },
    {
      "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": 0.6548824,
      "accuracy_std_dev": 0.13698361654533725,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.6155487378787879,
      "avg_duration_per_task_ms": 631265.5565656566,
      "avg_output_tokens_per_task": 40882.59494949495,
      "last_run_timestamp": "2026-10-10T08:22:10.781Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Parasail",
      "accuracy": 0.652862,
      "accuracy_std_dev": 0.023361257312054087,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.007757086868686868,
      "avg_duration_per_task_ms": 821683.2373737374,
      "avg_output_tokens_per_task": 53689.36767676767,
      "last_run_timestamp": "2026-10-10T16:23:01.112Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "ibm-granite/granite-4.2-8b-20260831",
      "model_slug": "ibm-granite/granite-4.2-8b",
      "model_name": "IBM: Granite 4.2 8B",
      "provider_name": "DeepInfra",
      "accuracy": 0.6523569999999999,
      "accuracy_std_dev": 0.0025249999999999995,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.01652235606060606,
      "avg_duration_per_task_ms": 1917916.494949495,
      "avg_output_tokens_per_task": 65966.15909090909,
      "last_run_timestamp": "2026-10-09T00:15:37.020Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.6-sonnet-20260217",
      "model_slug": "anthropic/claude-sonnet-4.6",
      "model_name": "Anthropic: Claude Sonnet 4.6",
      "provider_name": "Azure",
      "accuracy": 0.6515153,
      "accuracy_std_dev": 0.14316487590330945,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.6191192621212122,
      "avg_duration_per_task_ms": 634077.3176767677,
      "avg_output_tokens_per_task": 41194.81414141414,
      "last_run_timestamp": "2026-10-10T08:22:09.021Z"
    },
    {
      "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": 0.6515151250000001,
      "accuracy_std_dev": 0.019686800377648354,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.007399155303030303,
      "avg_duration_per_task_ms": 666816.1003787878,
      "avg_output_tokens_per_task": 57307.84595959596,
      "last_run_timestamp": "2026-10-10T16:27:33.647Z"
    },
    {
      "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": "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": "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": 0.6495510000000001,
      "accuracy_std_dev": 0.00792134559276387,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.0032256414141414147,
      "avg_duration_per_task_ms": 40985.67760942761,
      "avg_output_tokens_per_task": 3060.60101010101,
      "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": "Novita",
      "accuracy": 0.649551,
      "accuracy_std_dev": 0.009444634067377443,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.002492604377104377,
      "avg_duration_per_task_ms": 61525.246632996634,
      "avg_output_tokens_per_task": 2691.0799663299663,
      "last_run_timestamp": "2026-08-29T17:03:33.297Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4.1-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": "ibm-granite/granite-4.2-8b-20260831",
      "model_slug": "ibm-granite/granite-4.2-8b",
      "model_name": "IBM: Granite 4.2 8B",
      "provider_name": "CoreWeave",
      "accuracy": 0.647493,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 113,
      "avg_cost_per_task_usd": 0.005189893805309735,
      "avg_duration_per_task_ms": 619234.2654867256,
      "avg_output_tokens_per_task": 34365.4778761062,
      "last_run_timestamp": "2026-10-01T08:28:32.911Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4.1-mini-2025-04-14",
      "model_slug": "openai/gpt-4.1-mini",
      "model_name": "OpenAI: GPT-4.1 Mini",
      "provider_name": "OpenAI",
      "accuracy": 0.6473065000000001,
      "accuracy_std_dev": 0.009707729665065862,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.00413114393939394,
      "avg_duration_per_task_ms": 27377.72601010101,
      "avg_output_tokens_per_task": 2388.9305555555557,
      "last_run_timestamp": "2026-10-02T23:15:53.868Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-chat-v3-0324",
      "model_slug": "deepseek/deepseek-chat-v3-0324",
      "model_name": "DeepSeek: DeepSeek V3 0324",
      "provider_name": "GMICloud",
      "accuracy": 0.646465,
      "accuracy_std_dev": 0.004761657064510171,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0029047154882154884,
      "avg_duration_per_task_ms": 53981.06734006734,
      "avg_output_tokens_per_task": 2363.939393939394,
      "last_run_timestamp": "2026-08-29T16:33:11.860Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4.1-2025-04-14",
      "model_slug": "openai/gpt-4.1",
      "model_name": "OpenAI: GPT-4.1",
      "provider_name": "auto-routing",
      "accuracy": 0.6464646666666667,
      "accuracy_std_dev": 0.013537798302858865,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.015368191919191919,
      "avg_duration_per_task_ms": 14060.520202020201,
      "avg_output_tokens_per_task": 1727.3535353535353,
      "last_run_timestamp": "2026-10-01T19:12:45.033Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "DeepInfra",
      "accuracy": 0.6461939736191261,
      "accuracy_std_dev": 0.08774312195800593,
      "run_count": 8,
      "total_questions": 1213,
      "avg_cost_per_task_usd": 0.046983095630667764,
      "avg_duration_per_task_ms": 297281.52844187967,
      "avg_output_tokens_per_task": 17813.622423742785,
      "last_run_timestamp": "2026-10-10T16:23:45.151Z"
    },
    {
      "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": 0.6457914,
      "accuracy_std_dev": 0.017547304854022453,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.007042358080808081,
      "avg_duration_per_task_ms": 238513.37575757576,
      "avg_output_tokens_per_task": 11322.903535353535,
      "last_run_timestamp": "2026-10-03T10:20:18.247Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "DekaLLM",
      "accuracy": 0.6447479677419355,
      "accuracy_std_dev": 0.02407747186802543,
      "run_count": 3,
      "total_questions": 403,
      "avg_cost_per_task_usd": 0.003165789081885856,
      "avg_duration_per_task_ms": 173618.99503722086,
      "avg_output_tokens_per_task": 10324.468982630273,
      "last_run_timestamp": "2026-10-09T03:20:25.108Z"
    },
    {
      "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": 0.6445006666666666,
      "accuracy_std_dev": 0.013998443861928202,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.002421077441077441,
      "avg_duration_per_task_ms": 66682.89730639731,
      "avg_output_tokens_per_task": 2836.0850168350166,
      "last_run_timestamp": "2026-08-29T16:11:23.388Z"
    },
    {
      "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/qwen3-32b-04-28",
      "model_slug": "qwen/qwen3-32b",
      "model_name": "Qwen: Qwen3 32B",
      "provider_name": "SiliconFlow",
      "accuracy": 0.643098,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.018197469696969695,
      "avg_duration_per_task_ms": 1435371.2424242424,
      "avg_output_tokens_per_task": 32217.19191919192,
      "last_run_timestamp": "2026-07-18T17:46:02.370Z"
    },
    {
      "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": 0.6428570571428571,
      "accuracy_std_dev": 0.0682565,
      "run_count": 2,
      "total_questions": 175,
      "avg_cost_per_task_usd": 0.09669735428571428,
      "avg_duration_per_task_ms": 1267820.5257142857,
      "avg_output_tokens_per_task": 34805.99428571429,
      "last_run_timestamp": "2026-08-01T18:16:39.158Z"
    },
    {
      "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": 0.6428055813449023,
      "accuracy_std_dev": 0.011146034629373067,
      "run_count": 7,
      "total_questions": 1383,
      "avg_cost_per_task_usd": 0.0026296478669558935,
      "avg_duration_per_task_ms": 140713.84381778742,
      "avg_output_tokens_per_task": 9405.384671005062,
      "last_run_timestamp": "2026-10-03T17:16:11.892Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "Ambient",
      "accuracy": 0.641414,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.00284089898989899,
      "avg_duration_per_task_ms": 209352.64646464647,
      "avg_output_tokens_per_task": 9266.040404040405,
      "last_run_timestamp": "2026-10-09T15:00:06.492Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.6413729024390243,
      "accuracy_std_dev": 0.03637250000000003,
      "run_count": 2,
      "total_questions": 369,
      "avg_cost_per_task_usd": 0.027402341463414633,
      "avg_duration_per_task_ms": 991300.8970189702,
      "avg_output_tokens_per_task": 32627.39024390244,
      "last_run_timestamp": "2026-08-01T17:59:38.872Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "SiliconFlow",
      "accuracy": 0.6406363579175706,
      "accuracy_std_dev": 0.016263383886158854,
      "run_count": 7,
      "total_questions": 1383,
      "avg_cost_per_task_usd": 0.002550553145336226,
      "avg_duration_per_task_ms": 225461.96384671004,
      "avg_output_tokens_per_task": 8275.081706435285,
      "last_run_timestamp": "2026-10-03T17:03:50.406Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-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": "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": 0.6402118571428571,
      "accuracy_std_dev": 0.01348951470511788,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.004506121212121212,
      "avg_duration_per_task_ms": 72990.16883116883,
      "avg_output_tokens_per_task": 8472.02886002886,
      "last_run_timestamp": "2026-10-03T16:51:25.541Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "WandB",
      "accuracy": 0.639316,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 195,
      "avg_cost_per_task_usd": 0.0024375846153846152,
      "avg_duration_per_task_ms": 97724.68205128206,
      "avg_output_tokens_per_task": 7853.553846153846,
      "last_run_timestamp": "2026-07-18T08:33:31.052Z"
    },
    {
      "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": 0.638468,
      "accuracy_std_dev": 0.016031912152017304,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.009180828282828282,
      "avg_duration_per_task_ms": 658413.9090909091,
      "avg_output_tokens_per_task": 48820.804924242424,
      "last_run_timestamp": "2026-10-10T16:28:07.050Z"
    },
    {
      "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": 0.6377313888888888,
      "accuracy_std_dev": 0.020028968176674062,
      "run_count": 6,
      "total_questions": 1152,
      "avg_cost_per_task_usd": 0.006115868055555556,
      "avg_duration_per_task_ms": 899818.5989583334,
      "avg_output_tokens_per_task": 35427.116319444445,
      "last_run_timestamp": "2026-10-10T08:14:41.978Z"
    },
    {
      "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": 0.6371981347826087,
      "accuracy_std_dev": 0.014429669830892581,
      "run_count": 7,
      "total_questions": 1380,
      "avg_cost_per_task_usd": 0.0032497253623188405,
      "avg_duration_per_task_ms": 218812.71086956523,
      "avg_output_tokens_per_task": 10563.355072463768,
      "last_run_timestamp": "2026-10-03T16:41:22.866Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-32b-04-28",
      "model_slug": "qwen/qwen3-32b",
      "model_name": "Qwen: Qwen3 32B",
      "provider_name": "Groq",
      "accuracy": 0.6361772567678425,
      "accuracy_std_dev": 0.013327375155318935,
      "run_count": 7,
      "total_questions": 1219,
      "avg_cost_per_task_usd": 0.007504426579163248,
      "avg_duration_per_task_ms": 35368.206726825265,
      "avg_output_tokens_per_task": 12428.807219031993,
      "last_run_timestamp": "2026-08-29T16:23:45.123Z"
    },
    {
      "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": 0.6356561135888502,
      "accuracy_std_dev": 0.014057249530380225,
      "run_count": 8,
      "total_questions": 1435,
      "avg_cost_per_task_usd": 0.01605337770034843,
      "avg_duration_per_task_ms": 91943.85017421603,
      "avg_output_tokens_per_task": 53289.633449477355,
      "last_run_timestamp": "2026-10-10T16:28:40.646Z"
    },
    {
      "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/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "StreamLake",
      "accuracy": 0.6340491250000001,
      "accuracy_std_dev": 0.012397945832248788,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.0016212228535353534,
      "avg_duration_per_task_ms": 88876.8813131313,
      "avg_output_tokens_per_task": 8195.30366161616,
      "last_run_timestamp": "2026-10-03T16:47:32.559Z"
    },
    {
      "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": 0.632856543563069,
      "accuracy_std_dev": 0.17446938590265632,
      "run_count": 4,
      "total_questions": 769,
      "avg_cost_per_task_usd": 0.06782110143042913,
      "avg_duration_per_task_ms": 1537627.533159948,
      "avg_output_tokens_per_task": 34138.841352405725,
      "last_run_timestamp": "2026-08-22T17:02:13.071Z"
    },
    {
      "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": 0.6328437272727273,
      "accuracy_std_dev": 0.03335048624993999,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.006873024334251608,
      "avg_duration_per_task_ms": 112649.99265381084,
      "avg_output_tokens_per_task": 9586.507805325988,
      "last_run_timestamp": "2026-10-03T10:32:36.729Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3-nano-30b-a3b",
      "model_slug": "nvidia/nemotron-3-nano-30b-a3b",
      "model_name": "NVIDIA: Nemotron 3 Nano 30B A3B",
      "provider_name": "Crusoe",
      "accuracy": 0.6325426993295513,
      "accuracy_std_dev": 0.010534965676735725,
      "run_count": 10,
      "total_questions": 1939,
      "avg_cost_per_task_usd": 0.01766013512119649,
      "avg_duration_per_task_ms": 478451.96183599794,
      "avg_output_tokens_per_task": 88282.73388344508,
      "last_run_timestamp": "2026-10-03T17:08:02.072Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "inclusionai/ling-3.0-flash-fin-20260827",
      "model_slug": "inclusionai/ling-3.0-flash-fin",
      "model_name": "inclusionAI: Ling 3.0 Flash Fin",
      "provider_name": "Novita",
      "accuracy": 0.632155,
      "accuracy_std_dev": 0.005892000000000008,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.005966434343434343,
      "avg_duration_per_task_ms": 235473.01515151514,
      "avg_output_tokens_per_task": 48249.11363636364,
      "last_run_timestamp": "2026-10-01T08:24:52.614Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "SiliconFlow",
      "accuracy": 0.631776382667417,
      "accuracy_std_dev": 0.016262402453375572,
      "run_count": 9,
      "total_questions": 1777,
      "avg_cost_per_task_usd": 0.0017607208778840744,
      "avg_duration_per_task_ms": 208780.86043894204,
      "avg_output_tokens_per_task": 9645.604952166574,
      "last_run_timestamp": "2026-10-10T16:20:25.117Z"
    },
    {
      "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": 0.63118499609375,
      "accuracy_std_dev": 0.12107794090401254,
      "run_count": 3,
      "total_questions": 512,
      "avg_cost_per_task_usd": 0.00688423046875,
      "avg_duration_per_task_ms": 160958.9140625,
      "avg_output_tokens_per_task": 8535.359375,
      "last_run_timestamp": "2026-08-08T16:46:11.805Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.6-sonnet-20260217",
      "model_slug": "anthropic/claude-sonnet-4.6",
      "model_name": "Anthropic: Claude Sonnet 4.6",
      "provider_name": "Anthropic",
      "accuracy": 0.6310326666666666,
      "accuracy_std_dev": 0.13772289129161577,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.615210808080808,
      "avg_duration_per_task_ms": 628271.4768518518,
      "avg_output_tokens_per_task": 40892.35058922559,
      "last_run_timestamp": "2026-10-10T08:24:11.887Z"
    },
    {
      "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-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Phala",
      "accuracy": 0.6305803258928572,
      "accuracy_std_dev": 0.21954975394425835,
      "run_count": 8,
      "total_questions": 1344,
      "avg_cost_per_task_usd": 0.0069273124999999994,
      "avg_duration_per_task_ms": 192378.60044642858,
      "avg_output_tokens_per_task": 11473.340773809523,
      "last_run_timestamp": "2026-10-10T16:21:31.775Z"
    },
    {
      "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": 0.6297015953307392,
      "accuracy_std_dev": 0.01592646788319292,
      "run_count": 8,
      "total_questions": 1542,
      "avg_cost_per_task_usd": 0.023244770428015562,
      "avg_duration_per_task_ms": 1042943.9273670558,
      "avg_output_tokens_per_task": 96652.03437094683,
      "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-3.5-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": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Darkbloom",
      "accuracy": 0.62899825,
      "accuracy_std_dev": 0.007126819079189552,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.001126895202020202,
      "avg_duration_per_task_ms": 367383.1767676768,
      "avg_output_tokens_per_task": 11681.671717171717,
      "last_run_timestamp": "2026-10-10T16:22:34.748Z"
    },
    {
      "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": 0.628788,
      "accuracy_std_dev": 0.066498,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.00551144696969697,
      "avg_duration_per_task_ms": 2569506.212121212,
      "avg_output_tokens_per_task": 42356.12373737374,
      "last_run_timestamp": "2026-10-10T16:26:33.881Z"
    },
    {
      "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": 0.6285929916344916,
      "accuracy_std_dev": 0.03714754479584586,
      "run_count": 8,
      "total_questions": 1554,
      "avg_cost_per_task_usd": 0.06302590733590734,
      "avg_duration_per_task_ms": 1075507.0637065638,
      "avg_output_tokens_per_task": 25670.111969111967,
      "last_run_timestamp": "2026-10-10T16:24:08.120Z"
    },
    {
      "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": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "Mara",
      "accuracy": 0.628499,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 131,
      "avg_cost_per_task_usd": 0.0358695572519084,
      "avg_duration_per_task_ms": 72030.85496183205,
      "avg_output_tokens_per_task": 7685.93893129771,
      "last_run_timestamp": "2026-10-10T16:20:00.366Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "Darkbloom",
      "accuracy": 0.6279425772773797,
      "accuracy_std_dev": 0.06414256831964248,
      "run_count": 5,
      "total_questions": 977,
      "avg_cost_per_task_usd": 0.012322606960081882,
      "avg_duration_per_task_ms": 1444518.9590583418,
      "avg_output_tokens_per_task": 53882.452405322416,
      "last_run_timestamp": "2026-10-10T16:20:00.388Z"
    },
    {
      "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": 0.626263,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.012936333333333333,
      "avg_duration_per_task_ms": 385474.66161616164,
      "avg_output_tokens_per_task": 30893.848484848484,
      "last_run_timestamp": "2026-07-18T16:20:00.500Z"
    },
    {
      "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": 0.6261091711026616,
      "accuracy_std_dev": 0.05551308764606775,
      "run_count": 4,
      "total_questions": 789,
      "avg_cost_per_task_usd": 0.0414238669201521,
      "avg_duration_per_task_ms": 392853.54372623574,
      "avg_output_tokens_per_task": 20615.842839036755,
      "last_run_timestamp": "2026-08-22T10:29:25.897Z"
    },
    {
      "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": 0.623263890625,
      "accuracy_std_dev": 0.046540499999999985,
      "run_count": 2,
      "total_questions": 384,
      "avg_cost_per_task_usd": 0.03095198958333333,
      "avg_duration_per_task_ms": 834783.7890625,
      "avg_output_tokens_per_task": 38538.341145833336,
      "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-vl-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 VL 30B A3B Instruct",
      "provider_name": "SiliconFlow",
      "accuracy": 0.623188452685422,
      "accuracy_std_dev": 0.0187657309831867,
      "run_count": 6,
      "total_questions": 1173,
      "avg_cost_per_task_usd": 0.014858657289002557,
      "avg_duration_per_task_ms": 724190.4782608695,
      "avg_output_tokens_per_task": 15645.849104859335,
      "last_run_timestamp": "2026-10-03T11:06:38.068Z"
    },
    {
      "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": 0.6212125000000001,
      "accuracy_std_dev": 0.011784500000000031,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.008206156565656566,
      "avg_duration_per_task_ms": 339475.95454545453,
      "avg_output_tokens_per_task": 15670.888888888889,
      "last_run_timestamp": "2026-08-01T18:08:49.019Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.621212,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.00552190909090909,
      "avg_duration_per_task_ms": 174055.64141414143,
      "avg_output_tokens_per_task": 10415.944444444445,
      "last_run_timestamp": "2026-10-04T03:09:40.101Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "DigitalOcean",
      "accuracy": 0.6208752999999999,
      "accuracy_std_dev": 0.12687124759381063,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.002729046969696969,
      "avg_duration_per_task_ms": 232884.66262626264,
      "avg_output_tokens_per_task": 6365.909595959596,
      "last_run_timestamp": "2026-10-10T16:19:37.313Z"
    },
    {
      "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": 0.614478,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.009234040404040404,
      "avg_duration_per_task_ms": 990972.4696969697,
      "avg_output_tokens_per_task": 16689.560606060608,
      "last_run_timestamp": "2026-07-18T08:03:56.103Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-14b-04-28",
      "model_slug": "qwen/qwen3-14b",
      "model_name": "Qwen: Qwen3 14B",
      "provider_name": "DeepInfra",
      "accuracy": 0.6128760749682338,
      "accuracy_std_dev": 0.007225322047320813,
      "run_count": 4,
      "total_questions": 787,
      "avg_cost_per_task_usd": 0.0033133367217280814,
      "avg_duration_per_task_ms": 404332.3011435832,
      "avg_output_tokens_per_task": 13398.312579415502,
      "last_run_timestamp": "2026-10-08T22:16:42.056Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-32b-04-28",
      "model_slug": "qwen/qwen3-32b",
      "model_name": "Qwen: Qwen3 32B",
      "provider_name": "Nebius",
      "accuracy": 0.6108306666666666,
      "accuracy_std_dev": 0.019131244624668038,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.005598868686868688,
      "avg_duration_per_task_ms": 615283.2525252525,
      "avg_output_tokens_per_task": 18654.404882154882,
      "last_run_timestamp": "2026-08-29T19:06:01.882Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "Google",
      "accuracy": 0.6102353048048047,
      "accuracy_std_dev": 0.004626267235105652,
      "run_count": 7,
      "total_questions": 1332,
      "avg_cost_per_task_usd": 0.004039552552552552,
      "avg_duration_per_task_ms": 36383.835585585584,
      "avg_output_tokens_per_task": 2146.3363363363364,
      "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-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "Novita",
      "accuracy": 0.6096681428571429,
      "accuracy_std_dev": 0.010277069224089844,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.0036427301587301586,
      "avg_duration_per_task_ms": 34466.44155844156,
      "avg_output_tokens_per_task": 2150.969696969697,
      "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-vl-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 VL 30B A3B Instruct",
      "provider_name": "DeepInfra",
      "accuracy": 0.6096471504551988,
      "accuracy_std_dev": 0.016445311419549056,
      "run_count": 11,
      "total_questions": 2087,
      "avg_cost_per_task_usd": 0.008342372783900336,
      "avg_duration_per_task_ms": 542298.5462386201,
      "avg_output_tokens_per_task": 13701.597987541927,
      "last_run_timestamp": "2026-10-08T22:50:36.237Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-32b-04-28",
      "model_slug": "qwen/qwen3-32b",
      "model_name": "Qwen: Qwen3 32B",
      "provider_name": "auto-routing",
      "accuracy": 0.6084389949367088,
      "accuracy_std_dev": 0.010283734794972733,
      "run_count": 6,
      "total_questions": 1185,
      "avg_cost_per_task_usd": 0.008408201687763715,
      "avg_duration_per_task_ms": 279347.2447257384,
      "avg_output_tokens_per_task": 17442.14936708861,
      "last_run_timestamp": "2026-08-29T19:52:22.642Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-30b-a3b-04-28",
      "model_slug": "qwen/qwen3-30b-a3b",
      "model_name": "Qwen: Qwen3 30B A3B",
      "provider_name": "DeepInfra",
      "accuracy": 0.6077443333333332,
      "accuracy_std_dev": 0.001374572014200141,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0077788232323232325,
      "avg_duration_per_task_ms": 287812.98148148146,
      "avg_output_tokens_per_task": 15362.217171717171,
      "last_run_timestamp": "2026-10-08T22:57:53.933Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "CoreWeave",
      "accuracy": 0.607744,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.003953611111111112,
      "avg_duration_per_task_ms": 57224.242424242424,
      "avg_output_tokens_per_task": 2094.1161616161617,
      "last_run_timestamp": "2026-08-01T17:44:33.510Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "WandB",
      "accuracy": 0.607744,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.004114661616161616,
      "avg_duration_per_task_ms": 51474.35353535353,
      "avg_output_tokens_per_task": 2201.479797979798,
      "last_run_timestamp": "2026-07-18T08:26:39.830Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-30b-a3b-04-28",
      "model_slug": "qwen/qwen3-30b-a3b",
      "model_name": "Qwen: Qwen3 30B A3B",
      "provider_name": "Alibaba",
      "accuracy": 0.6060605,
      "accuracy_std_dev": 0.001683499999999949,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.008235411616161616,
      "avg_duration_per_task_ms": 152037.7323232323,
      "avg_output_tokens_per_task": 15653.954545454546,
      "last_run_timestamp": "2026-08-01T17:51:29.206Z"
    },
    {
      "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": 0.605239731707317,
      "accuracy_std_dev": 0.09170649999999997,
      "run_count": 2,
      "total_questions": 369,
      "avg_cost_per_task_usd": 0.03298509756097561,
      "avg_duration_per_task_ms": 869238.1111111111,
      "avg_output_tokens_per_task": 34362.55826558266,
      "last_run_timestamp": "2026-08-01T18:05:37.080Z"
    },
    {
      "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": 0.6046178571428571,
      "accuracy_std_dev": 0.006995035043894045,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.0034120865800865803,
      "avg_duration_per_task_ms": 53141.78787878788,
      "avg_output_tokens_per_task": 2114.74531024531,
      "last_run_timestamp": "2026-10-03T16:57:23.266Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "Alibaba",
      "accuracy": 0.6036557142857142,
      "accuracy_std_dev": 0.0048341465405498454,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.011036557720057719,
      "avg_duration_per_task_ms": 33894.917748917745,
      "avg_output_tokens_per_task": 2102.8852813852814,
      "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": "DeepInfra",
      "accuracy": 0.6026559059760956,
      "accuracy_std_dev": 0.018857781098506346,
      "run_count": 7,
      "total_questions": 1255,
      "avg_cost_per_task_usd": 0.002205658167330677,
      "avg_duration_per_task_ms": 47812.13147410358,
      "avg_output_tokens_per_task": 2066.2470119521913,
      "last_run_timestamp": "2026-10-08T23:15:07.689Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "Venice",
      "accuracy": 0.6002888571428572,
      "accuracy_std_dev": 0.011234048798548233,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.003266399711399712,
      "avg_duration_per_task_ms": 60828.217893217894,
      "avg_output_tokens_per_task": 2084.8766233766232,
      "last_run_timestamp": "2026-10-03T16:59:44.383Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Nebius",
      "accuracy": 0.599327,
      "accuracy_std_dev": 0.0033669999999999534,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.014991068181818182,
      "avg_duration_per_task_ms": 284433.01515151514,
      "avg_output_tokens_per_task": 12302.075757575758,
      "last_run_timestamp": "2026-08-15T15:26:03.892Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.599327,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.012440297979797978,
      "avg_duration_per_task_ms": 230792.08080808082,
      "avg_output_tokens_per_task": 41242.71717171717,
      "last_run_timestamp": "2026-10-04T03:37:16.537Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.5980126829268292,
      "accuracy_std_dev": 0.061359,
      "run_count": 2,
      "total_questions": 369,
      "avg_cost_per_task_usd": 0.03206006233062331,
      "avg_duration_per_task_ms": 831219.6937669377,
      "avg_output_tokens_per_task": 40208.840108401084,
      "last_run_timestamp": "2026-08-01T18:07:49.346Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-14b-04-28",
      "model_slug": "qwen/qwen3-14b",
      "model_name": "Qwen: Qwen3 14B",
      "provider_name": "NextBit",
      "accuracy": 0.5973626666666666,
      "accuracy_std_dev": 0.014397638598819675,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.003072445286195286,
      "avg_duration_per_task_ms": 655562.0867003367,
      "avg_output_tokens_per_task": 13595.005050505051,
      "last_run_timestamp": "2026-10-07T13:51:26.485Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.5968335954198474,
      "accuracy_std_dev": 0.034825528389147715,
      "run_count": 6,
      "total_questions": 1179,
      "avg_cost_per_task_usd": 0.024714866836301953,
      "avg_duration_per_task_ms": 1074343.121289228,
      "avg_output_tokens_per_task": 64653.8074639525,
      "last_run_timestamp": "2026-08-30T00:32:50.853Z"
    },
    {
      "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": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "Darkbloom",
      "accuracy": 0.5903911367346939,
      "accuracy_std_dev": 0.19755781409865822,
      "run_count": 10,
      "total_questions": 1960,
      "avg_cost_per_task_usd": 0.024625327040816332,
      "avg_duration_per_task_ms": 733683.4362244898,
      "avg_output_tokens_per_task": 34746.054081632654,
      "last_run_timestamp": "2026-10-10T16:19:57.541Z"
    },
    {
      "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": 0.5878479469525959,
      "accuracy_std_dev": 0.06702947204028986,
      "run_count": 10,
      "total_questions": 1772,
      "avg_cost_per_task_usd": 0.009048836343115122,
      "avg_duration_per_task_ms": 528814.486455982,
      "avg_output_tokens_per_task": 45050.80643340858,
      "last_run_timestamp": "2026-10-03T17:04:03.802Z"
    },
    {
      "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": 0.587542,
      "accuracy_std_dev": 0.205387,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.0584044595959596,
      "avg_duration_per_task_ms": 1074676.5176767677,
      "avg_output_tokens_per_task": 23210.881313131315,
      "last_run_timestamp": "2026-10-10T16:11:14.316Z"
    },
    {
      "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": "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": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Novita",
      "accuracy": 0.5860806989010988,
      "accuracy_std_dev": 0.014672256934467207,
      "run_count": 7,
      "total_questions": 1365,
      "avg_cost_per_task_usd": 0.006546478388278388,
      "avg_duration_per_task_ms": 728895.4901098901,
      "avg_output_tokens_per_task": 43384.46666666667,
      "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": 0.5858587500000001,
      "accuracy_std_dev": 0.016059605768433433,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.010333723484848484,
      "avg_duration_per_task_ms": 11960.345328282829,
      "avg_output_tokens_per_task": 13323.82702020202,
      "last_run_timestamp": "2026-10-10T16:19:40.664Z"
    },
    {
      "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": 0.582492,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.010260318181818183,
      "avg_duration_per_task_ms": 317852.45454545453,
      "avg_output_tokens_per_task": 40768.580808080806,
      "last_run_timestamp": "2026-10-10T16:21:33.628Z"
    },
    {
      "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": 0.5786679990347491,
      "accuracy_std_dev": 0.10049636372121881,
      "run_count": 8,
      "total_questions": 1036,
      "avg_cost_per_task_usd": 0.0965155164092664,
      "avg_duration_per_task_ms": 1156721.8156370656,
      "avg_output_tokens_per_task": 27492.188223938225,
      "last_run_timestamp": "2026-10-10T16:27:46.291Z"
    },
    {
      "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": 0.5750844,
      "accuracy_std_dev": 0.02978589572666902,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.0254069303030303,
      "avg_duration_per_task_ms": 1275373.0656565656,
      "avg_output_tokens_per_task": 63407.47878787879,
      "last_run_timestamp": "2026-08-30T00:49:28.653Z"
    },
    {
      "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": 0.5654762371428571,
      "accuracy_std_dev": 0.030454147358725023,
      "run_count": 7,
      "total_questions": 700,
      "avg_cost_per_task_usd": 0.032299544285714285,
      "avg_duration_per_task_ms": 59076.84,
      "avg_output_tokens_per_task": 6910.772857142857,
      "last_run_timestamp": "2026-10-10T16:21:24.133Z"
    },
    {
      "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": 0.564814935483871,
      "accuracy_std_dev": 0.015030477400564811,
      "run_count": 3,
      "total_questions": 558,
      "avg_cost_per_task_usd": 0.0051593978494623655,
      "avg_duration_per_task_ms": 615212.1326164874,
      "avg_output_tokens_per_task": 14268.974910394265,
      "last_run_timestamp": "2026-10-05T18:22:40.566Z"
    },
    {
      "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": 0.5633876206896551,
      "accuracy_std_dev": 0.09194318097738406,
      "run_count": 5,
      "total_questions": 986,
      "avg_cost_per_task_usd": 0.029487100405679513,
      "avg_duration_per_task_ms": 2640133.111561866,
      "avg_output_tokens_per_task": 23360.7738336714,
      "last_run_timestamp": "2026-08-22T18:19:05.094Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-32b-04-28",
      "model_slug": "qwen/qwen3-32b",
      "model_name": "Qwen: Qwen3 32B",
      "provider_name": "DeepInfra",
      "accuracy": 0.5630111428571428,
      "accuracy_std_dev": 0.01687958484110801,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.004603796536796536,
      "avg_duration_per_task_ms": 419928.0497835498,
      "avg_output_tokens_per_task": 16214.744588744588,
      "last_run_timestamp": "2026-10-08T23:30:03.872Z"
    },
    {
      "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": 0.5602958771331057,
      "accuracy_std_dev": 0.05661860675431709,
      "run_count": 5,
      "total_questions": 586,
      "avg_cost_per_task_usd": 0.05058588907849829,
      "avg_duration_per_task_ms": 611241.4914675768,
      "avg_output_tokens_per_task": 24217.037542662118,
      "last_run_timestamp": "2026-08-29T19:07:49.385Z"
    },
    {
      "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": 0.5560472579013906,
      "accuracy_std_dev": 0.1010106402129715,
      "run_count": 8,
      "total_questions": 1582,
      "avg_cost_per_task_usd": 0.009295692793931732,
      "avg_duration_per_task_ms": 331772.0916561315,
      "avg_output_tokens_per_task": 27138.45195954488,
      "last_run_timestamp": "2026-10-10T16:20:03.061Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-2.5-flash-lite",
      "model_slug": "google/gemini-2.5-flash-lite",
      "model_name": "Google: Gemini 2.5 Flash Lite",
      "provider_name": "Google",
      "accuracy": 0.5505053333333333,
      "accuracy_std_dev": 0.016890890559378122,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.017397131313131312,
      "avg_duration_per_task_ms": 129348.1271043771,
      "avg_output_tokens_per_task": 43400.656565656565,
      "last_run_timestamp": "2026-10-01T19:27:01.395Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-2.5-flash-lite",
      "model_slug": "google/gemini-2.5-flash-lite",
      "model_name": "Google: Gemini 2.5 Flash Lite",
      "provider_name": "auto-routing",
      "accuracy": 0.5488216666666668,
      "accuracy_std_dev": 0.006298902038362641,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0171891531986532,
      "avg_duration_per_task_ms": 121387.12121212122,
      "avg_output_tokens_per_task": 42892.78114478115,
      "last_run_timestamp": "2026-10-01T19:15:36.386Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4o-2024-05-13",
      "model_slug": "openai/gpt-4o-2024-05-13",
      "model_name": "OpenAI: GPT-4o (2024-05-13)",
      "provider_name": "OpenAI",
      "accuracy": 0.542088,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.023943484848484846,
      "avg_duration_per_task_ms": 10793.646464646465,
      "avg_output_tokens_per_task": 1334.611111111111,
      "last_run_timestamp": "2026-08-01T08:53:34.329Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-vl-8b-instruct",
      "model_slug": "qwen/qwen3-vl-8b-instruct",
      "model_name": "Qwen: Qwen3 VL 8B Instruct",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.538721,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.005473,
      "avg_duration_per_task_ms": 117831.5808080808,
      "avg_output_tokens_per_task": 11819.661616161617,
      "last_run_timestamp": "2026-10-04T03:06:12.419Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-chat-v3",
      "model_slug": "deepseek/deepseek-chat",
      "model_name": "DeepSeek: DeepSeek V3",
      "provider_name": "DeepInfra",
      "accuracy": 0.535985,
      "accuracy_std_dev": 0.011835000612589788,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.0018682651515151514,
      "avg_duration_per_task_ms": 92043.92676767676,
      "avg_output_tokens_per_task": 1839.9911616161617,
      "last_run_timestamp": "2026-10-08T22:10:22.349Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-vl-8b-instruct",
      "model_slug": "qwen/qwen3-vl-8b-instruct",
      "model_name": "Qwen: Qwen3 VL 8B Instruct",
      "provider_name": "Alibaba",
      "accuracy": 0.5353534999999999,
      "accuracy_std_dev": 0.0016835000000000044,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.005200381313131313,
      "avg_duration_per_task_ms": 113966.90151515152,
      "avg_output_tokens_per_task": 11220.5,
      "last_run_timestamp": "2026-08-01T17:48:17.532Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-14b-04-28",
      "model_slug": "qwen/qwen3-14b",
      "model_name": "Qwen: Qwen3 14B",
      "provider_name": "Alibaba",
      "accuracy": 0.5331086666666667,
      "accuracy_std_dev": 0.05754091454106565,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.005119237373737374,
      "avg_duration_per_task_ms": 101325.84511784512,
      "avg_output_tokens_per_task": 5420.193602693603,
      "last_run_timestamp": "2026-10-01T16:05:56.758Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.7-flash-20260119",
      "model_slug": "z-ai/glm-4.7-flash",
      "model_name": "Z.ai: GLM 4.7 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.5282142748691099,
      "accuracy_std_dev": 0.013706854874680634,
      "run_count": 6,
      "total_questions": 1146,
      "avg_cost_per_task_usd": 0.01651366666666667,
      "avg_duration_per_task_ms": 725814.7731239093,
      "avg_output_tokens_per_task": 41825.98516579407,
      "last_run_timestamp": "2026-08-30T00:11:29.373Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4o-2024-08-06",
      "model_slug": "openai/gpt-4o-2024-08-06",
      "model_name": "OpenAI: GPT-4o (2024-08-06)",
      "provider_name": "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": "deepseek/deepseek-chat-v3-0324",
      "model_slug": "deepseek/deepseek-chat-v3-0324",
      "model_name": "DeepSeek: DeepSeek V3 0324",
      "provider_name": "DeepInfra",
      "accuracy": 0.517469628909552,
      "accuracy_std_dev": 0.07510589055164418,
      "run_count": 6,
      "total_questions": 1183,
      "avg_cost_per_task_usd": 0.003801546914623838,
      "avg_duration_per_task_ms": 137539.94759087067,
      "avg_output_tokens_per_task": 4071.0912933220625,
      "last_run_timestamp": "2026-08-29T20:34:10.117Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4o",
      "model_slug": "openai/gpt-4o",
      "model_name": "OpenAI: GPT-4o",
      "provider_name": "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": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.006161969696969697,
      "avg_duration_per_task_ms": 165410.29292929292,
      "avg_output_tokens_per_task": 12691.05303030303,
      "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": "auto-routing",
      "accuracy": 0.5080647741935485,
      "accuracy_std_dev": 0.12443414088451235,
      "run_count": 6,
      "total_questions": 1116,
      "avg_cost_per_task_usd": 0.004283293010752688,
      "avg_duration_per_task_ms": 107626.95967741935,
      "avg_output_tokens_per_task": 3879.1989247311826,
      "last_run_timestamp": "2026-08-29T20:57:58.443Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4o-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": "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": 0.49601803668941985,
      "accuracy_std_dev": 0.017030085894772614,
      "run_count": 6,
      "total_questions": 1172,
      "avg_cost_per_task_usd": 0.01113635409556314,
      "avg_duration_per_task_ms": 463652.90699658706,
      "avg_output_tokens_per_task": 27776.748293515357,
      "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": 0.49551055555555557,
      "accuracy_std_dev": 0.015490728216948283,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.002358911335578002,
      "avg_duration_per_task_ms": 10696.016273849607,
      "avg_output_tokens_per_task": 2339.1021324354656,
      "last_run_timestamp": "2026-10-10T16:21:44.739Z"
    },
    {
      "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": 0.4939061636197441,
      "accuracy_std_dev": 0.024075339672231702,
      "run_count": 6,
      "total_questions": 1094,
      "avg_cost_per_task_usd": 0.010896170018281536,
      "avg_duration_per_task_ms": 477480.3711151737,
      "avg_output_tokens_per_task": 26830.819012797074,
      "last_run_timestamp": "2026-08-30T00:23:23.003Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "SambaNova",
      "accuracy": 0.4855029999999999,
      "accuracy_std_dev": 0.0052309263679263035,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.0023659646464646465,
      "avg_duration_per_task_ms": 14180.718294051627,
      "avg_output_tokens_per_task": 2226.1846240179575,
      "last_run_timestamp": "2026-10-10T16:21:33.523Z"
    },
    {
      "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": 0.48439725021570323,
      "accuracy_std_dev": 0.017142665996636056,
      "run_count": 12,
      "total_questions": 2318,
      "avg_cost_per_task_usd": 0.0011154288179465058,
      "avg_duration_per_task_ms": 60423.7131147541,
      "avg_output_tokens_per_task": 2138.6691113028473,
      "last_run_timestamp": "2026-10-10T16:22:21.682Z"
    },
    {
      "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": 0.48063949999999994,
      "accuracy_std_dev": 0.003857437653417109,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.04852650126262626,
      "avg_duration_per_task_ms": 45991.458333333336,
      "avg_output_tokens_per_task": 17009.132575757576,
      "last_run_timestamp": "2026-08-15T14:46:10.940Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "CoreWeave",
      "accuracy": 0.4797977142857143,
      "accuracy_std_dev": 0.009264662382473756,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.002598592352092352,
      "avg_duration_per_task_ms": 54248.330447330445,
      "avg_output_tokens_per_task": 2851.073593073593,
      "last_run_timestamp": "2026-10-10T16:22:32.082Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen-2.5-72b-instruct",
      "model_slug": "qwen/qwen-2.5-72b-instruct",
      "model_name": "Qwen2.5 72B Instruct",
      "provider_name": "DeepInfra",
      "accuracy": 0.4765793529411765,
      "accuracy_std_dev": 0.006944000000000006,
      "run_count": 2,
      "total_questions": 306,
      "avg_cost_per_task_usd": 0.0011232875816993464,
      "avg_duration_per_task_ms": 126677.51307189543,
      "avg_output_tokens_per_task": 2069.3627450980393,
      "last_run_timestamp": "2026-10-01T16:04:45.916Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "Parasail",
      "accuracy": 0.4757993750000001,
      "accuracy_std_dev": 0.01160075218399113,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.0013241029040404041,
      "avg_duration_per_task_ms": 47333.95138888889,
      "avg_output_tokens_per_task": 2371.179924242424,
      "last_run_timestamp": "2026-10-10T16:25:26.229Z"
    },
    {
      "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": "Together",
      "accuracy": 0.4756825555555556,
      "accuracy_std_dev": 0.012142897971069449,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.003145421997755331,
      "avg_duration_per_task_ms": 58008.31144781145,
      "avg_output_tokens_per_task": 2218.3967452300785,
      "last_run_timestamp": "2026-10-10T16:27:29.219Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "Nebius",
      "accuracy": 0.4741861666666667,
      "accuracy_std_dev": 0.012808772416026279,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.0010600345117845118,
      "avg_duration_per_task_ms": 199985.06313131313,
      "avg_output_tokens_per_task": 2395.604377104377,
      "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": "Novita",
      "accuracy": 0.47383958607594934,
      "accuracy_std_dev": 0.017544611925299883,
      "run_count": 8,
      "total_questions": 1580,
      "avg_cost_per_task_usd": 0.0013290139240506329,
      "avg_duration_per_task_ms": 95497.43164556962,
      "avg_output_tokens_per_task": 3050.5791139240505,
      "last_run_timestamp": "2026-10-10T16:26:55.311Z"
    },
    {
      "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": 0.47264275,
      "accuracy_std_dev": 0.009698346982218158,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.0013737335858585856,
      "avg_duration_per_task_ms": 72305.53409090909,
      "avg_output_tokens_per_task": 2290.848484848485,
      "last_run_timestamp": "2026-10-10T16:23:00.867Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.469697,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.006110727272727273,
      "avg_duration_per_task_ms": 42879.368686868685,
      "avg_output_tokens_per_task": 2318.919191919192,
      "last_run_timestamp": "2026-10-04T03:02:23.034Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "WandB",
      "accuracy": 0.469697,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0022648939393939394,
      "avg_duration_per_task_ms": 33028.21717171717,
      "avg_output_tokens_per_task": 2380.9646464646466,
      "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": "Crusoe",
      "accuracy": 0.46742483824768327,
      "accuracy_std_dev": 0.006948085005956674,
      "run_count": 6,
      "total_questions": 1187,
      "avg_cost_per_task_usd": 0.002137747262005055,
      "avg_duration_per_task_ms": 34994.02274641955,
      "avg_output_tokens_per_task": 2587.217354675653,
      "last_run_timestamp": "2026-08-29T16:31:42.504Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "DeepInfra",
      "accuracy": 0.46286933333333335,
      "accuracy_std_dev": 0.017560541816495547,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.0007922811447811448,
      "avg_duration_per_task_ms": 158149.69472502806,
      "avg_output_tokens_per_task": 2225.2008978675644,
      "last_run_timestamp": "2026-10-10T16:20:22.051Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen2.5-vl-72b-instruct",
      "model_slug": "qwen/qwen2.5-vl-72b-instruct",
      "model_name": "Qwen: Qwen2.5 VL 72B Instruct",
      "provider_name": "Parasail",
      "accuracy": 0.462542,
      "accuracy_std_dev": 0.005471000000000004,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.001783989898989899,
      "avg_duration_per_task_ms": 27644.777777777777,
      "avg_output_tokens_per_task": 1397.5252525252524,
      "last_run_timestamp": "2026-09-01T13:03:42.840Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "Ionstream",
      "accuracy": 0.46128619685039374,
      "accuracy_std_dev": 0.075159,
      "run_count": 2,
      "total_questions": 254,
      "avg_cost_per_task_usd": 0.00731248031496063,
      "avg_duration_per_task_ms": 1016477.8543307086,
      "avg_output_tokens_per_task": 20586.740157480315,
      "last_run_timestamp": "2026-08-08T10:11:59.748Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen-2.5-72b-instruct",
      "model_slug": "qwen/qwen-2.5-72b-instruct",
      "model_name": "Qwen2.5 72B Instruct",
      "provider_name": "Novita",
      "accuracy": 0.461279,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0011232171717171719,
      "avg_duration_per_task_ms": 136788.7626262626,
      "avg_output_tokens_per_task": 2036.2272727272727,
      "last_run_timestamp": "2026-10-01T16:27:00.375Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen-2.5-72b-instruct",
      "model_slug": "qwen/qwen-2.5-72b-instruct",
      "model_name": "Qwen2.5 72B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.45959600000000006,
      "accuracy_std_dev": 0.01346799999999998,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.0010278964646464647,
      "avg_duration_per_task_ms": 96379.95959595959,
      "avg_output_tokens_per_task": 1848.388888888889,
      "last_run_timestamp": "2026-10-01T16:47:11.191Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-vl-8b-instruct",
      "model_slug": "qwen/qwen3-vl-8b-instruct",
      "model_name": "Qwen: Qwen3 VL 8B Instruct",
      "provider_name": "Parasail",
      "accuracy": 0.45862900000000006,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 141,
      "avg_cost_per_task_usd": 0.0011888297872340425,
      "avg_duration_per_task_ms": 278701.0070921986,
      "avg_output_tokens_per_task": 7719.879432624113,
      "last_run_timestamp": "2026-08-01T18:13:08.069Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen2.5-vl-72b-instruct",
      "model_slug": "qwen/qwen2.5-vl-72b-instruct",
      "model_name": "Qwen: Qwen2.5 VL 72B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.4444445,
      "accuracy_std_dev": 0.015151500000000012,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.0015239469696969696,
      "avg_duration_per_task_ms": 59255.792929292926,
      "avg_output_tokens_per_task": 1592.9924242424242,
      "last_run_timestamp": "2026-09-01T13:21:35.512Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen2.5-vl-72b-instruct",
      "model_slug": "qwen/qwen2.5-vl-72b-instruct",
      "model_name": "Qwen: Qwen2.5 VL 72B Instruct",
      "provider_name": "Nebius",
      "accuracy": 0.4402355,
      "accuracy_std_dev": 0.002525500000000014,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.0013545,
      "avg_duration_per_task_ms": 68001.7196969697,
      "avg_output_tokens_per_task": 1535.1843434343434,
      "last_run_timestamp": "2026-09-01T12:41:25.955Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4o-mini",
      "model_slug": "openai/gpt-4o-mini",
      "model_name": "OpenAI: GPT-4o-mini",
      "provider_name": "auto-routing",
      "accuracy": 0.4320986666666667,
      "accuracy_std_dev": 0.00781631474687543,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0010060690235690235,
      "avg_duration_per_task_ms": 20918.510101010103,
      "avg_output_tokens_per_task": 1483.5286195286196,
      "last_run_timestamp": "2026-10-01T19:16:07.087Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4o-mini",
      "model_slug": "openai/gpt-4o-mini",
      "model_name": "OpenAI: GPT-4o-mini",
      "provider_name": "Azure",
      "accuracy": 0.4315373333333334,
      "accuracy_std_dev": 0.010588221658469791,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.0010723358585858589,
      "avg_duration_per_task_ms": 21765.176767676767,
      "avg_output_tokens_per_task": 1505.858585858586,
      "last_run_timestamp": "2026-10-01T19:05:18.003Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4o-mini",
      "model_slug": "openai/gpt-4o-mini",
      "model_name": "OpenAI: GPT-4o-mini",
      "provider_name": "OpenAI",
      "accuracy": 0.4305555,
      "accuracy_std_dev": 0.007469952225416168,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.0010262436868686867,
      "avg_duration_per_task_ms": 17602.7702020202,
      "avg_output_tokens_per_task": 1517.9292929292928,
      "last_run_timestamp": "2026-10-02T23:19:57.879Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "AkashML",
      "accuracy": 0.40740725,
      "accuracy_std_dev": 0.3006428663691981,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.017353968434343434,
      "avg_duration_per_task_ms": 189056.44444444444,
      "avg_output_tokens_per_task": 19318.219696969696,
      "last_run_timestamp": "2026-08-15T15:24:35.389Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-3-12b-it",
      "model_slug": "google/gemma-3-12b-it",
      "model_name": "Google: Gemma 3 12B",
      "provider_name": "DeepInfra",
      "accuracy": 0.405724,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0003259848484848485,
      "avg_duration_per_task_ms": 29056.348484848484,
      "avg_output_tokens_per_task": 1917.1515151515152,
      "last_run_timestamp": "2026-10-08T22:01:16.852Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "DigitalOcean",
      "accuracy": 0.3994878450704225,
      "accuracy_std_dev": 0.0901063557941392,
      "run_count": 4,
      "total_questions": 781,
      "avg_cost_per_task_usd": 0.016465559539052498,
      "avg_duration_per_task_ms": 3369255.4609475033,
      "avg_output_tokens_per_task": 16587.451984635085,
      "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": "google/gemma-3-12b-it",
      "model_slug": "google/gemma-3-12b-it",
      "model_name": "Google: Gemma 3 12B",
      "provider_name": "NextBit",
      "accuracy": 0.37878799999999996,
      "accuracy_std_dev": 0.01683499999999999,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.000344469696969697,
      "avg_duration_per_task_ms": 46431.74494949495,
      "avg_output_tokens_per_task": 2036.358585858586,
      "last_run_timestamp": "2026-10-07T13:35:26.756Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.36698714615384614,
      "accuracy_std_dev": 0.09273293342139835,
      "run_count": 8,
      "total_questions": 1560,
      "avg_cost_per_task_usd": 0.0036227448717948718,
      "avg_duration_per_task_ms": 36782.153846153844,
      "avg_output_tokens_per_task": 1520.774358974359,
      "last_run_timestamp": "2026-10-10T16:29:29.248Z"
    },
    {
      "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": "mistralai/mistral-nemo",
      "model_slug": "mistralai/mistral-nemo",
      "model_name": "Mistral: Mistral Nemo",
      "provider_name": "Mistral",
      "accuracy": 0.350168,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.00011657070707070707,
      "avg_duration_per_task_ms": 7587.606060606061,
      "avg_output_tokens_per_task": 355.9040404040404,
      "last_run_timestamp": "2026-07-18T09:09:45.839Z"
    },
    {
      "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": "mistralai/mistral-nemo",
      "model_slug": "mistralai/mistral-nemo",
      "model_name": "Mistral: Mistral Nemo",
      "provider_name": "Novita",
      "accuracy": 0.344383,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 181,
      "avg_cost_per_task_usd": 0.00011365193370165745,
      "avg_duration_per_task_ms": 62171.872928176796,
      "avg_output_tokens_per_task": 485.27624309392263,
      "last_run_timestamp": "2026-07-18T08:00:15.379Z"
    },
    {
      "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": "mistralai/mistral-nemo",
      "model_slug": "mistralai/mistral-nemo",
      "model_name": "Mistral: Mistral Nemo",
      "provider_name": "auto-routing",
      "accuracy": 0.338384,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.00013078787878787878,
      "avg_duration_per_task_ms": 16627.883838383837,
      "avg_output_tokens_per_task": 667.8888888888889,
      "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": "DeepInfra",
      "accuracy": 0.301347,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.00002756060606060606,
      "avg_duration_per_task_ms": 15713.474747474747,
      "avg_output_tokens_per_task": 431.8787878787879,
      "last_run_timestamp": "2026-07-18T08:00:15.193Z"
    },
    {
      "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": 0.29292928571428567,
      "accuracy_std_dev": 0.07781645629981633,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.008772483405483405,
      "avg_duration_per_task_ms": 593759.7330447331,
      "avg_output_tokens_per_task": 21699.420634920636,
      "last_run_timestamp": "2026-10-10T16:20:00.361Z"
    },
    {
      "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": 0.289562,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.00002257070707070707,
      "avg_duration_per_task_ms": 19325.131313131315,
      "avg_output_tokens_per_task": 285.6515151515151,
      "last_run_timestamp": "2026-07-18T08:04:03.449Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "sao10k/l3-lunaris-8b",
      "model_slug": "sao10k/l3-lunaris-8b",
      "model_name": "Sao10K: Llama 3 8B Lunaris",
      "provider_name": "DeepInfra",
      "accuracy": 0.2891415,
      "accuracy_std_dev": 0.003232916137174003,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.00006112626262626263,
      "avg_duration_per_task_ms": 10017.040404040405,
      "avg_output_tokens_per_task": 577.7083333333334,
      "last_run_timestamp": "2026-10-08T22:00:27.722Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.1-8b-instruct",
      "model_slug": "meta-llama/llama-3.1-8b-instruct",
      "model_name": "Meta: Llama 3.1 8B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.284932625,
      "accuracy_std_dev": 0.018475559591914253,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.0014604040404040406,
      "avg_duration_per_task_ms": 111179.26388888889,
      "avg_output_tokens_per_task": 19382.94191919192,
      "last_run_timestamp": "2026-10-10T16:10:32.137Z"
    },
    {
      "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": 0.279356125,
      "accuracy_std_dev": 0.011390538589521344,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.0023889917929292926,
      "avg_duration_per_task_ms": 48269.878787878784,
      "avg_output_tokens_per_task": 29863.64267676768,
      "last_run_timestamp": "2026-10-10T16:10:51.751Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Coder 30B A3B Instruct",
      "provider_name": "Novita",
      "accuracy": 0.275533,
      "accuracy_std_dev": 0.009152576176501717,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0007405521885521885,
      "avg_duration_per_task_ms": 48360.410774410775,
      "avg_output_tokens_per_task": 2532.1531986531986,
      "last_run_timestamp": "2026-08-15T09:28:07.208Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "sao10k/l3-lunaris-8b",
      "model_slug": "sao10k/l3-lunaris-8b",
      "model_name": "Sao10K: Llama 3 8B Lunaris",
      "provider_name": "Novita",
      "accuracy": 0.2739361560283688,
      "accuracy_std_dev": 0.008153796102974918,
      "run_count": 3,
      "total_questions": 564,
      "avg_cost_per_task_usd": 0.00006610283687943263,
      "avg_duration_per_task_ms": 8879.929078014184,
      "avg_output_tokens_per_task": 551.2074468085107,
      "last_run_timestamp": "2026-10-01T16:01:04.877Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": "CoreWeave",
      "accuracy": 0.27249684030418253,
      "accuracy_std_dev": 0.016430604143810795,
      "run_count": 7,
      "total_questions": 1315,
      "avg_cost_per_task_usd": 0.008003044866920151,
      "avg_duration_per_task_ms": 324114.34372623573,
      "avg_output_tokens_per_task": 35682.14980988593,
      "last_run_timestamp": "2026-10-10T16:10:51.751Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "sao10k/l3-lunaris-8b",
      "model_slug": "sao10k/l3-lunaris-8b",
      "model_name": "Sao10K: Llama 3 8B Lunaris",
      "provider_name": "Parasail",
      "accuracy": 0.2700681612244898,
      "accuracy_std_dev": 0.016752437809465226,
      "run_count": 5,
      "total_questions": 980,
      "avg_cost_per_task_usd": 0.00005762346938775511,
      "avg_duration_per_task_ms": 9734.870408163266,
      "avg_output_tokens_per_task": 576.8010204081633,
      "last_run_timestamp": "2026-10-01T20:47:49.128Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.1-8b-instruct",
      "model_slug": "meta-llama/llama-3.1-8b-instruct",
      "model_name": "Meta: Llama 3.1 8B Instruct",
      "provider_name": "DeepInfra",
      "accuracy": 0.26565141116751273,
      "accuracy_std_dev": 0.016394730800616743,
      "run_count": 9,
      "total_questions": 1773,
      "avg_cost_per_task_usd": 0.00036760180485053583,
      "avg_duration_per_task_ms": 403603.9740552736,
      "avg_output_tokens_per_task": 8952.823463056966,
      "last_run_timestamp": "2026-10-10T16:11:11.296Z"
    },
    {
      "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": 0.25645355555555555,
      "accuracy_std_dev": 0.014994751949421879,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.0004083338945005611,
      "avg_duration_per_task_ms": 68671.33221099888,
      "avg_output_tokens_per_task": 7843.052188552188,
      "last_run_timestamp": "2026-10-10T16:11:11.295Z"
    },
    {
      "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": 0.254209,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.001291050505050505,
      "avg_duration_per_task_ms": 40150.21212121212,
      "avg_output_tokens_per_task": 5062.732323232323,
      "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": "Cloudflare",
      "accuracy": 0.23148166666666667,
      "accuracy_std_dev": 0.10429433789947032,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.0013719949494949493,
      "avg_duration_per_task_ms": 502712.0462962963,
      "avg_output_tokens_per_task": 4456.498316498317,
      "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": 0.1910775,
      "accuracy_std_dev": 0.11363615432122239,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.0014224511784511782,
      "avg_duration_per_task_ms": 46218.05892255892,
      "avg_output_tokens_per_task": 1225.9587542087543,
      "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": 0.17174282233502539,
      "accuracy_std_dev": 0.01788397279378631,
      "run_count": 3,
      "total_questions": 591,
      "avg_cost_per_task_usd": 0.019897788494077834,
      "avg_duration_per_task_ms": 1784246.4196277496,
      "avg_output_tokens_per_task": 110294.06598984772,
      "last_run_timestamp": "2026-10-08T23:13:58.126Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.2-3b-instruct",
      "model_slug": "meta-llama/llama-3.2-3b-instruct",
      "model_name": "Meta: Llama 3.2 3B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.116162,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0001537979797979798,
      "avg_duration_per_task_ms": 29722.909090909092,
      "avg_output_tokens_per_task": 1836.7929292929293,
      "last_run_timestamp": "2026-08-01T09:03:02.970Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.1-flash-lite-image-20260630",
      "model_slug": "google/gemini-3.1-flash-lite-image",
      "model_name": "Google: Nano Banana 2 Lite (Gemini 3.1 Flash Lite Image)",
      "provider_name": "Google AI Studio",
      "accuracy": 0.1161615,
      "accuracy_std_dev": 0.0016835000000000044,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.017850737373737373,
      "avg_duration_per_task_ms": 46853.982323232325,
      "avg_output_tokens_per_task": 11776.444444444445,
      "last_run_timestamp": "2026-10-01T16:00:58.389Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.1-flash-lite-image-20260630",
      "model_slug": "google/gemini-3.1-flash-lite-image",
      "model_name": "Google: Nano Banana 2 Lite (Gemini 3.1 Flash Lite Image)",
      "provider_name": "Google",
      "accuracy": 0.112795,
      "accuracy_std_dev": 0.006733999999999997,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.018038613636363634,
      "avg_duration_per_task_ms": 21125.79292929293,
      "avg_output_tokens_per_task": 11795.222222222223,
      "last_run_timestamp": "2026-10-01T16:00:58.364Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.1-flash-lite-image-20260630",
      "model_slug": "google/gemini-3.1-flash-lite-image",
      "model_name": "Google: Nano Banana 2 Lite (Gemini 3.1 Flash Lite Image)",
      "provider_name": "auto-routing",
      "accuracy": 0.102694,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.017984406565656568,
      "avg_duration_per_task_ms": 21554.590909090908,
      "avg_output_tokens_per_task": 11812.823232323231,
      "last_run_timestamp": "2026-10-01T16:02:13.023Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.8368423578947368,
      "accuracy_std_dev": 0.014073500000000017,
      "run_count": 2,
      "total_questions": 95,
      "avg_cost_per_task_usd": 0.008221831578947368,
      "avg_duration_per_task_ms": 31565.694736842106,
      "avg_output_tokens_per_task": 9295.157894736842,
      "last_run_timestamp": "2026-10-10T11:02:59.339Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Wafer",
      "accuracy": 0.826667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.00673608,
      "avg_duration_per_task_ms": 350369.78,
      "avg_output_tokens_per_task": 14092.86,
      "last_run_timestamp": "2026-10-10T20:45:37.638Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Cerebras",
      "accuracy": 0.82,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.28041234,
      "avg_duration_per_task_ms": 27407.28,
      "avg_output_tokens_per_task": 12578.56,
      "last_run_timestamp": "2026-10-10T11:23:44.621Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Reka",
      "accuracy": 0.8115940000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 46,
      "avg_cost_per_task_usd": 0.05019476086956522,
      "avg_duration_per_task_ms": 222941.58695652173,
      "avg_output_tokens_per_task": 15139.565217391304,
      "last_run_timestamp": "2026-10-10T19:11:12.554Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.8053333,
      "accuracy_std_dev": 0.019043900162781787,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.8753633840000001,
      "avg_duration_per_task_ms": 132992.252,
      "avg_output_tokens_per_task": 5787.606,
      "last_run_timestamp": "2026-10-10T11:24:18.153Z"
    },
    {
      "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": 0.8,
      "accuracy_std_dev": 0.02683269150495342,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.9213041,
      "avg_duration_per_task_ms": 126365.092,
      "avg_output_tokens_per_task": 5777.006,
      "last_run_timestamp": "2026-10-10T11:19:59.852Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-5-fable-20260609",
      "model_slug": "anthropic/claude-fable-5",
      "model_name": "Anthropic: Claude Fable 5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.8,
      "accuracy_std_dev": 0.033110611481316166,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.7322859599999999,
      "avg_duration_per_task_ms": 146639.07333333333,
      "avg_output_tokens_per_task": 5472.34,
      "last_run_timestamp": "2026-08-15T08:09:14.229Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "WandB",
      "accuracy": 0.8,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.030123700000000003,
      "avg_duration_per_task_ms": 294712.06,
      "avg_output_tokens_per_task": 14851.14,
      "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": "anthropic/claude-5-fable-20260609",
      "model_slug": "anthropic/claude-fable-5",
      "model_name": "Anthropic: Claude Fable 5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.7991667499999999,
      "accuracy_std_dev": 0.021714744735720465,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.8790712225,
      "avg_duration_per_task_ms": 145295.465,
      "avg_output_tokens_per_task": 5765.4325,
      "last_run_timestamp": "2026-10-10T11:04:02.050Z"
    },
    {
      "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": "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": 0.7966665,
      "accuracy_std_dev": 0.0033335000000000448,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01733308,
      "avg_duration_per_task_ms": 494326.98,
      "avg_output_tokens_per_task": 23551.69,
      "last_run_timestamp": "2026-10-10T13:10:11.999Z"
    },
    {
      "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": 0.7955556666666667,
      "accuracy_std_dev": 0.00831485358994507,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.05401102000000001,
      "avg_duration_per_task_ms": 879734.1466666666,
      "avg_output_tokens_per_task": 37622.22,
      "last_run_timestamp": "2026-10-10T13:25:18.383Z"
    },
    {
      "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": 0.7953334,
      "accuracy_std_dev": 0.01739731013806443,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.811625548,
      "avg_duration_per_task_ms": 134771.192,
      "avg_output_tokens_per_task": 5750.948,
      "last_run_timestamp": "2026-10-10T11:20:49.389Z"
    },
    {
      "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": 0.7921053684210526,
      "accuracy_std_dev": 0.021335845450199346,
      "run_count": 19,
      "total_questions": 950,
      "avg_cost_per_task_usd": 0.9452779715789473,
      "avg_duration_per_task_ms": 96650.52210526315,
      "avg_output_tokens_per_task": 5665.656842105263,
      "last_run_timestamp": "2026-10-10T11:36:10.174Z"
    },
    {
      "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": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Baidu",
      "accuracy": 0.7908167142857143,
      "accuracy_std_dev": 0.016180499999999987,
      "run_count": 2,
      "total_questions": 98,
      "avg_cost_per_task_usd": 0.030663948979591835,
      "avg_duration_per_task_ms": 84618.54081632652,
      "avg_output_tokens_per_task": 15463.275510204081,
      "last_run_timestamp": "2026-10-10T11:01:59.660Z"
    },
    {
      "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": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "DeepSeek",
      "accuracy": 0.79,
      "accuracy_std_dev": 0.01666640000120001,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.016075525,
      "avg_duration_per_task_ms": 177598.75,
      "avg_output_tokens_per_task": 11878.23,
      "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": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Nebius",
      "accuracy": 0.7897090469798659,
      "accuracy_std_dev": 0.0266343317585071,
      "run_count": 3,
      "total_questions": 149,
      "avg_cost_per_task_usd": 0.24385301342281882,
      "avg_duration_per_task_ms": 51187.906040268455,
      "avg_output_tokens_per_task": 4637.738255033557,
      "last_run_timestamp": "2026-10-10T11:25:15.915Z"
    },
    {
      "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": 0.7874807758887171,
      "accuracy_std_dev": 0.014215753702334737,
      "run_count": 13,
      "total_questions": 647,
      "avg_cost_per_task_usd": 0.10405211437403399,
      "avg_duration_per_task_ms": 139136.8485316847,
      "avg_output_tokens_per_task": 11436.242658423493,
      "last_run_timestamp": "2026-10-10T11:58:08.220Z"
    },
    {
      "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": 0.7873333,
      "accuracy_std_dev": 0.02279383055148914,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.487091924,
      "avg_duration_per_task_ms": 126152.578,
      "avg_output_tokens_per_task": 7141.378,
      "last_run_timestamp": "2026-10-10T11:33:13.840Z"
    },
    {
      "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": 0.7871621216216216,
      "accuracy_std_dev": 0.015496260115123086,
      "run_count": 3,
      "total_questions": 148,
      "avg_cost_per_task_usd": 0.06734989864864864,
      "avg_duration_per_task_ms": 463154.4662162162,
      "avg_output_tokens_per_task": 9087.885135135135,
      "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": "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": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.14809964,
      "avg_duration_per_task_ms": 328566.04,
      "avg_output_tokens_per_task": 11598.26,
      "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": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "DeepInfra",
      "accuracy": 0.7866669999999999,
      "accuracy_std_dev": 0.040000000000000036,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.07300471,
      "avg_duration_per_task_ms": 912319.43,
      "avg_output_tokens_per_task": 13438.04,
      "last_run_timestamp": "2026-10-10T12:49:31.998Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "Darkbloom",
      "accuracy": 0.7862320000000002,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 46,
      "avg_cost_per_task_usd": 0.029582717391304348,
      "avg_duration_per_task_ms": 2435445.304347826,
      "avg_output_tokens_per_task": 108721.58695652174,
      "last_run_timestamp": "2026-10-03T16:12:07.492Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "DeepInfra",
      "accuracy": 0.7860360641891891,
      "accuracy_std_dev": 0.025554613476174253,
      "run_count": 6,
      "total_questions": 296,
      "avg_cost_per_task_usd": 0.09440423986486486,
      "avg_duration_per_task_ms": 381879.5878378378,
      "avg_output_tokens_per_task": 16484.91891891892,
      "last_run_timestamp": "2026-10-10T11:07:21.694Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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-sonnet-5.5-20260928",
      "model_slug": "anthropic/claude-sonnet-5.5",
      "model_name": "Anthropic: Claude Sonnet 5.5",
      "provider_name": "Google",
      "accuracy": 0.785926,
      "accuracy_std_dev": 0.017900825710316006,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.18531832888888886,
      "avg_duration_per_task_ms": 64060.64,
      "avg_output_tokens_per_task": 6337.593333333333,
      "last_run_timestamp": "2026-10-10T11:33:06.422Z"
    },
    {
      "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": 0.7853334,
      "accuracy_std_dev": 0.008844443331267377,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.08491233599999999,
      "avg_duration_per_task_ms": 41767.208,
      "avg_output_tokens_per_task": 4731.868,
      "last_run_timestamp": "2026-10-10T11:37:21.818Z"
    },
    {
      "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": 0.7850339489795918,
      "accuracy_std_dev": 0.024279488005516096,
      "run_count": 10,
      "total_questions": 490,
      "avg_cost_per_task_usd": 0.08008161836734694,
      "avg_duration_per_task_ms": 1138745.7020408164,
      "avg_output_tokens_per_task": 26323.910204081632,
      "last_run_timestamp": "2026-10-10T19:24:40.467Z"
    },
    {
      "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": "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": 0.7844446666666666,
      "accuracy_std_dev": 0.013698571612479267,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.18011569333333333,
      "avg_duration_per_task_ms": 71496.32666666666,
      "avg_output_tokens_per_task": 6301.473333333333,
      "last_run_timestamp": "2026-10-10T11:00:15.933Z"
    },
    {
      "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": 0.7844443333333334,
      "accuracy_std_dev": 0.006285236475706792,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.15203402,
      "avg_duration_per_task_ms": 476027.4533333333,
      "avg_output_tokens_per_task": 12653,
      "last_run_timestamp": "2026-10-10T21:40:31.943Z"
    },
    {
      "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": 0.78416675,
      "accuracy_std_dev": 0.021327508971689602,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.09952602499999999,
      "avg_duration_per_task_ms": 76992.08,
      "avg_output_tokens_per_task": 4752.155,
      "last_run_timestamp": "2026-10-10T11:49:24.601Z"
    },
    {
      "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": 0.7837446052249636,
      "accuracy_std_dev": 0.019731934402518834,
      "run_count": 14,
      "total_questions": 689,
      "avg_cost_per_task_usd": 0.08364079680696661,
      "avg_duration_per_task_ms": 85376.37300435414,
      "avg_output_tokens_per_task": 4386.013062409289,
      "last_run_timestamp": "2026-10-10T11:59:03.438Z"
    },
    {
      "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": 0.7837037777777778,
      "accuracy_std_dev": 0.02157059167672393,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.06625726222222222,
      "avg_duration_per_task_ms": 303688.90222222224,
      "avg_output_tokens_per_task": 15392.597777777777,
      "last_run_timestamp": "2026-10-10T20:56:17.811Z"
    },
    {
      "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": 0.7827380586734692,
      "accuracy_std_dev": 0.0387434346968538,
      "run_count": 8,
      "total_questions": 392,
      "avg_cost_per_task_usd": 0.11838895663265307,
      "avg_duration_per_task_ms": 3017033.7933673467,
      "avg_output_tokens_per_task": 26124.95918367347,
      "last_run_timestamp": "2026-10-10T19:28:36.252Z"
    },
    {
      "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": 0.7827005274261605,
      "accuracy_std_dev": 0.02374291531720566,
      "run_count": 5,
      "total_questions": 237,
      "avg_cost_per_task_usd": 0.10862543037974683,
      "avg_duration_per_task_ms": 71286.46835443038,
      "avg_output_tokens_per_task": 8041.662447257384,
      "last_run_timestamp": "2026-08-29T17:15:26.910Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "Phala",
      "accuracy": 0.782500125,
      "accuracy_std_dev": 0.028662343424943032,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.19210731,
      "avg_duration_per_task_ms": 737716.5375,
      "avg_output_tokens_per_task": 31714.73,
      "last_run_timestamp": "2026-10-10T18:35:22.054Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Novita",
      "accuracy": 0.7822223333333334,
      "accuracy_std_dev": 0.021998563367234276,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.011777086666666667,
      "avg_duration_per_task_ms": 550585.4066666666,
      "avg_output_tokens_per_task": 22732.366666666665,
      "last_run_timestamp": "2026-10-10T13:14:05.134Z"
    },
    {
      "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": 0.7822221666666667,
      "accuracy_std_dev": 0.012570738116974498,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.1844831233333333,
      "avg_duration_per_task_ms": 71285.40333333334,
      "avg_output_tokens_per_task": 6291.4,
      "last_run_timestamp": "2026-10-10T11:08:41.713Z"
    },
    {
      "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": 0.7822221666666667,
      "accuracy_std_dev": 0.017915990515148437,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.07431559666666666,
      "avg_duration_per_task_ms": 47439.64666666667,
      "avg_output_tokens_per_task": 4714.83,
      "last_run_timestamp": "2026-10-10T11:39:52.102Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "ModelRun",
      "accuracy": 0.7822220000000001,
      "accuracy_std_dev": 0.020608136952831674,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.0806628,
      "avg_duration_per_task_ms": 57888.54,
      "avg_output_tokens_per_task": 12516.78,
      "last_run_timestamp": "2026-10-10T11:19:52.284Z"
    },
    {
      "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": 0.78166675,
      "accuracy_std_dev": 0.022985462799071505,
      "run_count": 20,
      "total_questions": 1000,
      "avg_cost_per_task_usd": 0.47495438199999995,
      "avg_duration_per_task_ms": 107875.898,
      "avg_output_tokens_per_task": 7230.244,
      "last_run_timestamp": "2026-10-10T11:20:52.353Z"
    },
    {
      "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": 0.7816666250000001,
      "accuracy_std_dev": 0.02303379461756953,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.48336481500000006,
      "avg_duration_per_task_ms": 127661.045,
      "avg_output_tokens_per_task": 7945.53,
      "last_run_timestamp": "2026-10-10T10:58:55.051Z"
    },
    {
      "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": 0.7813332000000001,
      "accuracy_std_dev": 0.03357245431838428,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.03929822,
      "avg_duration_per_task_ms": 339374.972,
      "avg_output_tokens_per_task": 12667.496,
      "last_run_timestamp": "2026-10-10T11:49:13.044Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Crusoe",
      "accuracy": 0.7811448181818181,
      "accuracy_std_dev": 0.0011565000000000047,
      "run_count": 2,
      "total_questions": 99,
      "avg_cost_per_task_usd": 0.035013020202020205,
      "avg_duration_per_task_ms": 141478.38383838383,
      "avg_output_tokens_per_task": 23821.878787878788,
      "last_run_timestamp": "2026-10-10T11:21:33.807Z"
    },
    {
      "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": 0.7803333,
      "accuracy_std_dev": 0.024333256165379916,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.13113940600000001,
      "avg_duration_per_task_ms": 299273.458,
      "avg_output_tokens_per_task": 19639.796,
      "last_run_timestamp": "2026-10-10T12:00:37.936Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "amazon/nova-micro-v1",
      "model_slug": "amazon/nova-micro-v1",
      "model_name": "Amazon: Nova Micro 1.0",
      "provider_name": "auto-routing",
      "accuracy": 0.7800003333333333,
      "accuracy_std_dev": 0.018856180831641284,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 1.2757919799999997,
      "avg_duration_per_task_ms": 102078.39333333333,
      "avg_output_tokens_per_task": 5685.653333333334,
      "last_run_timestamp": "2026-10-01T19:04:24.509Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-sonnet-5.5-20260928",
      "model_slug": "anthropic/claude-sonnet-5.5",
      "model_name": "Anthropic: Claude Sonnet 5.5",
      "provider_name": "auto-routing",
      "accuracy": 0.78,
      "accuracy_std_dev": 0.033333,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.15629982,
      "avg_duration_per_task_ms": 82269.59,
      "avg_output_tokens_per_task": 6332.78,
      "last_run_timestamp": "2026-10-10T11:25:14.954Z"
    },
    {
      "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": 0.78,
      "accuracy_std_dev": 0.029514395734962978,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.086451392,
      "avg_duration_per_task_ms": 268938.36,
      "avg_output_tokens_per_task": 20045.732,
      "last_run_timestamp": "2026-10-10T11:31:04.105Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Near AI",
      "accuracy": 0.78,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.04487794,
      "avg_duration_per_task_ms": 324016.02,
      "avg_output_tokens_per_task": 12761.91,
      "last_run_timestamp": "2026-10-10T11:24:10.824Z"
    },
    {
      "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": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7796420402684564,
      "accuracy_std_dev": 0.01821392930104136,
      "run_count": 6,
      "total_questions": 298,
      "avg_cost_per_task_usd": 0.15591147986577178,
      "avg_duration_per_task_ms": 329823.3187919463,
      "avg_output_tokens_per_task": 20298.120805369126,
      "last_run_timestamp": "2026-10-10T11:36:39.823Z"
    },
    {
      "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": 0.7782795483870968,
      "accuracy_std_dev": 0.024653522574764702,
      "run_count": 31,
      "total_questions": 1550,
      "avg_cost_per_task_usd": 0.5512626245161292,
      "avg_duration_per_task_ms": 114385.19612903227,
      "avg_output_tokens_per_task": 7228.9838709677415,
      "last_run_timestamp": "2026-10-10T11:19:53.369Z"
    },
    {
      "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": 0.7780952857142857,
      "accuracy_std_dev": 0.025125589310582973,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.15667032,
      "avg_duration_per_task_ms": 579481.5971428572,
      "avg_output_tokens_per_task": 12471.537142857143,
      "last_run_timestamp": "2026-10-10T12:12:40.300Z"
    },
    {
      "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": 0.7780951428571429,
      "accuracy_std_dev": 0.020381833538918,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.08139844285714284,
      "avg_duration_per_task_ms": 136861.11714285714,
      "avg_output_tokens_per_task": 4631.26,
      "last_run_timestamp": "2026-10-10T12:31:22.913Z"
    },
    {
      "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": 0.7780950000000001,
      "accuracy_std_dev": 0.016605340233964306,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.09330751714285715,
      "avg_duration_per_task_ms": 138037.12,
      "avg_output_tokens_per_task": 4580.668571428571,
      "last_run_timestamp": "2026-10-10T11:56:05.563Z"
    },
    {
      "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": 0.778,
      "accuracy_std_dev": 0.021578397978842947,
      "run_count": 30,
      "total_questions": 1500,
      "avg_cost_per_task_usd": 0.48919124333333336,
      "avg_duration_per_task_ms": 124186.09133333333,
      "avg_output_tokens_per_task": 7191.306666666666,
      "last_run_timestamp": "2026-10-10T11:23:03.579Z"
    },
    {
      "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": "Alibaba",
      "accuracy": 0.777037,
      "accuracy_std_dev": 0.017529048595593164,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.12491927333333332,
      "avg_duration_per_task_ms": 86437.00444444445,
      "avg_output_tokens_per_task": 4516.131111111111,
      "last_run_timestamp": "2026-10-10T11:52:54.491Z"
    },
    {
      "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": 0.7770368888888889,
      "accuracy_std_dev": 0.04333181447195696,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.11806641555555555,
      "avg_duration_per_task_ms": 199701.67777777778,
      "avg_output_tokens_per_task": 7078.295555555555,
      "last_run_timestamp": "2026-10-10T11:52:41.395Z"
    },
    {
      "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": 0.7769698181818182,
      "accuracy_std_dev": 0.020570475514096246,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.4460740036363637,
      "avg_duration_per_task_ms": 130580.85818181818,
      "avg_output_tokens_per_task": 7285.6563636363635,
      "last_run_timestamp": "2026-10-10T11:20:01.076Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.8-opus-20260528",
      "model_slug": "anthropic/claude-opus-4.8",
      "model_name": "Anthropic: Claude Opus 4.8",
      "provider_name": "Anthropic",
      "accuracy": 0.7769230769230769,
      "accuracy_std_dev": 0.0234780981226253,
      "run_count": 13,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.4968872015384615,
      "avg_duration_per_task_ms": 128094.2323076923,
      "avg_output_tokens_per_task": 8089.72,
      "last_run_timestamp": "2026-10-10T11:29:58.841Z"
    },
    {
      "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": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "auto-routing",
      "accuracy": 0.7761905714285714,
      "accuracy_std_dev": 0.025161623589773956,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.07583003714285713,
      "avg_duration_per_task_ms": 286315.0457142857,
      "avg_output_tokens_per_task": 13935.274285714286,
      "last_run_timestamp": "2026-10-10T11:31:34.623Z"
    },
    {
      "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": 0.776,
      "accuracy_std_dev": 0.02004423996064706,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.114664068,
      "avg_duration_per_task_ms": 365492.12,
      "avg_output_tokens_per_task": 20755.204,
      "last_run_timestamp": "2026-10-10T12:02:52.502Z"
    },
    {
      "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": 0.7759258888888889,
      "accuracy_std_dev": 0.02243107168116209,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.18105771111111107,
      "avg_duration_per_task_ms": 381492.62,
      "avg_output_tokens_per_task": 28609.035555555554,
      "last_run_timestamp": "2026-10-10T18:21:48.109Z"
    },
    {
      "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": 0.7755556666666666,
      "accuracy_std_dev": 0.027932599656236,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.17062818666666668,
      "avg_duration_per_task_ms": 84211.09333333334,
      "avg_output_tokens_per_task": 6269.78,
      "last_run_timestamp": "2026-10-10T11:20:55.496Z"
    },
    {
      "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": 0.7752381428571429,
      "accuracy_std_dev": 0.01841823507465649,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.06935552285714286,
      "avg_duration_per_task_ms": 37222.497142857144,
      "avg_output_tokens_per_task": 4756.58,
      "last_run_timestamp": "2026-10-10T11:29:01.092Z"
    },
    {
      "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": 0.7750876842105263,
      "accuracy_std_dev": 0.028681883484458735,
      "run_count": 19,
      "total_questions": 950,
      "avg_cost_per_task_usd": 0.5084564494736843,
      "avg_duration_per_task_ms": 128153.6252631579,
      "avg_output_tokens_per_task": 7952.632631578947,
      "last_run_timestamp": "2026-10-10T11:20:59.788Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Io Net",
      "accuracy": 0.7749997500000001,
      "accuracy_std_dev": 0.014433785597946898,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.056739235,
      "avg_duration_per_task_ms": 137409.03,
      "avg_output_tokens_per_task": 4567.525,
      "last_run_timestamp": "2026-10-10T11:51:41.085Z"
    },
    {
      "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": 0.7748147777777777,
      "accuracy_std_dev": 0.0399725818246507,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.11767610666666668,
      "avg_duration_per_task_ms": 282128.4533333333,
      "avg_output_tokens_per_task": 12261.444444444445,
      "last_run_timestamp": "2026-10-10T22:52:53.325Z"
    },
    {
      "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": 0.029275825597155438,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.02169746888888889,
      "avg_duration_per_task_ms": 300435.5488888889,
      "avg_output_tokens_per_task": 11767.76,
      "last_run_timestamp": "2026-10-10T19:46:14.463Z"
    },
    {
      "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": 0.7747070301507537,
      "accuracy_std_dev": 0.01996857802899347,
      "run_count": 4,
      "total_questions": 199,
      "avg_cost_per_task_usd": 0.06471272864321607,
      "avg_duration_per_task_ms": 1090207.1457286433,
      "avg_output_tokens_per_task": 32780.115577889446,
      "last_run_timestamp": "2026-10-10T11:21:41.112Z"
    },
    {
      "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": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7746668,
      "accuracy_std_dev": 0.0165463013679795,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.153314064,
      "avg_duration_per_task_ms": 285965.152,
      "avg_output_tokens_per_task": 19906.316,
      "last_run_timestamp": "2026-10-10T11:39:25.912Z"
    },
    {
      "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": 0.7744443333333334,
      "accuracy_std_dev": 0.019404651346749013,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.08295856333333333,
      "avg_duration_per_task_ms": 90230.37333333334,
      "avg_output_tokens_per_task": 4538.593333333333,
      "last_run_timestamp": "2026-10-10T11:21:34.698Z"
    },
    {
      "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": 0.7743589230769231,
      "accuracy_std_dev": 0.017314924986894253,
      "run_count": 13,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.4353327676923078,
      "avg_duration_per_task_ms": 95008.13230769231,
      "avg_output_tokens_per_task": 5676.784615384616,
      "last_run_timestamp": "2026-10-10T11:19:17.870Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "DeepInfra",
      "accuracy": 0.7742857142857142,
      "accuracy_std_dev": 0.026945723560596418,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.21623159142857146,
      "avg_duration_per_task_ms": 562306.5571428571,
      "avg_output_tokens_per_task": 12602.945714285714,
      "last_run_timestamp": "2026-10-10T11:06:43.832Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "auto-routing",
      "accuracy": 0.774166625,
      "accuracy_std_dev": 0.017460369812360068,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.10507661,
      "avg_duration_per_task_ms": 319305.445,
      "avg_output_tokens_per_task": 20731.5575,
      "last_run_timestamp": "2026-10-10T11:20:55.408Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "StreamLake",
      "accuracy": 0.774,
      "accuracy_std_dev": 0.026574933888911183,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.04397801800000001,
      "avg_duration_per_task_ms": 214737.652,
      "avg_output_tokens_per_task": 9570.176,
      "last_run_timestamp": "2026-10-10T19:06:48.528Z"
    },
    {
      "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": 0.77399995,
      "accuracy_std_dev": 0.031048382675229646,
      "run_count": 20,
      "total_questions": 1000,
      "avg_cost_per_task_usd": 0.444216434,
      "avg_duration_per_task_ms": 79552.85,
      "avg_output_tokens_per_task": 5673.027,
      "last_run_timestamp": "2026-10-10T11:48:23.299Z"
    },
    {
      "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": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7733334,
      "accuracy_std_dev": 0.01837863493951607,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.09612659999999999,
      "avg_duration_per_task_ms": 122579.768,
      "avg_output_tokens_per_task": 5303.732,
      "last_run_timestamp": "2026-10-10T11:43:14.702Z"
    },
    {
      "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": 0.7733333,
      "accuracy_std_dev": 0.027648644733693543,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.4732467780000001,
      "avg_duration_per_task_ms": 113874.772,
      "avg_output_tokens_per_task": 7213.052,
      "last_run_timestamp": "2026-10-10T11:27:45.281Z"
    },
    {
      "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": 0.77333325,
      "accuracy_std_dev": 0.004713927363409425,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.007096445,
      "avg_duration_per_task_ms": 164831.185,
      "avg_output_tokens_per_task": 17141.055,
      "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": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "WandB",
      "accuracy": 0.773333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.11157742000000001,
      "avg_duration_per_task_ms": 52237.66,
      "avg_output_tokens_per_task": 12874.88,
      "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": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "GMICloud",
      "accuracy": 0.773333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.26629643999999997,
      "avg_duration_per_task_ms": 438741.58,
      "avg_output_tokens_per_task": 27876.64,
      "last_run_timestamp": "2026-08-29T16:58:30.455Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "DeepInfra",
      "accuracy": 0.773333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.042562220000000005,
      "avg_duration_per_task_ms": 321425.02,
      "avg_output_tokens_per_task": 8985.7,
      "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": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "StreamLake",
      "accuracy": 0.773333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.057549260000000005,
      "avg_duration_per_task_ms": 140710.34,
      "avg_output_tokens_per_task": 4743.94,
      "last_run_timestamp": "2026-09-11T04:09:34.110Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "DeepInfra",
      "accuracy": 0.7732656294416244,
      "accuracy_std_dev": 0.03435466996639615,
      "run_count": 4,
      "total_questions": 197,
      "avg_cost_per_task_usd": 0.010553837563451776,
      "avg_duration_per_task_ms": 1118057.5126903553,
      "avg_output_tokens_per_task": 17178.984771573603,
      "last_run_timestamp": "2026-10-10T11:52:41.434Z"
    },
    {
      "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": 0.772857,
      "accuracy_std_dev": 0.03077969484207776,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.11466776285714285,
      "avg_duration_per_task_ms": 170975.38857142857,
      "avg_output_tokens_per_task": 9449.94,
      "last_run_timestamp": "2026-10-10T19:58:43.743Z"
    },
    {
      "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": 0.7725,
      "accuracy_std_dev": 0.019561299662854723,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.17478934,
      "avg_duration_per_task_ms": 739323.4375,
      "avg_output_tokens_per_task": 28891.6325,
      "last_run_timestamp": "2026-10-10T18:38:21.632Z"
    },
    {
      "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": 0.7722223333333333,
      "accuracy_std_dev": 0.010482248369929469,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.11473854333333333,
      "avg_duration_per_task_ms": 271731.52,
      "avg_output_tokens_per_task": 23286.896666666667,
      "last_run_timestamp": "2026-10-10T11:07:22.053Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta/muse-glimmer-30b-20260810",
      "model_slug": "meta/muse-glimmer-30b",
      "model_name": "Meta: Muse Glimmer 30B",
      "provider_name": "DeepInfra",
      "accuracy": 0.7722222222222223,
      "accuracy_std_dev": 0.03095202181433487,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.03593436666666667,
      "avg_duration_per_task_ms": 446735.1666666667,
      "avg_output_tokens_per_task": 13815.786666666667,
      "last_run_timestamp": "2026-10-08T22:42:47.098Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "BaseTen",
      "accuracy": 0.772109,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 49,
      "avg_cost_per_task_usd": 0.06601828571428571,
      "avg_duration_per_task_ms": 84701.85714285714,
      "avg_output_tokens_per_task": 5801.673469387755,
      "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": "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": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Friendli",
      "accuracy": 0.772,
      "accuracy_std_dev": 0.010666750001757801,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.029716368000000003,
      "avg_duration_per_task_ms": 111311.872,
      "avg_output_tokens_per_task": 12081.752,
      "last_run_timestamp": "2026-10-10T18:37:53.300Z"
    },
    {
      "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": 0.7719999000000001,
      "accuracy_std_dev": 0.019955421431029706,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.43238272600000005,
      "avg_duration_per_task_ms": 99679.176,
      "avg_output_tokens_per_task": 5611.424,
      "last_run_timestamp": "2026-10-10T11:06:20.481Z"
    },
    {
      "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": 0.771905,
      "accuracy_std_dev": 0.035315195576004266,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.1764812885714286,
      "avg_duration_per_task_ms": 513263.16857142857,
      "avg_output_tokens_per_task": 25304.974285714285,
      "last_run_timestamp": "2026-10-10T18:50:47.547Z"
    },
    {
      "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": 0.7717949230769231,
      "accuracy_std_dev": 0.0290184402604263,
      "run_count": 13,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.5512982523076922,
      "avg_duration_per_task_ms": 183923.44307692308,
      "avg_output_tokens_per_task": 10555.14,
      "last_run_timestamp": "2026-10-10T11:29:26.517Z"
    },
    {
      "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": 0.7711111111111111,
      "accuracy_std_dev": 0.02828443711240216,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.18611659777777775,
      "avg_duration_per_task_ms": 180574.86666666667,
      "avg_output_tokens_per_task": 9565.442222222222,
      "last_run_timestamp": "2026-10-10T18:32:40.177Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.1-20251113",
      "model_slug": "openai/gpt-5.1",
      "model_name": "OpenAI: GPT-5.1",
      "provider_name": "auto-routing",
      "accuracy": 0.771111,
      "accuracy_std_dev": 0.04085521560339636,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.4149127733333333,
      "avg_duration_per_task_ms": 472514.2266666667,
      "avg_output_tokens_per_task": 36470.113333333335,
      "last_run_timestamp": "2026-10-01T16:00:45.071Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "StreamLake",
      "accuracy": 0.77083325,
      "accuracy_std_dev": 0.021327567581360528,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.08374891499999998,
      "avg_duration_per_task_ms": 335021.265,
      "avg_output_tokens_per_task": 20027.295,
      "last_run_timestamp": "2026-10-10T12:20:33.626Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "SambaNova",
      "accuracy": 0.770833,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 48,
      "avg_cost_per_task_usd": 0.07522506250000001,
      "avg_duration_per_task_ms": 171662.70833333334,
      "avg_output_tokens_per_task": 11489.854166666666,
      "last_run_timestamp": "2026-10-10T18:58:05.434Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta/muse-glimmer-30b-20260810",
      "model_slug": "meta/muse-glimmer-30b",
      "model_name": "Meta: Muse Glimmer 30B",
      "provider_name": "Parasail",
      "accuracy": 0.7706935637583894,
      "accuracy_std_dev": 0.013644816256244217,
      "run_count": 3,
      "total_questions": 149,
      "avg_cost_per_task_usd": 0.09587820805369127,
      "avg_duration_per_task_ms": 288021.6644295302,
      "avg_output_tokens_per_task": 13114.979865771813,
      "last_run_timestamp": "2026-08-29T08:48:12.957Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "NextBit",
      "accuracy": 0.7706666,
      "accuracy_std_dev": 0.0204828947622156,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.10916058,
      "avg_duration_per_task_ms": 258128.42,
      "avg_output_tokens_per_task": 11359.708,
      "last_run_timestamp": "2026-10-03T16:07:05.245Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Novita",
      "accuracy": 0.7704761428571428,
      "accuracy_std_dev": 0.021338487899491252,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.008933357142857144,
      "avg_duration_per_task_ms": 211052.8457142857,
      "avg_output_tokens_per_task": 4263.237142857143,
      "last_run_timestamp": "2026-10-10T12:13:04.257Z"
    },
    {
      "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": 0.770476,
      "accuracy_std_dev": 0.02332836876300735,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.07879788857142857,
      "avg_duration_per_task_ms": 228548.80857142858,
      "avg_output_tokens_per_task": 9318.671428571428,
      "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": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "AkashML",
      "accuracy": 0.7704113333333333,
      "accuracy_std_dev": 0.022998343056737182,
      "run_count": 11,
      "total_questions": 543,
      "avg_cost_per_task_usd": 0.05714349907918969,
      "avg_duration_per_task_ms": 253713.67034990792,
      "avg_output_tokens_per_task": 11037.460405156538,
      "last_run_timestamp": "2026-10-10T11:10:08.248Z"
    },
    {
      "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": 0.7700001249999999,
      "accuracy_std_dev": 0.03395262221993134,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.071411165,
      "avg_duration_per_task_ms": 255968.725,
      "avg_output_tokens_per_task": 11897.3,
      "last_run_timestamp": "2026-10-10T11:38:42.065Z"
    },
    {
      "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": 0.7700001,
      "accuracy_std_dev": 0.020923293772491944,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.5362019739999999,
      "avg_duration_per_task_ms": 175302.906,
      "avg_output_tokens_per_task": 10636.738,
      "last_run_timestamp": "2026-10-10T11:19:29.150Z"
    },
    {
      "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": 0.77,
      "accuracy_std_dev": 0.016666999999999987,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.08223393,
      "avg_duration_per_task_ms": 429468.14,
      "avg_output_tokens_per_task": 14260.94,
      "last_run_timestamp": "2026-10-10T11:20:22.675Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Baidu",
      "accuracy": 0.7696296666666667,
      "accuracy_std_dev": 0.017245185808090188,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.04810054666666667,
      "avg_duration_per_task_ms": 199896.38444444444,
      "avg_output_tokens_per_task": 13265.613333333333,
      "last_run_timestamp": "2026-10-10T19:37:47.264Z"
    },
    {
      "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": 0.7696296666666667,
      "accuracy_std_dev": 0.030161915139607597,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.011042515555555554,
      "avg_duration_per_task_ms": 262078.49333333335,
      "avg_output_tokens_per_task": 6314.64,
      "last_run_timestamp": "2026-10-10T11:50:30.268Z"
    },
    {
      "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": "Modal",
      "accuracy": 0.7695238571428571,
      "accuracy_std_dev": 0.03282612614727386,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.08224388857142857,
      "avg_duration_per_task_ms": 49807.68,
      "avg_output_tokens_per_task": 4863.0028571428575,
      "last_run_timestamp": "2026-10-10T11:40:28.908Z"
    },
    {
      "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": 0.7694871538461538,
      "accuracy_std_dev": 0.0298099088245868,
      "run_count": 13,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.052062487692307695,
      "avg_duration_per_task_ms": 95666.30307692308,
      "avg_output_tokens_per_task": 4338.936923076923,
      "last_run_timestamp": "2026-10-10T11:52:30.774Z"
    },
    {
      "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": 0.7693334,
      "accuracy_std_dev": 0.016110772105644107,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.15629164,
      "avg_duration_per_task_ms": 313958.944,
      "avg_output_tokens_per_task": 15113.092,
      "last_run_timestamp": "2026-10-10T21:14:14.916Z"
    },
    {
      "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": 0.7693334,
      "accuracy_std_dev": 0.027194796889110983,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.047570167999999996,
      "avg_duration_per_task_ms": 114696.804,
      "avg_output_tokens_per_task": 4956.572,
      "last_run_timestamp": "2026-10-08T23:21:59.850Z"
    },
    {
      "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": 0.7693333333333333,
      "accuracy_std_dev": 0.019444111745090222,
      "run_count": 15,
      "total_questions": 750,
      "avg_cost_per_task_usd": 0.06786480933333333,
      "avg_duration_per_task_ms": 55615.890666666666,
      "avg_output_tokens_per_task": 4658.706666666667,
      "last_run_timestamp": "2026-10-10T11:39:51.191Z"
    },
    {
      "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": 0.7689700975609757,
      "accuracy_std_dev": 0.020642326587863087,
      "run_count": 5,
      "total_questions": 246,
      "avg_cost_per_task_usd": 0.06932281300813009,
      "avg_duration_per_task_ms": 206430.17479674798,
      "avg_output_tokens_per_task": 10777.69105691057,
      "last_run_timestamp": "2026-10-10T19:04:05.984Z"
    },
    {
      "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": 0.7688889999999999,
      "accuracy_std_dev": 0.029979235614004583,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.028244313333333337,
      "avg_duration_per_task_ms": 171006.77333333335,
      "avg_output_tokens_per_task": 15371.96,
      "last_run_timestamp": "2026-10-10T11:30:53.804Z"
    },
    {
      "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": 0.7688889999999999,
      "accuracy_std_dev": 0.026574014970769236,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.06194356999999999,
      "avg_duration_per_task_ms": 131938.69666666666,
      "avg_output_tokens_per_task": 6466.09,
      "last_run_timestamp": "2026-10-10T11:43:14.705Z"
    },
    {
      "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": 0.7688888888888888,
      "accuracy_std_dev": 0.02703454580283221,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.09845500888888889,
      "avg_duration_per_task_ms": 231449.82222222222,
      "avg_output_tokens_per_task": 11662.893333333333,
      "last_run_timestamp": "2026-10-10T18:58:28.525Z"
    },
    {
      "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": 0.7688888333333334,
      "accuracy_std_dev": 0.017916093876890582,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.060097880000000006,
      "avg_duration_per_task_ms": 48342.426666666666,
      "avg_output_tokens_per_task": 4654.916666666667,
      "last_run_timestamp": "2026-10-10T11:20:35.509Z"
    },
    {
      "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": "Relace",
      "accuracy": 0.7688886666666668,
      "accuracy_std_dev": 0.011331263428624758,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.025211093333333334,
      "avg_duration_per_task_ms": 135029.63333333333,
      "avg_output_tokens_per_task": 13854.966666666667,
      "last_run_timestamp": "2026-10-10T18:35:38.956Z"
    },
    {
      "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": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "DeepInfra",
      "accuracy": 0.768678143678161,
      "accuracy_std_dev": 0.0271195517707225,
      "run_count": 14,
      "total_questions": 696,
      "avg_cost_per_task_usd": 0.03471193247126437,
      "avg_duration_per_task_ms": 522555.5474137931,
      "avg_output_tokens_per_task": 12696.757183908046,
      "last_run_timestamp": "2026-10-10T19:16:07.502Z"
    },
    {
      "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": 0.7685808243243244,
      "accuracy_std_dev": 0.028647198085308245,
      "run_count": 6,
      "total_questions": 296,
      "avg_cost_per_task_usd": 0.07983985472972972,
      "avg_duration_per_task_ms": 256413,
      "avg_output_tokens_per_task": 10008.462837837838,
      "last_run_timestamp": "2026-10-10T19:17:08.161Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "CoreWeave",
      "accuracy": 0.7684563154362417,
      "accuracy_std_dev": 0.006017520216464235,
      "run_count": 3,
      "total_questions": 149,
      "avg_cost_per_task_usd": 0.1581116510067114,
      "avg_duration_per_task_ms": 537705.5637583892,
      "avg_output_tokens_per_task": 10979.71812080537,
      "last_run_timestamp": "2026-08-08T08:27:49.545Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.7683335,
      "accuracy_std_dev": 0.014433698997485036,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.03325037,
      "avg_duration_per_task_ms": 273019.26,
      "avg_output_tokens_per_task": 39890.13,
      "last_run_timestamp": "2026-10-10T11:29:19.064Z"
    },
    {
      "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": 0.7683109103139014,
      "accuracy_std_dev": 0.04045368552087958,
      "run_count": 9,
      "total_questions": 446,
      "avg_cost_per_task_usd": 0.027139372197309417,
      "avg_duration_per_task_ms": 484571.0381165919,
      "avg_output_tokens_per_task": 12770.852017937219,
      "last_run_timestamp": "2026-10-10T19:09:44.232Z"
    },
    {
      "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": 0.7680952857142856,
      "accuracy_std_dev": 0.02110869827626438,
      "run_count": 14,
      "total_questions": 700,
      "avg_cost_per_task_usd": 0.017966814285714287,
      "avg_duration_per_task_ms": 193611,
      "avg_output_tokens_per_task": 19085.56142857143,
      "last_run_timestamp": "2026-10-10T11:05:58.545Z"
    },
    {
      "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": 0.7680458965517241,
      "accuracy_std_dev": 0.020370441569782412,
      "run_count": 29,
      "total_questions": 1450,
      "avg_cost_per_task_usd": 0.5334861268965518,
      "avg_duration_per_task_ms": 121001.88,
      "avg_output_tokens_per_task": 8118.9544827586205,
      "last_run_timestamp": "2026-10-10T11:33:22.028Z"
    },
    {
      "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": 0.768,
      "accuracy_std_dev": 0.011469922022402766,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.09600952000000001,
      "avg_duration_per_task_ms": 63394.68,
      "avg_output_tokens_per_task": 4831.208,
      "last_run_timestamp": "2026-10-10T12:14:36.151Z"
    },
    {
      "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": 0.7678787272727273,
      "accuracy_std_dev": 0.02555520860657607,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.08920051818181818,
      "avg_duration_per_task_ms": 190464.52,
      "avg_output_tokens_per_task": 11544.347272727273,
      "last_run_timestamp": "2026-10-10T20:24:33.030Z"
    },
    {
      "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": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "DekaLLM",
      "accuracy": 0.7677778333333333,
      "accuracy_std_dev": 0.01822341121027811,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.058372879999999995,
      "avg_duration_per_task_ms": 136563.96666666667,
      "avg_output_tokens_per_task": 14796.806666666667,
      "last_run_timestamp": "2026-10-10T11:05:27.120Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Together",
      "accuracy": 0.7675,
      "accuracy_std_dev": 0.0239067401008586,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.026842017500000006,
      "avg_duration_per_task_ms": 165041.6625,
      "avg_output_tokens_per_task": 8909.3125,
      "last_run_timestamp": "2026-10-10T18:58:24.558Z"
    },
    {
      "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": 0.7673335,
      "accuracy_std_dev": 0.020965394769715172,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.49187341399999995,
      "avg_duration_per_task_ms": 131219.752,
      "avg_output_tokens_per_task": 8161.99,
      "last_run_timestamp": "2026-10-10T11:05:25.784Z"
    },
    {
      "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": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "ModelRun",
      "accuracy": 0.7666669999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.0668136,
      "avg_duration_per_task_ms": 100184.26,
      "avg_output_tokens_per_task": 11544.38,
      "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": "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-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Venice",
      "accuracy": 0.7666667777777777,
      "accuracy_std_dev": 0.028458217152628754,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.13082103333333334,
      "avg_duration_per_task_ms": 266780,
      "avg_output_tokens_per_task": 11720.468888888889,
      "last_run_timestamp": "2026-10-10T18:44:45.575Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "NextBit",
      "accuracy": 0.76666675,
      "accuracy_std_dev": 0.024267078752653787,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.036940722499999995,
      "avg_duration_per_task_ms": 209470.17,
      "avg_output_tokens_per_task": 17358.2575,
      "last_run_timestamp": "2026-10-03T08:39:48.671Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Novita",
      "accuracy": 0.7666667142857143,
      "accuracy_std_dev": 0.02108204682875446,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.058299279999999995,
      "avg_duration_per_task_ms": 114981.15428571429,
      "avg_output_tokens_per_task": 4748.437142857143,
      "last_run_timestamp": "2026-10-10T11:50:44.207Z"
    },
    {
      "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": 0.7666666666666667,
      "accuracy_std_dev": 0.005443174461857933,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.18197839999999998,
      "avg_duration_per_task_ms": 75193.68666666666,
      "avg_output_tokens_per_task": 6381.386666666666,
      "last_run_timestamp": "2026-10-10T11:01:03.350Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.1-pro-preview-20260219",
      "model_slug": "google/gemini-3.1-pro-preview",
      "model_name": "Google: Gemini 3.1 Pro Preview",
      "provider_name": "Google",
      "accuracy": 0.7666666666666667,
      "accuracy_std_dev": 0.014401440259764175,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.3460653,
      "avg_duration_per_task_ms": 170899.55333333334,
      "avg_output_tokens_per_task": 12611.153333333334,
      "last_run_timestamp": "2026-10-01T19:04:40.361Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.7-opus-20260416",
      "model_slug": "anthropic/claude-opus-4.7",
      "model_name": "Anthropic: Claude Opus 4.7",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.766666625,
      "accuracy_std_dev": 0.03480106160714606,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.434086625,
      "avg_duration_per_task_ms": 94242.2325,
      "avg_output_tokens_per_task": 5557.605,
      "last_run_timestamp": "2026-10-10T11:08:25.975Z"
    },
    {
      "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": 0.766666625,
      "accuracy_std_dev": 0.036055535868079636,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.07008643250000002,
      "avg_duration_per_task_ms": 305737.3925,
      "avg_output_tokens_per_task": 12550.4925,
      "last_run_timestamp": "2026-10-10T11:39:51.190Z"
    },
    {
      "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": 0.7666666,
      "accuracy_std_dev": 0.03319954484386795,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.032209988,
      "avg_duration_per_task_ms": 207977.044,
      "avg_output_tokens_per_task": 14953.416,
      "last_run_timestamp": "2026-10-10T11:06:17.699Z"
    },
    {
      "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": 0.7666666,
      "accuracy_std_dev": 0.022705633174170663,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.013304756000000001,
      "avg_duration_per_task_ms": 229500.18,
      "avg_output_tokens_per_task": 4888.664,
      "last_run_timestamp": "2026-10-10T12:10:15.915Z"
    },
    {
      "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": 0.7666665714285714,
      "accuracy_std_dev": 0.01851637627967774,
      "run_count": 14,
      "total_questions": 700,
      "avg_cost_per_task_usd": 0.019599598571428567,
      "avg_duration_per_task_ms": 240323.45,
      "avg_output_tokens_per_task": 16987.242857142857,
      "last_run_timestamp": "2026-10-10T12:00:58.928Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "Anthropic",
      "accuracy": 0.7666665,
      "accuracy_std_dev": 0.013333499999999998,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.012833540000000001,
      "avg_duration_per_task_ms": 88757.68,
      "avg_output_tokens_per_task": 11232.71,
      "last_run_timestamp": "2026-10-10T11:24:21.404Z"
    },
    {
      "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": "Xiaomi",
      "accuracy": 0.7666665,
      "accuracy_std_dev": 0.013333499999999998,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01488279,
      "avg_duration_per_task_ms": 714098.21,
      "avg_output_tokens_per_task": 29412.46,
      "last_run_timestamp": "2026-10-10T13:27:56.959Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "DekaLLM",
      "accuracy": 0.7666665,
      "accuracy_std_dev": 0.026666499999999982,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01352679,
      "avg_duration_per_task_ms": 90848.6,
      "avg_output_tokens_per_task": 4507.38,
      "last_run_timestamp": "2026-10-10T11:18:50.119Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7664834065934065,
      "accuracy_std_dev": 0.03806138445886618,
      "run_count": 11,
      "total_questions": 546,
      "avg_cost_per_task_usd": 0.10488499816849818,
      "avg_duration_per_task_ms": 804500.0018315018,
      "avg_output_tokens_per_task": 10329.6336996337,
      "last_run_timestamp": "2026-10-10T19:42:22.453Z"
    },
    {
      "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": 0.76633335,
      "accuracy_std_dev": 0.03073014101379133,
      "run_count": 20,
      "total_questions": 1000,
      "avg_cost_per_task_usd": 0.567469223,
      "avg_duration_per_task_ms": 181854.152,
      "avg_output_tokens_per_task": 10594.497,
      "last_run_timestamp": "2026-10-10T11:40:33.909Z"
    },
    {
      "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": 0.7662836379310344,
      "accuracy_std_dev": 0.02598916979066786,
      "run_count": 7,
      "total_questions": 348,
      "avg_cost_per_task_usd": 0.10895231896551724,
      "avg_duration_per_task_ms": 46529.12356321839,
      "avg_output_tokens_per_task": 4239.747126436781,
      "last_run_timestamp": "2026-10-10T11:35:23.041Z"
    },
    {
      "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-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Phala",
      "accuracy": 0.7655556666666666,
      "accuracy_std_dev": 0.020518134106741345,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.05405915333333334,
      "avg_duration_per_task_ms": 120906.59666666666,
      "avg_output_tokens_per_task": 4450.843333333333,
      "last_run_timestamp": "2026-10-10T11:50:02.644Z"
    },
    {
      "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": 0.7655172758620691,
      "accuracy_std_dev": 0.02543532653057561,
      "run_count": 29,
      "total_questions": 1450,
      "avg_cost_per_task_usd": 0.5846374510344827,
      "avg_duration_per_task_ms": 118911.25586206897,
      "avg_output_tokens_per_task": 8140.452413793103,
      "last_run_timestamp": "2026-10-10T11:25:12.451Z"
    },
    {
      "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": 0.7654544545454546,
      "accuracy_std_dev": 0.028258289861283148,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.16681981636363638,
      "avg_duration_per_task_ms": 582078.4036363637,
      "avg_output_tokens_per_task": 36559.66,
      "last_run_timestamp": "2026-10-10T11:50:56.932Z"
    },
    {
      "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": 0.7653332,
      "accuracy_std_dev": 0.02646590427247858,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.019831184,
      "avg_duration_per_task_ms": 375594.772,
      "avg_output_tokens_per_task": 14077.236,
      "last_run_timestamp": "2026-10-10T11:21:32.952Z"
    },
    {
      "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": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "AkashML",
      "accuracy": 0.7649122421052632,
      "accuracy_std_dev": 0.0039184999999999914,
      "run_count": 2,
      "total_questions": 95,
      "avg_cost_per_task_usd": 0.25060463157894736,
      "avg_duration_per_task_ms": 358896.81052631576,
      "avg_output_tokens_per_task": 6743.5684210526315,
      "last_run_timestamp": "2026-10-10T11:20:06.376Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Moonshot AI",
      "accuracy": 0.7648808794642857,
      "accuracy_std_dev": 0.021171058001169644,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.10855453124999999,
      "avg_duration_per_task_ms": 356320.9129464286,
      "avg_output_tokens_per_task": 11844.839285714286,
      "last_run_timestamp": "2026-10-10T20:34:18.024Z"
    },
    {
      "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": 0.7644445,
      "accuracy_std_dev": 0.016178189863207813,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.16990947999999997,
      "avg_duration_per_task_ms": 205513.85333333333,
      "avg_output_tokens_per_task": 11240.563333333334,
      "last_run_timestamp": "2026-10-10T18:54:48.885Z"
    },
    {
      "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": 0.7644445,
      "accuracy_std_dev": 0.019876128745725773,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.07212067999999999,
      "avg_duration_per_task_ms": 124793.37,
      "avg_output_tokens_per_task": 4621.306666666666,
      "last_run_timestamp": "2026-10-10T12:09:46.550Z"
    },
    {
      "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": 0.7644445,
      "accuracy_std_dev": 0.023622535799175632,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.050448443333333336,
      "avg_duration_per_task_ms": 138327.16,
      "avg_output_tokens_per_task": 4647.42,
      "last_run_timestamp": "2026-10-10T11:23:27.114Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Decart",
      "accuracy": 0.7644443333333334,
      "accuracy_std_dev": 0.012570472951413688,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.02870791333333333,
      "avg_duration_per_task_ms": 1126711.7933333332,
      "avg_output_tokens_per_task": 60958.193333333336,
      "last_run_timestamp": "2026-10-10T12:23:30.561Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "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": 0.764,
      "accuracy_std_dev": 0.02425793799151117,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.416801944,
      "avg_duration_per_task_ms": 97505.054,
      "avg_output_tokens_per_task": 5629.594,
      "last_run_timestamp": "2026-10-10T11:22:55.917Z"
    },
    {
      "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": 0.764,
      "accuracy_std_dev": 0.02515740668669966,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.23303213200000003,
      "avg_duration_per_task_ms": 355571.66,
      "avg_output_tokens_per_task": 17605.012,
      "last_run_timestamp": "2026-10-10T11:21:47.208Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "GMICloud",
      "accuracy": 0.7638094285714286,
      "accuracy_std_dev": 0.03917523513233161,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.009541142857142856,
      "avg_duration_per_task_ms": 210820.91142857142,
      "avg_output_tokens_per_task": 15468.797142857144,
      "last_run_timestamp": "2026-10-10T18:57:39.891Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Relace",
      "accuracy": 0.7638094285714286,
      "accuracy_std_dev": 0.03525091916068954,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.006661174285714286,
      "avg_duration_per_task_ms": 122379.59714285714,
      "avg_output_tokens_per_task": 4276.557142857143,
      "last_run_timestamp": "2026-10-10T12:05:49.353Z"
    },
    {
      "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": 0.7637778000000002,
      "accuracy_std_dev": 0.022089610541911643,
      "run_count": 30,
      "total_questions": 1500,
      "avg_cost_per_task_usd": 0.6385702406666667,
      "avg_duration_per_task_ms": 72802.16,
      "avg_output_tokens_per_task": 5665.454,
      "last_run_timestamp": "2026-10-10T11:34:56.275Z"
    },
    {
      "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": 0.76375,
      "accuracy_std_dev": 0.012849189342133631,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.1187352025,
      "avg_duration_per_task_ms": 490603.2525,
      "avg_output_tokens_per_task": 21506.57,
      "last_run_timestamp": "2026-10-10T12:38:14.941Z"
    },
    {
      "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": 0.7637037777777778,
      "accuracy_std_dev": 0.023749974515905763,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.04213422666666667,
      "avg_duration_per_task_ms": 238270.13333333333,
      "avg_output_tokens_per_task": 12104.153333333334,
      "last_run_timestamp": "2026-10-10T19:15:26.797Z"
    },
    {
      "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": 0.7634529596412556,
      "accuracy_std_dev": 0.02656307081059885,
      "run_count": 9,
      "total_questions": 446,
      "avg_cost_per_task_usd": 0.2075218811659193,
      "avg_duration_per_task_ms": 487178.55829596415,
      "avg_output_tokens_per_task": 20355.004484304933,
      "last_run_timestamp": "2026-10-10T13:15:05.878Z"
    },
    {
      "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": 0.7634227449664429,
      "accuracy_std_dev": 0.018780119728052862,
      "run_count": 6,
      "total_questions": 298,
      "avg_cost_per_task_usd": 0.020238607382550334,
      "avg_duration_per_task_ms": 117111.80872483221,
      "avg_output_tokens_per_task": 14479.771812080537,
      "last_run_timestamp": "2026-10-10T11:09:17.233Z"
    },
    {
      "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": "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": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "SiliconFlow",
      "accuracy": 0.763333375,
      "accuracy_std_dev": 0.016666625002812498,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.014786937500000001,
      "avg_duration_per_task_ms": 384155.3975,
      "avg_output_tokens_per_task": 15743.1525,
      "last_run_timestamp": "2026-10-10T19:26:17.283Z"
    },
    {
      "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": 0.763333375,
      "accuracy_std_dev": 0.019148607435904456,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.09197625250000002,
      "avg_duration_per_task_ms": 241657.31,
      "avg_output_tokens_per_task": 16454.4975,
      "last_run_timestamp": "2026-10-10T19:19:35.243Z"
    },
    {
      "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": 0.7633332142857143,
      "accuracy_std_dev": 0.017772848027975442,
      "run_count": 14,
      "total_questions": 700,
      "avg_cost_per_task_usd": 0.009256418571428572,
      "avg_duration_per_task_ms": 359533.8657142857,
      "avg_output_tokens_per_task": 14645.47,
      "last_run_timestamp": "2026-10-10T11:45:08.503Z"
    },
    {
      "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": "auto-routing",
      "accuracy": 0.762963,
      "accuracy_std_dev": 0.033717927809664976,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.04426393111111111,
      "avg_duration_per_task_ms": 210488.84444444443,
      "avg_output_tokens_per_task": 9459.097777777777,
      "last_run_timestamp": "2026-10-10T19:42:04.863Z"
    },
    {
      "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": 0.7628572857142857,
      "accuracy_std_dev": 0.02869817284733489,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.14805788285714286,
      "avg_duration_per_task_ms": 317351.07714285713,
      "avg_output_tokens_per_task": 20174.72857142857,
      "last_run_timestamp": "2026-10-10T12:09:30.514Z"
    },
    {
      "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": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "CoreWeave",
      "accuracy": 0.7627551734693877,
      "accuracy_std_dev": 0.020230852904660254,
      "run_count": 8,
      "total_questions": 392,
      "avg_cost_per_task_usd": 0.025161864795918365,
      "avg_duration_per_task_ms": 402528.1301020408,
      "avg_output_tokens_per_task": 12698.543367346938,
      "last_run_timestamp": "2026-10-10T18:47:04.172Z"
    },
    {
      "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": "anthropic/claude-4.6-opus-20260205",
      "model_slug": "anthropic/claude-opus-4.6",
      "model_name": "Anthropic: Claude Opus 4.6",
      "provider_name": "Anthropic",
      "accuracy": 0.7625640769230769,
      "accuracy_std_dev": 0.027739709669218682,
      "run_count": 13,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.5190420169230768,
      "avg_duration_per_task_ms": 219683.05076923076,
      "avg_output_tokens_per_task": 10161.08,
      "last_run_timestamp": "2026-10-10T11:39:24.594Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "DeepInfra",
      "accuracy": 0.7622223333333333,
      "accuracy_std_dev": 0.030469494967188978,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.029655940000000006,
      "avg_duration_per_task_ms": 256385.88,
      "avg_output_tokens_per_task": 18856.086666666666,
      "last_run_timestamp": "2026-10-08T22:27:05.549Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Cohere",
      "accuracy": 0.7621439748743719,
      "accuracy_std_dev": 0.01479216708895287,
      "run_count": 4,
      "total_questions": 199,
      "avg_cost_per_task_usd": 0.02848052763819095,
      "avg_duration_per_task_ms": 102090.86432160804,
      "avg_output_tokens_per_task": 17372.989949748742,
      "last_run_timestamp": "2026-10-10T11:32:54.658Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "DeepInfra",
      "accuracy": 0.7620649764309765,
      "accuracy_std_dev": 0.008722071727647192,
      "run_count": 6,
      "total_questions": 297,
      "avg_cost_per_task_usd": 0.007020558922558921,
      "avg_duration_per_task_ms": 240106.04377104377,
      "avg_output_tokens_per_task": 12275.050505050505,
      "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": "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": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "Darkbloom",
      "accuracy": 0.76166675,
      "accuracy_std_dev": 0.021794590307860812,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.06705333,
      "avg_duration_per_task_ms": 1165177.905,
      "avg_output_tokens_per_task": 61424.67,
      "last_run_timestamp": "2026-10-10T19:06:52.156Z"
    },
    {
      "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": 0.761666625,
      "accuracy_std_dev": 0.02533126362292996,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.5473962499999999,
      "avg_duration_per_task_ms": 178661.2125,
      "avg_output_tokens_per_task": 10775.3575,
      "last_run_timestamp": "2026-10-10T11:09:36.722Z"
    },
    {
      "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": 0.7611111666666667,
      "accuracy_std_dev": 0.022580443304008813,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.13447535,
      "avg_duration_per_task_ms": 192048.67666666667,
      "avg_output_tokens_per_task": 8291.013333333334,
      "last_run_timestamp": "2026-10-10T14:25:49.953Z"
    },
    {
      "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": 0.7610694385964911,
      "accuracy_std_dev": 0.023000953436759987,
      "run_count": 8,
      "total_questions": 399,
      "avg_cost_per_task_usd": 0.011744431077694234,
      "avg_duration_per_task_ms": 283947.4812030075,
      "avg_output_tokens_per_task": 13467.954887218046,
      "last_run_timestamp": "2026-10-10T18:43:43.892Z"
    },
    {
      "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": 0.76083325,
      "accuracy_std_dev": 0.048812266503282735,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0205383975,
      "avg_duration_per_task_ms": 237661.9975,
      "avg_output_tokens_per_task": 12406.4725,
      "last_run_timestamp": "2026-10-10T19:04:06.069Z"
    },
    {
      "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": 0.7606667,
      "accuracy_std_dev": 0.02298790890033282,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.049680242,
      "avg_duration_per_task_ms": 131451.166,
      "avg_output_tokens_per_task": 4591.58,
      "last_run_timestamp": "2026-10-10T12:34:34.275Z"
    },
    {
      "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": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "CoreWeave",
      "accuracy": 0.7600001249999999,
      "accuracy_std_dev": 0.03282949216191708,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.087410225,
      "avg_duration_per_task_ms": 68616.515,
      "avg_output_tokens_per_task": 12187.7175,
      "last_run_timestamp": "2026-10-10T19:27:59.354Z"
    },
    {
      "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": 0.76,
      "accuracy_std_dev": 0.03144669210743797,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.01082417,
      "avg_duration_per_task_ms": 1179559.2475,
      "avg_output_tokens_per_task": 12522.405,
      "last_run_timestamp": "2026-10-10T19:51:16.100Z"
    },
    {
      "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": 0.01989890611886173,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.006725167272727272,
      "avg_duration_per_task_ms": 258933.9818181818,
      "avg_output_tokens_per_task": 15286.843636363636,
      "last_run_timestamp": "2026-10-10T11:41:36.161Z"
    },
    {
      "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": "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": "Relace",
      "accuracy": 0.76,
      "accuracy_std_dev": 0.010886348913509402,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.06629763999999999,
      "avg_duration_per_task_ms": 203915.74666666667,
      "avg_output_tokens_per_task": 10663.093333333334,
      "last_run_timestamp": "2026-10-10T18:57:23.421Z"
    },
    {
      "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": "DekaLLM",
      "accuracy": 0.76,
      "accuracy_std_dev": 0.014401594564491825,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.019761673333333334,
      "avg_duration_per_task_ms": 291193.25333333336,
      "avg_output_tokens_per_task": 17811.393333333333,
      "last_run_timestamp": "2026-10-10T11:23:23.352Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "GLM 5.3",
      "provider_name": "digitalocean",
      "accuracy": 0.76,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.01597332,
      "avg_duration_per_task_ms": 42256.58,
      "avg_output_tokens_per_task": 1764.66,
      "last_run_timestamp": "2026-10-01T23:50:45.025Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "GMICloud",
      "accuracy": 0.76,
      "accuracy_std_dev": 0.01699689514293714,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.021459479999999996,
      "avg_duration_per_task_ms": 117845.22,
      "avg_output_tokens_per_task": 8829.08,
      "last_run_timestamp": "2026-10-10T18:45:26.349Z"
    },
    {
      "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": 0.76,
      "accuracy_std_dev": 0.01914880327200632,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.09582124500000001,
      "avg_duration_per_task_ms": 338140.68,
      "avg_output_tokens_per_task": 12863.7625,
      "last_run_timestamp": "2026-10-10T19:58:55.313Z"
    },
    {
      "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": 0,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.30033978,
      "avg_duration_per_task_ms": 256464.59,
      "avg_output_tokens_per_task": 10538.18,
      "last_run_timestamp": "2026-10-10T11:11:00.029Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3-ultra-550b-a55b-20260604",
      "model_slug": "nvidia/nemotron-3-ultra-550b-a55b",
      "model_name": "NVIDIA: Nemotron 3 Ultra",
      "provider_name": "Venice",
      "accuracy": 0.76,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.19103682,
      "avg_duration_per_task_ms": 233230.24,
      "avg_output_tokens_per_task": 15236.66,
      "last_run_timestamp": "2026-08-04T03:12:00.236Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "Novita",
      "accuracy": 0.76,
      "accuracy_std_dev": 0.01333300000000004,
      "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-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "Baidu",
      "accuracy": 0.76,
      "accuracy_std_dev": 0.015071811643668545,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.05839733999999999,
      "avg_duration_per_task_ms": 223770.25333333333,
      "avg_output_tokens_per_task": 10458.826666666666,
      "last_run_timestamp": "2026-10-10T19:41:57.202Z"
    },
    {
      "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": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "DeepInfra",
      "accuracy": 0.7598205185185184,
      "accuracy_std_dev": 0.017395146832761534,
      "run_count": 6,
      "total_questions": 297,
      "avg_cost_per_task_usd": 0.1637978249158249,
      "avg_duration_per_task_ms": 852122.5521885522,
      "avg_output_tokens_per_task": 31760.62962962963,
      "last_run_timestamp": "2026-10-10T18:46:35.608Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Morph",
      "accuracy": 0.7597053453815261,
      "accuracy_std_dev": 0.039111697810373804,
      "run_count": 10,
      "total_questions": 498,
      "avg_cost_per_task_usd": 0.11686768072289157,
      "avg_duration_per_task_ms": 385580.7168674699,
      "avg_output_tokens_per_task": 18897.180722891568,
      "last_run_timestamp": "2026-10-10T12:15:32.028Z"
    },
    {
      "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": 0.7593333,
      "accuracy_std_dev": 0.040381551894522326,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.19732587599999998,
      "avg_duration_per_task_ms": 103738.506,
      "avg_output_tokens_per_task": 7652.948,
      "last_run_timestamp": "2026-10-10T11:23:34.691Z"
    },
    {
      "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": "z-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": 0.7591521098265895,
      "accuracy_std_dev": 0.026787431210497644,
      "run_count": 7,
      "total_questions": 346,
      "avg_cost_per_task_usd": 0.022639277456647403,
      "avg_duration_per_task_ms": 278720.62716763007,
      "avg_output_tokens_per_task": 4187.803468208092,
      "last_run_timestamp": "2026-10-10T11:48:33.640Z"
    },
    {
      "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": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "InferenceNet",
      "accuracy": 0.7590474285714286,
      "accuracy_std_dev": 0.01256277841035793,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.03711470857142857,
      "avg_duration_per_task_ms": 81451.03428571428,
      "avg_output_tokens_per_task": 4616.405714285715,
      "last_run_timestamp": "2026-10-10T11:07:20.511Z"
    },
    {
      "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": 0.7588833197969542,
      "accuracy_std_dev": 0.017045916820458813,
      "run_count": 8,
      "total_questions": 394,
      "avg_cost_per_task_usd": 0.03293115228426395,
      "avg_duration_per_task_ms": 176423.84517766497,
      "avg_output_tokens_per_task": 16897.4923857868,
      "last_run_timestamp": "2026-10-10T11:29:01.096Z"
    },
    {
      "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": 0.7586667,
      "accuracy_std_dev": 0.030521374661210805,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.20059342000000002,
      "avg_duration_per_task_ms": 127826.938,
      "avg_output_tokens_per_task": 7898.85,
      "last_run_timestamp": "2026-10-10T11:32:27.637Z"
    },
    {
      "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": 0.7586666,
      "accuracy_std_dev": 0.027776885675683684,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.548834584,
      "avg_duration_per_task_ms": 192613.822,
      "avg_output_tokens_per_task": 10699.43,
      "last_run_timestamp": "2026-10-10T11:37:23.818Z"
    },
    {
      "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": 0.7586666,
      "accuracy_std_dev": 0.011469759275590758,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.065548452,
      "avg_duration_per_task_ms": 298543.66,
      "avg_output_tokens_per_task": 14030.988,
      "last_run_timestamp": "2026-10-10T11:18:50.146Z"
    },
    {
      "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": 0.7585170300601203,
      "accuracy_std_dev": 0.026101515127670283,
      "run_count": 10,
      "total_questions": 499,
      "avg_cost_per_task_usd": 0.013800801603206412,
      "avg_duration_per_task_ms": 219369.76553106212,
      "avg_output_tokens_per_task": 16017.749498997997,
      "last_run_timestamp": "2026-10-10T11:53:37.852Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta/muse-glimmer-30b-20260810",
      "model_slug": "meta/muse-glimmer-30b",
      "model_name": "Meta: Muse Glimmer 30B",
      "provider_name": "Phala",
      "accuracy": 0.758333375,
      "accuracy_std_dev": 0.012360224066916245,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.034803532500000005,
      "avg_duration_per_task_ms": 275650.13,
      "avg_output_tokens_per_task": 13717.48,
      "last_run_timestamp": "2026-10-03T08:20:09.018Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.1-20251113",
      "model_slug": "openai/gpt-5.1",
      "model_name": "OpenAI: GPT-5.1",
      "provider_name": "OpenAI",
      "accuracy": 0.7583332500000001,
      "accuracy_std_dev": 0.005527984369324845,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.40697005500000005,
      "avg_duration_per_task_ms": 402356.405,
      "avg_output_tokens_per_task": 35595.7,
      "last_run_timestamp": "2026-10-02T23:12:57.355Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Baidu",
      "accuracy": 0.7583332500000001,
      "accuracy_std_dev": 0.002886895683515417,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.061411685,
      "avg_duration_per_task_ms": 71222.875,
      "avg_output_tokens_per_task": 4382.8,
      "last_run_timestamp": "2026-10-10T12:11:29.952Z"
    },
    {
      "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": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Reka",
      "accuracy": 0.7582160060362173,
      "accuracy_std_dev": 0.027807515956122388,
      "run_count": 10,
      "total_questions": 497,
      "avg_cost_per_task_usd": 0.06698872032193161,
      "avg_duration_per_task_ms": 262502.1629778672,
      "avg_output_tokens_per_task": 11819.241448692153,
      "last_run_timestamp": "2026-10-10T11:04:01.954Z"
    },
    {
      "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": 0.7580952857142857,
      "accuracy_std_dev": 0.026358758071298158,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.08864299142857142,
      "avg_duration_per_task_ms": 293845.9,
      "avg_output_tokens_per_task": 12060.377142857144,
      "last_run_timestamp": "2026-10-10T11:33:11.671Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "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": "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": 0.7579564673366834,
      "accuracy_std_dev": 0.015677192643694233,
      "run_count": 4,
      "total_questions": 199,
      "avg_cost_per_task_usd": 0.011453532663316581,
      "avg_duration_per_task_ms": 323217.94974874373,
      "avg_output_tokens_per_task": 15027.713567839195,
      "last_run_timestamp": "2026-10-10T11:55:38.365Z"
    },
    {
      "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": 0.7579487692307693,
      "accuracy_std_dev": 0.02358978428229489,
      "run_count": 13,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.027215110769230763,
      "avg_duration_per_task_ms": 172637.00615384616,
      "avg_output_tokens_per_task": 17994.436923076923,
      "last_run_timestamp": "2026-10-10T11:30:09.582Z"
    },
    {
      "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": 0.757778,
      "accuracy_std_dev": 0.03095182238038124,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.06718082666666668,
      "avg_duration_per_task_ms": 287456.31333333335,
      "avg_output_tokens_per_task": 13536.3,
      "last_run_timestamp": "2026-10-10T12:17:59.841Z"
    },
    {
      "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": 0.7577778333333334,
      "accuracy_std_dev": 0.013698760873605408,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.19346874333333333,
      "avg_duration_per_task_ms": 319237.7366666667,
      "avg_output_tokens_per_task": 12183.71,
      "last_run_timestamp": "2026-10-10T11:32:01.671Z"
    },
    {
      "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": 0.7577776666666667,
      "accuracy_std_dev": 0.02199916945907024,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.33451158000000003,
      "avg_duration_per_task_ms": 129909.81333333334,
      "avg_output_tokens_per_task": 7169.106666666667,
      "last_run_timestamp": "2026-10-10T11:07:11.549Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3-flash-preview-20251217",
      "model_slug": "google/gemini-3-flash-preview",
      "model_name": "Google: Gemini 3 Flash Preview",
      "provider_name": "Google AI Studio",
      "accuracy": 0.7577776666666667,
      "accuracy_std_dev": 0.04629468145358481,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.12420428666666666,
      "avg_duration_per_task_ms": 116279.86,
      "avg_output_tokens_per_task": 19029.046666666665,
      "last_run_timestamp": "2026-10-01T19:05:00.104Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.1-pro-preview-20260219",
      "model_slug": "google/gemini-3.1-pro-preview",
      "model_name": "Google: Gemini 3.1 Pro Preview",
      "provider_name": "auto-routing",
      "accuracy": 0.7577776666666667,
      "accuracy_std_dev": 0.017497908167041637,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.35916456,
      "avg_duration_per_task_ms": 165675.65333333332,
      "avg_output_tokens_per_task": 12758.866666666667,
      "last_run_timestamp": "2026-10-01T19:06:01.702Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "InferenceNet",
      "accuracy": 0.7574073333333333,
      "accuracy_std_dev": 0.015053593332564244,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.34690624666666664,
      "avg_duration_per_task_ms": 307425.4088888889,
      "avg_output_tokens_per_task": 10583.915555555555,
      "last_run_timestamp": "2026-10-10T11:25:46.010Z"
    },
    {
      "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": 0.19399629919678715,
      "avg_duration_per_task_ms": 481017.5502008032,
      "avg_output_tokens_per_task": 28021.941767068274,
      "last_run_timestamp": "2026-10-10T19:09:38.586Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "NextBit",
      "accuracy": 0.7573333999999999,
      "accuracy_std_dev": 0.03991088294989224,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.023026752,
      "avg_duration_per_task_ms": 161149.652,
      "avg_output_tokens_per_task": 18473.992,
      "last_run_timestamp": "2026-10-03T08:55:31.322Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Parasail",
      "accuracy": 0.7571839109195402,
      "accuracy_std_dev": 0.02316220409317909,
      "run_count": 14,
      "total_questions": 696,
      "avg_cost_per_task_usd": 0.020125860632183912,
      "avg_duration_per_task_ms": 462996.8103448276,
      "avg_output_tokens_per_task": 17023.745689655174,
      "last_run_timestamp": "2026-10-10T11:45:16.493Z"
    },
    {
      "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": 0.757037,
      "accuracy_std_dev": 0.032827576317615806,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.16180529777777777,
      "avg_duration_per_task_ms": 235632.8177777778,
      "avg_output_tokens_per_task": 9773.54,
      "last_run_timestamp": "2026-10-10T20:11:50.535Z"
    },
    {
      "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": 0.7566667,
      "accuracy_std_dev": 0.028166280077603446,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.534804308,
      "avg_duration_per_task_ms": 188676.386,
      "avg_output_tokens_per_task": 10302.236,
      "last_run_timestamp": "2026-10-10T11:08:59.134Z"
    },
    {
      "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": 0.7566666666666667,
      "accuracy_std_dev": 0.01875779344402985,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.02229521,
      "avg_duration_per_task_ms": 131709.92666666667,
      "avg_output_tokens_per_task": 15842.956666666667,
      "last_run_timestamp": "2026-10-10T11:09:54.049Z"
    },
    {
      "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": 0.7566666666666667,
      "accuracy_std_dev": 0.01990715486675305,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.0064209,
      "avg_duration_per_task_ms": 136204.27333333335,
      "avg_output_tokens_per_task": 4260.18,
      "last_run_timestamp": "2026-10-10T11:23:25.227Z"
    },
    {
      "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": 0.756551724137931,
      "accuracy_std_dev": 0.029588262127353878,
      "run_count": 29,
      "total_questions": 1450,
      "avg_cost_per_task_usd": 0.2020337827586207,
      "avg_duration_per_task_ms": 119704.51379310346,
      "avg_output_tokens_per_task": 7806.491724137931,
      "last_run_timestamp": "2026-10-10T11:48:54.053Z"
    },
    {
      "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": 0.756538611018364,
      "accuracy_std_dev": 0.027065118181665127,
      "run_count": 12,
      "total_questions": 599,
      "avg_cost_per_task_usd": 0.08304781302170285,
      "avg_duration_per_task_ms": 142342.55425709515,
      "avg_output_tokens_per_task": 12749.095158597664,
      "last_run_timestamp": "2026-10-10T18:35:14.365Z"
    },
    {
      "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": 0.7564913157894737,
      "accuracy_std_dev": 0.03429274061902074,
      "run_count": 19,
      "total_questions": 950,
      "avg_cost_per_task_usd": 0.19956802315789474,
      "avg_duration_per_task_ms": 102733.17894736842,
      "avg_output_tokens_per_task": 7681.796842105263,
      "last_run_timestamp": "2026-10-10T11:31:19.613Z"
    },
    {
      "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": 0.7564102307692308,
      "accuracy_std_dev": 0.020358506100376164,
      "run_count": 13,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.026013500000000002,
      "avg_duration_per_task_ms": 296382.0492307692,
      "avg_output_tokens_per_task": 17325.696923076925,
      "last_run_timestamp": "2026-10-10T13:17:24.675Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "DeepInfra",
      "accuracy": 0.7563351081871345,
      "accuracy_std_dev": 0.013338594901982123,
      "run_count": 7,
      "total_questions": 342,
      "avg_cost_per_task_usd": 0.033182684210526316,
      "avg_duration_per_task_ms": 2146298.769005848,
      "avg_output_tokens_per_task": 50213.23391812865,
      "last_run_timestamp": "2026-10-09T04:05:56.641Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "CoreWeave",
      "accuracy": 0.7561907142857143,
      "accuracy_std_dev": 0.01846738376330619,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.14821341999999998,
      "avg_duration_per_task_ms": 147978.52,
      "avg_output_tokens_per_task": 20008.071428571428,
      "last_run_timestamp": "2026-10-10T11:08:09.342Z"
    },
    {
      "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": 0.7561904285714286,
      "accuracy_std_dev": 0.02541274877951248,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.07065733999999999,
      "avg_duration_per_task_ms": 161225.76571428572,
      "avg_output_tokens_per_task": 4721.2,
      "last_run_timestamp": "2026-10-10T11:58:35.890Z"
    },
    {
      "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": 0.7561520067114095,
      "accuracy_std_dev": 0.03687511483673266,
      "run_count": 3,
      "total_questions": 149,
      "avg_cost_per_task_usd": 0.05932429530201341,
      "avg_duration_per_task_ms": 336709.1543624161,
      "avg_output_tokens_per_task": 47869.12080536913,
      "last_run_timestamp": "2026-10-10T18:58:53.075Z"
    },
    {
      "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": 0.756,
      "accuracy_std_dev": 0.036090468630928046,
      "run_count": 30,
      "total_questions": 1500,
      "avg_cost_per_task_usd": 0.22400043066666664,
      "avg_duration_per_task_ms": 97447.95733333334,
      "avg_output_tokens_per_task": 7557.712666666666,
      "last_run_timestamp": "2026-10-10T11:27:47.581Z"
    },
    {
      "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": 0.75583325,
      "accuracy_std_dev": 0.024650453569610042,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.012750845,
      "avg_duration_per_task_ms": 348532.645,
      "avg_output_tokens_per_task": 4259.125,
      "last_run_timestamp": "2026-10-10T11:53:43.435Z"
    },
    {
      "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": 0.7557575454545454,
      "accuracy_std_dev": 0.021133944253321,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.022080854545454547,
      "avg_duration_per_task_ms": 128740.06909090909,
      "avg_output_tokens_per_task": 17525.272727272728,
      "last_run_timestamp": "2026-10-10T11:50:52.789Z"
    },
    {
      "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": 0.7557575454545454,
      "accuracy_std_dev": 0.02365967382148038,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.022588227272727274,
      "avg_duration_per_task_ms": 122776.59454545454,
      "avg_output_tokens_per_task": 15090.52909090909,
      "last_run_timestamp": "2026-10-10T11:23:47.008Z"
    },
    {
      "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": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "DeepInfra",
      "accuracy": 0.7556891325301206,
      "accuracy_std_dev": 0.028686868714448425,
      "run_count": 5,
      "total_questions": 249,
      "avg_cost_per_task_usd": 0.05997379919678716,
      "avg_duration_per_task_ms": 208407.72289156626,
      "avg_output_tokens_per_task": 7587.301204819277,
      "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/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "DeepInfra",
      "accuracy": 0.7555803571428571,
      "accuracy_std_dev": 0.023882484930784247,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.044577171874999995,
      "avg_duration_per_task_ms": 507800.79910714284,
      "avg_output_tokens_per_task": 22983.540178571428,
      "last_run_timestamp": "2026-10-09T00:21:32.259Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Morph",
      "accuracy": 0.7555556666666666,
      "accuracy_std_dev": 0.00831485358994506,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.027460826666666667,
      "avg_duration_per_task_ms": 642340.9866666667,
      "avg_output_tokens_per_task": 17001.633333333335,
      "last_run_timestamp": "2026-08-08T08:29:47.435Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Azure",
      "accuracy": 0.7555555,
      "accuracy_std_dev": 0.02819693530480457,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.025840563333333334,
      "avg_duration_per_task_ms": 269041.9766666667,
      "avg_output_tokens_per_task": 13228.333333333334,
      "last_run_timestamp": "2026-10-10T18:35:19.634Z"
    },
    {
      "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": 0.7554166250000001,
      "accuracy_std_dev": 0.041664622033979574,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.1590986875,
      "avg_duration_per_task_ms": 337073.18,
      "avg_output_tokens_per_task": 22844.935,
      "last_run_timestamp": "2026-10-10T11:23:09.879Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Modal",
      "accuracy": 0.7554165,
      "accuracy_std_dev": 0.031576074364144766,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0160830725,
      "avg_duration_per_task_ms": 75201.88,
      "avg_output_tokens_per_task": 4275.875,
      "last_run_timestamp": "2026-10-10T11:55:27.055Z"
    },
    {
      "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": 0.7552381428571429,
      "accuracy_std_dev": 0.023289474817292592,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.00790055142857143,
      "avg_duration_per_task_ms": 177661.23142857142,
      "avg_output_tokens_per_task": 4430.751428571429,
      "last_run_timestamp": "2026-10-10T12:18:32.503Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Z.AI",
      "accuracy": 0.7552380000000001,
      "accuracy_std_dev": 0.025872597329441617,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.008791902857142856,
      "avg_duration_per_task_ms": 194073.0285714286,
      "avg_output_tokens_per_task": 4458.371428571429,
      "last_run_timestamp": "2026-10-10T11:41:33.492Z"
    },
    {
      "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": 0.7551516363636364,
      "accuracy_std_dev": 0.02221853890651896,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.24386067272727271,
      "avg_duration_per_task_ms": 305161.86363636365,
      "avg_output_tokens_per_task": 10007.39090909091,
      "last_run_timestamp": "2026-10-10T11:43:15.517Z"
    },
    {
      "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": 0.755027365630713,
      "accuracy_std_dev": 0.03406230364037725,
      "run_count": 11,
      "total_questions": 547,
      "avg_cost_per_task_usd": 0.19381868007312616,
      "avg_duration_per_task_ms": 569519.8921389397,
      "avg_output_tokens_per_task": 27223.491773308957,
      "last_run_timestamp": "2026-10-10T11:06:26.011Z"
    },
    {
      "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": 0.755000125,
      "accuracy_std_dev": 0.023033866975811414,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.19533066500000001,
      "avg_duration_per_task_ms": 121599.4775,
      "avg_output_tokens_per_task": 7673.8175,
      "last_run_timestamp": "2026-10-10T11:29:16.870Z"
    },
    {
      "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": 0.755,
      "accuracy_std_dev": 0.023033658947288434,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.018529425000000002,
      "avg_duration_per_task_ms": 140249.55,
      "avg_output_tokens_per_task": 4584.6825,
      "last_run_timestamp": "2026-10-10T11:55:21.479Z"
    },
    {
      "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": 0.7548146666666667,
      "accuracy_std_dev": 0.03201517730007999,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.09135532,
      "avg_duration_per_task_ms": 231928.58,
      "avg_output_tokens_per_task": 11085.384444444444,
      "last_run_timestamp": "2026-10-10T20:08:22.445Z"
    },
    {
      "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": 0.7546667,
      "accuracy_std_dev": 0.02166412087784778,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.34589102000000005,
      "avg_duration_per_task_ms": 220634.988,
      "avg_output_tokens_per_task": 11962.75,
      "last_run_timestamp": "2026-10-10T11:51:26.821Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Sail Research",
      "accuracy": 0.7544342825688073,
      "accuracy_std_dev": 0.02504935915479907,
      "run_count": 11,
      "total_questions": 545,
      "avg_cost_per_task_usd": 0.1000418,
      "avg_duration_per_task_ms": 796849.0642201835,
      "avg_output_tokens_per_task": 21296.05137614679,
      "last_run_timestamp": "2026-10-10T11:45:05.221Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Ionstream",
      "accuracy": 0.7542953608247424,
      "accuracy_std_dev": 0.02652500000000002,
      "run_count": 2,
      "total_questions": 97,
      "avg_cost_per_task_usd": 0.021299783505154636,
      "avg_duration_per_task_ms": 633495.5257731959,
      "avg_output_tokens_per_task": 26267.113402061856,
      "last_run_timestamp": "2026-08-08T08:30:53.738Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Novita",
      "accuracy": 0.75416675,
      "accuracy_std_dev": 0.03169964729989123,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.014154627500000003,
      "avg_duration_per_task_ms": 169480.67,
      "avg_output_tokens_per_task": 14972.985,
      "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": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "SiliconFlow",
      "accuracy": 0.75416675,
      "accuracy_std_dev": 0.030264911385753292,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0725496575,
      "avg_duration_per_task_ms": 303727.6275,
      "avg_output_tokens_per_task": 11490.4825,
      "last_run_timestamp": "2026-10-10T21:35:36.905Z"
    },
    {
      "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": 0.75416675,
      "accuracy_std_dev": 0.012219036835507957,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.1638202875,
      "avg_duration_per_task_ms": 453197.5425,
      "avg_output_tokens_per_task": 9528.3525,
      "last_run_timestamp": "2026-10-03T16:16:10.656Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Venice",
      "accuracy": 0.7540001,
      "accuracy_std_dev": 0.032584956993833826,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.020762876000000003,
      "avg_duration_per_task_ms": 187591.642,
      "avg_output_tokens_per_task": 14436.292,
      "last_run_timestamp": "2026-10-10T11:41:11.074Z"
    },
    {
      "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": "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": 0.7537931724137933,
      "accuracy_std_dev": 0.024380773394523443,
      "run_count": 29,
      "total_questions": 1450,
      "avg_cost_per_task_usd": 0.3712446551724138,
      "avg_duration_per_task_ms": 167809.39172413794,
      "avg_output_tokens_per_task": 11853.664827586206,
      "last_run_timestamp": "2026-10-10T11:36:16.238Z"
    },
    {
      "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": 0.7533785608108108,
      "accuracy_std_dev": 0.03423001508788255,
      "run_count": 6,
      "total_questions": 296,
      "avg_cost_per_task_usd": 0.056939652027027034,
      "avg_duration_per_task_ms": 407763.0641891892,
      "avg_output_tokens_per_task": 12744.533783783783,
      "last_run_timestamp": "2026-10-10T11:23:27.030Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Decart",
      "accuracy": 0.7533336666666666,
      "accuracy_std_dev": 0.024944382578492966,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.3148796866666667,
      "avg_duration_per_task_ms": 281526.8,
      "avg_output_tokens_per_task": 9656.493333333334,
      "last_run_timestamp": "2026-10-10T11:41:13.622Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "x-ai/grok-4.6-20260810",
      "model_slug": "x-ai/grok-4.6",
      "model_name": "SpaceXAI: Grok 4.6",
      "provider_name": "auto-routing",
      "accuracy": 0.7533335,
      "accuracy_std_dev": 0.00666650000000002,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.33087047999999997,
      "avg_duration_per_task_ms": 241177.17,
      "avg_output_tokens_per_task": 11982.77,
      "last_run_timestamp": "2026-10-01T16:01:02.676Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Near AI",
      "accuracy": 0.7533335,
      "accuracy_std_dev": 0.053333500000000034,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.00958291,
      "avg_duration_per_task_ms": 393540.68,
      "avg_output_tokens_per_task": 4161.75,
      "last_run_timestamp": "2026-10-10T11:31:07.384Z"
    },
    {
      "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/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "CoreWeave",
      "accuracy": 0.7533333333333333,
      "accuracy_std_dev": 0.03678447069076843,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.08935323333333334,
      "avg_duration_per_task_ms": 248145.7288888889,
      "avg_output_tokens_per_task": 12283.197777777777,
      "last_run_timestamp": "2026-10-10T11:25:27.264Z"
    },
    {
      "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": 0.7533332857142857,
      "accuracy_std_dev": 0.017089883763830138,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.1783802742857143,
      "avg_duration_per_task_ms": 316437.2057142857,
      "avg_output_tokens_per_task": 10826.988571428572,
      "last_run_timestamp": "2026-10-10T11:07:50.254Z"
    },
    {
      "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": 0.7533332500000001,
      "accuracy_std_dev": 0.026246544747594865,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.34973291500000003,
      "avg_duration_per_task_ms": 104206.33,
      "avg_output_tokens_per_task": 7119.515,
      "last_run_timestamp": "2026-10-10T11:23:32.977Z"
    },
    {
      "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": 0.7533332500000001,
      "accuracy_std_dev": 0.027487471895165275,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.1073622525,
      "avg_duration_per_task_ms": 273662.3775,
      "avg_output_tokens_per_task": 12208.0525,
      "last_run_timestamp": "2026-10-10T11:27:45.281Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7533332222222222,
      "accuracy_std_dev": 0.02591535128399458,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.013880666666666666,
      "avg_duration_per_task_ms": 210103.7488888889,
      "avg_output_tokens_per_task": 14896.853333333333,
      "last_run_timestamp": "2026-10-10T18:54:17.152Z"
    },
    {
      "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": "Phala",
      "accuracy": 0.7533332,
      "accuracy_std_dev": 0.02732525083800698,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.011721984,
      "avg_duration_per_task_ms": 506375.232,
      "avg_output_tokens_per_task": 6019.948,
      "last_run_timestamp": "2026-10-10T11:38:18.867Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "auto-routing",
      "accuracy": 0.753333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.0119157,
      "avg_duration_per_task_ms": 95782.12,
      "avg_output_tokens_per_task": 11167.88,
      "last_run_timestamp": "2026-10-10T11:10:14.496Z"
    },
    {
      "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": "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": "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": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "StreamLake",
      "accuracy": 0.7526158878923768,
      "accuracy_std_dev": 0.04664293923964891,
      "run_count": 9,
      "total_questions": 446,
      "avg_cost_per_task_usd": 0.07574351569506728,
      "avg_duration_per_task_ms": 360044.43497757846,
      "avg_output_tokens_per_task": 12747.226457399103,
      "last_run_timestamp": "2026-10-10T19:37:33.554Z"
    },
    {
      "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": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Venice",
      "accuracy": 0.7525,
      "accuracy_std_dev": 0.03044809791678292,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.021822374999999998,
      "avg_duration_per_task_ms": 485502.945,
      "avg_output_tokens_per_task": 14781.975,
      "last_run_timestamp": "2026-10-10T18:37:34.913Z"
    },
    {
      "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": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "CoreWeave",
      "accuracy": 0.752,
      "accuracy_std_dev": 0.022469515152757562,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.019928192,
      "avg_duration_per_task_ms": 74879.332,
      "avg_output_tokens_per_task": 4253.16,
      "last_run_timestamp": "2026-10-10T11:24:18.065Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "AkashML",
      "accuracy": 0.7517005306122448,
      "accuracy_std_dev": 0.012796557796532647,
      "run_count": 5,
      "total_questions": 245,
      "avg_cost_per_task_usd": 0.09461716326530611,
      "avg_duration_per_task_ms": 209290.9224489796,
      "avg_output_tokens_per_task": 7253.518367346939,
      "last_run_timestamp": "2026-08-15T08:10:05.152Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "DeepInfra",
      "accuracy": 0.7515597513368985,
      "accuracy_std_dev": 0.021149574595143886,
      "run_count": 15,
      "total_questions": 748,
      "avg_cost_per_task_usd": 0.00707783155080214,
      "avg_duration_per_task_ms": 201238.5909090909,
      "avg_output_tokens_per_task": 13061.36898395722,
      "last_run_timestamp": "2026-10-10T11:51:08.689Z"
    },
    {
      "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": 0.7514815555555555,
      "accuracy_std_dev": 0.0333005142973669,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.07395462222222221,
      "avg_duration_per_task_ms": 378942.0577777778,
      "avg_output_tokens_per_task": 9157.224444444444,
      "last_run_timestamp": "2026-10-10T18:30:31.154Z"
    },
    {
      "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": 0.7513333999999999,
      "accuracy_std_dev": 0.030521588717496353,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.062658464,
      "avg_duration_per_task_ms": 1700876.012,
      "avg_output_tokens_per_task": 8868.604,
      "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": "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-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Ionstream",
      "accuracy": 0.751111,
      "accuracy_std_dev": 0.01257094435593444,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.03708808,
      "avg_duration_per_task_ms": 176459.57333333333,
      "avg_output_tokens_per_task": 25948.413333333334,
      "last_run_timestamp": "2026-10-10T11:21:01.694Z"
    },
    {
      "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": 0.750833375,
      "accuracy_std_dev": 0.028467763202162117,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.022724927500000002,
      "avg_duration_per_task_ms": 216195.8375,
      "avg_output_tokens_per_task": 13169.8975,
      "last_run_timestamp": "2026-10-10T18:54:11.948Z"
    },
    {
      "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": 0.7506667,
      "accuracy_std_dev": 0.026365013855676238,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.49418161999999993,
      "avg_duration_per_task_ms": 195069.658,
      "avg_output_tokens_per_task": 10103.428,
      "last_run_timestamp": "2026-10-10T11:51:44.657Z"
    },
    {
      "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": 0.7505569242761692,
      "accuracy_std_dev": 0.02142411426983423,
      "run_count": 9,
      "total_questions": 449,
      "avg_cost_per_task_usd": 0.028247770601336302,
      "avg_duration_per_task_ms": 598851.2494432072,
      "avg_output_tokens_per_task": 22337.24053452116,
      "last_run_timestamp": "2026-10-10T19:09:44.921Z"
    },
    {
      "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": 0.7505555,
      "accuracy_std_dev": 0.013181957837008396,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.36349930333333336,
      "avg_duration_per_task_ms": 169671.99166666667,
      "avg_output_tokens_per_task": 7150.9766666666665,
      "last_run_timestamp": "2026-10-10T11:07:06.201Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Fireworks",
      "accuracy": 0.7505555,
      "accuracy_std_dev": 0.03274012621004996,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.02124632666666667,
      "avg_duration_per_task_ms": 354797.86666666664,
      "avg_output_tokens_per_task": 18494.39,
      "last_run_timestamp": "2026-09-19T08:13:24.062Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7503702222222223,
      "accuracy_std_dev": 0.036799392403733415,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.07642833777777779,
      "avg_duration_per_task_ms": 230421.34666666668,
      "avg_output_tokens_per_task": 9496.562222222223,
      "last_run_timestamp": "2026-10-10T19:41:42.109Z"
    },
    {
      "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": 0.7503701111111112,
      "accuracy_std_dev": 0.02516343723405248,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.0877466288888889,
      "avg_duration_per_task_ms": 251071.52,
      "avg_output_tokens_per_task": 9774.766666666666,
      "last_run_timestamp": "2026-10-10T20:31:30.609Z"
    },
    {
      "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": 0.7503029090909091,
      "accuracy_std_dev": 0.02076579266903998,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.03820286181818182,
      "avg_duration_per_task_ms": 292584.13636363635,
      "avg_output_tokens_per_task": 23337.65090909091,
      "last_run_timestamp": "2026-10-10T18:44:59.530Z"
    },
    {
      "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": 0.7500001999999999,
      "accuracy_std_dev": 0.026832766040048864,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.015060424,
      "avg_duration_per_task_ms": 269962.708,
      "avg_output_tokens_per_task": 16418.532,
      "last_run_timestamp": "2026-10-10T11:41:33.490Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.75,
      "accuracy_std_dev": 0.016666999999999987,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01247142,
      "avg_duration_per_task_ms": 96762.37,
      "avg_output_tokens_per_task": 10932.5,
      "last_run_timestamp": "2026-10-10T11:07:19.680Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "BaseTen",
      "accuracy": 0.75,
      "accuracy_std_dev": 0.0033330000000000304,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.28014248,
      "avg_duration_per_task_ms": 223020.56,
      "avg_output_tokens_per_task": 18715.38,
      "last_run_timestamp": "2026-09-12T08:12:13.731Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "Chutes",
      "accuracy": 0.75,
      "accuracy_std_dev": 0.010000000000000009,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.05066206,
      "avg_duration_per_task_ms": 368063.6,
      "avg_output_tokens_per_task": 8379.6,
      "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": "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": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "DeepInfra",
      "accuracy": 0.7498134675615213,
      "accuracy_std_dev": 0.030550014400594683,
      "run_count": 9,
      "total_questions": 447,
      "avg_cost_per_task_usd": 0.06635837807606264,
      "avg_duration_per_task_ms": 426889.067114094,
      "avg_output_tokens_per_task": 10337.557046979866,
      "last_run_timestamp": "2026-10-10T18:57:24.715Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "Venice",
      "accuracy": 0.7496296666666666,
      "accuracy_std_dev": 0.025553012220784384,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.32866541777777775,
      "avg_duration_per_task_ms": 573042.4666666667,
      "avg_output_tokens_per_task": 26324.751111111113,
      "last_run_timestamp": "2026-10-10T18:29:59.611Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Fireworks",
      "accuracy": 0.7495746530612246,
      "accuracy_std_dev": 0.021234638556127484,
      "run_count": 8,
      "total_questions": 392,
      "avg_cost_per_task_usd": 0.014195257653061224,
      "avg_duration_per_task_ms": 56974.566326530614,
      "avg_output_tokens_per_task": 3915.466836734694,
      "last_run_timestamp": "2026-10-10T11:44:02.286Z"
    },
    {
      "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": 0.7495237142857143,
      "accuracy_std_dev": 0.027338579222746164,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.17267997714285713,
      "avg_duration_per_task_ms": 322644.1828571429,
      "avg_output_tokens_per_task": 11143.974285714286,
      "last_run_timestamp": "2026-10-10T11:20:38.332Z"
    },
    {
      "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": 0.7493333,
      "accuracy_std_dev": 0.028237051310467966,
      "run_count": 20,
      "total_questions": 1000,
      "avg_cost_per_task_usd": 0.5616704749999999,
      "avg_duration_per_task_ms": 202133.878,
      "avg_output_tokens_per_task": 9880.147,
      "last_run_timestamp": "2026-10-10T11:58:23.290Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "StreamLake",
      "accuracy": 0.7492560625,
      "accuracy_std_dev": 0.02625846130626927,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.005948455357142857,
      "avg_duration_per_task_ms": 261080.47767857142,
      "avg_output_tokens_per_task": 12566.60267857143,
      "last_run_timestamp": "2026-10-10T18:32:35.317Z"
    },
    {
      "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": 0.7491665000000001,
      "accuracy_std_dev": 0.022095128925851498,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.11723742000000001,
      "avg_duration_per_task_ms": 336463.235,
      "avg_output_tokens_per_task": 9787.0025,
      "last_run_timestamp": "2026-10-10T19:45:51.536Z"
    },
    {
      "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": 0.7490475714285715,
      "accuracy_std_dev": 0.017155993645346933,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.034581794285714285,
      "avg_duration_per_task_ms": 410319.02857142856,
      "avg_output_tokens_per_task": 9408.788571428571,
      "last_run_timestamp": "2026-10-09T02:05:38.689Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-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": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Mancer 2",
      "accuracy": 0.7488888333333333,
      "accuracy_std_dev": 0.02948100036982387,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.09071532,
      "avg_duration_per_task_ms": 524416.5966666667,
      "avg_output_tokens_per_task": 12656.81,
      "last_run_timestamp": "2026-10-10T11:02:27.901Z"
    },
    {
      "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": "anthropic/claude-4.5-sonnet-20250929",
      "model_slug": "anthropic/claude-sonnet-4.5",
      "model_name": "Anthropic: Claude Sonnet 4.5",
      "provider_name": "Anthropic",
      "accuracy": 0.7487180769230769,
      "accuracy_std_dev": 0.026458784394024375,
      "run_count": 13,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.35342018461538466,
      "avg_duration_per_task_ms": 231535.48615384614,
      "avg_output_tokens_per_task": 10881.203076923077,
      "last_run_timestamp": "2026-10-10T11:05:56.995Z"
    },
    {
      "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": 0.7486667,
      "accuracy_std_dev": 0.030265335507970156,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.548401042,
      "avg_duration_per_task_ms": 183718.584,
      "avg_output_tokens_per_task": 10523.854,
      "last_run_timestamp": "2026-10-10T11:37:47.013Z"
    },
    {
      "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": 0.7486667,
      "accuracy_std_dev": 0.021509684642272205,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.32253603399999997,
      "avg_duration_per_task_ms": 198893.474,
      "avg_output_tokens_per_task": 11527.52,
      "last_run_timestamp": "2026-10-10T11:52:35.504Z"
    },
    {
      "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": 0.748484909090909,
      "accuracy_std_dev": 0.0478288811635331,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.13337880545454545,
      "avg_duration_per_task_ms": 874645.5763636364,
      "avg_output_tokens_per_task": 32352.327272727274,
      "last_run_timestamp": "2026-10-10T19:56:25.635Z"
    },
    {
      "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": 0.7484848181818182,
      "accuracy_std_dev": 0.0229345089085429,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.009084052727272729,
      "avg_duration_per_task_ms": 248840.61454545453,
      "avg_output_tokens_per_task": 4233.045454545455,
      "last_run_timestamp": "2026-10-10T11:56:23.494Z"
    },
    {
      "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": 0.7483335,
      "accuracy_std_dev": 0.05644579011981318,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.02282431,
      "avg_duration_per_task_ms": 395794.725,
      "avg_output_tokens_per_task": 15096.88,
      "last_run_timestamp": "2026-10-10T12:00:54.577Z"
    },
    {
      "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": 0.748333375,
      "accuracy_std_dev": 0.026822334704204524,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.34900722500000003,
      "avg_duration_per_task_ms": 202391.0825,
      "avg_output_tokens_per_task": 11926.3975,
      "last_run_timestamp": "2026-10-10T11:49:25.841Z"
    },
    {
      "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": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Ionstream",
      "accuracy": 0.7483132064777328,
      "accuracy_std_dev": 0.03485977243528707,
      "run_count": 5,
      "total_questions": 247,
      "avg_cost_per_task_usd": 0.10924008502024292,
      "avg_duration_per_task_ms": 442802.26720647776,
      "avg_output_tokens_per_task": 28877.655870445345,
      "last_run_timestamp": "2026-10-10T11:44:46.399Z"
    },
    {
      "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": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Together",
      "accuracy": 0.748,
      "accuracy_std_dev": 0.029333454545961698,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.07858999599999998,
      "avg_duration_per_task_ms": 166655.096,
      "avg_output_tokens_per_task": 8640.336,
      "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": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7477778333333334,
      "accuracy_std_dev": 0.015112569375817262,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.01356969,
      "avg_duration_per_task_ms": 102348.01,
      "avg_output_tokens_per_task": 11130.603333333333,
      "last_run_timestamp": "2026-10-10T11:22:10.457Z"
    },
    {
      "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": 0.7476922307692307,
      "accuracy_std_dev": 0.027029079235729832,
      "run_count": 13,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.33929932,
      "avg_duration_per_task_ms": 209258.88923076924,
      "avg_output_tokens_per_task": 11836.412307692308,
      "last_run_timestamp": "2026-10-10T11:52:21.526Z"
    },
    {
      "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": "anthropic/claude-4.6-opus-20260205",
      "model_slug": "anthropic/claude-opus-4.6",
      "model_name": "Anthropic: Claude Opus 4.6",
      "provider_name": "Azure",
      "accuracy": 0.7473333,
      "accuracy_std_dev": 0.02866662093114572,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.520711906,
      "avg_duration_per_task_ms": 226639.988,
      "avg_output_tokens_per_task": 10065.72,
      "last_run_timestamp": "2026-10-10T11:53:44.033Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "Modal",
      "accuracy": 0.7472221666666666,
      "accuracy_std_dev": 0.011613686586906374,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.17439429333333334,
      "avg_duration_per_task_ms": 161420.59666666668,
      "avg_output_tokens_per_task": 12499.9,
      "last_run_timestamp": "2026-10-10T11:19:48.441Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7470688291457286,
      "accuracy_std_dev": 0.06803424962002137,
      "run_count": 8,
      "total_questions": 398,
      "avg_cost_per_task_usd": 0.2510165728643216,
      "avg_duration_per_task_ms": 432358.3894472362,
      "avg_output_tokens_per_task": 28211.77135678392,
      "last_run_timestamp": "2026-10-10T18:34:57.651Z"
    },
    {
      "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": 0.7466670000000001,
      "accuracy_std_dev": 0.03999999999999998,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.16298501999999998,
      "avg_duration_per_task_ms": 1681916.16,
      "avg_output_tokens_per_task": 38161.67,
      "last_run_timestamp": "2026-10-10T12:00:26.892Z"
    },
    {
      "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": 0.746667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.02836242,
      "avg_duration_per_task_ms": 881974.84,
      "avg_output_tokens_per_task": 13870.46,
      "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": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "Makora",
      "accuracy": 0.746667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.02949668,
      "avg_duration_per_task_ms": 109584.32,
      "avg_output_tokens_per_task": 40655.62,
      "last_run_timestamp": "2026-10-10T18:33:47.508Z"
    },
    {
      "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": "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": 0.746667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.38669936,
      "avg_duration_per_task_ms": 338232.82,
      "avg_output_tokens_per_task": 9439.24,
      "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-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "InferenceNet",
      "accuracy": 0.7466667142857143,
      "accuracy_std_dev": 0.030446278476379096,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.007662862857142856,
      "avg_duration_per_task_ms": 184997.26571428572,
      "avg_output_tokens_per_task": 4375.4742857142855,
      "last_run_timestamp": "2026-10-10T11:52:30.897Z"
    },
    {
      "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": 0.7466667,
      "accuracy_std_dev": 0.025820018075322887,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.085612296,
      "avg_duration_per_task_ms": 852424.342,
      "avg_output_tokens_per_task": 36599.922,
      "last_run_timestamp": "2026-10-10T19:38:18.068Z"
    },
    {
      "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": "Baidu",
      "accuracy": 0.746666625,
      "accuracy_std_dev": 0.01632993162062767,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.00887094,
      "avg_duration_per_task_ms": 153911.1775,
      "avg_output_tokens_per_task": 13256.88,
      "last_run_timestamp": "2026-10-10T18:34:58.542Z"
    },
    {
      "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": "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": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Makora",
      "accuracy": 0.7466665,
      "accuracy_std_dev": 0.00666650000000002,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01003968,
      "avg_duration_per_task_ms": 293152.56,
      "avg_output_tokens_per_task": 17237.02,
      "last_run_timestamp": "2026-10-10T11:52:52.331Z"
    },
    {
      "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/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "AkashML",
      "accuracy": 0.7465427067238912,
      "accuracy_std_dev": 0.015385569261983973,
      "run_count": 14,
      "total_questions": 699,
      "avg_cost_per_task_usd": 0.048925171673819746,
      "avg_duration_per_task_ms": 263945.0042918455,
      "avg_output_tokens_per_task": 28450.080114449214,
      "last_run_timestamp": "2026-10-05T17:03:10.825Z"
    },
    {
      "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": 0.7465276875000001,
      "accuracy_std_dev": 0.001809000000000005,
      "run_count": 2,
      "total_questions": 96,
      "avg_cost_per_task_usd": 0.03102420833333333,
      "avg_duration_per_task_ms": 572971.2916666666,
      "avg_output_tokens_per_task": 10510.135416666666,
      "last_run_timestamp": "2026-10-10T12:09:56.410Z"
    },
    {
      "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": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "auto-routing",
      "accuracy": 0.7462797633928571,
      "accuracy_std_dev": 0.024545314608085133,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.03578149553571429,
      "avg_duration_per_task_ms": 601980.8191964285,
      "avg_output_tokens_per_task": 15021.482142857143,
      "last_run_timestamp": "2026-10-10T19:19:24.410Z"
    },
    {
      "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": 0.07188528888888888,
      "avg_duration_per_task_ms": 265497.6822222222,
      "avg_output_tokens_per_task": 9840.462222222222,
      "last_run_timestamp": "2026-10-10T20:32:33.459Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7459258888888889,
      "accuracy_std_dev": 0.038160653111241984,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.11357590222222222,
      "avg_duration_per_task_ms": 276977.56444444443,
      "avg_output_tokens_per_task": 9323.804444444444,
      "last_run_timestamp": "2026-10-10T19:45:37.580Z"
    },
    {
      "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": 0.74583325,
      "accuracy_std_dev": 0.030810710170125915,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.11509349499999999,
      "avg_duration_per_task_ms": 368700.52,
      "avg_output_tokens_per_task": 8722.7075,
      "last_run_timestamp": "2026-10-10T20:05:26.252Z"
    },
    {
      "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": 0.7455555,
      "accuracy_std_dev": 0.03536397169318889,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.04841413333333334,
      "avg_duration_per_task_ms": 491827.81,
      "avg_output_tokens_per_task": 9312.496666666666,
      "last_run_timestamp": "2026-10-10T19:27:19.554Z"
    },
    {
      "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": 0.7454953513513514,
      "accuracy_std_dev": 0.009666043692569685,
      "run_count": 3,
      "total_questions": 148,
      "avg_cost_per_task_usd": 0.27390454729729735,
      "avg_duration_per_task_ms": 475940.8445945946,
      "avg_output_tokens_per_task": 13709.68918918919,
      "last_run_timestamp": "2026-10-10T11:24:20.724Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Inceptron",
      "accuracy": 0.7453938040201006,
      "accuracy_std_dev": 0.03820312878139041,
      "run_count": 4,
      "total_questions": 199,
      "avg_cost_per_task_usd": 0.016367582914572864,
      "avg_duration_per_task_ms": 153008.0201005025,
      "avg_output_tokens_per_task": 3923.0502512562816,
      "last_run_timestamp": "2026-10-10T11:50:28.035Z"
    },
    {
      "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": 0.7453333,
      "accuracy_std_dev": 0.036246163322619415,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.030892698000000003,
      "avg_duration_per_task_ms": 266107.998,
      "avg_output_tokens_per_task": 17044.59,
      "last_run_timestamp": "2026-10-10T11:29:00.636Z"
    },
    {
      "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": "Cloudflare",
      "accuracy": 0.7451852222222222,
      "accuracy_std_dev": 0.02325970157818395,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.07844153111111112,
      "avg_duration_per_task_ms": 265823.6488888889,
      "avg_output_tokens_per_task": 11310.011111111111,
      "last_run_timestamp": "2026-10-10T22:19:57.322Z"
    },
    {
      "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": 0.7450001666666666,
      "accuracy_std_dev": 0.03916974692462142,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.025603288333333335,
      "avg_duration_per_task_ms": 399238.37333333335,
      "avg_output_tokens_per_task": 15346.756666666666,
      "last_run_timestamp": "2026-10-10T13:03:25.768Z"
    },
    {
      "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": 0.7449999642857142,
      "accuracy_std_dev": 0.03317215855576367,
      "run_count": 28,
      "total_questions": 1400,
      "avg_cost_per_task_usd": 0.37590654428571424,
      "avg_duration_per_task_ms": 201970.09785714286,
      "avg_output_tokens_per_task": 11978.649285714286,
      "last_run_timestamp": "2026-10-10T11:56:02.029Z"
    },
    {
      "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": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "DeepInfra",
      "accuracy": 0.7447618571428571,
      "accuracy_std_dev": 0.016218418466392492,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.013583602857142859,
      "avg_duration_per_task_ms": 386744.47428571427,
      "avg_output_tokens_per_task": 16391.625714285714,
      "last_run_timestamp": "2026-10-10T11:53:49.565Z"
    },
    {
      "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": 0.7447618571428571,
      "accuracy_std_dev": 0.01621859463198753,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.05547540571428572,
      "avg_duration_per_task_ms": 220891.45428571428,
      "avg_output_tokens_per_task": 9176.125714285714,
      "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-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "GMICloud",
      "accuracy": 0.7447618571428571,
      "accuracy_std_dev": 0.026358618706647914,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.008384540000000001,
      "avg_duration_per_task_ms": 237385.73142857142,
      "avg_output_tokens_per_task": 5177.305714285714,
      "last_run_timestamp": "2026-10-10T12:19:01.192Z"
    },
    {
      "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": "amazon/nova-micro-v1",
      "model_slug": "amazon/nova-micro-v1",
      "model_name": "Amazon: Nova Micro 1.0",
      "provider_name": "Anthropic",
      "accuracy": 0.7444443333333334,
      "accuracy_std_dev": 0.00831485358994506,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 1.2319787333333332,
      "avg_duration_per_task_ms": 98531.56,
      "avg_output_tokens_per_task": 5524.42,
      "last_run_timestamp": "2026-10-01T19:04:26.700Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "Reka",
      "accuracy": 0.7444443333333334,
      "accuracy_std_dev": 0.006285236475706844,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.04041968666666667,
      "avg_duration_per_task_ms": 291588.68666666665,
      "avg_output_tokens_per_task": 24330.48,
      "last_run_timestamp": "2026-09-30T11:25:05.946Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-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": 0.7444443333333334,
      "accuracy_std_dev": 0.03499547766339085,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.028980353333333334,
      "avg_duration_per_task_ms": 164219.27333333335,
      "avg_output_tokens_per_task": 9416.14,
      "last_run_timestamp": "2026-10-10T11:38:09.393Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Phala",
      "accuracy": 0.74416675,
      "accuracy_std_dev": 0.034146227627477405,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.127937195,
      "avg_duration_per_task_ms": 400891.875,
      "avg_output_tokens_per_task": 11822.0075,
      "last_run_timestamp": "2026-10-10T20:11:03.050Z"
    },
    {
      "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": 0.74416675,
      "accuracy_std_dev": 0.006400770358909954,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.08563862,
      "avg_duration_per_task_ms": 309336.805,
      "avg_output_tokens_per_task": 9480.84,
      "last_run_timestamp": "2026-08-22T16:12:20.461Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Reka",
      "accuracy": 0.7439999,
      "accuracy_std_dev": 0.022548769862012447,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.014404088000000002,
      "avg_duration_per_task_ms": 119393.714,
      "avg_output_tokens_per_task": 15524.802,
      "last_run_timestamp": "2026-10-10T11:03:23.025Z"
    },
    {
      "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": "anthropic/claude-opus-5.5-20260921",
      "model_slug": "anthropic/claude-opus-5.5",
      "model_name": "Anthropic: Claude Opus 5.5",
      "provider_name": "Azure",
      "accuracy": 0.7433335,
      "accuracy_std_dev": 0.023333285714832345,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.353090755,
      "avg_duration_per_task_ms": 88187.84,
      "avg_output_tokens_per_task": 6970.8425,
      "last_run_timestamp": "2026-10-10T11:20:56.943Z"
    },
    {
      "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-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7433333333333333,
      "accuracy_std_dev": 0.04741472657718158,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.08355239,
      "avg_duration_per_task_ms": 182015.24166666667,
      "avg_output_tokens_per_task": 8140.348333333333,
      "last_run_timestamp": "2026-10-10T12:46:20.944Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Cloudflare",
      "accuracy": 0.7433333333333333,
      "accuracy_std_dev": 0.017532868606008445,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.012350623333333333,
      "avg_duration_per_task_ms": 125234.99333333333,
      "avg_output_tokens_per_task": 4198.8,
      "last_run_timestamp": "2026-10-03T08:23:09.816Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "Z.AI",
      "accuracy": 0.742963,
      "accuracy_std_dev": 0.05087977976673161,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.0923280022222222,
      "avg_duration_per_task_ms": 241983.94666666666,
      "avg_output_tokens_per_task": 11240.468888888889,
      "last_run_timestamp": "2026-10-10T20:09:18.127Z"
    },
    {
      "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": 0.742962888888889,
      "accuracy_std_dev": 0.04298856070177132,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.009493242222222222,
      "avg_duration_per_task_ms": 317851.38666666666,
      "avg_output_tokens_per_task": 15469.368888888888,
      "last_run_timestamp": "2026-10-10T11:49:09.972Z"
    },
    {
      "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": 0.742962888888889,
      "accuracy_std_dev": 0.04098905672234545,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.04189076222222222,
      "avg_duration_per_task_ms": 206004.44666666666,
      "avg_output_tokens_per_task": 19717.306666666667,
      "last_run_timestamp": "2026-10-10T11:33:22.065Z"
    },
    {
      "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": 0.74291675,
      "accuracy_std_dev": 0.0298810059057171,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0853285525,
      "avg_duration_per_task_ms": 188726.3925,
      "avg_output_tokens_per_task": 10064.6375,
      "last_run_timestamp": "2026-10-10T19:51:36.138Z"
    },
    {
      "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": 0.7426966853932584,
      "accuracy_std_dev": 0.023739075903780084,
      "run_count": 9,
      "total_questions": 445,
      "avg_cost_per_task_usd": 0.022410094382022478,
      "avg_duration_per_task_ms": 289993.2921348315,
      "avg_output_tokens_per_task": 16658.757303370785,
      "last_run_timestamp": "2026-10-10T11:23:23.870Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "AkashML",
      "accuracy": 0.7426668,
      "accuracy_std_dev": 0.03690245640821219,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.011645404,
      "avg_duration_per_task_ms": 585840.916,
      "avg_output_tokens_per_task": 18794.172,
      "last_run_timestamp": "2026-08-29T08:43:07.210Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Phala",
      "accuracy": 0.7426666000000001,
      "accuracy_std_dev": 0.021746139377829814,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.024568152000000003,
      "avg_duration_per_task_ms": 145960.132,
      "avg_output_tokens_per_task": 17370.136,
      "last_run_timestamp": "2026-10-10T11:10:23.832Z"
    },
    {
      "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": 0.7425,
      "accuracy_std_dev": 0.020258742970628775,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.010828945,
      "avg_duration_per_task_ms": 216783.6,
      "avg_output_tokens_per_task": 14322.1425,
      "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": "qwen/qwen3.5-122b-a10b-20260224",
      "model_slug": "qwen/qwen3.5-122b-a10b",
      "model_name": "Qwen: Qwen3.5-122B-A10B",
      "provider_name": "Alibaba",
      "accuracy": 0.7423934827586207,
      "accuracy_std_dev": 0.04075671767696707,
      "run_count": 10,
      "total_questions": 493,
      "avg_cost_per_task_usd": 0.11995382961460448,
      "avg_duration_per_task_ms": 491199.4949290061,
      "avg_output_tokens_per_task": 23541.371196754564,
      "last_run_timestamp": "2026-10-03T08:18:33.162Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Friendli",
      "accuracy": 0.742293063583815,
      "accuracy_std_dev": 0.017156108636605785,
      "run_count": 7,
      "total_questions": 346,
      "avg_cost_per_task_usd": 0.019260367052023122,
      "avg_duration_per_task_ms": 95216.28323699422,
      "avg_output_tokens_per_task": 4012.3959537572255,
      "last_run_timestamp": "2026-10-10T11:31:00.690Z"
    },
    {
      "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": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "auto-routing",
      "accuracy": 0.74222225,
      "accuracy_std_dev": 0.028458392848522442,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.023521068333333332,
      "avg_duration_per_task_ms": 166328.99,
      "avg_output_tokens_per_task": 7628.006666666667,
      "last_run_timestamp": "2026-10-10T18:45:14.922Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "Google",
      "accuracy": 0.742222,
      "accuracy_std_dev": 0.015713539024251252,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.01313277,
      "avg_duration_per_task_ms": 70981.14,
      "avg_output_tokens_per_task": 10941.743333333334,
      "last_run_timestamp": "2026-10-10T11:35:03.364Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "CoreWeave",
      "accuracy": 0.742222,
      "accuracy_std_dev": 0.020608136952831695,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.044708606666666664,
      "avg_duration_per_task_ms": 149318.53333333333,
      "avg_output_tokens_per_task": 38342.77333333333,
      "last_run_timestamp": "2026-10-10T11:04:02.145Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.7422219999999999,
      "accuracy_std_dev": 0.036243279496572404,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.06723134,
      "avg_duration_per_task_ms": 1036437.88,
      "avg_output_tokens_per_task": 42501.62,
      "last_run_timestamp": "2026-10-10T13:56:29.042Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Fireworks",
      "accuracy": 0.7421700402684563,
      "accuracy_std_dev": 0.02926487036106602,
      "run_count": 6,
      "total_questions": 298,
      "avg_cost_per_task_usd": 0.06028264093959732,
      "avg_duration_per_task_ms": 213032.25167785236,
      "avg_output_tokens_per_task": 32783.35906040268,
      "last_run_timestamp": "2026-10-10T11:29:57.297Z"
    },
    {
      "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": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "InferenceNet",
      "accuracy": 0.7420002,
      "accuracy_std_dev": 0.025612585448564165,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.015137903999999999,
      "avg_duration_per_task_ms": 534442.24,
      "avg_output_tokens_per_task": 22548.816,
      "last_run_timestamp": "2026-10-10T11:30:00.933Z"
    },
    {
      "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": 0.7420002,
      "accuracy_std_dev": 0.031874700521887275,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.014078051999999999,
      "avg_duration_per_task_ms": 39974.624,
      "avg_output_tokens_per_task": 3995.512,
      "last_run_timestamp": "2026-10-10T11:44:25.722Z"
    },
    {
      "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": 0.7420000999999999,
      "accuracy_std_dev": 0.01606935109112998,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.015730394,
      "avg_duration_per_task_ms": 180471.802,
      "avg_output_tokens_per_task": 4250.502,
      "last_run_timestamp": "2026-10-10T11:54:55.212Z"
    },
    {
      "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": 0.7419048571428571,
      "accuracy_std_dev": 0.03181571083343471,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.07701604857142856,
      "avg_duration_per_task_ms": 557828.82,
      "avg_output_tokens_per_task": 11690.26,
      "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": "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": 0.020933986849907486,
      "run_count": 9,
      "total_questions": 446,
      "avg_cost_per_task_usd": 0.11384247085201793,
      "avg_duration_per_task_ms": 107222.65470852017,
      "avg_output_tokens_per_task": 10059.766816143498,
      "last_run_timestamp": "2026-10-10T18:37:34.818Z"
    },
    {
      "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": 0.7416667499999999,
      "accuracy_std_dev": 0.02843115955140591,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.29442077,
      "avg_duration_per_task_ms": 504048.0925,
      "avg_output_tokens_per_task": 9618.08,
      "last_run_timestamp": "2026-10-10T21:46:04.155Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "Azure",
      "accuracy": 0.7416665,
      "accuracy_std_dev": 0.011902170673872925,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.012876185,
      "avg_duration_per_task_ms": 78684.405,
      "avg_output_tokens_per_task": 10797.005,
      "last_run_timestamp": "2026-10-10T11:31:34.924Z"
    },
    {
      "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": 0.7416665,
      "accuracy_std_dev": 0.008660350267165872,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.34675104999999995,
      "avg_duration_per_task_ms": 129028.095,
      "avg_output_tokens_per_task": 7166.795,
      "last_run_timestamp": "2026-10-10T11:05:10.528Z"
    },
    {
      "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": 0.7416106107382551,
      "accuracy_std_dev": 0.01944857756081121,
      "run_count": 6,
      "total_questions": 298,
      "avg_cost_per_task_usd": 0.04701292617449664,
      "avg_duration_per_task_ms": 840147.7315436242,
      "avg_output_tokens_per_task": 14235.224832214764,
      "last_run_timestamp": "2026-10-10T19:13:25.278Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "auto-routing",
      "accuracy": 0.7414813333333333,
      "accuracy_std_dev": 0.021029672681977533,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.07095983555555556,
      "avg_duration_per_task_ms": 335910.37111111113,
      "avg_output_tokens_per_task": 12802.966666666667,
      "last_run_timestamp": "2026-10-10T20:35:57.442Z"
    },
    {
      "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": 0.7412059597989951,
      "accuracy_std_dev": 0.030598471473006233,
      "run_count": 8,
      "total_questions": 398,
      "avg_cost_per_task_usd": 0.09609109798994975,
      "avg_duration_per_task_ms": 367126.6331658292,
      "avg_output_tokens_per_task": 11257.263819095477,
      "last_run_timestamp": "2026-10-10T19:52:16.177Z"
    },
    {
      "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": 0.741147673992674,
      "accuracy_std_dev": 0.022891877368502254,
      "run_count": 11,
      "total_questions": 546,
      "avg_cost_per_task_usd": 0.14074178937728937,
      "avg_duration_per_task_ms": 591418.4395604396,
      "avg_output_tokens_per_task": 27023.639194139196,
      "last_run_timestamp": "2026-10-10T19:09:43.739Z"
    },
    {
      "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": 0.7410256153846154,
      "accuracy_std_dev": 0.029911083221231823,
      "run_count": 13,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.06005055230769231,
      "avg_duration_per_task_ms": 271297.1076923077,
      "avg_output_tokens_per_task": 9692.147692307692,
      "last_run_timestamp": "2026-10-10T11:37:01.698Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "GMICloud",
      "accuracy": 0.7407407777777778,
      "accuracy_std_dev": 0.0330273719470441,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.1090063111111111,
      "avg_duration_per_task_ms": 324070.6777777778,
      "avg_output_tokens_per_task": 18436.48,
      "last_run_timestamp": "2026-10-10T13:19:36.237Z"
    },
    {
      "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": 0.03670968370812553,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.046896531111111105,
      "avg_duration_per_task_ms": 552618.62,
      "avg_output_tokens_per_task": 9268.293333333333,
      "last_run_timestamp": "2026-10-10T22:42:55.824Z"
    },
    {
      "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": "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": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7402943534136547,
      "accuracy_std_dev": 0.02302499266101948,
      "run_count": 5,
      "total_questions": 249,
      "avg_cost_per_task_usd": 0.06053981927710843,
      "avg_duration_per_task_ms": 161020.49799196786,
      "avg_output_tokens_per_task": 6902.349397590361,
      "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-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": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "ModelRun",
      "accuracy": 0.7400000909090909,
      "accuracy_std_dev": 0.025584014897284242,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.0668473418181818,
      "avg_duration_per_task_ms": 146402.66363636364,
      "avg_output_tokens_per_task": 9276.552727272727,
      "last_run_timestamp": "2026-10-10T18:47:02.691Z"
    },
    {
      "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": "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": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "Together",
      "accuracy": 0.74,
      "accuracy_std_dev": 0.01805352125394564,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.22917437000000002,
      "avg_duration_per_task_ms": 262336.34,
      "avg_output_tokens_per_task": 12064.356666666667,
      "last_run_timestamp": "2026-10-10T11:08:41.912Z"
    },
    {
      "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-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Chutes",
      "accuracy": 0.74,
      "accuracy_std_dev": 0.0354338036978815,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.09767998,
      "avg_duration_per_task_ms": 450597.1575,
      "avg_output_tokens_per_task": 9690.325,
      "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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "WandB",
      "accuracy": 0.74,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.12237688000000001,
      "avg_duration_per_task_ms": 90244.42,
      "avg_output_tokens_per_task": 9936.14,
      "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-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.74,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.03319066,
      "avg_duration_per_task_ms": 77162.14,
      "avg_output_tokens_per_task": 4453.24,
      "last_run_timestamp": "2026-10-04T03:00:16.404Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Morph",
      "accuracy": 0.7399998571428572,
      "accuracy_std_dev": 0.023094052007442304,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.01886730285714286,
      "avg_duration_per_task_ms": 206247.68,
      "avg_output_tokens_per_task": 4550.928571428572,
      "last_run_timestamp": "2026-10-10T11:50:34.352Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Inceptron",
      "accuracy": 0.7398129458917835,
      "accuracy_std_dev": 0.024891144232638256,
      "run_count": 10,
      "total_questions": 499,
      "avg_cost_per_task_usd": 0.011375460921843689,
      "avg_duration_per_task_ms": 475329.3386773547,
      "avg_output_tokens_per_task": 14172.182364729459,
      "last_run_timestamp": "2026-10-10T11:52:20.821Z"
    },
    {
      "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": 0.7396491578947368,
      "accuracy_std_dev": 0.01866585079556883,
      "run_count": 19,
      "total_questions": 950,
      "avg_cost_per_task_usd": 0.47321189052631585,
      "avg_duration_per_task_ms": 648402.0189473684,
      "avg_output_tokens_per_task": 9813.902105263158,
      "last_run_timestamp": "2026-10-10T11:08:03.918Z"
    },
    {
      "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": 0.739166625,
      "accuracy_std_dev": 0.031523896360291105,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.5017313675,
      "avg_duration_per_task_ms": 190760.71,
      "avg_output_tokens_per_task": 9915.5125,
      "last_run_timestamp": "2026-10-10T11:44:47.020Z"
    },
    {
      "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": 0.7391303686200378,
      "accuracy_std_dev": 0.02855825218538939,
      "run_count": 11,
      "total_questions": 529,
      "avg_cost_per_task_usd": 0.020223423440453688,
      "avg_duration_per_task_ms": 99917.1190926276,
      "avg_output_tokens_per_task": 6754.173913043478,
      "last_run_timestamp": "2026-10-10T18:46:12.945Z"
    },
    {
      "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": 0.7390713237704918,
      "accuracy_std_dev": 0.02181217991306695,
      "run_count": 5,
      "total_questions": 244,
      "avg_cost_per_task_usd": 0.12843605327868854,
      "avg_duration_per_task_ms": 201024.64344262294,
      "avg_output_tokens_per_task": 10510.745901639344,
      "last_run_timestamp": "2026-10-10T19:21:45.805Z"
    },
    {
      "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": 0.7390475714285715,
      "accuracy_std_dev": 0.026290010003025565,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.43027001428571426,
      "avg_duration_per_task_ms": 716198.2485714286,
      "avg_output_tokens_per_task": 9746.097142857143,
      "last_run_timestamp": "2026-10-10T11:29:35.890Z"
    },
    {
      "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": 0.7390474285714286,
      "accuracy_std_dev": 0.01965692829409479,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.014954228571428571,
      "avg_duration_per_task_ms": 79171.07714285715,
      "avg_output_tokens_per_task": 4448.311428571428,
      "last_run_timestamp": "2026-10-10T11:22:45.589Z"
    },
    {
      "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": "anthropic/claude-opus-5.5-20260921",
      "model_slug": "anthropic/claude-opus-5.5",
      "model_name": "Anthropic: Claude Opus 5.5",
      "provider_name": "Google",
      "accuracy": 0.7388888333333333,
      "accuracy_std_dev": 0.01383327086431679,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.35860266999999996,
      "avg_duration_per_task_ms": 86556.65666666666,
      "avg_output_tokens_per_task": 7118.746666666667,
      "last_run_timestamp": "2026-10-10T11:09:18.472Z"
    },
    {
      "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": 0.7388888333333333,
      "accuracy_std_dev": 0.0319528259544007,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.02934656666666667,
      "avg_duration_per_task_ms": 451531.085,
      "avg_output_tokens_per_task": 14709.348333333333,
      "last_run_timestamp": "2026-10-10T13:27:05.588Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.4-20260305",
      "model_slug": "openai/gpt-5.4",
      "model_name": "OpenAI: GPT-5.4",
      "provider_name": "Azure",
      "accuracy": 0.7385184444444445,
      "accuracy_std_dev": 0.017992647486688756,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.3103866155555556,
      "avg_duration_per_task_ms": 191198.25333333333,
      "avg_output_tokens_per_task": 10879.702222222222,
      "last_run_timestamp": "2026-10-01T19:05:18.005Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Fireworks",
      "accuracy": 0.7384840804020101,
      "accuracy_std_dev": 0.03165357517370423,
      "run_count": 16,
      "total_questions": 796,
      "avg_cost_per_task_usd": 0.10895770226130652,
      "avg_duration_per_task_ms": 110251.2148241206,
      "avg_output_tokens_per_task": 8058.163316582914,
      "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": "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": 0.73833325,
      "accuracy_std_dev": 0.027436952982018623,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.35042460250000007,
      "avg_duration_per_task_ms": 221570.3825,
      "avg_output_tokens_per_task": 10876.18,
      "last_run_timestamp": "2026-10-10T11:29:08.503Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "AkashML",
      "accuracy": 0.7382865983935742,
      "accuracy_std_dev": 0.027461257286584682,
      "run_count": 5,
      "total_questions": 249,
      "avg_cost_per_task_usd": 0.07498479919678715,
      "avg_duration_per_task_ms": 91080.29718875502,
      "avg_output_tokens_per_task": 4367.955823293173,
      "last_run_timestamp": "2026-09-26T08:11:26.495Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "Novita",
      "accuracy": 0.7377779166666666,
      "accuracy_std_dev": 0.037940818112903,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.009713255,
      "avg_duration_per_task_ms": 395149.92,
      "avg_output_tokens_per_task": 14577.848333333333,
      "last_run_timestamp": "2026-10-10T14:25:02.047Z"
    },
    {
      "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": 0.7377778333333334,
      "accuracy_std_dev": 0.02484511527723086,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.06635720666666667,
      "avg_duration_per_task_ms": 282495.5133333333,
      "avg_output_tokens_per_task": 8158.293333333333,
      "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": "AtlasCloud",
      "accuracy": 0.7373332,
      "accuracy_std_dev": 0.03690245640821219,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.028322468,
      "avg_duration_per_task_ms": 157393.044,
      "avg_output_tokens_per_task": 16186.18,
      "last_run_timestamp": "2026-10-10T11:38:15.210Z"
    },
    {
      "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": 0.7372221666666667,
      "accuracy_std_dev": 0.02588554461095142,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.02299656,
      "avg_duration_per_task_ms": 397547.93166666664,
      "avg_output_tokens_per_task": 12291.753333333334,
      "last_run_timestamp": "2026-10-10T14:27:59.534Z"
    },
    {
      "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": 0.7371428571428571,
      "accuracy_std_dev": 0.038190332503054716,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.34757304285714286,
      "avg_duration_per_task_ms": 395834.59714285715,
      "avg_output_tokens_per_task": 9470.625714285714,
      "last_run_timestamp": "2026-10-10T11:27:02.343Z"
    },
    {
      "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": 0.7369697272727272,
      "accuracy_std_dev": 0.10754797319428362,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.14869326363636362,
      "avg_duration_per_task_ms": 623801.8454545455,
      "avg_output_tokens_per_task": 27614.49090909091,
      "last_run_timestamp": "2026-10-10T11:09:25.323Z"
    },
    {
      "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": 0.7368726405570062,
      "accuracy_std_dev": 0.02640283262471465,
      "run_count": 23,
      "total_questions": 1149,
      "avg_cost_per_task_usd": 0.12460780939947784,
      "avg_duration_per_task_ms": 171402.6083550914,
      "avg_output_tokens_per_task": 9190.991296779808,
      "last_run_timestamp": "2026-10-10T11:58:34.689Z"
    },
    {
      "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": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "StreamLake",
      "accuracy": 0.7366668333333333,
      "accuracy_std_dev": 0.033054360390205026,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.016296663333333333,
      "avg_duration_per_task_ms": 149274.76666666666,
      "avg_output_tokens_per_task": 15030.68,
      "last_run_timestamp": "2026-10-10T11:18:50.119Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Makora",
      "accuracy": 0.7366668,
      "accuracy_std_dev": 0.024221175309220654,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.04493056000000001,
      "avg_duration_per_task_ms": 138671.204,
      "avg_output_tokens_per_task": 28727.112,
      "last_run_timestamp": "2026-10-10T11:07:22.568Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "Venice",
      "accuracy": 0.7366668,
      "accuracy_std_dev": 0.022900761755015906,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.063241744,
      "avg_duration_per_task_ms": 1530537.112,
      "avg_output_tokens_per_task": 25222.716,
      "last_run_timestamp": "2026-10-03T08:06:43.536Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "Novita",
      "accuracy": 0.7366667499999999,
      "accuracy_std_dev": 0.029814193116492343,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.049640665,
      "avg_duration_per_task_ms": 368265.4225,
      "avg_output_tokens_per_task": 10154.2325,
      "last_run_timestamp": "2026-10-10T20:13:30.041Z"
    },
    {
      "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": 0.7366665833333333,
      "accuracy_std_dev": 0.04004635046097278,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.09680749833333332,
      "avg_duration_per_task_ms": 150044.80833333332,
      "avg_output_tokens_per_task": 9307.375,
      "last_run_timestamp": "2026-10-10T11:40:53.090Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Fireworks",
      "accuracy": 0.7366665,
      "accuracy_std_dev": 0.01666650000000003,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.17216302,
      "avg_duration_per_task_ms": 234612.69,
      "avg_output_tokens_per_task": 17080.8,
      "last_run_timestamp": "2026-09-19T08:12:02.180Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Phala",
      "accuracy": 0.7366665,
      "accuracy_std_dev": 0.0033334999999999892,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.020005850000000002,
      "avg_duration_per_task_ms": 329585.74,
      "avg_output_tokens_per_task": 11689.47,
      "last_run_timestamp": "2026-08-29T16:43:42.382Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Modal",
      "accuracy": 0.7362964444444444,
      "accuracy_std_dev": 0.017808576717170822,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.38587963333333336,
      "avg_duration_per_task_ms": 178574.04,
      "avg_output_tokens_per_task": 10909.295555555556,
      "last_run_timestamp": "2026-10-10T11:25:32.854Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Inceptron",
      "accuracy": 0.7362542886597938,
      "accuracy_std_dev": 0.030839009988703172,
      "run_count": 12,
      "total_questions": 582,
      "avg_cost_per_task_usd": 0.058931000000000004,
      "avg_duration_per_task_ms": 124849.64604810996,
      "avg_output_tokens_per_task": 7792.945017182131,
      "last_run_timestamp": "2026-10-10T11:39:50.484Z"
    },
    {
      "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": 0.73625,
      "accuracy_std_dev": 0.03453823565629835,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0753320925,
      "avg_duration_per_task_ms": 225286.185,
      "avg_output_tokens_per_task": 11323.5925,
      "last_run_timestamp": "2026-10-10T18:54:49.288Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "CoreWeave",
      "accuracy": 0.736,
      "accuracy_std_dev": 0.014360191892868317,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.021830644,
      "avg_duration_per_task_ms": 202173.696,
      "avg_output_tokens_per_task": 8712.768,
      "last_run_timestamp": "2026-10-10T18:44:32.341Z"
    },
    {
      "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": 0.736,
      "accuracy_std_dev": 0.03542738426133094,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.42442627200000005,
      "avg_duration_per_task_ms": 39355.976,
      "avg_output_tokens_per_task": 9202.272,
      "last_run_timestamp": "2026-08-15T14:16:11.865Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Phala",
      "accuracy": 0.7356322298850575,
      "accuracy_std_dev": 0.05623144531563914,
      "run_count": 7,
      "total_questions": 348,
      "avg_cost_per_task_usd": 0.29622710057471263,
      "avg_duration_per_task_ms": 387662.2270114943,
      "avg_output_tokens_per_task": 10990.514367816091,
      "last_run_timestamp": "2026-10-10T11:22:20.552Z"
    },
    {
      "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": "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": 0.007911693333333334,
      "avg_duration_per_task_ms": 368979.4766666667,
      "avg_output_tokens_per_task": 14227.998333333333,
      "last_run_timestamp": "2026-10-10T13:14:11.957Z"
    },
    {
      "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": 0.7354429620253164,
      "accuracy_std_dev": 0.021701362929548943,
      "run_count": 8,
      "total_questions": 395,
      "avg_cost_per_task_usd": 0.020455908860759497,
      "avg_duration_per_task_ms": 288283.3063291139,
      "avg_output_tokens_per_task": 10845.830379746836,
      "last_run_timestamp": "2026-10-10T11:51:18.823Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "GMICloud",
      "accuracy": 0.7354165,
      "accuracy_std_dev": 0.0429125354005097,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.006874592499999999,
      "avg_duration_per_task_ms": 421319.53,
      "avg_output_tokens_per_task": 14336.97,
      "last_run_timestamp": "2026-10-10T15:31:23.439Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "SiliconFlow",
      "accuracy": 0.735,
      "accuracy_std_dev": 0.031578860722008335,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.05864864000000001,
      "avg_duration_per_task_ms": 688996.4525,
      "avg_output_tokens_per_task": 14948.475,
      "last_run_timestamp": "2026-10-10T19:35:42.927Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Baidu",
      "accuracy": 0.735,
      "accuracy_std_dev": 0.011902170673872925,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.07856429000000001,
      "avg_duration_per_task_ms": 285372.9925,
      "avg_output_tokens_per_task": 13300.165,
      "last_run_timestamp": "2026-10-10T19:11:19.715Z"
    },
    {
      "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": 0.73499995,
      "accuracy_std_dev": 0.024843857950155427,
      "run_count": 20,
      "total_questions": 1000,
      "avg_cost_per_task_usd": 0.36867235800000003,
      "avg_duration_per_task_ms": 230064.043,
      "avg_output_tokens_per_task": 10896.078,
      "last_run_timestamp": "2026-10-10T11:35:23.046Z"
    },
    {
      "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": 0.029767504188449777,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.1267184125,
      "avg_duration_per_task_ms": 254157.8425,
      "avg_output_tokens_per_task": 9997.145,
      "last_run_timestamp": "2026-10-10T20:45:18.917Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.5-flash-lite-20260721",
      "model_slug": "google/gemini-3.5-flash-lite",
      "model_name": "Google: Gemini 3.5 Flash Lite",
      "provider_name": "Google AI Studio",
      "accuracy": 0.7349997500000001,
      "accuracy_std_dev": 0.002886895683515417,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.12295462999999998,
      "avg_duration_per_task_ms": 89986.35,
      "avg_output_tokens_per_task": 23350.75,
      "last_run_timestamp": "2026-10-01T19:05:46.177Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7349397469879518,
      "accuracy_std_dev": 0.035884660088817896,
      "run_count": 10,
      "total_questions": 498,
      "avg_cost_per_task_usd": 0.08050074096385541,
      "avg_duration_per_task_ms": 752519.5602409638,
      "avg_output_tokens_per_task": 14439.815261044178,
      "last_run_timestamp": "2026-10-10T21:54:37.922Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.8-flash-20260902",
      "model_slug": "google/gemini-3.8-flash",
      "model_name": "Google: Gemini 3.8 Flash",
      "provider_name": "Google AI Studio",
      "accuracy": 0.734694,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 49,
      "avg_cost_per_task_usd": 0.25339910204081634,
      "avg_duration_per_task_ms": 146531.16326530612,
      "avg_output_tokens_per_task": 29831.102040816328,
      "last_run_timestamp": "2026-10-01T08:03:06.861Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Minimax",
      "accuracy": 0.7344444166666667,
      "accuracy_std_dev": 0.024770657922423504,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.023368726666666666,
      "avg_duration_per_task_ms": 89065.79833333334,
      "avg_output_tokens_per_task": 7629.631666666667,
      "last_run_timestamp": "2026-10-10T18:55:15.771Z"
    },
    {
      "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": 0.7344444166666667,
      "accuracy_std_dev": 0.025507201840716597,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.07600889333333334,
      "avg_duration_per_task_ms": 184831.82166666666,
      "avg_output_tokens_per_task": 8645.84,
      "last_run_timestamp": "2026-10-10T11:53:30.354Z"
    },
    {
      "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": 0.7344443333333334,
      "accuracy_std_dev": 0.023543870268548078,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.3636283366666667,
      "avg_duration_per_task_ms": 194252.73333333334,
      "avg_output_tokens_per_task": 8764.83,
      "last_run_timestamp": "2026-10-10T11:18:57.064Z"
    },
    {
      "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": 0.7342857142857143,
      "accuracy_std_dev": 0.05044910324479597,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.17247626285714285,
      "avg_duration_per_task_ms": 282624.91142857145,
      "avg_output_tokens_per_task": 9879.474285714286,
      "last_run_timestamp": "2026-10-10T11:32:26.500Z"
    },
    {
      "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": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7338218205128205,
      "accuracy_std_dev": 0.06370553922722423,
      "run_count": 11,
      "total_questions": 546,
      "avg_cost_per_task_usd": 0.15103331501831502,
      "avg_duration_per_task_ms": 1333108.2106227106,
      "avg_output_tokens_per_task": 27702.578754578753,
      "last_run_timestamp": "2026-10-10T21:21:31.949Z"
    },
    {
      "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": 0.7334079284116332,
      "accuracy_std_dev": 0.030021614385655343,
      "run_count": 9,
      "total_questions": 447,
      "avg_cost_per_task_usd": 0.07689493288590601,
      "avg_duration_per_task_ms": 215056.84116331095,
      "avg_output_tokens_per_task": 8337.460850111856,
      "last_run_timestamp": "2026-10-10T11:38:34.891Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.5-flash-20260519",
      "model_slug": "google/gemini-3.5-flash",
      "model_name": "Google: Gemini 3.5 Flash",
      "provider_name": "Google AI Studio",
      "accuracy": 0.7333336666666668,
      "accuracy_std_dev": 0.03771236166328251,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.44470873333333333,
      "avg_duration_per_task_ms": 130557.96,
      "avg_output_tokens_per_task": 20380.893333333333,
      "last_run_timestamp": "2026-10-01T19:04:52.658Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "StreamLake",
      "accuracy": 0.7333335714285714,
      "accuracy_std_dev": 0.023904731551109535,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.019937122857142855,
      "avg_duration_per_task_ms": 479765.5571428572,
      "avg_output_tokens_per_task": 14687.234285714285,
      "last_run_timestamp": "2026-08-29T16:31:42.782Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Decart",
      "accuracy": 0.7333335,
      "accuracy_std_dev": 0.033333500000000016,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.009175609999999999,
      "avg_duration_per_task_ms": 106003.625,
      "avg_output_tokens_per_task": 5053.315,
      "last_run_timestamp": "2026-10-10T11:59:02.396Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Decart",
      "accuracy": 0.7333333333333333,
      "accuracy_std_dev": 0.005443174461857933,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.008480786666666665,
      "avg_duration_per_task_ms": 397846.0266666667,
      "avg_output_tokens_per_task": 19127.326666666668,
      "last_run_timestamp": "2026-08-22T08:38:00.859Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Venice",
      "accuracy": 0.7333333333333333,
      "accuracy_std_dev": 0.05192585882668569,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.12086406999999999,
      "avg_duration_per_task_ms": 154571.83,
      "avg_output_tokens_per_task": 9530.663333333334,
      "last_run_timestamp": "2026-10-10T12:38:40.745Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "InferenceNet",
      "accuracy": 0.7333333333333333,
      "accuracy_std_dev": 0.05760637824832321,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.04898614,
      "avg_duration_per_task_ms": 106829.63333333333,
      "avg_output_tokens_per_task": 9897.326666666666,
      "last_run_timestamp": "2026-10-10T11:51:48.927Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-sol-20260709",
      "model_slug": "openai/gpt-5.6-sol",
      "model_name": "OpenAI: GPT-5.6 Sol",
      "provider_name": "OpenAI",
      "accuracy": 0.73333325,
      "accuracy_std_dev": 0.012472414008021857,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.17039939499999998,
      "avg_duration_per_task_ms": 114141.445,
      "avg_output_tokens_per_task": 5083.04,
      "last_run_timestamp": "2026-10-02T23:08:37.533Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "NextBit",
      "accuracy": 0.7333330000000001,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01186947,
      "avg_duration_per_task_ms": 176514.74,
      "avg_output_tokens_per_task": 14201.61,
      "last_run_timestamp": "2026-08-29T16:50:07.898Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7333330000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.058005,
      "avg_duration_per_task_ms": 191243.62,
      "avg_output_tokens_per_task": 30812.32,
      "last_run_timestamp": "2026-09-12T08:09:12.993Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.2-20251211",
      "model_slug": "openai/gpt-5.2",
      "model_name": "OpenAI: GPT-5.2",
      "provider_name": "auto-routing",
      "accuracy": 0.7333330000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.2412942,
      "avg_duration_per_task_ms": 207134.84,
      "avg_output_tokens_per_task": 12538.34,
      "last_run_timestamp": "2026-08-01T08:06:12.346Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "ModelRun",
      "accuracy": 0.7328571428571429,
      "accuracy_std_dev": 0.02503509736024079,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.06745117142857142,
      "avg_duration_per_task_ms": 76562.92857142857,
      "avg_output_tokens_per_task": 8496.594285714285,
      "last_run_timestamp": "2026-09-26T08:08:19.642Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Moonshot AI",
      "accuracy": 0.7326620894854586,
      "accuracy_std_dev": 0.026574598734064515,
      "run_count": 9,
      "total_questions": 447,
      "avg_cost_per_task_usd": 0.36663018791946306,
      "avg_duration_per_task_ms": 400926.4272930649,
      "avg_output_tokens_per_task": 9409.45413870246,
      "last_run_timestamp": "2026-10-10T11:29:27.268Z"
    },
    {
      "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": 0.7322223333333333,
      "accuracy_std_dev": 0.035468537615125265,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.022583826666666664,
      "avg_duration_per_task_ms": 117756.11,
      "avg_output_tokens_per_task": 7162.1,
      "last_run_timestamp": "2026-10-10T19:14:49.200Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "SambaNova",
      "accuracy": 0.7319933969849246,
      "accuracy_std_dev": 0.06840954592516678,
      "run_count": 4,
      "total_questions": 199,
      "avg_cost_per_task_usd": 0.07448894472361808,
      "avg_duration_per_task_ms": 91970.93467336684,
      "avg_output_tokens_per_task": 13988.37688442211,
      "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": "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": "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": 0.7315942608695652,
      "accuracy_std_dev": 0.033605750226408763,
      "run_count": 23,
      "total_questions": 1150,
      "avg_cost_per_task_usd": 0.35594851652173914,
      "avg_duration_per_task_ms": 206193.49739130435,
      "avg_output_tokens_per_task": 10313.84,
      "last_run_timestamp": "2026-10-10T11:22:45.965Z"
    },
    {
      "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": 0.731293,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 49,
      "avg_cost_per_task_usd": 0.12389479591836736,
      "avg_duration_per_task_ms": 269993.51020408166,
      "avg_output_tokens_per_task": 23771,
      "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": "openai/gpt-5.6-terra-20260709",
      "model_slug": "openai/gpt-5.6-terra",
      "model_name": "OpenAI: GPT-5.6 Terra",
      "provider_name": "auto-routing",
      "accuracy": 0.7311113333333332,
      "accuracy_std_dev": 0.011331263428624758,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.10536965333333334,
      "avg_duration_per_task_ms": 88766.42666666667,
      "avg_output_tokens_per_task": 4017.0533333333333,
      "last_run_timestamp": "2026-10-01T19:04:40.360Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.5-flash-20260519",
      "model_slug": "google/gemini-3.5-flash",
      "model_name": "Google: Gemini 3.5 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.7311110000000001,
      "accuracy_std_dev": 0.022662199996175744,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.3988358066666667,
      "avg_duration_per_task_ms": 172929.82666666666,
      "avg_output_tokens_per_task": 27148.5,
      "last_run_timestamp": "2026-10-01T19:04:40.362Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "SambaNova",
      "accuracy": 0.7311110000000001,
      "accuracy_std_dev": 0.013698950130089036,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.16557658,
      "avg_duration_per_task_ms": 484024.5333333333,
      "avg_output_tokens_per_task": 29783.94,
      "last_run_timestamp": "2026-10-10T19:22:43.594Z"
    },
    {
      "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": 0.7302118929765886,
      "accuracy_std_dev": 0.012689587626825722,
      "run_count": 6,
      "total_questions": 299,
      "avg_cost_per_task_usd": 0.2335497525083612,
      "avg_duration_per_task_ms": 376307.7892976589,
      "avg_output_tokens_per_task": 9829.47491638796,
      "last_run_timestamp": "2026-10-10T11:21:19.129Z"
    },
    {
      "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": 0.7301412182628062,
      "accuracy_std_dev": 0.02931592753958457,
      "run_count": 9,
      "total_questions": 449,
      "avg_cost_per_task_usd": 0.06766333184855233,
      "avg_duration_per_task_ms": 212965.04008908686,
      "avg_output_tokens_per_task": 11590.710467706014,
      "last_run_timestamp": "2026-10-10T19:18:15.537Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Crusoe",
      "accuracy": 0.7300001666666667,
      "accuracy_std_dev": 0.03605552816336059,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.11662036,
      "avg_duration_per_task_ms": 82058.93666666666,
      "avg_output_tokens_per_task": 9160.97,
      "last_run_timestamp": "2026-09-12T08:12:28.966Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Ambient",
      "accuracy": 0.730000125,
      "accuracy_std_dev": 0.04988867885461965,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.07825786500000001,
      "avg_duration_per_task_ms": 212986.7025,
      "avg_output_tokens_per_task": 8293.9625,
      "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-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Parasail",
      "accuracy": 0.7299998000000001,
      "accuracy_std_dev": 0.02290067442150996,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.045081952,
      "avg_duration_per_task_ms": 177241.02,
      "avg_output_tokens_per_task": 31491.432,
      "last_run_timestamp": "2026-10-10T11:06:58.135Z"
    },
    {
      "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": 0.7296296666666666,
      "accuracy_std_dev": 0.012216782182264232,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.25703444444444445,
      "avg_duration_per_task_ms": 284813.7555555556,
      "avg_output_tokens_per_task": 10375.755555555555,
      "last_run_timestamp": "2026-10-10T11:00:29.672Z"
    },
    {
      "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": 0.728889,
      "accuracy_std_dev": 0.020608136952831695,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.06727646666666667,
      "avg_duration_per_task_ms": 501138.4733333333,
      "avg_output_tokens_per_task": 9493.433333333332,
      "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-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "Novita",
      "accuracy": 0.728889,
      "accuracy_std_dev": 0.022443443323657393,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.05784399555555556,
      "avg_duration_per_task_ms": 318648.68,
      "avg_output_tokens_per_task": 9856.32,
      "last_run_timestamp": "2026-10-10T20:00:56.451Z"
    },
    {
      "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": 0.7288888333333333,
      "accuracy_std_dev": 0.025867535305969056,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.07605682166666666,
      "avg_duration_per_task_ms": 645935.9233333333,
      "avg_output_tokens_per_task": 8649.321666666667,
      "last_run_timestamp": "2026-10-10T11:19:35.083Z"
    },
    {
      "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": 0.7288603933823529,
      "accuracy_std_dev": 0.04323775063491326,
      "run_count": 11,
      "total_questions": 544,
      "avg_cost_per_task_usd": 0.021905998161764705,
      "avg_duration_per_task_ms": 100677.26838235294,
      "avg_output_tokens_per_task": 7100.75,
      "last_run_timestamp": "2026-10-10T19:08:09.142Z"
    },
    {
      "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": 0.728,
      "accuracy_std_dev": 0.027776952878240627,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.04150964,
      "avg_duration_per_task_ms": 140081.308,
      "avg_output_tokens_per_task": 18264.704,
      "last_run_timestamp": "2026-10-10T11:28:56.274Z"
    },
    {
      "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": 0.7279999,
      "accuracy_std_dev": 0.03448671628453485,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.30246312399999997,
      "avg_duration_per_task_ms": 332768.642,
      "avg_output_tokens_per_task": 9378.478,
      "last_run_timestamp": "2026-10-10T11:50:42.586Z"
    },
    {
      "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": 0.7278787272727273,
      "accuracy_std_dev": 0.027970785626470766,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.02821320545454546,
      "avg_duration_per_task_ms": 208497.22363636363,
      "avg_output_tokens_per_task": 8992.374545454546,
      "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": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "DeepInfra",
      "accuracy": 0.7278566673427991,
      "accuracy_std_dev": 0.03572243647191496,
      "run_count": 10,
      "total_questions": 493,
      "avg_cost_per_task_usd": 0.10080718052738337,
      "avg_duration_per_task_ms": 324790.75456389453,
      "avg_output_tokens_per_task": 14787.588235294117,
      "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": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "SiliconFlow",
      "accuracy": 0.72777775,
      "accuracy_std_dev": 0.027057441185759493,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.07957451833333333,
      "avg_duration_per_task_ms": 222728.50333333333,
      "avg_output_tokens_per_task": 8680.433333333332,
      "last_run_timestamp": "2026-10-10T12:28:42.500Z"
    },
    {
      "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": 0.7274997500000001,
      "accuracy_std_dev": 0.035502329736053946,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.064519175,
      "avg_duration_per_task_ms": 370870.4875,
      "avg_output_tokens_per_task": 11564.955,
      "last_run_timestamp": "2026-10-10T19:34:22.239Z"
    },
    {
      "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": 0.7274305937499999,
      "accuracy_std_dev": 0.02334167860109178,
      "run_count": 6,
      "total_questions": 288,
      "avg_cost_per_task_usd": 0.14647149305555557,
      "avg_duration_per_task_ms": 714189.5451388889,
      "avg_output_tokens_per_task": 29008.288194444445,
      "last_run_timestamp": "2026-10-10T18:53:41.160Z"
    },
    {
      "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": 0.7273333,
      "accuracy_std_dev": 0.03495384321659065,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.33412321800000006,
      "avg_duration_per_task_ms": 205827.788,
      "avg_output_tokens_per_task": 10339.92,
      "last_run_timestamp": "2026-10-10T11:19:03.850Z"
    },
    {
      "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": 0.7273159627507164,
      "accuracy_std_dev": 0.03210912735135148,
      "run_count": 7,
      "total_questions": 349,
      "avg_cost_per_task_usd": 0.054678793696275076,
      "avg_duration_per_task_ms": 920429.4957020057,
      "avg_output_tokens_per_task": 14154.799426934098,
      "last_run_timestamp": "2026-10-10T19:33:17.084Z"
    },
    {
      "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": 0.7272728080808081,
      "accuracy_std_dev": 0.03401092910650194,
      "run_count": 4,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.12744196464646465,
      "avg_duration_per_task_ms": 652406.3232323233,
      "avg_output_tokens_per_task": 27175.661616161615,
      "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": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "Together",
      "accuracy": 0.7269664179775281,
      "accuracy_std_dev": 0.05285482315214426,
      "run_count": 9,
      "total_questions": 445,
      "avg_cost_per_task_usd": 0.061720067415730336,
      "avg_duration_per_task_ms": 941730.6539325842,
      "avg_output_tokens_per_task": 65555.35730337078,
      "last_run_timestamp": "2026-10-10T19:11:36.943Z"
    },
    {
      "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": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Mancer 2",
      "accuracy": 0.726666625,
      "accuracy_std_dev": 0.045582550268544374,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0456602475,
      "avg_duration_per_task_ms": 292041.375,
      "avg_output_tokens_per_task": 14472.8075,
      "last_run_timestamp": "2026-10-10T11:51:12.282Z"
    },
    {
      "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": 0.7266666,
      "accuracy_std_dev": 0.03155246072559159,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.3544039280000001,
      "avg_duration_per_task_ms": 237853.414,
      "avg_output_tokens_per_task": 10864.908,
      "last_run_timestamp": "2026-10-10T11:43:14.700Z"
    },
    {
      "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": 0.7266666,
      "accuracy_std_dev": 0.020221023165013202,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.044841616,
      "avg_duration_per_task_ms": 129167.684,
      "avg_output_tokens_per_task": 26225.984,
      "last_run_timestamp": "2026-10-10T11:48:30.799Z"
    },
    {
      "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": 0.7266665714285715,
      "accuracy_std_dev": 0.035989231194333456,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.11753475142857144,
      "avg_duration_per_task_ms": 678895.0285714286,
      "avg_output_tokens_per_task": 31918.602857142858,
      "last_run_timestamp": "2026-10-10T11:08:14.449Z"
    },
    {
      "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": "Relace",
      "accuracy": 0.7261110000000001,
      "accuracy_std_dev": 0.026134816662835056,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.013091720000000001,
      "avg_duration_per_task_ms": 286219.44333333336,
      "avg_output_tokens_per_task": 19785.066666666666,
      "last_run_timestamp": "2026-10-10T11:40:53.088Z"
    },
    {
      "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": 0.07136424727272728,
      "avg_duration_per_task_ms": 198968.38727272727,
      "avg_output_tokens_per_task": 8450.036363636364,
      "last_run_timestamp": "2026-10-10T12:55:32.994Z"
    },
    {
      "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": 0.725714,
      "accuracy_std_dev": 0.027004721179823374,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.04600672857142857,
      "avg_duration_per_task_ms": 172219.76,
      "avg_output_tokens_per_task": 8290.077142857142,
      "last_run_timestamp": "2026-08-29T17:36:31.073Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-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": 0.07936439,
      "avg_duration_per_task_ms": 253012.47333333333,
      "avg_output_tokens_per_task": 9156.243333333334,
      "last_run_timestamp": "2026-10-10T11:53:49.659Z"
    },
    {
      "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": 0.7254545454545455,
      "accuracy_std_dev": 0.02738706202293217,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.038856060000000005,
      "avg_duration_per_task_ms": 146486.05272727273,
      "avg_output_tokens_per_task": 8570.81090909091,
      "last_run_timestamp": "2026-10-10T11:39:07.303Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "auto-routing",
      "accuracy": 0.7251851111111112,
      "accuracy_std_dev": 0.030107077312834407,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.29754660666666666,
      "avg_duration_per_task_ms": 378777.12666666665,
      "avg_output_tokens_per_task": 9437.744444444445,
      "last_run_timestamp": "2026-10-10T11:49:11.146Z"
    },
    {
      "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": 0.7251207652173913,
      "accuracy_std_dev": 0.03173202093295975,
      "run_count": 7,
      "total_questions": 345,
      "avg_cost_per_task_usd": 0.020070034782608694,
      "avg_duration_per_task_ms": 306185.0115942029,
      "avg_output_tokens_per_task": 7045.43768115942,
      "last_run_timestamp": "2026-10-09T02:01:31.527Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "Novita",
      "accuracy": 0.7247618571428571,
      "accuracy_std_dev": 0.02776646862894973,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.05254494857142857,
      "avg_duration_per_task_ms": 272263.16,
      "avg_output_tokens_per_task": 8601.185714285713,
      "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-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "DeepInfra",
      "accuracy": 0.7247618571428571,
      "accuracy_std_dev": 0.03161563234689254,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.034889062857142855,
      "avg_duration_per_task_ms": 356175.16285714286,
      "avg_output_tokens_per_task": 8342.902857142857,
      "last_run_timestamp": "2026-08-29T16:11:10.137Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Baidu",
      "accuracy": 0.7245833125,
      "accuracy_std_dev": 0.036187558051281155,
      "run_count": 16,
      "total_questions": 800,
      "avg_cost_per_task_usd": 0.06812713125000001,
      "avg_duration_per_task_ms": 121769.8325,
      "avg_output_tokens_per_task": 8642.055,
      "last_run_timestamp": "2026-10-10T12:45:34.305Z"
    },
    {
      "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": 0.7244897346938776,
      "accuracy_std_dev": 0.009027500000000022,
      "run_count": 2,
      "total_questions": 98,
      "avg_cost_per_task_usd": 0.08190116326530611,
      "avg_duration_per_task_ms": 1045226.0918367347,
      "avg_output_tokens_per_task": 15985.285714285714,
      "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": "moonshotai/kimi-k2-thinking-20251106",
      "model_slug": "moonshotai/kimi-k2-thinking",
      "model_name": "MoonshotAI: Kimi K2 Thinking",
      "provider_name": "Novita",
      "accuracy": 0.7244446666666666,
      "accuracy_std_dev": 0.01911642045177101,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.05245255333333333,
      "avg_duration_per_task_ms": 276706.37333333335,
      "avg_output_tokens_per_task": 8448.833333333334,
      "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": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "GMICloud",
      "accuracy": 0.7244445555555555,
      "accuracy_std_dev": 0.025337407080840913,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.054274675555555554,
      "avg_duration_per_task_ms": 301393.08222222223,
      "avg_output_tokens_per_task": 11200.555555555555,
      "last_run_timestamp": "2026-10-10T18:58:00.575Z"
    },
    {
      "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": 0.7244444444444444,
      "accuracy_std_dev": 0.017497922277364575,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.10756234222222222,
      "avg_duration_per_task_ms": 333076.56666666665,
      "avg_output_tokens_per_task": 14814.204444444444,
      "last_run_timestamp": "2026-10-10T19:04:05.899Z"
    },
    {
      "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": 0.7240565707547171,
      "accuracy_std_dev": 0.03163321895945818,
      "run_count": 17,
      "total_questions": 848,
      "avg_cost_per_task_usd": 0.11481138679245283,
      "avg_duration_per_task_ms": 107730.78773584905,
      "avg_output_tokens_per_task": 9300.52358490566,
      "last_run_timestamp": "2026-10-10T11:50:06.494Z"
    },
    {
      "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": 0.7238493974895398,
      "accuracy_std_dev": 0.03039350529715622,
      "run_count": 24,
      "total_questions": 1195,
      "avg_cost_per_task_usd": 0.11536971046025105,
      "avg_duration_per_task_ms": 139092.2150627615,
      "avg_output_tokens_per_task": 8422.310460251047,
      "last_run_timestamp": "2026-10-10T12:27:41.998Z"
    },
    {
      "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": 0.7238442007299271,
      "accuracy_std_dev": 0.026621319896989543,
      "run_count": 11,
      "total_questions": 548,
      "avg_cost_per_task_usd": 0.04390420437956204,
      "avg_duration_per_task_ms": 215259.2098540146,
      "avg_output_tokens_per_task": 9833.600364963504,
      "last_run_timestamp": "2026-10-10T12:00:06.272Z"
    },
    {
      "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-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "StreamLake",
      "accuracy": 0.7233333333333334,
      "accuracy_std_dev": 0.023959997528287768,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.06615269666666666,
      "avg_duration_per_task_ms": 207301.43333333332,
      "avg_output_tokens_per_task": 9384.513333333334,
      "last_run_timestamp": "2026-10-10T11:22:15.856Z"
    },
    {
      "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-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "Groq",
      "accuracy": 0.7230575714285715,
      "accuracy_std_dev": 0.02915508369237475,
      "run_count": 8,
      "total_questions": 399,
      "avg_cost_per_task_usd": 0.13198787969924813,
      "avg_duration_per_task_ms": 71794.78947368421,
      "avg_output_tokens_per_task": 10252.233082706767,
      "last_run_timestamp": "2026-10-10T19:19:07.184Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Z.AI",
      "accuracy": 0.7230303636363636,
      "accuracy_std_dev": 0.04122986471936161,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.12495049272727274,
      "avg_duration_per_task_ms": 180996.27272727274,
      "avg_output_tokens_per_task": 12125.23090909091,
      "last_run_timestamp": "2026-10-10T12:09:55.160Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "auto-routing",
      "accuracy": 0.722963,
      "accuracy_std_dev": 0.01808372541510429,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.051240302222222214,
      "avg_duration_per_task_ms": 386538.15555555554,
      "avg_output_tokens_per_task": 9392.624444444444,
      "last_run_timestamp": "2026-10-10T20:11:19.446Z"
    },
    {
      "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": 0.7228571428571429,
      "accuracy_std_dev": 0.026393263417711994,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.3014312628571429,
      "avg_duration_per_task_ms": 206224.79142857142,
      "avg_output_tokens_per_task": 11626.885714285714,
      "last_run_timestamp": "2026-10-10T12:42:09.199Z"
    },
    {
      "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": 0.7226666,
      "accuracy_std_dev": 0.01959597237801686,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.033446764,
      "avg_duration_per_task_ms": 104528.44,
      "avg_output_tokens_per_task": 23630.596,
      "last_run_timestamp": "2026-10-10T11:30:58.693Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Io Net",
      "accuracy": 0.7222223333333333,
      "accuracy_std_dev": 0.011330936570096713,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.018792733333333332,
      "avg_duration_per_task_ms": 301192.17333333334,
      "avg_output_tokens_per_task": 19692.3,
      "last_run_timestamp": "2026-10-01T16:00:06.404Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-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": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "Chutes",
      "accuracy": 0.7222221111111111,
      "accuracy_std_dev": 0.027917975508621698,
      "run_count": 4,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.13745050505050505,
      "avg_duration_per_task_ms": 712939.898989899,
      "avg_output_tokens_per_task": 27938.80808080808,
      "last_run_timestamp": "2026-08-15T14:15:18.888Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-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": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "Alibaba",
      "accuracy": 0.7219429246231156,
      "accuracy_std_dev": 0.09608921942840087,
      "run_count": 8,
      "total_questions": 398,
      "avg_cost_per_task_usd": 0.07445455527638191,
      "avg_duration_per_task_ms": 193168.42211055275,
      "avg_output_tokens_per_task": 23981.597989949747,
      "last_run_timestamp": "2026-10-10T18:47:18.238Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Nebius",
      "accuracy": 0.7216665,
      "accuracy_std_dev": 0.011666500000000024,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.36281153000000005,
      "avg_duration_per_task_ms": 114354.7,
      "avg_output_tokens_per_task": 9018.87,
      "last_run_timestamp": "2026-10-10T11:34:33.998Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Morph",
      "accuracy": 0.7211113333333333,
      "accuracy_std_dev": 0.015595100530472852,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.017302566666666668,
      "avg_duration_per_task_ms": 811343.51,
      "avg_output_tokens_per_task": 20213.606666666667,
      "last_run_timestamp": "2026-10-10T11:19:51.484Z"
    },
    {
      "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": 0.7211000606060606,
      "accuracy_std_dev": 0.07754356597114187,
      "run_count": 12,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.04577266666666666,
      "avg_duration_per_task_ms": 266350.4158249158,
      "avg_output_tokens_per_task": 25957.685185185186,
      "last_run_timestamp": "2026-10-10T19:43:28.866Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Makora",
      "accuracy": 0.7210882448979592,
      "accuracy_std_dev": 0.0076780598851532815,
      "run_count": 5,
      "total_questions": 245,
      "avg_cost_per_task_usd": 0.009489710204081632,
      "avg_duration_per_task_ms": 145499.9755102041,
      "avg_output_tokens_per_task": 16260.81224489796,
      "last_run_timestamp": "2026-09-26T08:06:43.933Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Phala",
      "accuracy": 0.7210078506375227,
      "accuracy_std_dev": 0.03602466228677285,
      "run_count": 11,
      "total_questions": 549,
      "avg_cost_per_task_usd": 0.10133114936247721,
      "avg_duration_per_task_ms": 207370.83424408015,
      "avg_output_tokens_per_task": 10427.30965391621,
      "last_run_timestamp": "2026-10-10T12:12:49.617Z"
    },
    {
      "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": 0.72083325,
      "accuracy_std_dev": 0.03748153063760203,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.06457228749999999,
      "avg_duration_per_task_ms": 75803.1425,
      "avg_output_tokens_per_task": 9417.155,
      "last_run_timestamp": "2026-10-10T11:05:58.282Z"
    },
    {
      "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": 0.024634474601437903,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.06025161999999999,
      "avg_duration_per_task_ms": 307034.87555555557,
      "avg_output_tokens_per_task": 9651.193333333333,
      "last_run_timestamp": "2026-10-10T19:54:45.316Z"
    },
    {
      "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": 0.7205555833333334,
      "accuracy_std_dev": 0.037164138347019274,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.046787416666666665,
      "avg_duration_per_task_ms": 151030.90833333333,
      "avg_output_tokens_per_task": 8556.955,
      "last_run_timestamp": "2026-10-10T11:51:25.350Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-sol-pro-20260709",
      "model_slug": "openai/gpt-5.6-sol-pro",
      "model_name": "OpenAI: GPT-5.6 Sol Pro",
      "provider_name": "Azure",
      "accuracy": 0.7205384141414142,
      "accuracy_std_dev": 0.051112796809483045,
      "run_count": 4,
      "total_questions": 198,
      "avg_cost_per_task_usd": 1.6761988181818182,
      "avg_duration_per_task_ms": 313695.66161616164,
      "avg_output_tokens_per_task": 23404.97474747475,
      "last_run_timestamp": "2026-10-01T16:01:13.647Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-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": "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": "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": "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-v3.1-terminus",
      "model_slug": "deepseek/deepseek-v3.1-terminus",
      "model_name": "DeepSeek: DeepSeek V3.1 Terminus",
      "provider_name": "Novita",
      "accuracy": 0.7192392617449664,
      "accuracy_std_dev": 0.027973240079682514,
      "run_count": 6,
      "total_questions": 298,
      "avg_cost_per_task_usd": 0.06318058053691275,
      "avg_duration_per_task_ms": 493682.4899328859,
      "avg_output_tokens_per_task": 14513.312080536913,
      "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": "openai/gpt-5.6-sol-20260709",
      "model_slug": "openai/gpt-5.6-sol",
      "model_name": "OpenAI: GPT-5.6 Sol",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7188551919191919,
      "accuracy_std_dev": 0.027525589315898408,
      "run_count": 4,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.29904142424242425,
      "avg_duration_per_task_ms": 154139.7323232323,
      "avg_output_tokens_per_task": 5036.570707070707,
      "last_run_timestamp": "2026-10-01T19:05:12.160Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Decart",
      "accuracy": 0.7184344343434343,
      "accuracy_std_dev": 0.09707188521882894,
      "run_count": 8,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.07242429797979798,
      "avg_duration_per_task_ms": 168379.42676767678,
      "avg_output_tokens_per_task": 14277.45202020202,
      "last_run_timestamp": "2026-10-03T16:16:30.145Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "StreamLake",
      "accuracy": 0.7180952857142857,
      "accuracy_std_dev": 0.033752526067220005,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.13201746285714283,
      "avg_duration_per_task_ms": 330154.5142857143,
      "avg_output_tokens_per_task": 10313.082857142857,
      "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": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "auto-routing",
      "accuracy": 0.7175758181818181,
      "accuracy_std_dev": 0.03143352307451446,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.009597009090909091,
      "avg_duration_per_task_ms": 350866.65454545454,
      "avg_output_tokens_per_task": 14254.325454545455,
      "last_run_timestamp": "2026-10-10T13:51:45.436Z"
    },
    {
      "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": 0.7172549411764706,
      "accuracy_std_dev": 0.014106710908060687,
      "run_count": 17,
      "total_questions": 850,
      "avg_cost_per_task_usd": 0.037945944705882355,
      "avg_duration_per_task_ms": 187615.06352941177,
      "avg_output_tokens_per_task": 9902.724705882352,
      "last_run_timestamp": "2026-10-10T19:20:40.215Z"
    },
    {
      "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": 0.7170904591700134,
      "accuracy_std_dev": 0.030739071735857975,
      "run_count": 15,
      "total_questions": 747,
      "avg_cost_per_task_usd": 0.437738623828648,
      "avg_duration_per_task_ms": 237840.68808567603,
      "avg_output_tokens_per_task": 10036.461847389559,
      "last_run_timestamp": "2026-10-10T11:37:21.818Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-terra-20260709",
      "model_slug": "openai/gpt-5.6-terra",
      "model_name": "OpenAI: GPT-5.6 Terra",
      "provider_name": "Azure",
      "accuracy": 0.7170371111111111,
      "accuracy_std_dev": 0.016956299841680886,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.12728777111111111,
      "avg_duration_per_task_ms": 101323.44222222222,
      "avg_output_tokens_per_task": 3885.56,
      "last_run_timestamp": "2026-10-01T19:04:52.651Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "GMICloud",
      "accuracy": 0.717037,
      "accuracy_std_dev": 0.018083848299887195,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.025146933333333333,
      "avg_duration_per_task_ms": 216796.8177777778,
      "avg_output_tokens_per_task": 10040.375555555556,
      "last_run_timestamp": "2026-10-10T19:16:26.718Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Relace",
      "accuracy": 0.716667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.015077799999999999,
      "avg_duration_per_task_ms": 476878.72,
      "avg_output_tokens_per_task": 26097.54,
      "last_run_timestamp": "2026-09-28T09:00:02.615Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "StreamLake",
      "accuracy": 0.7161110833333333,
      "accuracy_std_dev": 0.034151986458717,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.03611448666666667,
      "avg_duration_per_task_ms": 183171.97833333333,
      "avg_output_tokens_per_task": 9167.023333333333,
      "last_run_timestamp": "2026-10-10T12:11:35.312Z"
    },
    {
      "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": 0.7157575454545455,
      "accuracy_std_dev": 0.03731083147233555,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.1024602309090909,
      "avg_duration_per_task_ms": 114222.2890909091,
      "avg_output_tokens_per_task": 9174.52,
      "last_run_timestamp": "2026-10-10T11:10:05.169Z"
    },
    {
      "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": "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": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7155558333333333,
      "accuracy_std_dev": 0.0289741860421575,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.04852732,
      "avg_duration_per_task_ms": 969869.3766666667,
      "avg_output_tokens_per_task": 8596.123333333333,
      "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": "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": "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": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "Wafer",
      "accuracy": 0.715278,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 48,
      "avg_cost_per_task_usd": 0.009582875,
      "avg_duration_per_task_ms": 164124.5,
      "avg_output_tokens_per_task": 18445.333333333332,
      "last_run_timestamp": "2026-10-10T11:44:02.317Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-terra-20260709",
      "model_slug": "openai/gpt-5.6-terra",
      "model_name": "OpenAI: GPT-5.6 Terra",
      "provider_name": "OpenAI",
      "accuracy": 0.715,
      "accuracy_std_dev": 0.029202378139459823,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.09407665500000001,
      "avg_duration_per_task_ms": 83159.84,
      "avg_output_tokens_per_task": 3660.855,
      "last_run_timestamp": "2026-10-02T23:10:36.500Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7148361502890174,
      "accuracy_std_dev": 0.08759138393281216,
      "run_count": 7,
      "total_questions": 346,
      "avg_cost_per_task_usd": 0.00943905780346821,
      "avg_duration_per_task_ms": 208132.0809248555,
      "avg_output_tokens_per_task": 4098.601156069364,
      "last_run_timestamp": "2026-10-10T11:51:58.383Z"
    },
    {
      "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": 0.7142424545454545,
      "accuracy_std_dev": 0.0319117780689303,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.1133730290909091,
      "avg_duration_per_task_ms": 98039.63272727272,
      "avg_output_tokens_per_task": 10090.498181818182,
      "last_run_timestamp": "2026-10-10T11:41:14.865Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "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-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "Venice",
      "accuracy": 0.7140741111111111,
      "accuracy_std_dev": 0.02141748193231379,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.07203324000000001,
      "avg_duration_per_task_ms": 281427.5711111111,
      "avg_output_tokens_per_task": 10764.266666666666,
      "last_run_timestamp": "2026-10-10T19:48:33.421Z"
    },
    {
      "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": 0.7139392727272728,
      "accuracy_std_dev": 0.061723041464698515,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.2911906890909091,
      "avg_duration_per_task_ms": 424451.22363636363,
      "avg_output_tokens_per_task": 9948.334545454545,
      "last_run_timestamp": "2026-10-10T11:29:58.824Z"
    },
    {
      "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": 0.7133736491935484,
      "accuracy_std_dev": 0.10276651148311886,
      "run_count": 10,
      "total_questions": 496,
      "avg_cost_per_task_usd": 0.2060511633064516,
      "avg_duration_per_task_ms": 1153803.5564516129,
      "avg_output_tokens_per_task": 55616.0504032258,
      "last_run_timestamp": "2026-10-10T19:16:51.412Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "inclusionai/ling-3.0-flash-vl-20260910",
      "model_slug": "inclusionai/ling-3.0-flash-vl",
      "model_name": "inclusionAI: Ling 3.0 Flash VL",
      "provider_name": "DeepInfra",
      "accuracy": 0.713333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.006318460000000001,
      "avg_duration_per_task_ms": 234030.92,
      "avg_output_tokens_per_task": 14773.36,
      "last_run_timestamp": "2026-10-01T08:05:57.971Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "GMICloud",
      "accuracy": 0.713333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.09868048,
      "avg_duration_per_task_ms": 318461,
      "avg_output_tokens_per_task": 11794.24,
      "last_run_timestamp": "2026-08-29T17:18:09.782Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-flash-20260826",
      "model_slug": "qwen/qwen3.8-flash",
      "model_name": "Qwen: Qwen3.8 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.713333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.01665106,
      "avg_duration_per_task_ms": 327560.6,
      "avg_output_tokens_per_task": 17227.24,
      "last_run_timestamp": "2026-10-01T08:00:41.459Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "Novita",
      "accuracy": 0.7129095028901734,
      "accuracy_std_dev": 0.03295967347643888,
      "run_count": 7,
      "total_questions": 346,
      "avg_cost_per_task_usd": 0.05974209248554914,
      "avg_duration_per_task_ms": 481297.3583815029,
      "avg_output_tokens_per_task": 14073.28612716763,
      "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": "google/gemini-3.8-flash-20260902",
      "model_slug": "google/gemini-3.8-flash",
      "model_name": "Google: Gemini 3.8 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.712766,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 47,
      "avg_cost_per_task_usd": 0.21022357446808512,
      "avg_duration_per_task_ms": 261220.25531914894,
      "avg_output_tokens_per_task": 27609.17021276596,
      "last_run_timestamp": "2026-10-01T08:02:55.088Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5-mini-2025-08-07",
      "model_slug": "openai/gpt-5-mini",
      "model_name": "OpenAI: GPT-5 Mini",
      "provider_name": "OpenAI",
      "accuracy": 0.712121101010101,
      "accuracy_std_dev": 0.029111223892169177,
      "run_count": 4,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.12218062626262625,
      "avg_duration_per_task_ms": 537818.2474747475,
      "avg_output_tokens_per_task": 55422.06060606061,
      "last_run_timestamp": "2026-10-02T23:27:09.304Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "CoreWeave",
      "accuracy": 0.7116666666666667,
      "accuracy_std_dev": 0.03762638455776595,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.08100557333333333,
      "avg_duration_per_task_ms": 79109.17,
      "avg_output_tokens_per_task": 8517.581666666667,
      "last_run_timestamp": "2026-10-10T11:36:38.992Z"
    },
    {
      "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": 0.7114282857142857,
      "accuracy_std_dev": 0.029323450716420958,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.2380128942857143,
      "avg_duration_per_task_ms": 294855.26,
      "avg_output_tokens_per_task": 21043.11142857143,
      "last_run_timestamp": "2026-10-10T11:21:52.972Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Cloudflare",
      "accuracy": 0.7113333333333334,
      "accuracy_std_dev": 0.03114610968037938,
      "run_count": 15,
      "total_questions": 750,
      "avg_cost_per_task_usd": 0.13518174266666666,
      "avg_duration_per_task_ms": 137112.55866666668,
      "avg_output_tokens_per_task": 8948.754666666666,
      "last_run_timestamp": "2026-10-10T11:51:03.166Z"
    },
    {
      "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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Decart",
      "accuracy": 0.7104444,
      "accuracy_std_dev": 0.028359365645937846,
      "run_count": 15,
      "total_questions": 750,
      "avg_cost_per_task_usd": 0.12106834666666669,
      "avg_duration_per_task_ms": 90872.85466666667,
      "avg_output_tokens_per_task": 10614.206666666667,
      "last_run_timestamp": "2026-10-10T12:07:01.508Z"
    },
    {
      "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": "auto-routing",
      "accuracy": 0.710000125,
      "accuracy_std_dev": 0.03144646685812851,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.043149095,
      "avg_duration_per_task_ms": 372456.3975,
      "avg_output_tokens_per_task": 9746.7325,
      "last_run_timestamp": "2026-10-10T20:36:24.064Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-terra-20260709",
      "model_slug": "openai/gpt-5.6-terra",
      "model_name": "OpenAI: GPT-5.6 Terra",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.71,
      "accuracy_std_dev": 0.01795033271279393,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.12391315,
      "avg_duration_per_task_ms": 76197.77,
      "avg_output_tokens_per_task": 4134.025,
      "last_run_timestamp": "2026-10-01T19:04:38.304Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "Z.AI",
      "accuracy": 0.7096297777777777,
      "accuracy_std_dev": 0.043331717602641394,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.06758097555555556,
      "avg_duration_per_task_ms": 341751.31333333335,
      "avg_output_tokens_per_task": 10011.704444444444,
      "last_run_timestamp": "2026-10-10T19:49:16.061Z"
    },
    {
      "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": 0.7095516447368422,
      "accuracy_std_dev": 0.03406448599188871,
      "run_count": 28,
      "total_questions": 1368,
      "avg_cost_per_task_usd": 0.09692296418128656,
      "avg_duration_per_task_ms": 73548.15350877192,
      "avg_output_tokens_per_task": 8046.236111111111,
      "last_run_timestamp": "2026-10-10T11:52:20.355Z"
    },
    {
      "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": 0.7095158747913188,
      "accuracy_std_dev": 0.03442095606966778,
      "run_count": 12,
      "total_questions": 599,
      "avg_cost_per_task_usd": 0.07650697662771284,
      "avg_duration_per_task_ms": 239472.96160267113,
      "avg_output_tokens_per_task": 8830.707846410685,
      "last_run_timestamp": "2026-10-10T11:59:03.435Z"
    },
    {
      "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": 0.7088887777777778,
      "accuracy_std_dev": 0.013333370373096707,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.026421224444444447,
      "avg_duration_per_task_ms": 218382.79777777777,
      "avg_output_tokens_per_task": 10082.562222222223,
      "last_run_timestamp": "2026-10-10T18:51:03.760Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Io Net",
      "accuracy": 0.7088886666666667,
      "accuracy_std_dev": 0.01749773883169545,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.11639628666666667,
      "avg_duration_per_task_ms": 755791.72,
      "avg_output_tokens_per_task": 27174.273333333334,
      "last_run_timestamp": "2026-08-15T08:15:24.293Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "GMICloud",
      "accuracy": 0.7084847272727273,
      "accuracy_std_dev": 0.04178733689026097,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.09325543272727273,
      "avg_duration_per_task_ms": 214793.19090909092,
      "avg_output_tokens_per_task": 8277.278181818181,
      "last_run_timestamp": "2026-10-10T12:11:48.158Z"
    },
    {
      "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": "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-thinking-20251106",
      "model_slug": "moonshotai/kimi-k2-thinking",
      "model_name": "MoonshotAI: Kimi K2 Thinking",
      "provider_name": "auto-routing",
      "accuracy": 0.7066665000000001,
      "accuracy_std_dev": 0.013333499999999998,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.054561849999999995,
      "avg_duration_per_task_ms": 281726.22,
      "avg_output_tokens_per_task": 8888.58,
      "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": "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": "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": 0.7043380136986301,
      "accuracy_std_dev": 0.016697582419819515,
      "run_count": 3,
      "total_questions": 146,
      "avg_cost_per_task_usd": 0.027861123287671232,
      "avg_duration_per_task_ms": 317404.41780821915,
      "avg_output_tokens_per_task": 16264.835616438357,
      "last_run_timestamp": "2026-10-10T20:55:29.743Z"
    },
    {
      "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": 0.7042423636363636,
      "accuracy_std_dev": 0.1528256864263529,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.05846522,
      "avg_duration_per_task_ms": 152693.06363636363,
      "avg_output_tokens_per_task": 18465.061818181817,
      "last_run_timestamp": "2026-10-10T19:04:13.865Z"
    },
    {
      "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/qwen3.7-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": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "Phala",
      "accuracy": 0.702381,
      "accuracy_std_dev": 0.02809942180391412,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.1988400742857143,
      "avg_duration_per_task_ms": 1709336.88,
      "avg_output_tokens_per_task": 13446.757142857143,
      "last_run_timestamp": "2026-10-10T18:42:31.220Z"
    },
    {
      "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": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7016668749999999,
      "accuracy_std_dev": 0.05052503642363232,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.030146600000000003,
      "avg_duration_per_task_ms": 298672.2075,
      "avg_output_tokens_per_task": 10079.875,
      "last_run_timestamp": "2026-10-10T18:56:27.493Z"
    },
    {
      "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": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "DeepInfra",
      "accuracy": 0.7011108333333333,
      "accuracy_std_dev": 0.020608011147906124,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.051610086666666666,
      "avg_duration_per_task_ms": 398260.51,
      "avg_output_tokens_per_task": 9237.883333333333,
      "last_run_timestamp": "2026-10-03T16:12:07.463Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "Venice",
      "accuracy": 0.7008929330357142,
      "accuracy_std_dev": 0.031656430987076,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.04512097991071429,
      "avg_duration_per_task_ms": 429480.1316964286,
      "avg_output_tokens_per_task": 9115.185267857143,
      "last_run_timestamp": "2026-10-10T18:33:03.632Z"
    },
    {
      "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": 0.70000025,
      "accuracy_std_dev": 0.03162288201109283,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.12919751,
      "avg_duration_per_task_ms": 506623.765,
      "avg_output_tokens_per_task": 10892.11,
      "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": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.7,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.12726106,
      "avg_duration_per_task_ms": 181226.44,
      "avg_output_tokens_per_task": 9838.56,
      "last_run_timestamp": "2026-10-04T03:00:16.405Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.2-20251211",
      "model_slug": "openai/gpt-5.2",
      "model_name": "OpenAI: GPT-5.2",
      "provider_name": "Azure",
      "accuracy": 0.7,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.31292506,
      "avg_duration_per_task_ms": 261751.42,
      "avg_output_tokens_per_task": 14948.8,
      "last_run_timestamp": "2026-08-01T08:06:08.193Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6.1-sol-20260929",
      "model_slug": "openai/gpt-6.1-sol",
      "model_name": "OpenAI: GPT-6.1 Sol",
      "provider_name": "auto-routing",
      "accuracy": 0.7,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.1012818,
      "avg_duration_per_task_ms": 144103.72,
      "avg_output_tokens_per_task": 3638.14,
      "last_run_timestamp": "2026-10-01T08:03:00.822Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-flash-20260826",
      "model_slug": "qwen/qwen3.8-flash",
      "model_name": "Qwen: Qwen3.8 Flash",
      "provider_name": "Alibaba",
      "accuracy": 0.7,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.019795359999999998,
      "avg_duration_per_task_ms": 396615.94,
      "avg_output_tokens_per_task": 23218.96,
      "last_run_timestamp": "2026-10-01T08:03:06.120Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.01539542,
      "avg_duration_per_task_ms": 2335452.5,
      "avg_output_tokens_per_task": 25378.6,
      "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": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "auto-routing",
      "accuracy": 0.6998645487804879,
      "accuracy_std_dev": 0.05531710950158186,
      "run_count": 10,
      "total_questions": 492,
      "avg_cost_per_task_usd": 0.031559408536585364,
      "avg_duration_per_task_ms": 1019512.3414634146,
      "avg_output_tokens_per_task": 43385.81707317073,
      "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": "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-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Venice",
      "accuracy": 0.6985507420289855,
      "accuracy_std_dev": 0.027723151851321442,
      "run_count": 7,
      "total_questions": 345,
      "avg_cost_per_task_usd": 0.01719656231884058,
      "avg_duration_per_task_ms": 309641.62898550724,
      "avg_output_tokens_per_task": 9142.672463768116,
      "last_run_timestamp": "2026-08-29T17:28:21.827Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "StreamLake",
      "accuracy": 0.6977778333333333,
      "accuracy_std_dev": 0.026573791972647694,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.04297996,
      "avg_duration_per_task_ms": 184061.30666666667,
      "avg_output_tokens_per_task": 8458.556666666667,
      "last_run_timestamp": "2026-08-29T17:03:30.001Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-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": "deepseek/deepseek-v3.2-exp",
      "model_slug": "deepseek/deepseek-v3.2-exp",
      "model_name": "DeepSeek: DeepSeek V3.2 Exp",
      "provider_name": "AtlasCloud",
      "accuracy": 0.6972790204081633,
      "accuracy_std_dev": 0.04361100000000001,
      "run_count": 2,
      "total_questions": 98,
      "avg_cost_per_task_usd": 0.08154698979591836,
      "avg_duration_per_task_ms": 436567.05102040817,
      "avg_output_tokens_per_task": 13545.377551020409,
      "last_run_timestamp": "2026-08-01T17:38:10.889Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "AtlasCloud",
      "accuracy": 0.6970371111111111,
      "accuracy_std_dev": 0.05417024882195113,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.10031135333333334,
      "avg_duration_per_task_ms": 244307.61111111112,
      "avg_output_tokens_per_task": 10714.342222222222,
      "last_run_timestamp": "2026-10-10T18:48:20.929Z"
    },
    {
      "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": 0.69666675,
      "accuracy_std_dev": 0.12892723184489574,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.04194705,
      "avg_duration_per_task_ms": 293196.69,
      "avg_output_tokens_per_task": 10603.27,
      "last_run_timestamp": "2026-10-10T20:59:28.749Z"
    },
    {
      "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": "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": 0.696561865470852,
      "accuracy_std_dev": 0.012961611737724062,
      "run_count": 9,
      "total_questions": 446,
      "avg_cost_per_task_usd": 0.021608876681614344,
      "avg_duration_per_task_ms": 698943.9192825112,
      "avg_output_tokens_per_task": 19226.64125560538,
      "last_run_timestamp": "2026-10-10T18:35:46.234Z"
    },
    {
      "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": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "AtlasCloud",
      "accuracy": 0.6958332500000001,
      "accuracy_std_dev": 0.08906522812628674,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0512019375,
      "avg_duration_per_task_ms": 570157.7625,
      "avg_output_tokens_per_task": 19195.9675,
      "last_run_timestamp": "2026-10-10T19:12:44.434Z"
    },
    {
      "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": 0.6951850555555555,
      "accuracy_std_dev": 0.023810538468576885,
      "run_count": 18,
      "total_questions": 900,
      "avg_cost_per_task_usd": 0.10653520444444445,
      "avg_duration_per_task_ms": 77015.53444444445,
      "avg_output_tokens_per_task": 9317.83111111111,
      "last_run_timestamp": "2026-10-10T11:34:42.663Z"
    },
    {
      "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": 0.6940235606060606,
      "accuracy_std_dev": 0.024175874322025556,
      "run_count": 8,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.034200517676767676,
      "avg_duration_per_task_ms": 689874.9823232323,
      "avg_output_tokens_per_task": 19910.09595959596,
      "last_run_timestamp": "2026-10-10T19:24:00.790Z"
    },
    {
      "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": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Nebius",
      "accuracy": 0.69333325,
      "accuracy_std_dev": 0.05676452335030659,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.22464821499999998,
      "avg_duration_per_task_ms": 688281.77,
      "avg_output_tokens_per_task": 10611.7,
      "last_run_timestamp": "2026-10-10T14:01:53.590Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.2-exp",
      "model_slug": "deepseek/deepseek-v3.2-exp",
      "model_name": "DeepSeek: DeepSeek V3.2 Exp",
      "provider_name": "SiliconFlow",
      "accuracy": 0.693333,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.08382193,
      "avg_duration_per_task_ms": 858486.8,
      "avg_output_tokens_per_task": 15204.12,
      "last_run_timestamp": "2026-08-01T17:38:16.661Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "Ionstream",
      "accuracy": 0.693333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.015592,
      "avg_duration_per_task_ms": 1084747.56,
      "avg_output_tokens_per_task": 17745.48,
      "last_run_timestamp": "2026-07-31T03:44:20.825Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "Morph",
      "accuracy": 0.693333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.060288180000000004,
      "avg_duration_per_task_ms": 350636.38,
      "avg_output_tokens_per_task": 10540.6,
      "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-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": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "DeepInfra",
      "accuracy": 0.6929823609022557,
      "accuracy_std_dev": 0.026094683447675196,
      "run_count": 8,
      "total_questions": 399,
      "avg_cost_per_task_usd": 0.03234121052631579,
      "avg_duration_per_task_ms": 241069.54887218046,
      "avg_output_tokens_per_task": 9648.423558897242,
      "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": "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": 0.6924997500000001,
      "accuracy_std_dev": 0.0419573344802491,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.019822750000000004,
      "avg_duration_per_task_ms": 438051.1125,
      "avg_output_tokens_per_task": 20830,
      "last_run_timestamp": "2026-10-10T20:24:13.976Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.7-flash-20260119",
      "model_slug": "z-ai/glm-4.7-flash",
      "model_name": "Z.ai: GLM 4.7 Flash",
      "provider_name": "Venice",
      "accuracy": 0.6923076113360324,
      "accuracy_std_dev": 0.004408577938519381,
      "run_count": 5,
      "total_questions": 247,
      "avg_cost_per_task_usd": 0.009435951417004049,
      "avg_duration_per_task_ms": 259361.54655870446,
      "avg_output_tokens_per_task": 9080.562753036438,
      "last_run_timestamp": "2026-08-23T00:00:19.944Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-astra-pro-20260903",
      "model_slug": "openai/gpt-6-astra-pro",
      "model_name": "OpenAI: GPT-6 Astra Pro",
      "provider_name": "Azure",
      "accuracy": 0.69,
      "accuracy_std_dev": 0.009999999999999953,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 2.99530482,
      "avg_duration_per_task_ms": 320001.13,
      "avg_output_tokens_per_task": 12475.35,
      "last_run_timestamp": "2026-10-01T08:05:20.373Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.5-air",
      "model_slug": "z-ai/glm-4.5-air",
      "model_name": "Z.ai: GLM 4.5 Air",
      "provider_name": "auto-routing",
      "accuracy": 0.69,
      "accuracy_std_dev": 0.003332999999999975,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01726069,
      "avg_duration_per_task_ms": 298145.74,
      "avg_output_tokens_per_task": 10516.76,
      "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": "openai/gpt-5-2025-08-07",
      "model_slug": "openai/gpt-5",
      "model_name": "OpenAI: GPT-5",
      "provider_name": "auto-routing",
      "accuracy": 0.6891892297297297,
      "accuracy_std_dev": 0.03170486152346707,
      "run_count": 3,
      "total_questions": 148,
      "avg_cost_per_task_usd": 0.6622049662162162,
      "avg_duration_per_task_ms": 777791.8783783783,
      "avg_output_tokens_per_task": 59355.4527027027,
      "last_run_timestamp": "2026-10-01T16:01:02.610Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "DekaLLM",
      "accuracy": 0.6888886666666667,
      "accuracy_std_dev": 0.021998731726069,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.015484113333333334,
      "avg_duration_per_task_ms": 264745.86666666664,
      "avg_output_tokens_per_task": 18035.506666666668,
      "last_run_timestamp": "2026-10-10T20:39:08.934Z"
    },
    {
      "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": 0.6887609855907781,
      "accuracy_std_dev": 0.02815688345390403,
      "run_count": 7,
      "total_questions": 347,
      "avg_cost_per_task_usd": 0.019089083573487032,
      "avg_duration_per_task_ms": 454304.7060518732,
      "avg_output_tokens_per_task": 9047.129682997118,
      "last_run_timestamp": "2026-10-10T19:16:30.073Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Friendli",
      "accuracy": 0.688485,
      "accuracy_std_dev": 0.07169463814489388,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.11210284363636364,
      "avg_duration_per_task_ms": 69179.89090909091,
      "avg_output_tokens_per_task": 8846.072727272727,
      "last_run_timestamp": "2026-10-10T11:52:26.136Z"
    },
    {
      "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": 0.6879166250000001,
      "accuracy_std_dev": 0.07218680699569953,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.05169213000000001,
      "avg_duration_per_task_ms": 304751.5525,
      "avg_output_tokens_per_task": 8046.7025,
      "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": "moonshotai/kimi-k2-thinking-20251106",
      "model_slug": "moonshotai/kimi-k2-thinking",
      "model_name": "MoonshotAI: Kimi K2 Thinking",
      "provider_name": "Google",
      "accuracy": 0.6877776666666667,
      "accuracy_std_dev": 0.010304212256267955,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.11416634666666667,
      "avg_duration_per_task_ms": 162037.17333333334,
      "avg_output_tokens_per_task": 8889.293333333333,
      "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": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Minimax",
      "accuracy": 0.6876191428571429,
      "accuracy_std_dev": 0.0358884483600765,
      "run_count": 14,
      "total_questions": 700,
      "avg_cost_per_task_usd": 0.03518511571428571,
      "avg_duration_per_task_ms": 165038.81714285715,
      "avg_output_tokens_per_task": 10124.521428571428,
      "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": 0.6872909230769231,
      "accuracy_std_dev": 0.021147758508094125,
      "run_count": 6,
      "total_questions": 299,
      "avg_cost_per_task_usd": 0.006207953177257525,
      "avg_duration_per_task_ms": 252164.59866220737,
      "avg_output_tokens_per_task": 9519.789297658863,
      "last_run_timestamp": "2026-08-30T00:01:09.439Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "inclusionai/ling-3.0-flash-20260723",
      "model_slug": "inclusionai/ling-3.0-flash",
      "model_name": "inclusionAI: Ling 3.0 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.686667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.00219582,
      "avg_duration_per_task_ms": 78121.92,
      "avg_output_tokens_per_task": 11659.46,
      "last_run_timestamp": "2026-10-01T16:00:49.880Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.6866665000000001,
      "accuracy_std_dev": 0.006666499999999964,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.02072933,
      "avg_duration_per_task_ms": 450665.84,
      "avg_output_tokens_per_task": 10600.05,
      "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": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "GMICloud",
      "accuracy": 0.6855555,
      "accuracy_std_dev": 0.052973362745660516,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.018873013333333334,
      "avg_duration_per_task_ms": 269968.75333333336,
      "avg_output_tokens_per_task": 10405.67,
      "last_run_timestamp": "2026-10-10T19:21:53.439Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-luna-pro-20260922",
      "model_slug": "openai/gpt-6-luna-pro",
      "model_name": "OpenAI: GPT-6 Luna Pro",
      "provider_name": "Azure",
      "accuracy": 0.6850001666666666,
      "accuracy_std_dev": 0.0143694957278334,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.04138883,
      "avg_duration_per_task_ms": 242662.93333333332,
      "avg_output_tokens_per_task": 30866.096666666668,
      "last_run_timestamp": "2026-10-01T08:05:57.699Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "inclusionai/ling-3.0-flash-20260723",
      "model_slug": "inclusionai/ling-3.0-flash",
      "model_name": "inclusionAI: Ling 3.0 Flash",
      "provider_name": "DeepInfra",
      "accuracy": 0.6833333333333333,
      "accuracy_std_dev": 0.07779388246014092,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.00909372,
      "avg_duration_per_task_ms": 253755.2,
      "avg_output_tokens_per_task": 15941.146666666667,
      "last_run_timestamp": "2026-10-08T22:38:42.670Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6.1-sol-20260929",
      "model_slug": "openai/gpt-6.1-sol",
      "model_name": "OpenAI: GPT-6.1 Sol",
      "provider_name": "Azure",
      "accuracy": 0.6833333333333333,
      "accuracy_std_dev": 0.012619766977070334,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.10710260999999997,
      "avg_duration_per_task_ms": 135398.91,
      "avg_output_tokens_per_task": 4027.3166666666666,
      "last_run_timestamp": "2026-10-01T08:06:09.665Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "Novita",
      "accuracy": 0.6824557368421051,
      "accuracy_std_dev": 0.05773499999999998,
      "run_count": 2,
      "total_questions": 95,
      "avg_cost_per_task_usd": 0.19861939999999997,
      "avg_duration_per_task_ms": 328265.04210526316,
      "avg_output_tokens_per_task": 23629.65263157895,
      "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": "minimax/minimax-m2.1",
      "model_slug": "minimax/minimax-m2.1",
      "model_name": "MiniMax: MiniMax M2.1",
      "provider_name": "Minimax",
      "accuracy": 0.6822223333333333,
      "accuracy_std_dev": 0.058584163368230795,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.02942621,
      "avg_duration_per_task_ms": 175208.89333333334,
      "avg_output_tokens_per_task": 10644.23,
      "last_run_timestamp": "2026-08-01T17:38:12.859Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6.1-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": "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": 0.681992448275862,
      "accuracy_std_dev": 0.10159207419370662,
      "run_count": 7,
      "total_questions": 348,
      "avg_cost_per_task_usd": 0.07432972701149426,
      "avg_duration_per_task_ms": 819407.2413793104,
      "avg_output_tokens_per_task": 46149.24425287356,
      "last_run_timestamp": "2026-10-10T18:48:14.818Z"
    },
    {
      "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": 0.6819047142857143,
      "accuracy_std_dev": 0.025376904886779485,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.007673354285714286,
      "avg_duration_per_task_ms": 259027.58,
      "avg_output_tokens_per_task": 11768.4,
      "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": "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": "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": 0.68,
      "accuracy_std_dev": 0.01440159456449179,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.015966806666666666,
      "avg_duration_per_task_ms": 259752.2,
      "avg_output_tokens_per_task": 10038.846666666666,
      "last_run_timestamp": "2026-08-01T17:38:31.690Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "Novita",
      "accuracy": 0.6777777777777778,
      "accuracy_std_dev": 0.056655654704103615,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.02349592666666667,
      "avg_duration_per_task_ms": 202633.27333333335,
      "avg_output_tokens_per_task": 9395.406666666666,
      "last_run_timestamp": "2026-10-10T19:22:03.608Z"
    },
    {
      "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": 0.6777777777777778,
      "accuracy_std_dev": 0.07894361482135037,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.07500947555555555,
      "avg_duration_per_task_ms": 278493.0911111111,
      "avg_output_tokens_per_task": 25748.14,
      "last_run_timestamp": "2026-10-10T18:46:46.591Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "Fireworks",
      "accuracy": 0.6777776666666667,
      "accuracy_std_dev": 0.013698571612479227,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.0340762,
      "avg_duration_per_task_ms": 244093.44,
      "avg_output_tokens_per_task": 27086.806666666667,
      "last_run_timestamp": "2026-09-19T08:01:53.393Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-astra-20260903",
      "model_slug": "openai/gpt-6-astra",
      "model_name": "OpenAI: GPT-6 Astra",
      "provider_name": "OpenAI",
      "accuracy": 0.6766665000000001,
      "accuracy_std_dev": 0.0033335000000000448,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.5967202500000001,
      "avg_duration_per_task_ms": 137724.15,
      "avg_output_tokens_per_task": 4015.19,
      "last_run_timestamp": "2026-10-02T23:06:01.512Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.6766665000000001,
      "accuracy_std_dev": 0.0033335000000000448,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.04451592,
      "avg_duration_per_task_ms": 466424.76,
      "avg_output_tokens_per_task": 10206.78,
      "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-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": 0.6761905714285714,
      "accuracy_std_dev": 0.01880810362474601,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.009157782857142858,
      "avg_duration_per_task_ms": 249382.73428571428,
      "avg_output_tokens_per_task": 10084.637142857142,
      "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": "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": 0.6757575454545455,
      "accuracy_std_dev": 0.03827318025687446,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.017890945454545452,
      "avg_duration_per_task_ms": 338670.8781818182,
      "avg_output_tokens_per_task": 20078.554545454546,
      "last_run_timestamp": "2026-10-10T21:10:48.057Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Nebius",
      "accuracy": 0.675675972972973,
      "accuracy_std_dev": 0.02284471405428887,
      "run_count": 3,
      "total_questions": 148,
      "avg_cost_per_task_usd": 0.04042408108108108,
      "avg_duration_per_task_ms": 518660.3783783784,
      "avg_output_tokens_per_task": 20990.527027027027,
      "last_run_timestamp": "2026-09-26T08:11:26.490Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "thinkingmachines/inkling-20260715",
      "model_slug": "thinkingmachines/inkling",
      "model_name": "Thinking Machines: Inkling",
      "provider_name": "DeepInfra",
      "accuracy": 0.6755556666666667,
      "accuracy_std_dev": 0.006285236475706844,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.07120372000000001,
      "avg_duration_per_task_ms": 144319.38666666666,
      "avg_output_tokens_per_task": 2561.846666666667,
      "last_run_timestamp": "2026-10-08T22:25:52.178Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Tenstorrent",
      "accuracy": 0.675,
      "accuracy_std_dev": 0.011667000000000038,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.1079227,
      "avg_duration_per_task_ms": 145751.42,
      "avg_output_tokens_per_task": 18113.26,
      "last_run_timestamp": "2026-10-04T05:20:00.959Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Friendli",
      "accuracy": 0.6741667499999999,
      "accuracy_std_dev": 0.0918596083103858,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.090455615,
      "avg_duration_per_task_ms": 88078.0775,
      "avg_output_tokens_per_task": 8488.9925,
      "last_run_timestamp": "2026-10-10T18:41:45.088Z"
    },
    {
      "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": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "DigitalOcean",
      "accuracy": 0.6733333636363636,
      "accuracy_std_dev": 0.06381806526834732,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.06596764181818182,
      "avg_duration_per_task_ms": 779394.0927272728,
      "avg_output_tokens_per_task": 18514.336363636365,
      "last_run_timestamp": "2026-10-05T20:43:11.175Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "DekaLLM",
      "accuracy": 0.6733330000000001,
      "accuracy_std_dev": 0.016329931618554488,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.008951993333333332,
      "avg_duration_per_task_ms": 824834.48,
      "avg_output_tokens_per_task": 18340.86,
      "last_run_timestamp": "2026-10-10T19:34:42.395Z"
    },
    {
      "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": 0.673333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.09000018000000001,
      "avg_duration_per_task_ms": 1674604.48,
      "avg_output_tokens_per_task": 17624.46,
      "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-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": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Mancer 2",
      "accuracy": 0.672916875,
      "accuracy_std_dev": 0.0946110438485348,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.056751722500000004,
      "avg_duration_per_task_ms": 524628.53,
      "avg_output_tokens_per_task": 16401.87,
      "last_run_timestamp": "2026-10-10T19:25:46.224Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.6726727747747747,
      "accuracy_std_dev": 0.04092304282098771,
      "run_count": 9,
      "total_questions": 444,
      "avg_cost_per_task_usd": 0.023259461711711713,
      "avg_duration_per_task_ms": 273637.62387387385,
      "avg_output_tokens_per_task": 17560.18018018018,
      "last_run_timestamp": "2026-10-10T19:28:16.526Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "thinkingmachines/inkling-20260715",
      "model_slug": "thinkingmachines/inkling",
      "model_name": "Thinking Machines: Inkling",
      "provider_name": "auto-routing",
      "accuracy": 0.672,
      "accuracy_std_dev": 0.028409779224766934,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.07962264,
      "avg_duration_per_task_ms": 80557.024,
      "avg_output_tokens_per_task": 3531.236,
      "last_run_timestamp": "2026-10-01T16:00:45.064Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-sol-20260922",
      "model_slug": "openai/gpt-6-sol",
      "model_name": "OpenAI: GPT-6 Sol",
      "provider_name": "Azure",
      "accuracy": 0.6706668,
      "accuracy_std_dev": 0.015548970672041286,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.142799016,
      "avg_duration_per_task_ms": 115124.684,
      "avg_output_tokens_per_task": 5068.812,
      "last_run_timestamp": "2026-10-01T08:04:38.837Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "CoreWeave",
      "accuracy": 0.67,
      "accuracy_std_dev": 0.017384450270284638,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.017308236,
      "avg_duration_per_task_ms": 120715.156,
      "avg_output_tokens_per_task": 21418.432,
      "last_run_timestamp": "2026-10-10T11:31:03.201Z"
    },
    {
      "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": 0.6698850000000001,
      "accuracy_std_dev": 0.030556564488017048,
      "run_count": 29,
      "total_questions": 1450,
      "avg_cost_per_task_usd": 0.12515819931034483,
      "avg_duration_per_task_ms": 120179.14275862068,
      "avg_output_tokens_per_task": 12653.900689655173,
      "last_run_timestamp": "2026-10-10T11:29:00.634Z"
    },
    {
      "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": 0.6685714285714286,
      "accuracy_std_dev": 0.01875964453559928,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.029877088571428574,
      "avg_duration_per_task_ms": 242438.32571428572,
      "avg_output_tokens_per_task": 11057.357142857143,
      "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-4.5-haiku-20251001",
      "model_slug": "anthropic/claude-haiku-4.5",
      "model_name": "Anthropic: Claude Haiku 4.5",
      "provider_name": "auto-routing",
      "accuracy": 0.6679999,
      "accuracy_std_dev": 0.03383610563421859,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.11777596400000001,
      "avg_duration_per_task_ms": 124529.096,
      "avg_output_tokens_per_task": 12738.322,
      "last_run_timestamp": "2026-10-10T11:22:45.678Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-luna-pro-20260922",
      "model_slug": "openai/gpt-6-luna-pro",
      "model_name": "OpenAI: GPT-6 Luna Pro",
      "provider_name": "auto-routing",
      "accuracy": 0.666667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.038762819999999996,
      "avg_duration_per_task_ms": 314594.08,
      "avg_output_tokens_per_task": 31952.48,
      "last_run_timestamp": "2026-10-01T08:00:24.890Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.5-haiku-20251001",
      "model_slug": "anthropic/claude-haiku-4.5",
      "model_name": "Anthropic: Claude Haiku 4.5",
      "provider_name": "Anthropic",
      "accuracy": 0.6663334,
      "accuracy_std_dev": 0.02651835925995422,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.114373532,
      "avg_duration_per_task_ms": 125231.99,
      "avg_output_tokens_per_task": 12239.046,
      "last_run_timestamp": "2026-10-10T11:19:31.087Z"
    },
    {
      "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": "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": 0.6628353908045977,
      "accuracy_std_dev": 0.06360316641360503,
      "run_count": 7,
      "total_questions": 348,
      "avg_cost_per_task_usd": 0.024118077586206897,
      "avg_duration_per_task_ms": 468405.7327586207,
      "avg_output_tokens_per_task": 16303.336206896553,
      "last_run_timestamp": "2026-10-10T19:08:25.349Z"
    },
    {
      "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Parasail",
      "accuracy": 0.6619998,
      "accuracy_std_dev": 0.015289659543626185,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.02312782,
      "avg_duration_per_task_ms": 165165.768,
      "avg_output_tokens_per_task": 17110.856,
      "last_run_timestamp": "2026-10-10T18:59:10.821Z"
    },
    {
      "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": 0.6600001428571428,
      "accuracy_std_dev": 0.024688651721727035,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.02290525142857143,
      "avg_duration_per_task_ms": 311367.0028571429,
      "avg_output_tokens_per_task": 10406.782857142856,
      "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": "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": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "SiliconFlow",
      "accuracy": 0.6588888333333334,
      "accuracy_std_dev": 0.02354406690595791,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.022377286666666666,
      "avg_duration_per_task_ms": 254720.16666666666,
      "avg_output_tokens_per_task": 9788.826666666666,
      "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": "DigitalOcean",
      "accuracy": 0.6580951428571429,
      "accuracy_std_dev": 0.018418057788319512,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.026633637142857144,
      "avg_duration_per_task_ms": 383567.4857142857,
      "avg_output_tokens_per_task": 10552.454285714286,
      "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": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "DigitalOcean",
      "accuracy": 0.6579147260273973,
      "accuracy_std_dev": 0.11208961231097506,
      "run_count": 9,
      "total_questions": 438,
      "avg_cost_per_task_usd": 0.36986931735159817,
      "avg_duration_per_task_ms": 1949962.3150684931,
      "avg_output_tokens_per_task": 62327.19634703196,
      "last_run_timestamp": "2026-10-10T11:38:58.690Z"
    },
    {
      "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": 0.6573333,
      "accuracy_std_dev": 0.028263311647375403,
      "run_count": 30,
      "total_questions": 1500,
      "avg_cost_per_task_usd": 0.12181550133333331,
      "avg_duration_per_task_ms": 120760.41533333334,
      "avg_output_tokens_per_task": 12566.805333333334,
      "last_run_timestamp": "2026-10-10T11:31:06.764Z"
    },
    {
      "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": 0.6561111666666667,
      "accuracy_std_dev": 0.04057057332359941,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.021655356666666667,
      "avg_duration_per_task_ms": 227541.03,
      "avg_output_tokens_per_task": 28375,
      "last_run_timestamp": "2026-10-10T11:34:56.282Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "Phala",
      "accuracy": 0.6557823102040816,
      "accuracy_std_dev": 0.009318408855593308,
      "run_count": 5,
      "total_questions": 245,
      "avg_cost_per_task_usd": 0.014797057142857142,
      "avg_duration_per_task_ms": 277907.2775510204,
      "avg_output_tokens_per_task": 20245.579591836733,
      "last_run_timestamp": "2026-10-10T11:18:54.565Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "AtlasCloud",
      "accuracy": 0.6540179419642858,
      "accuracy_std_dev": 0.10628475367719596,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.09064835491071428,
      "avg_duration_per_task_ms": 220062.35267857142,
      "avg_output_tokens_per_task": 23673.90401785714,
      "last_run_timestamp": "2026-10-10T18:54:19.769Z"
    },
    {
      "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": "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": 0.649933124497992,
      "accuracy_std_dev": 0.01992940230513702,
      "run_count": 5,
      "total_questions": 249,
      "avg_cost_per_task_usd": 0.01677981124497992,
      "avg_duration_per_task_ms": 331039.8714859438,
      "avg_output_tokens_per_task": 22769,
      "last_run_timestamp": "2026-10-10T13:36:34.540Z"
    },
    {
      "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": 0.6495237142857143,
      "accuracy_std_dev": 0.031240416581794835,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.021353940000000002,
      "avg_duration_per_task_ms": 237904.40285714285,
      "avg_output_tokens_per_task": 10393.82,
      "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": 0.6480123486238532,
      "accuracy_std_dev": 0.024081519327223473,
      "run_count": 11,
      "total_questions": 545,
      "avg_cost_per_task_usd": 0.01123864036697248,
      "avg_duration_per_task_ms": 311115.2660550459,
      "avg_output_tokens_per_task": 17090.746788990826,
      "last_run_timestamp": "2026-10-10T18:41:07.171Z"
    },
    {
      "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": 0.6480001000000001,
      "accuracy_std_dev": 0.02343796919722355,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.116462472,
      "avg_duration_per_task_ms": 144546.822,
      "avg_output_tokens_per_task": 12620.218,
      "last_run_timestamp": "2026-10-10T11:05:43.973Z"
    },
    {
      "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": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Novita",
      "accuracy": 0.6476190000000001,
      "accuracy_std_dev": 0.02121044324316276,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.02434872285714286,
      "avg_duration_per_task_ms": 154447.84,
      "avg_output_tokens_per_task": 10538.351428571428,
      "last_run_timestamp": "2026-08-29T16:49:08.574Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Inceptron",
      "accuracy": 0.6473332,
      "accuracy_std_dev": 0.027840964224681578,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.022036860000000002,
      "avg_duration_per_task_ms": 301895.372,
      "avg_output_tokens_per_task": 10556.636,
      "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": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "Parasail",
      "accuracy": 0.6471734736842106,
      "accuracy_std_dev": 0.08517975279659948,
      "run_count": 7,
      "total_questions": 342,
      "avg_cost_per_task_usd": 0.10989668713450293,
      "avg_duration_per_task_ms": 1476007.2105263157,
      "avg_output_tokens_per_task": 80545.20760233919,
      "last_run_timestamp": "2026-10-10T19:03:17.197Z"
    },
    {
      "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": "Groq",
      "accuracy": 0.6470371111111111,
      "accuracy_std_dev": 0.03160113383846446,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.03364751555555555,
      "avg_duration_per_task_ms": 80453.21555555555,
      "avg_output_tokens_per_task": 19157.16,
      "last_run_timestamp": "2026-10-10T18:38:13.047Z"
    },
    {
      "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": "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": 0.6458334054054053,
      "accuracy_std_dev": 0.03153139496156523,
      "run_count": 6,
      "total_questions": 296,
      "avg_cost_per_task_usd": 0.013519689189189188,
      "avg_duration_per_task_ms": 1091817.445945946,
      "avg_output_tokens_per_task": 22703.493243243243,
      "last_run_timestamp": "2026-10-10T14:47:01.449Z"
    },
    {
      "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Mara",
      "accuracy": 0.644208,
      "accuracy_std_dev": 0.03904293739234052,
      "run_count": 3,
      "total_questions": 141,
      "avg_cost_per_task_usd": 0.024240567375886524,
      "avg_duration_per_task_ms": 70321.02836879433,
      "avg_output_tokens_per_task": 14782.985815602837,
      "last_run_timestamp": "2026-10-10T18:42:24.416Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.4-nano-20260317",
      "model_slug": "openai/gpt-5.4-nano",
      "model_name": "OpenAI: GPT-5.4 Nano",
      "provider_name": "Azure",
      "accuracy": 0.6433335,
      "accuracy_std_dev": 0.014272415770172448,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.031060726666666667,
      "avg_duration_per_task_ms": 229987.72666666665,
      "avg_output_tokens_per_task": 16358.803333333333,
      "last_run_timestamp": "2026-10-01T19:05:16.543Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-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": "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": 0.642570281124498,
      "accuracy_std_dev": 0.017333401911915625,
      "run_count": 5,
      "total_questions": 249,
      "avg_cost_per_task_usd": 0.014325088353413656,
      "avg_duration_per_task_ms": 874847.4096385542,
      "avg_output_tokens_per_task": 23432.22891566265,
      "last_run_timestamp": "2026-10-10T18:57:22.987Z"
    },
    {
      "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": 0.64166675,
      "accuracy_std_dev": 0.028625974026004797,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.017551492499999998,
      "avg_duration_per_task_ms": 331507.07,
      "avg_output_tokens_per_task": 20713.525,
      "last_run_timestamp": "2026-10-10T19:31:53.275Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4-sonnet-20250522",
      "model_slug": "anthropic/claude-sonnet-4",
      "model_name": "Anthropic: Claude Sonnet 4",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.64,
      "accuracy_std_dev": 0.020000000000000018,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.29234208,
      "avg_duration_per_task_ms": 227202.07,
      "avg_output_tokens_per_task": 8702.01,
      "last_run_timestamp": "2026-09-01T13:00:51.629Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "ibm-granite/granite-4.2-8b-20260831",
      "model_slug": "ibm-granite/granite-4.2-8b",
      "model_name": "IBM: Granite 4.2 8B",
      "provider_name": "DeepInfra",
      "accuracy": 0.64,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.01684376,
      "avg_duration_per_task_ms": 1178898.4,
      "avg_output_tokens_per_task": 43776.78,
      "last_run_timestamp": "2026-10-09T00:39:20.903Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "GMICloud",
      "accuracy": 0.64,
      "accuracy_std_dev": 0.17706617541553699,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.1395266857142857,
      "avg_duration_per_task_ms": 380475.96857142856,
      "avg_output_tokens_per_task": 8284.165714285715,
      "last_run_timestamp": "2026-10-10T12:10:04.065Z"
    },
    {
      "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": 0.639765898245614,
      "accuracy_std_dev": 0.0817052994894864,
      "run_count": 6,
      "total_questions": 285,
      "avg_cost_per_task_usd": 0.13215583157894736,
      "avg_duration_per_task_ms": 352439.649122807,
      "avg_output_tokens_per_task": 10044.143859649123,
      "last_run_timestamp": "2026-10-10T18:48:59.120Z"
    },
    {
      "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": 0.6383332500000001,
      "accuracy_std_dev": 0.028431218171008805,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.216643625,
      "avg_duration_per_task_ms": 1122861.59,
      "avg_output_tokens_per_task": 30036.18,
      "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": "inclusionai/ling-3.0-flash-fin-20260827",
      "model_slug": "inclusionai/ling-3.0-flash-fin",
      "model_name": "inclusionAI: Ling 3.0 Flash Fin",
      "provider_name": "DeepInfra",
      "accuracy": 0.638298,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 47,
      "avg_cost_per_task_usd": 0.020538510638297872,
      "avg_duration_per_task_ms": 686659.4680851063,
      "avg_output_tokens_per_task": 47094.31914893617,
      "last_run_timestamp": "2026-10-09T00:23:16.829Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "DeepInfra",
      "accuracy": 0.6378857133182845,
      "accuracy_std_dev": 0.03816624687028554,
      "run_count": 18,
      "total_questions": 886,
      "avg_cost_per_task_usd": 0.021538330699774268,
      "avg_duration_per_task_ms": 306224.9401805869,
      "avg_output_tokens_per_task": 18801.29683972912,
      "last_run_timestamp": "2026-10-10T19:10:56.880Z"
    },
    {
      "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": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Parasail",
      "accuracy": 0.6346803232323233,
      "accuracy_std_dev": 0.01547599999999999,
      "run_count": 2,
      "total_questions": 99,
      "avg_cost_per_task_usd": 0.019095151515151516,
      "avg_duration_per_task_ms": 172977.9898989899,
      "avg_output_tokens_per_task": 9245.858585858587,
      "last_run_timestamp": "2026-08-01T17:40:52.342Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-9b-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": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "Venice",
      "accuracy": 0.631638420338983,
      "accuracy_std_dev": 0.06716132817452815,
      "run_count": 12,
      "total_questions": 590,
      "avg_cost_per_task_usd": 0.023603845762711864,
      "avg_duration_per_task_ms": 254670.11864406778,
      "avg_output_tokens_per_task": 24695.162711864406,
      "last_run_timestamp": "2026-10-10T19:04:00.310Z"
    },
    {
      "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": 0.6314284285714286,
      "accuracy_std_dev": 0.018418242461965675,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.028592208571428572,
      "avg_duration_per_task_ms": 225611.1542857143,
      "avg_output_tokens_per_task": 10242.23142857143,
      "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": "minimax/minimax-m2.1",
      "model_slug": "minimax/minimax-m2.1",
      "model_name": "MiniMax: MiniMax M2.1",
      "provider_name": "auto-routing",
      "accuracy": 0.631111,
      "accuracy_std_dev": 0.013698950130089036,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.02504141333333333,
      "avg_duration_per_task_ms": 168170.98,
      "avg_output_tokens_per_task": 11245.873333333333,
      "last_run_timestamp": "2026-08-01T17:38:10.364Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "CoreWeave",
      "accuracy": 0.63,
      "accuracy_std_dev": 0.023333000000000048,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.07672618,
      "avg_duration_per_task_ms": 231628.06,
      "avg_output_tokens_per_task": 10580.58,
      "last_run_timestamp": "2026-08-01T17:40:48.165Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.6266667499999999,
      "accuracy_std_dev": 0.012472414008021857,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.116286155,
      "avg_duration_per_task_ms": 669417.115,
      "avg_output_tokens_per_task": 30726.725,
      "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": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.6266665714285714,
      "accuracy_std_dev": 0.04047336572843623,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.09534962571428571,
      "avg_duration_per_task_ms": 55134.33714285714,
      "avg_output_tokens_per_task": 3606.0057142857145,
      "last_run_timestamp": "2026-10-10T18:59:05.793Z"
    },
    {
      "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": 0.6266665,
      "accuracy_std_dev": 0.013333499999999998,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.015634330000000002,
      "avg_duration_per_task_ms": 100213.58,
      "avg_output_tokens_per_task": 19899.75,
      "last_run_timestamp": "2026-10-10T18:54:14.146Z"
    },
    {
      "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": 0.6266663333333334,
      "accuracy_std_dev": 0.018856180831641284,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.02545802,
      "avg_duration_per_task_ms": 175220.38666666666,
      "avg_output_tokens_per_task": 11691.16,
      "last_run_timestamp": "2026-08-01T17:38:10.202Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-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": "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": 0.6174497919463087,
      "accuracy_std_dev": 0.011290615681716505,
      "run_count": 6,
      "total_questions": 298,
      "avg_cost_per_task_usd": 0.11355312416107383,
      "avg_duration_per_task_ms": 671627.1577181208,
      "avg_output_tokens_per_task": 32151.05033557047,
      "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": "Cerebras",
      "accuracy": 0.61708325,
      "accuracy_std_dev": 0.02524513292077309,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.064975285,
      "avg_duration_per_task_ms": 31392.6825,
      "avg_output_tokens_per_task": 20231.8325,
      "last_run_timestamp": "2026-10-10T18:04:52.147Z"
    },
    {
      "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": 0.6155553333333333,
      "accuracy_std_dev": 0.04228062043326968,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.09065885333333334,
      "avg_duration_per_task_ms": 148176.83333333334,
      "avg_output_tokens_per_task": 3889.28,
      "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": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "NextBit",
      "accuracy": 0.613333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.07264287999999999,
      "avg_duration_per_task_ms": 1482714.84,
      "avg_output_tokens_per_task": 23767.86,
      "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "SambaNova",
      "accuracy": 0.6131220995475114,
      "accuracy_std_dev": 0.03551941611896371,
      "run_count": 9,
      "total_questions": 442,
      "avg_cost_per_task_usd": 0.041041904977375566,
      "avg_duration_per_task_ms": 84582.88235294117,
      "avg_output_tokens_per_task": 23671.278280542985,
      "last_run_timestamp": "2026-10-10T18:32:41.404Z"
    },
    {
      "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": "thinkingmachines/inkling-small-20260730",
      "model_slug": "thinkingmachines/inkling-small",
      "model_name": "Thinking Machines: Inkling Small",
      "provider_name": "DeepInfra",
      "accuracy": 0.6076188571428572,
      "accuracy_std_dev": 0.03274303997331158,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.046068719999999994,
      "avg_duration_per_task_ms": 191059.44,
      "avg_output_tokens_per_task": 3405.154285714286,
      "last_run_timestamp": "2026-10-08T22:20:40.622Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.6-20251208",
      "model_slug": "z-ai/glm-4.6v",
      "model_name": "Z.ai: GLM 4.6V",
      "provider_name": "auto-routing",
      "accuracy": 0.606667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.03212956,
      "avg_duration_per_task_ms": 335001.52,
      "avg_output_tokens_per_task": 8851.18,
      "last_run_timestamp": "2026-10-01T08:03:01.921Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "thinkingmachines/inkling-small-20260730",
      "model_slug": "thinkingmachines/inkling-small",
      "model_name": "Thinking Machines: Inkling Small",
      "provider_name": "auto-routing",
      "accuracy": 0.6066668000000001,
      "accuracy_std_dev": 0.026666399999999996,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.046513716,
      "avg_duration_per_task_ms": 209556.196,
      "avg_output_tokens_per_task": 3673.824,
      "last_run_timestamp": "2026-10-01T16:00:32.452Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-2.5-pro",
      "model_slug": "google/gemini-2.5-pro",
      "model_name": "Google: Gemini 2.5 Pro",
      "provider_name": "Google",
      "accuracy": 0.6060269955357143,
      "accuracy_std_dev": 0.032348691330028045,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.19772456473214287,
      "avg_duration_per_task_ms": 188670.52455357142,
      "avg_output_tokens_per_task": 13047.183035714286,
      "last_run_timestamp": "2026-10-01T19:04:55.725Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Phala",
      "accuracy": 0.605442,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 49,
      "avg_cost_per_task_usd": 0.035960428571428574,
      "avg_duration_per_task_ms": 291757.14285714284,
      "avg_output_tokens_per_task": 21024.51020408163,
      "last_run_timestamp": "2026-10-10T19:16:28.931Z"
    },
    {
      "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": 0.6033332222222223,
      "accuracy_std_dev": 0.027397264890973577,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.04748484000000001,
      "avg_duration_per_task_ms": 91610.96,
      "avg_output_tokens_per_task": 4621.971111111111,
      "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": "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": "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": 0.600000125,
      "accuracy_std_dev": 0.06790519535064586,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.036252275,
      "avg_duration_per_task_ms": 434224.3475,
      "avg_output_tokens_per_task": 19344.175,
      "last_run_timestamp": "2026-10-10T20:39:12.141Z"
    },
    {
      "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": 0.5979899899497488,
      "accuracy_std_dev": 0.022095125194471644,
      "run_count": 8,
      "total_questions": 398,
      "avg_cost_per_task_usd": 0.020116060301507536,
      "avg_duration_per_task_ms": 100330.36180904522,
      "avg_output_tokens_per_task": 4353.20351758794,
      "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": "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": "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": 0.5913978951612903,
      "accuracy_std_dev": 0.012203964545998971,
      "run_count": 5,
      "total_questions": 248,
      "avg_cost_per_task_usd": 0.156355625,
      "avg_duration_per_task_ms": 554104.3709677419,
      "avg_output_tokens_per_task": 27726.576612903227,
      "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-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": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "AtlasCloud",
      "accuracy": 0.5856268495412844,
      "accuracy_std_dev": 0.2377490081349595,
      "run_count": 11,
      "total_questions": 545,
      "avg_cost_per_task_usd": 0.019255034862385324,
      "avg_duration_per_task_ms": 330742.1412844037,
      "avg_output_tokens_per_task": 9878.405504587156,
      "last_run_timestamp": "2026-10-10T12:29:44.559Z"
    },
    {
      "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-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Mancer 2",
      "accuracy": 0.5806451129032257,
      "accuracy_std_dev": 0.09293035051607196,
      "run_count": 5,
      "total_questions": 248,
      "avg_cost_per_task_usd": 0.038077282258064525,
      "avg_duration_per_task_ms": 2550353.040322581,
      "avg_output_tokens_per_task": 78807.90322580645,
      "last_run_timestamp": "2026-10-10T19:59:40.914Z"
    },
    {
      "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": "qwen/qwen3-coder-next-2025-02-03",
      "model_slug": "qwen/qwen3-coder-next",
      "model_name": "Qwen: Qwen3 Coder Next",
      "provider_name": "Alibaba",
      "accuracy": 0.579037868556701,
      "accuracy_std_dev": 0.04024006578337068,
      "run_count": 8,
      "total_questions": 388,
      "avg_cost_per_task_usd": 0.07504983247422681,
      "avg_duration_per_task_ms": 90151.55927835051,
      "avg_output_tokens_per_task": 5664.682989690722,
      "last_run_timestamp": "2026-10-03T16:05:18.093Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "SiliconFlow",
      "accuracy": 0.57416675,
      "accuracy_std_dev": 0.12703624537484373,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0368084625,
      "avg_duration_per_task_ms": 399880.175,
      "avg_output_tokens_per_task": 8103.195,
      "last_run_timestamp": "2026-10-10T19:10:33.654Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2-0905",
      "model_slug": "moonshotai/kimi-k2-0905",
      "model_name": "MoonshotAI: Kimi K2 0905",
      "provider_name": "auto-routing",
      "accuracy": 0.572072,
      "accuracy_std_dev": 0.02796372233607127,
      "run_count": 3,
      "total_questions": 148,
      "avg_cost_per_task_usd": 0.09883952702702703,
      "avg_duration_per_task_ms": 105307.2027027027,
      "avg_output_tokens_per_task": 3660.560810810811,
      "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Novita",
      "accuracy": 0.5713488862876254,
      "accuracy_std_dev": 0.03528059746478538,
      "run_count": 6,
      "total_questions": 299,
      "avg_cost_per_task_usd": 0.011061280936454851,
      "avg_duration_per_task_ms": 252847.0033444816,
      "avg_output_tokens_per_task": 17983.45150501672,
      "last_run_timestamp": "2026-10-10T19:08:16.520Z"
    },
    {
      "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": 0.5712907976878613,
      "accuracy_std_dev": 0.043538142708785955,
      "run_count": 7,
      "total_questions": 346,
      "avg_cost_per_task_usd": 0.18383380635838154,
      "avg_duration_per_task_ms": 1966033.9104046242,
      "avg_output_tokens_per_task": 40033.85260115607,
      "last_run_timestamp": "2026-10-10T19:23:34.524Z"
    },
    {
      "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": "google/gemini-2.5-flash",
      "model_slug": "google/gemini-2.5-flash",
      "model_name": "Google: Gemini 2.5 Flash",
      "provider_name": "Google AI Studio",
      "accuracy": 0.5511113333333334,
      "accuracy_std_dev": 0.006285236475706844,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.035753513333333334,
      "avg_duration_per_task_ms": 61316.40666666667,
      "avg_output_tokens_per_task": 6659.566666666667,
      "last_run_timestamp": "2026-10-01T19:04:42.483Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Darkbloom",
      "accuracy": 0.5499152233502539,
      "accuracy_std_dev": 0.02918541097342131,
      "run_count": 4,
      "total_questions": 197,
      "avg_cost_per_task_usd": 0.0026568781725888327,
      "avg_duration_per_task_ms": 406357.32994923857,
      "avg_output_tokens_per_task": 6485.319796954314,
      "last_run_timestamp": "2026-10-10T19:11:30.565Z"
    },
    {
      "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": 0.54833325,
      "accuracy_std_dev": 0.09993065790305547,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0400913925,
      "avg_duration_per_task_ms": 97587.1625,
      "avg_output_tokens_per_task": 4289.61,
      "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": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "auto-routing",
      "accuracy": 0.5460636532663317,
      "accuracy_std_dev": 0.07718163585497526,
      "run_count": 8,
      "total_questions": 398,
      "avg_cost_per_task_usd": 0.05313004020100502,
      "avg_duration_per_task_ms": 465636.0527638191,
      "avg_output_tokens_per_task": 9251.170854271357,
      "last_run_timestamp": "2026-10-10T18:37:07.092Z"
    },
    {
      "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": 0.5416667142857143,
      "accuracy_std_dev": 0.10219251042847255,
      "run_count": 7,
      "total_questions": 336,
      "avg_cost_per_task_usd": 0.06745207738095238,
      "avg_duration_per_task_ms": 446328.3333333333,
      "avg_output_tokens_per_task": 28269.869047619046,
      "last_run_timestamp": "2026-08-29T08:10:31.240Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "auto-routing",
      "accuracy": 0.5377905988372093,
      "accuracy_std_dev": 0.016893313972681363,
      "run_count": 7,
      "total_questions": 344,
      "avg_cost_per_task_usd": 0.02223077034883721,
      "avg_duration_per_task_ms": 1431254.4069767443,
      "avg_output_tokens_per_task": 108185.83430232559,
      "last_run_timestamp": "2026-10-10T19:31:51.356Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Nebius",
      "accuracy": 0.5333335,
      "accuracy_std_dev": 0.026666500000000037,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.10219375,
      "avg_duration_per_task_ms": 375443.12,
      "avg_output_tokens_per_task": 17495.06,
      "last_run_timestamp": "2026-08-15T14:16:11.776Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "WandB",
      "accuracy": 0.526667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.02328844,
      "avg_duration_per_task_ms": 1276910.42,
      "avg_output_tokens_per_task": 153904.82,
      "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-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": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "SiliconFlow",
      "accuracy": 0.5223215892857143,
      "accuracy_std_dev": 0.017182300915926027,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.06601706026785713,
      "avg_duration_per_task_ms": 301611.93526785716,
      "avg_output_tokens_per_task": 4026.2254464285716,
      "last_run_timestamp": "2026-10-10T18:28:10.194Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4.1-2025-04-14",
      "model_slug": "openai/gpt-4.1",
      "model_name": "OpenAI: GPT-4.1",
      "provider_name": "Azure",
      "accuracy": 0.5221089591836734,
      "accuracy_std_dev": 0.023889008586586408,
      "run_count": 4,
      "total_questions": 196,
      "avg_cost_per_task_usd": 0.13644156632653062,
      "avg_duration_per_task_ms": 73125.92346938775,
      "avg_output_tokens_per_task": 2553.7397959183672,
      "last_run_timestamp": "2026-10-01T19:05:18.003Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "CoreWeave",
      "accuracy": 0.5188173145161291,
      "accuracy_std_dev": 0.2566565496323832,
      "run_count": 5,
      "total_questions": 248,
      "avg_cost_per_task_usd": 0.007278975806451614,
      "avg_duration_per_task_ms": 482880.30241935485,
      "avg_output_tokens_per_task": 15769.91935483871,
      "last_run_timestamp": "2026-10-10T18:54:19.893Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3-nano-30b-a3b",
      "model_slug": "nvidia/nemotron-3-nano-30b-a3b",
      "model_name": "NVIDIA: Nemotron 3 Nano 30B A3B",
      "provider_name": "auto-routing",
      "accuracy": 0.517500125,
      "accuracy_std_dev": 0.024930426803594342,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0282700725,
      "avg_duration_per_task_ms": 444815,
      "avg_output_tokens_per_task": 58343.125,
      "last_run_timestamp": "2026-10-03T16:09:19.689Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.511079063583815,
      "accuracy_std_dev": 0.04011160215484506,
      "run_count": 7,
      "total_questions": 346,
      "avg_cost_per_task_usd": 0.155033725433526,
      "avg_duration_per_task_ms": 1878039.6213872833,
      "avg_output_tokens_per_task": 35119.08670520231,
      "last_run_timestamp": "2026-10-10T19:22:46.804Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.1-terminus",
      "model_slug": "deepseek/deepseek-v3.1-terminus",
      "model_name": "DeepSeek: DeepSeek V3.1 Terminus",
      "provider_name": "DeepInfra",
      "accuracy": 0.5104164999999999,
      "accuracy_std_dev": 0.003472500000000045,
      "run_count": 2,
      "total_questions": 96,
      "avg_cost_per_task_usd": 0.024820281250000003,
      "avg_duration_per_task_ms": 480070.4583333333,
      "avg_output_tokens_per_task": 9578.8125,
      "last_run_timestamp": "2026-08-08T16:03:43.735Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4.1-2025-04-14",
      "model_slug": "openai/gpt-4.1",
      "model_name": "OpenAI: GPT-4.1",
      "provider_name": "OpenAI",
      "accuracy": 0.5083332500000001,
      "accuracy_std_dev": 0.026404802805313633,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.12635733,
      "avg_duration_per_task_ms": 53878.245,
      "avg_output_tokens_per_task": 2800.47,
      "last_run_timestamp": "2026-10-02T23:07:04.603Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Together",
      "accuracy": 0.5055554444444444,
      "accuracy_std_dev": 0.03471227413117765,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.02842421111111111,
      "avg_duration_per_task_ms": 279218.0311111111,
      "avg_output_tokens_per_task": 19993.748888888887,
      "last_run_timestamp": "2026-10-10T18:54:26.767Z"
    },
    {
      "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": 0.5045660136986302,
      "accuracy_std_dev": 0.02732875245760204,
      "run_count": 3,
      "total_questions": 146,
      "avg_cost_per_task_usd": 0.05861860273972602,
      "avg_duration_per_task_ms": 2642554.2671232875,
      "avg_output_tokens_per_task": 40736.36301369863,
      "last_run_timestamp": "2026-08-15T14:12:30.435Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-2.5-flash-lite",
      "model_slug": "google/gemini-2.5-flash-lite",
      "model_name": "Google: Gemini 2.5 Flash Lite",
      "provider_name": "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-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "DeepInfra",
      "accuracy": 0.5013551951219513,
      "accuracy_std_dev": 0.01245431828082131,
      "run_count": 5,
      "total_questions": 246,
      "avg_cost_per_task_usd": 0.020900845528455284,
      "avg_duration_per_task_ms": 1272391.512195122,
      "avg_output_tokens_per_task": 112484.65853658537,
      "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": "qwen/qwen3-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "CoreWeave",
      "accuracy": 0.49666650000000007,
      "accuracy_std_dev": 0.023333500000000007,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.23109727999999996,
      "avg_duration_per_task_ms": 80168.37,
      "avg_output_tokens_per_task": 2746.03,
      "last_run_timestamp": "2026-08-01T17:41:27.708Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "Mancer 2",
      "accuracy": 0.493333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.14235776,
      "avg_duration_per_task_ms": 443941.74,
      "avg_output_tokens_per_task": 8833.66,
      "last_run_timestamp": "2026-08-06T00:00:02.908Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.1-terminus",
      "model_slug": "deepseek/deepseek-v3.1-terminus",
      "model_name": "DeepSeek: DeepSeek V3.1 Terminus",
      "provider_name": "SiliconFlow",
      "accuracy": 0.48888899999999996,
      "accuracy_std_dev": 0.01872462563043652,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.06696013,
      "avg_duration_per_task_ms": 315550.43333333335,
      "avg_output_tokens_per_task": 4089.9266666666667,
      "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": "nvidia/nemotron-3-nano-30b-a3b",
      "model_slug": "nvidia/nemotron-3-nano-30b-a3b",
      "model_name": "NVIDIA: Nemotron 3 Nano 30B A3B",
      "provider_name": "Nebius",
      "accuracy": 0.48833333333333334,
      "accuracy_std_dev": 0.018831396165859726,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.03323320666666667,
      "avg_duration_per_task_ms": 665027.6466666666,
      "avg_output_tokens_per_task": 61694.416666666664,
      "last_run_timestamp": "2026-10-03T16:04:32.703Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3-nano-30b-a3b",
      "model_slug": "nvidia/nemotron-3-nano-30b-a3b",
      "model_name": "NVIDIA: Nemotron 3 Nano 30B A3B",
      "provider_name": "Crusoe",
      "accuracy": 0.4852150080645161,
      "accuracy_std_dev": 0.023420064714684302,
      "run_count": 10,
      "total_questions": 496,
      "avg_cost_per_task_usd": 0.028619552419354842,
      "avg_duration_per_task_ms": 386610.2439516129,
      "avg_output_tokens_per_task": 66355.3125,
      "last_run_timestamp": "2026-10-03T16:14:03.053Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-32b-04-28",
      "model_slug": "qwen/qwen3-32b",
      "model_name": "Qwen: Qwen3 32B",
      "provider_name": "Nebius",
      "accuracy": 0.4842343648648649,
      "accuracy_std_dev": 0.02401659838297026,
      "run_count": 6,
      "total_questions": 296,
      "avg_cost_per_task_usd": 0.023849905405405405,
      "avg_duration_per_task_ms": 836568.3851351351,
      "avg_output_tokens_per_task": 22266.58108108108,
      "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": "google/gemma-3-12b-it",
      "model_slug": "google/gemma-3-12b-it",
      "model_name": "Google: Gemma 3 12B",
      "provider_name": "DeepInfra",
      "accuracy": 0.48,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.011001320000000002,
      "avg_duration_per_task_ms": 80205.64,
      "avg_output_tokens_per_task": 3762.88,
      "last_run_timestamp": "2026-10-08T22:08:03.352Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-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": 0.48,
      "accuracy_std_dev": 0.0241687453956551,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.034828557142857146,
      "avg_duration_per_task_ms": 43808.188571428575,
      "avg_output_tokens_per_task": 3663.8714285714286,
      "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-vl-235b-a22b-instruct",
      "model_slug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_name": "Qwen: Qwen3 VL 235B A22B Instruct",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.48,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.07899766,
      "avg_duration_per_task_ms": 132704.5,
      "avg_output_tokens_per_task": 3161.38,
      "last_run_timestamp": "2026-10-04T03:00:16.401Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Crusoe",
      "accuracy": 0.47916700000000007,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 96,
      "avg_cost_per_task_usd": 0.04362447916666667,
      "avg_duration_per_task_ms": 66953.94791666667,
      "avg_output_tokens_per_task": 2828.96875,
      "last_run_timestamp": "2026-08-15T14:08:14.681Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.4776285234899329,
      "accuracy_std_dev": 0.010657414972163225,
      "run_count": 6,
      "total_questions": 298,
      "avg_cost_per_task_usd": 0.1647015033557047,
      "avg_duration_per_task_ms": 2471417.852348993,
      "avg_output_tokens_per_task": 44160.27181208054,
      "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": 0.47583349999999996,
      "accuracy_std_dev": 0.03650536870790377,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.02416941,
      "avg_duration_per_task_ms": 59754.7125,
      "avg_output_tokens_per_task": 3775.0125,
      "last_run_timestamp": "2026-10-03T16:10:49.378Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-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": 0.47333325000000004,
      "accuracy_std_dev": 0.025603927651583064,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0251383825,
      "avg_duration_per_task_ms": 126301.635,
      "avg_output_tokens_per_task": 3791.835,
      "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-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": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.03429832,
      "avg_duration_per_task_ms": 458175.58,
      "avg_output_tokens_per_task": 37835.14,
      "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": 0.4728994552845529,
      "accuracy_std_dev": 0.04759788478073367,
      "run_count": 5,
      "total_questions": 246,
      "avg_cost_per_task_usd": 0.02759002032520325,
      "avg_duration_per_task_ms": 2332100.9227642277,
      "avg_output_tokens_per_task": 138904.24390243902,
      "last_run_timestamp": "2026-10-10T19:06:47.154Z"
    },
    {
      "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": 0.4724461814516129,
      "accuracy_std_dev": 0.0356696342375416,
      "run_count": 5,
      "total_questions": 248,
      "avg_cost_per_task_usd": 0.015281346774193549,
      "avg_duration_per_task_ms": 122771.32258064517,
      "avg_output_tokens_per_task": 3293.3064516129034,
      "last_run_timestamp": "2026-10-10T18:31:25.736Z"
    },
    {
      "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": 0.4723047626112759,
      "accuracy_std_dev": 0.03812044880134972,
      "run_count": 7,
      "total_questions": 337,
      "avg_cost_per_task_usd": 0.05108310979228487,
      "avg_duration_per_task_ms": 212808.3293768546,
      "avg_output_tokens_per_task": 93380.04451038576,
      "last_run_timestamp": "2026-10-10T18:37:46.039Z"
    },
    {
      "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": 0.472222,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 48,
      "avg_cost_per_task_usd": 0.02306027083333333,
      "avg_duration_per_task_ms": 25624.729166666668,
      "avg_output_tokens_per_task": 1174.2916666666667,
      "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/qwen3-32b-04-28",
      "model_slug": "qwen/qwen3-32b",
      "model_name": "Qwen: Qwen3 32B",
      "provider_name": "auto-routing",
      "accuracy": 0.47047614285714284,
      "accuracy_std_dev": 0.020426164806293848,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.021544591428571434,
      "avg_duration_per_task_ms": 680725.9685714286,
      "avg_output_tokens_per_task": 21489.04857142857,
      "last_run_timestamp": "2026-08-29T16:16:50.456Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-next-80b-a3b-instruct-2509",
      "model_slug": "qwen/qwen3-next-80b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Next 80B A3B Instruct",
      "provider_name": "Parasail",
      "accuracy": 0.46952399999999994,
      "accuracy_std_dev": 0.034523096873170074,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.018877348571428573,
      "avg_duration_per_task_ms": 90973.66571428571,
      "avg_output_tokens_per_task": 3606.62,
      "last_run_timestamp": "2026-10-03T16:12:20.062Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "auto-routing",
      "accuracy": 0.46833325000000003,
      "accuracy_std_dev": 0.03295620410389368,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.09710940000000001,
      "avg_duration_per_task_ms": 99117.8225,
      "avg_output_tokens_per_task": 2749.945,
      "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-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": 0.465926,
      "accuracy_std_dev": 0.03230940398363575,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.022285648888888884,
      "avg_duration_per_task_ms": 222200.71555555557,
      "avg_output_tokens_per_task": 3598.428888888889,
      "last_run_timestamp": "2026-10-08T22:47:27.753Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Coder 30B A3B Instruct",
      "provider_name": "DigitalOcean",
      "accuracy": 0.46444433333333335,
      "accuracy_std_dev": 0.008314853589945066,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.033371,
      "avg_duration_per_task_ms": 305705.87333333335,
      "avg_output_tokens_per_task": 3594.2266666666665,
      "last_run_timestamp": "2026-08-15T08:15:42.180Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4o",
      "model_slug": "openai/gpt-4o",
      "model_name": "OpenAI: GPT-4o",
      "provider_name": "Azure",
      "accuracy": 0.463333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.26064434,
      "avg_duration_per_task_ms": 31467.4,
      "avg_output_tokens_per_task": 1819.56,
      "last_run_timestamp": "2026-08-01T08:06:06.230Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.45814811111111114,
      "accuracy_std_dev": 0.03865370011164504,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.05284759333333334,
      "avg_duration_per_task_ms": 162308.54,
      "avg_output_tokens_per_task": 3744.4333333333334,
      "last_run_timestamp": "2026-10-10T19:04:15.525Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Coder 30B A3B Instruct",
      "provider_name": "Novita",
      "accuracy": 0.45777799999999996,
      "accuracy_std_dev": 0.008314764498569207,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.00622334,
      "avg_duration_per_task_ms": 39170.17333333333,
      "avg_output_tokens_per_task": 1174.36,
      "last_run_timestamp": "2026-08-15T08:16:13.926Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "StreamLake",
      "accuracy": 0.4577777777777778,
      "accuracy_std_dev": 0.04455537267959145,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.05422914222222222,
      "avg_duration_per_task_ms": 126062.82666666666,
      "avg_output_tokens_per_task": 3580.0733333333333,
      "last_run_timestamp": "2026-10-03T16:19:54.632Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Google",
      "accuracy": 0.4570707575757576,
      "accuracy_std_dev": 0.044267226719521297,
      "run_count": 8,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.05378886363636363,
      "avg_duration_per_task_ms": 73384.54040404041,
      "avg_output_tokens_per_task": 3408.1742424242425,
      "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": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "Google",
      "accuracy": 0.4560182083333333,
      "accuracy_std_dev": 0.022912149746368182,
      "run_count": 3,
      "total_questions": 144,
      "avg_cost_per_task_usd": 0.08057918055555555,
      "avg_duration_per_task_ms": 52530.756944444445,
      "avg_output_tokens_per_task": 3473.4930555555557,
      "last_run_timestamp": "2026-08-15T14:15:58.674Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-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/qwen3-32b-04-28",
      "model_slug": "qwen/qwen3-32b",
      "model_name": "Qwen: Qwen3 32B",
      "provider_name": "DeepInfra",
      "accuracy": 0.45095228571428575,
      "accuracy_std_dev": 0.02741314960126298,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.020993765714285715,
      "avg_duration_per_task_ms": 566234.7285714286,
      "avg_output_tokens_per_task": 21154.594285714287,
      "last_run_timestamp": "2026-10-09T00:21:42.483Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-30b-a3b-04-28",
      "model_slug": "qwen/qwen3-30b-a3b",
      "model_name": "Qwen: Qwen3 30B A3B",
      "provider_name": "DeepInfra",
      "accuracy": 0.45,
      "accuracy_std_dev": 0.01784692109767581,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.03456686,
      "avg_duration_per_task_ms": 476526.7733333333,
      "avg_output_tokens_per_task": 23360.6,
      "last_run_timestamp": "2026-10-08T23:09:09.727Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "Novita",
      "accuracy": 0.44750012499999997,
      "accuracy_std_dev": 0.06501608615073176,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.16146158749999998,
      "avg_duration_per_task_ms": 120680.0125,
      "avg_output_tokens_per_task": 3281.1225,
      "last_run_timestamp": "2026-10-03T16:19:14.095Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Venice",
      "accuracy": 0.447037,
      "accuracy_std_dev": 0.02899323273915255,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.04068735111111112,
      "avg_duration_per_task_ms": 283005.94444444444,
      "avg_output_tokens_per_task": 3628.5377777777776,
      "last_run_timestamp": "2026-10-10T19:42:35.202Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-30b-a3b-04-28",
      "model_slug": "qwen/qwen3-30b-a3b",
      "model_name": "Qwen: Qwen3 30B A3B",
      "provider_name": "Alibaba",
      "accuracy": 0.4444443333333334,
      "accuracy_std_dev": 0.030952277087298266,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.03472188,
      "avg_duration_per_task_ms": 283560.81333333335,
      "avg_output_tokens_per_task": 21592.06,
      "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-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": 0.4441665,
      "accuracy_std_dev": 0.023555566693459103,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.03439918,
      "avg_duration_per_task_ms": 113438.9075,
      "avg_output_tokens_per_task": 3146.3175,
      "last_run_timestamp": "2026-10-03T16:18:20.377Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-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": "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": 0.44186027325581395,
      "accuracy_std_dev": 0.0336023614676719,
      "run_count": 7,
      "total_questions": 344,
      "avg_cost_per_task_usd": 0.05302273255813953,
      "avg_duration_per_task_ms": 228343.90697674418,
      "avg_output_tokens_per_task": 2116.357558139535,
      "last_run_timestamp": "2026-10-05T20:09:54.588Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-chat-v3-0324",
      "model_slug": "deepseek/deepseek-chat-v3-0324",
      "model_name": "DeepSeek: DeepSeek V3 0324",
      "provider_name": "SiliconFlow",
      "accuracy": 0.4416665,
      "accuracy_std_dev": 0.017873086533948188,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.104832345,
      "avg_duration_per_task_ms": 890298.59,
      "avg_output_tokens_per_task": 21334.155,
      "last_run_timestamp": "2026-08-29T16:10:09.206Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.4408060680100756,
      "accuracy_std_dev": 0.12174978780973233,
      "run_count": 16,
      "total_questions": 794,
      "avg_cost_per_task_usd": 0.03954040302267003,
      "avg_duration_per_task_ms": 655917.3727959698,
      "avg_output_tokens_per_task": 177113.59697733,
      "last_run_timestamp": "2026-10-10T18:58:33.701Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4o-2024-05-13",
      "model_slug": "openai/gpt-4o-2024-05-13",
      "model_name": "OpenAI: GPT-4o (2024-05-13)",
      "provider_name": "OpenAI",
      "accuracy": 0.44,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.7291186999999999,
      "avg_duration_per_task_ms": 35507.54,
      "avg_output_tokens_per_task": 3272.36,
      "last_run_timestamp": "2026-08-01T08:05:57.702Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "WandB",
      "accuracy": 0.44,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.2288897,
      "avg_duration_per_task_ms": 83836.68,
      "avg_output_tokens_per_task": 2734.56,
      "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": "Google",
      "accuracy": 0.43999974999999997,
      "accuracy_std_dev": 0.047081383440671115,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.050863820000000004,
      "avg_duration_per_task_ms": 37018.795,
      "avg_output_tokens_per_task": 2487.295,
      "last_run_timestamp": "2026-10-03T16:12:29.903Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "Venice",
      "accuracy": 0.43869733333333333,
      "accuracy_std_dev": 0.023050038409813497,
      "run_count": 7,
      "total_questions": 348,
      "avg_cost_per_task_usd": 0.01657819540229885,
      "avg_duration_per_task_ms": 200645.3132183908,
      "avg_output_tokens_per_task": 2785.6264367816093,
      "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/qwen3.6-35b-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": "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": "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": "auto-routing",
      "accuracy": 0.4362416308724832,
      "accuracy_std_dev": 0.02031037140436711,
      "run_count": 6,
      "total_questions": 298,
      "avg_cost_per_task_usd": 0.0464321610738255,
      "avg_duration_per_task_ms": 136873.4798657718,
      "avg_output_tokens_per_task": 2090.986577181208,
      "last_run_timestamp": "2026-08-29T16:33:29.885Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "BaseTen",
      "accuracy": 0.4358108783783784,
      "accuracy_std_dev": 0.3084304226783661,
      "run_count": 6,
      "total_questions": 296,
      "avg_cost_per_task_usd": 0.015068885135135136,
      "avg_duration_per_task_ms": 91458.4831081081,
      "avg_output_tokens_per_task": 15700.83445945946,
      "last_run_timestamp": "2026-10-10T18:32:50.715Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-chat-v3-0324",
      "model_slug": "deepseek/deepseek-chat-v3-0324",
      "model_name": "DeepSeek: DeepSeek V3 0324",
      "provider_name": "auto-routing",
      "accuracy": 0.435,
      "accuracy_std_dev": 0.020749805890658335,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.074401825,
      "avg_duration_per_task_ms": 778012.545,
      "avg_output_tokens_per_task": 20313.29,
      "last_run_timestamp": "2026-08-29T16:11:46.632Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-chat-v3",
      "model_slug": "deepseek/deepseek-chat",
      "model_name": "DeepSeek: DeepSeek V3",
      "provider_name": "DeepInfra",
      "accuracy": 0.43434354545454545,
      "accuracy_std_dev": 0.022037645087440712,
      "run_count": 4,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.06906736868686869,
      "avg_duration_per_task_ms": 323353.22222222225,
      "avg_output_tokens_per_task": 3378.3585858585857,
      "last_run_timestamp": "2026-10-08T22:54:33.845Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "Alibaba",
      "accuracy": 0.43416675,
      "accuracy_std_dev": 0.04222137919570487,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.400907955,
      "avg_duration_per_task_ms": 107994.6975,
      "avg_output_tokens_per_task": 3151.7075,
      "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": "openai/gpt-4o",
      "model_slug": "openai/gpt-4o",
      "model_name": "OpenAI: GPT-4o",
      "provider_name": "OpenAI",
      "accuracy": 0.43416662500000003,
      "accuracy_std_dev": 0.01613729505630899,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.2170375225,
      "avg_duration_per_task_ms": 46329.7025,
      "avg_output_tokens_per_task": 2265.4175,
      "last_run_timestamp": "2026-10-02T23:10:05.271Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-chat-v3-0324",
      "model_slug": "deepseek/deepseek-chat-v3-0324",
      "model_name": "DeepSeek: DeepSeek V3 0324",
      "provider_name": "DeepInfra",
      "accuracy": 0.4337431393442623,
      "accuracy_std_dev": 0.021677717946315293,
      "run_count": 5,
      "total_questions": 244,
      "avg_cost_per_task_usd": 0.032824139344262294,
      "avg_duration_per_task_ms": 478154.40573770495,
      "avg_output_tokens_per_task": 4694.745901639344,
      "last_run_timestamp": "2026-08-22T16:13:43.431Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Coder 30B A3B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.4311113333333333,
      "accuracy_std_dev": 0.006285236475706817,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.023601506666666664,
      "avg_duration_per_task_ms": 216917.58666666667,
      "avg_output_tokens_per_task": 3146.213333333333,
      "last_run_timestamp": "2026-08-15T08:02:48.904Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-14b-04-28",
      "model_slug": "qwen/qwen3-14b",
      "model_name": "Qwen: Qwen3 14B",
      "provider_name": "auto-routing",
      "accuracy": 0.431111,
      "accuracy_std_dev": 0.013698950130089036,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.029769013333333334,
      "avg_duration_per_task_ms": 630152.1266666667,
      "avg_output_tokens_per_task": 18760.18666666667,
      "last_run_timestamp": "2026-10-01T16:00:49.878Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta-llama/llama-4-maverick-17b-128e-instruct",
      "model_slug": "meta-llama/llama-4-maverick",
      "model_name": "Meta: Llama 4 Maverick",
      "provider_name": "Parasail",
      "accuracy": 0.43089426016260163,
      "accuracy_std_dev": 0.03483555910043644,
      "run_count": 5,
      "total_questions": 246,
      "avg_cost_per_task_usd": 0.04202976422764227,
      "avg_duration_per_task_ms": 58240.52032520325,
      "avg_output_tokens_per_task": 2134.682926829268,
      "last_run_timestamp": "2026-08-29T16:52:55.844Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-14b-04-28",
      "model_slug": "qwen/qwen3-14b",
      "model_name": "Qwen: Qwen3 14B",
      "provider_name": "DeepInfra",
      "accuracy": 0.43083325,
      "accuracy_std_dev": 0.031213896924086562,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.029063365,
      "avg_duration_per_task_ms": 761962.965,
      "avg_output_tokens_per_task": 19001.86,
      "last_run_timestamp": "2026-10-08T22:41:10.664Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4o-2024-08-06",
      "model_slug": "openai/gpt-4o-2024-08-06",
      "model_name": "OpenAI: GPT-4o (2024-08-06)",
      "provider_name": "OpenAI",
      "accuracy": 0.43,
      "accuracy_std_dev": 0.0033330000000000026,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.24811575000000002,
      "avg_duration_per_task_ms": 46287.18,
      "avg_output_tokens_per_task": 2565.8,
      "last_run_timestamp": "2026-10-02T23:09:02.629Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-30b-a3b-04-28",
      "model_slug": "qwen/qwen3-30b-a3b",
      "model_name": "Qwen: Qwen3 30B A3B",
      "provider_name": "auto-routing",
      "accuracy": 0.42666699999999996,
      "accuracy_std_dev": 0.04000000000000001,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.031438360000000005,
      "avg_duration_per_task_ms": 468393.89,
      "avg_output_tokens_per_task": 22584.78,
      "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/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Novita",
      "accuracy": 0.425833375,
      "accuracy_std_dev": 0.013616710018737092,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.024625232500000004,
      "avg_duration_per_task_ms": 228841.615,
      "avg_output_tokens_per_task": 3561.265,
      "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": "AtlasCloud",
      "accuracy": 0.4247621428571428,
      "accuracy_std_dev": 0.04058525139022997,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.05475465714285714,
      "avg_duration_per_task_ms": 141536.79714285713,
      "avg_output_tokens_per_task": 3538.397142857143,
      "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/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "DeepInfra",
      "accuracy": 0.42372316935483867,
      "accuracy_std_dev": 0.031047363202694044,
      "run_count": 10,
      "total_questions": 496,
      "avg_cost_per_task_usd": 0.025069254032258063,
      "avg_duration_per_task_ms": 357367.3891129032,
      "avg_output_tokens_per_task": 3605.673387096774,
      "last_run_timestamp": "2026-10-10T19:22:29.626Z"
    },
    {
      "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": 0.42144822950819677,
      "accuracy_std_dev": 0.025118027628776907,
      "run_count": 10,
      "total_questions": 488,
      "avg_cost_per_task_usd": 0.050889221311475406,
      "avg_duration_per_task_ms": 182960.04508196723,
      "avg_output_tokens_per_task": 4075.3032786885246,
      "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": "DeepInfra",
      "accuracy": 0.4170453954545455,
      "accuracy_std_dev": 0.02305814667528085,
      "run_count": 9,
      "total_questions": 440,
      "avg_cost_per_task_usd": 0.031998124999999995,
      "avg_duration_per_task_ms": 704757.9227272727,
      "avg_output_tokens_per_task": 7358.236363636363,
      "last_run_timestamp": "2026-10-10T18:59:12.194Z"
    },
    {
      "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": 0.4166667045454545,
      "accuracy_std_dev": 0.029461897667544087,
      "run_count": 9,
      "total_questions": 440,
      "avg_cost_per_task_usd": 0.011195077272727273,
      "avg_duration_per_task_ms": 112204.60909090909,
      "avg_output_tokens_per_task": 3271.315909090909,
      "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": "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": "openai/gpt-4.1-mini-2025-04-14",
      "model_slug": "openai/gpt-4.1-mini",
      "model_name": "OpenAI: GPT-4.1 Mini",
      "provider_name": "Azure",
      "accuracy": 0.40777783333333334,
      "accuracy_std_dev": 0.01781244190087991,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.025491406666666664,
      "avg_duration_per_task_ms": 73232.91333333333,
      "avg_output_tokens_per_task": 2860.7966666666666,
      "last_run_timestamp": "2026-10-01T19:05:00.133Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.406667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.06754938,
      "avg_duration_per_task_ms": 114529.78,
      "avg_output_tokens_per_task": 3789.7,
      "last_run_timestamp": "2026-10-04T03:00:16.402Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.40541675,
      "accuracy_std_dev": 0.021209148643627825,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.06368001,
      "avg_duration_per_task_ms": 101217.8175,
      "avg_output_tokens_per_task": 2945.865,
      "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": "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/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "auto-routing",
      "accuracy": 0.4051852222222222,
      "accuracy_std_dev": 0.12207173923538002,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.02358350666666667,
      "avg_duration_per_task_ms": 188827.18,
      "avg_output_tokens_per_task": 3348.0044444444443,
      "last_run_timestamp": "2026-10-10T18:35:31.926Z"
    },
    {
      "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": 0.40333325000000003,
      "accuracy_std_dev": 0.034960136029304874,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.05923553749999999,
      "avg_duration_per_task_ms": 320941.4325,
      "avg_output_tokens_per_task": 7834.6225,
      "last_run_timestamp": "2026-10-03T16:05:51.296Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "SiliconFlow",
      "accuracy": 0.402500125,
      "accuracy_std_dev": 0.02559013346212511,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.026921807500000002,
      "avg_duration_per_task_ms": 293558.7475,
      "avg_output_tokens_per_task": 4184.715,
      "last_run_timestamp": "2026-10-03T16:13:56.972Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_slug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_name": "Qwen: Qwen3 VL 235B A22B Instruct",
      "provider_name": "Venice",
      "accuracy": 0.400833375,
      "accuracy_std_dev": 0.029332849192234545,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0642667,
      "avg_duration_per_task_ms": 297648.6875,
      "avg_output_tokens_per_task": 11410.885,
      "last_run_timestamp": "2026-10-03T16:12:29.905Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_slug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_name": "Qwen: Qwen3 VL 235B A22B Instruct",
      "provider_name": "Novita",
      "accuracy": 0.40083312500000007,
      "accuracy_std_dev": 0.02459377320805766,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.07675446250000001,
      "avg_duration_per_task_ms": 122147.305,
      "avg_output_tokens_per_task": 3115.835,
      "last_run_timestamp": "2026-10-03T16:12:07.495Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4o-2024-05-13",
      "model_slug": "openai/gpt-4o-2024-05-13",
      "model_name": "OpenAI: GPT-4o (2024-05-13)",
      "provider_name": "auto-routing",
      "accuracy": 0.4,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 1.1315643,
      "avg_duration_per_task_ms": 71996.84,
      "avg_output_tokens_per_task": 3249.82,
      "last_run_timestamp": "2026-08-01T08:06:02.002Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Coder 30B A3B Instruct",
      "provider_name": "SiliconFlow",
      "accuracy": 0.4,
      "accuracy_std_dev": 0.019626323972325204,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.020620846666666668,
      "avg_duration_per_task_ms": 194089.08666666667,
      "avg_output_tokens_per_task": 3830.9066666666668,
      "last_run_timestamp": "2026-08-15T08:02:44.458Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Alibaba",
      "accuracy": 0.398333375,
      "accuracy_std_dev": 0.05074458509766707,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.048334209999999996,
      "avg_duration_per_task_ms": 87716.1125,
      "avg_output_tokens_per_task": 3723.8325,
      "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-coder-next-2025-02-03",
      "model_slug": "qwen/qwen3-coder-next",
      "model_name": "Qwen: Qwen3 Coder Next",
      "provider_name": "Ionstream",
      "accuracy": 0.39726041780821914,
      "accuracy_std_dev": 0.26615553243962864,
      "run_count": 3,
      "total_questions": 146,
      "avg_cost_per_task_usd": 0.011427219178082192,
      "avg_duration_per_task_ms": 86701.55479452055,
      "avg_output_tokens_per_task": 2538.931506849315,
      "last_run_timestamp": "2026-08-08T16:03:49.463Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-vl-8b-instruct",
      "model_slug": "qwen/qwen3-vl-8b-instruct",
      "model_name": "Qwen: Qwen3 VL 8B Instruct",
      "provider_name": "Parasail",
      "accuracy": 0.396667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.0528661,
      "avg_duration_per_task_ms": 135118.22,
      "avg_output_tokens_per_task": 7825.32,
      "last_run_timestamp": "2026-07-25T08:07:19.515Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.39521795358649786,
      "accuracy_std_dev": 0.015940999687598014,
      "run_count": 5,
      "total_questions": 237,
      "avg_cost_per_task_usd": 0.009386248945147679,
      "avg_duration_per_task_ms": 65821.63713080168,
      "avg_output_tokens_per_task": 1124.5696202531647,
      "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": "Parasail",
      "accuracy": 0.39249987500000005,
      "accuracy_std_dev": 0.016811927398706402,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0369896575,
      "avg_duration_per_task_ms": 191395.045,
      "avg_output_tokens_per_task": 3738.2225,
      "last_run_timestamp": "2026-10-03T16:08:50.061Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "CoreWeave",
      "accuracy": 0.39210027167630057,
      "accuracy_std_dev": 0.08499299536756556,
      "run_count": 7,
      "total_questions": 346,
      "avg_cost_per_task_usd": 0.3060756098265896,
      "avg_duration_per_task_ms": 324526.6416184971,
      "avg_output_tokens_per_task": 12470.086705202311,
      "last_run_timestamp": "2026-10-10T18:31:45.008Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_slug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_name": "Qwen: Qwen3 VL 235B A22B Instruct",
      "provider_name": "DeepInfra",
      "accuracy": 0.3912555560538117,
      "accuracy_std_dev": 0.029116472984579173,
      "run_count": 9,
      "total_questions": 446,
      "avg_cost_per_task_usd": 0.059476125560538114,
      "avg_duration_per_task_ms": 375172.7399103139,
      "avg_output_tokens_per_task": 3892.973094170404,
      "last_run_timestamp": "2026-10-08T23:21:09.489Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "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": 0.389166625,
      "accuracy_std_dev": 0.02355558880572453,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.025453972499999998,
      "avg_duration_per_task_ms": 227372.8075,
      "avg_output_tokens_per_task": 4499.2425,
      "last_run_timestamp": "2026-10-03T16:03:16.197Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "Ambient",
      "accuracy": 0.386667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.014510460000000001,
      "avg_duration_per_task_ms": 64332.54,
      "avg_output_tokens_per_task": 4106.96,
      "last_run_timestamp": "2026-10-09T15:37:38.473Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 VL 30B A3B Instruct",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.386667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.0592903,
      "avg_duration_per_task_ms": 203536.16,
      "avg_output_tokens_per_task": 5170.96,
      "last_run_timestamp": "2026-10-04T03:00:16.402Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.38666671428571425,
      "accuracy_std_dev": 0.008728715610219052,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.014610905714285714,
      "avg_duration_per_task_ms": 90807.17428571428,
      "avg_output_tokens_per_task": 601.44,
      "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": "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": 0.37776065177195683,
      "accuracy_std_dev": 0.01772313217238985,
      "run_count": 13,
      "total_questions": 649,
      "avg_cost_per_task_usd": 0.012601627118644071,
      "avg_duration_per_task_ms": 36693.28659476117,
      "avg_output_tokens_per_task": 655.8120184899846,
      "last_run_timestamp": "2026-10-10T18:35:37.313Z"
    },
    {
      "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": 0.37583325,
      "accuracy_std_dev": 0.03264919388265964,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.051693865,
      "avg_duration_per_task_ms": 72470.795,
      "avg_output_tokens_per_task": 4363.9325,
      "last_run_timestamp": "2026-10-03T16:05:34.166Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "WandB",
      "accuracy": 0.373333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.07157664,
      "avg_duration_per_task_ms": 16604.76,
      "avg_output_tokens_per_task": 863.76,
      "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": "WandB",
      "accuracy": 0.373333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.030392839999999997,
      "avg_duration_per_task_ms": 86392.54,
      "avg_output_tokens_per_task": 5907.74,
      "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": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Parasail",
      "accuracy": 0.37185199999999996,
      "accuracy_std_dev": 0.07211295476003061,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.006007911111111111,
      "avg_duration_per_task_ms": 147635.88666666666,
      "avg_output_tokens_per_task": 3781.9444444444443,
      "last_run_timestamp": "2026-10-10T19:09:43.617Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Chutes",
      "accuracy": 0.3718273705583756,
      "accuracy_std_dev": 0.06753307245238718,
      "run_count": 8,
      "total_questions": 394,
      "avg_cost_per_task_usd": 0.0041109720812182735,
      "avg_duration_per_task_ms": 291217.7411167513,
      "avg_output_tokens_per_task": 4238.002538071066,
      "last_run_timestamp": "2026-10-10T19:11:16.157Z"
    },
    {
      "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": 0.3714284285714286,
      "accuracy_std_dev": 0.017715366073046326,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.03264426571428572,
      "avg_duration_per_task_ms": 156269.61714285714,
      "avg_output_tokens_per_task": 4433.125714285714,
      "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": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "GMICloud",
      "accuracy": 0.36833325,
      "accuracy_std_dev": 0.17853565236861657,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.015193219999999999,
      "avg_duration_per_task_ms": 127682.09,
      "avg_output_tokens_per_task": 2917.505,
      "last_run_timestamp": "2026-10-03T16:12:29.879Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4o-2024-05-13",
      "model_slug": "openai/gpt-4o-2024-05-13",
      "model_name": "OpenAI: GPT-4o (2024-05-13)",
      "provider_name": "Azure",
      "accuracy": 0.36666699999999997,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.80181,
      "avg_duration_per_task_ms": 63430.68,
      "avg_output_tokens_per_task": 3064.6,
      "last_run_timestamp": "2026-08-01T08:06:07.633Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.36,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.041000339999999996,
      "avg_duration_per_task_ms": 322345.3,
      "avg_output_tokens_per_task": 4212.74,
      "last_run_timestamp": "2026-10-04T03:00:10.237Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Nebius",
      "accuracy": 0.35762146733668343,
      "accuracy_std_dev": 0.20389728295072376,
      "run_count": 8,
      "total_questions": 398,
      "avg_cost_per_task_usd": 0.12249766080402011,
      "avg_duration_per_task_ms": 301830.15075376886,
      "avg_output_tokens_per_task": 40654.94472361809,
      "last_run_timestamp": "2026-10-10T18:31:25.884Z"
    },
    {
      "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": 0.3575,
      "accuracy_std_dev": 0.033819176021156985,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.07891495999999999,
      "avg_duration_per_task_ms": 25792.25,
      "avg_output_tokens_per_task": 920.17,
      "last_run_timestamp": "2026-10-10T18:49:13.368Z"
    },
    {
      "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": 0.3567604417670683,
      "accuracy_std_dev": 0.024567950668299542,
      "run_count": 10,
      "total_questions": 498,
      "avg_cost_per_task_usd": 0.0795283875502008,
      "avg_duration_per_task_ms": 211758.7168674699,
      "avg_output_tokens_per_task": 7341.082329317269,
      "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": 0.3566663333333333,
      "accuracy_std_dev": 0.016996731711975962,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.043029019999999994,
      "avg_duration_per_task_ms": 111012.42666666667,
      "avg_output_tokens_per_task": 4511.693333333334,
      "last_run_timestamp": "2026-08-01T17:38:36.457Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 VL 30B A3B Instruct",
      "provider_name": "DeepInfra",
      "accuracy": 0.35592184615384614,
      "accuracy_std_dev": 0.02760835600602419,
      "run_count": 11,
      "total_questions": 546,
      "avg_cost_per_task_usd": 0.07156286996336997,
      "avg_duration_per_task_ms": 377710.5695970696,
      "avg_output_tokens_per_task": 7880.789377289378,
      "last_run_timestamp": "2026-10-08T23:34:52.306Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-chat-v3",
      "model_slug": "deepseek/deepseek-chat",
      "model_name": "DeepSeek: DeepSeek V3",
      "provider_name": "Novita",
      "accuracy": 0.34058,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 46,
      "avg_cost_per_task_usd": 0.11353834782608695,
      "avg_duration_per_task_ms": 397172.5,
      "avg_output_tokens_per_task": 14067.434782608696,
      "last_run_timestamp": "2026-08-01T08:05:57.857Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "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": 0.33666633333333335,
      "accuracy_std_dev": 0.004714045207910321,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.029078933333333334,
      "avg_duration_per_task_ms": 79568.04666666666,
      "avg_output_tokens_per_task": 3568.733333333333,
      "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": "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": 0.3340808340807174,
      "accuracy_std_dev": 0.05005079503642052,
      "run_count": 9,
      "total_questions": 446,
      "avg_cost_per_task_usd": 0.029606439461883405,
      "avg_duration_per_task_ms": 98144.11883408071,
      "avg_output_tokens_per_task": 1206.6883408071749,
      "last_run_timestamp": "2026-10-10T19:14:33.599Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.333333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.051354699999999996,
      "avg_duration_per_task_ms": 358714.54,
      "avg_output_tokens_per_task": 69285.14,
      "last_run_timestamp": "2026-10-04T03:00:10.240Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 VL 30B A3B Instruct",
      "provider_name": "Alibaba",
      "accuracy": 0.3296970909090909,
      "accuracy_std_dev": 0.023845300316279384,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.08440126545454546,
      "avg_duration_per_task_ms": 151545.54545454544,
      "avg_output_tokens_per_task": 8175.638181818182,
      "last_run_timestamp": "2026-10-03T08:23:09.856Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-vl-8b-instruct",
      "model_slug": "qwen/qwen3-vl-8b-instruct",
      "model_name": "Qwen: Qwen3 VL 8B Instruct",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.326667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.02872976,
      "avg_duration_per_task_ms": 133695.86,
      "avg_output_tokens_per_task": 3408.56,
      "last_run_timestamp": "2026-10-04T03:00:16.402Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.313333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.15186704,
      "avg_duration_per_task_ms": 32635.52,
      "avg_output_tokens_per_task": 1294.54,
      "last_run_timestamp": "2026-10-04T03:00:16.406Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "DeepInfra",
      "accuracy": 0.30896692982456136,
      "accuracy_std_dev": 0.03129504337294144,
      "run_count": 7,
      "total_questions": 342,
      "avg_cost_per_task_usd": 0.01647059649122807,
      "avg_duration_per_task_ms": 365014.59064327483,
      "avg_output_tokens_per_task": 4454.833333333333,
      "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": 0.3022222222222222,
      "accuracy_std_dev": 0.036243388499843904,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.16888293333333335,
      "avg_duration_per_task_ms": 128280.22444444445,
      "avg_output_tokens_per_task": 1888.62,
      "last_run_timestamp": "2026-10-10T19:11:13.456Z"
    },
    {
      "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": 0.2921211818181818,
      "accuracy_std_dev": 0.32697940222083444,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.017488016363636363,
      "avg_duration_per_task_ms": 53819.056363636366,
      "avg_output_tokens_per_task": 9497.68909090909,
      "last_run_timestamp": "2026-10-10T20:16:03.339Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4o-mini",
      "model_slug": "openai/gpt-4o-mini",
      "model_name": "OpenAI: GPT-4o-mini",
      "provider_name": "auto-routing",
      "accuracy": 0.2844443333333333,
      "accuracy_std_dev": 0.031894788459698065,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.02081186,
      "avg_duration_per_task_ms": 62184.12666666666,
      "avg_output_tokens_per_task": 3804.7533333333336,
      "last_run_timestamp": "2026-10-01T19:04:58.075Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4o-mini",
      "model_slug": "openai/gpt-4o-mini",
      "model_name": "OpenAI: GPT-4o-mini",
      "provider_name": "OpenAI",
      "accuracy": 0.28000025,
      "accuracy_std_dev": 0.027080189561513406,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.02535222,
      "avg_duration_per_task_ms": 67904.845,
      "avg_output_tokens_per_task": 4023.05,
      "last_run_timestamp": "2026-10-02T23:13:47.090Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 VL 30B A3B Instruct",
      "provider_name": "SiliconFlow",
      "accuracy": 0.2797619910714286,
      "accuracy_std_dev": 0.1899431271985133,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.15921619642857143,
      "avg_duration_per_task_ms": 942346.96875,
      "avg_output_tokens_per_task": 6098.834821428572,
      "last_run_timestamp": "2026-10-03T08:09:23.857Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen-2.5-72b-instruct",
      "model_slug": "qwen/qwen-2.5-72b-instruct",
      "model_name": "Qwen2.5 72B Instruct",
      "provider_name": "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": "mistralai/mistral-nemo",
      "model_slug": "mistralai/mistral-nemo",
      "model_name": "Mistral: Mistral Nemo",
      "provider_name": "auto-routing",
      "accuracy": 0.260417,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 48,
      "avg_cost_per_task_usd": 0.027361125,
      "avg_duration_per_task_ms": 394399.5416666667,
      "avg_output_tokens_per_task": 4477.166666666667,
      "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": 0.25,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.02790544,
      "avg_duration_per_task_ms": 464213.12,
      "avg_output_tokens_per_task": 5425.66,
      "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": "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": 0.2325,
      "accuracy_std_dev": 0.06851085072271398,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.013581069999999997,
      "avg_duration_per_task_ms": 21748.2225,
      "avg_output_tokens_per_task": 1125.92,
      "last_run_timestamp": "2026-10-03T16:13:42.255Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta-llama/llama-3.1-8b-instruct",
      "model_slug": "meta-llama/llama-3.1-8b-instruct",
      "model_name": "Meta: Llama 3.1 8B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.19666649999999997,
      "accuracy_std_dev": 0.023804971471312466,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.00425299,
      "avg_duration_per_task_ms": 15962.445,
      "avg_output_tokens_per_task": 1579.045,
      "last_run_timestamp": "2026-10-03T16:07:44.178Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "AtlasCloud",
      "accuracy": 0.18791944295302013,
      "accuracy_std_dev": 0.10708233389741122,
      "run_count": 6,
      "total_questions": 298,
      "avg_cost_per_task_usd": 0.20449519798657714,
      "avg_duration_per_task_ms": 608397.9966442953,
      "avg_output_tokens_per_task": 10862.70134228188,
      "last_run_timestamp": "2026-08-29T16:22:37.844Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "amazon/nova-micro-v1",
      "model_slug": "amazon/nova-micro-v1",
      "model_name": "Amazon: Nova Micro 1.0",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.18766967073170732,
      "accuracy_std_dev": 0.018396157170452745,
      "run_count": 5,
      "total_questions": 246,
      "avg_cost_per_task_usd": 0.011055426829268291,
      "avg_duration_per_task_ms": 138140.34552845528,
      "avg_output_tokens_per_task": 5390.59756097561,
      "last_run_timestamp": "2026-10-01T19:04:52.650Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-next-2025-02-03",
      "model_slug": "qwen/qwen3-coder-next",
      "model_name": "Qwen: Qwen3 Coder Next",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.186667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.01693508,
      "avg_duration_per_task_ms": 26941.4,
      "avg_output_tokens_per_task": 1041.58,
      "last_run_timestamp": "2026-10-04T03:00:16.405Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.1-terminus",
      "model_slug": "deepseek/deepseek-v3.1-terminus",
      "model_name": "DeepSeek: DeepSeek V3.1 Terminus",
      "provider_name": "AtlasCloud",
      "accuracy": 0.17999985714285713,
      "accuracy_std_dev": 0.12173356543044388,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.013454491428571429,
      "avg_duration_per_task_ms": 61907.58571428571,
      "avg_output_tokens_per_task": 1212.5742857142857,
      "last_run_timestamp": "2026-08-29T16:17:08.780Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen-2.5-7b-instruct",
      "model_slug": "qwen/qwen-2.5-7b-instruct",
      "model_name": "Qwen: Qwen2.5 7B Instruct",
      "provider_name": "Phala",
      "accuracy": 0.166667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.01877496,
      "avg_duration_per_task_ms": 63137.36,
      "avg_output_tokens_per_task": 2797.32,
      "last_run_timestamp": "2026-08-01T08:06:16.826Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen-2.5-7b-instruct",
      "model_slug": "qwen/qwen-2.5-7b-instruct",
      "model_name": "Qwen: Qwen2.5 7B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.166667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.016978299999999998,
      "avg_duration_per_task_ms": 57790.52,
      "avg_output_tokens_per_task": 2492.08,
      "last_run_timestamp": "2026-08-01T08:06:07.057Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4.1-nano-2025-04-14",
      "model_slug": "openai/gpt-4.1-nano",
      "model_name": "OpenAI: GPT-4.1 Nano",
      "provider_name": "Azure",
      "accuracy": 0.13333333333333333,
      "accuracy_std_dev": 0.021430473285072877,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.007961373333333334,
      "avg_duration_per_task_ms": 80728.30333333333,
      "avg_output_tokens_per_task": 2469.17,
      "last_run_timestamp": "2026-10-01T19:04:58.093Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4.1-nano-2025-04-14",
      "model_slug": "openai/gpt-4.1-nano",
      "model_name": "OpenAI: GPT-4.1 Nano",
      "provider_name": "OpenAI",
      "accuracy": 0.11833325,
      "accuracy_std_dev": 0.018484159885899604,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.007158584999999999,
      "avg_duration_per_task_ms": 55526.67,
      "avg_output_tokens_per_task": 2353.445,
      "last_run_timestamp": "2026-10-02T23:07:59.112Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4.1-nano-2025-04-14",
      "model_slug": "openai/gpt-4.1-nano",
      "model_name": "OpenAI: GPT-4.1 Nano",
      "provider_name": "auto-routing",
      "accuracy": 0.108889,
      "accuracy_std_dev": 0.008314764498569193,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.006775626666666668,
      "avg_duration_per_task_ms": 60620.74,
      "avg_output_tokens_per_task": 2327.0066666666667,
      "last_run_timestamp": "2026-10-01T19:04:36.194Z"
    }
  ]
}
