{
  "dataset": "benchmark-leaderboard",
  "title": "OpenRouter benchmark leaderboard",
  "description": "One row per benchmark, model version, and provider: the aggregate values displayed on the OpenRouter leaderboard pages (GPQA Diamond, tau2-bench airline). Derived from the benchmark-runs dataset.",
  "schema_version": "v1",
  "snapshot_date": "2026-11-14",
  "retrieved_at": "2026-11-14T01:46:03.748Z",
  "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": 2275,
  "rows": [
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-4-argon-20261013",
      "model_slug": "google/gemini-4-argon",
      "model_name": "Google: Gemini 4 Argon",
      "provider_name": "Google",
      "accuracy": 0.964646,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.09468125757575757,
      "avg_duration_per_task_ms": 101980.28282828283,
      "avg_output_tokens_per_task": 8439.414141414141,
      "last_run_timestamp": "2026-10-14T15:00:03.588Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.8-flash-20260902",
      "model_slug": "google/gemini-3.8-flash",
      "model_name": "Google: Gemini 3.8 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.95637,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 191,
      "avg_cost_per_task_usd": 0.06822796858638744,
      "avg_duration_per_task_ms": 132197.8062827225,
      "avg_output_tokens_per_task": 18049.068062827224,
      "last_run_timestamp": "2026-10-01T08:31:33.481Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-astra-pro-20260903",
      "model_slug": "openai/gpt-6-astra-pro",
      "model_name": "OpenAI: GPT-6 Astra Pro",
      "provider_name": "auto-routing",
      "accuracy": 0.954545,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.332555398989899,
      "avg_duration_per_task_ms": 75446.5505050505,
      "avg_output_tokens_per_task": 4530.388888888889,
      "last_run_timestamp": "2026-10-01T08:27:35.414Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.8-flash-20260902",
      "model_slug": "google/gemini-3.8-flash",
      "model_name": "Google: Gemini 3.8 Flash",
      "provider_name": "Google AI Studio",
      "accuracy": 0.952862,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.07423223737373738,
      "avg_duration_per_task_ms": 86748.28787878787,
      "avg_output_tokens_per_task": 19646.409090909092,
      "last_run_timestamp": "2026-10-01T08:29:11.347Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.1-pro-preview-20260219",
      "model_slug": "google/gemini-3.1-pro-preview",
      "model_name": "Google: Gemini 3.1 Pro Preview",
      "provider_name": "Google",
      "accuracy": 0.95202,
      "accuracy_std_dev": 0.0008420000000000094,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.20364760606060606,
      "avg_duration_per_task_ms": 135706.18686868687,
      "avg_output_tokens_per_task": 16847.588383838385,
      "last_run_timestamp": "2026-10-01T19:14:26.157Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-astra-pro-20260903",
      "model_slug": "openai/gpt-6-astra-pro",
      "model_name": "OpenAI: GPT-6 Astra Pro",
      "provider_name": "OpenAI",
      "accuracy": 0.9511785,
      "accuracy_std_dev": 0.0016835000000000044,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.33620659090909094,
      "avg_duration_per_task_ms": 70544.52272727272,
      "avg_output_tokens_per_task": 4588.830808080808,
      "last_run_timestamp": "2026-10-02T23:23:01.262Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-fable-5.2-20261109",
      "model_slug": "anthropic/claude-fable-5.2",
      "model_name": "Anthropic: Claude Fable 5.2",
      "provider_name": "Anthropic",
      "accuracy": 0.9511780000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.08733781818181818,
      "avg_duration_per_task_ms": 25865.449494949495,
      "avg_output_tokens_per_task": 1942.8484848484848,
      "last_run_timestamp": "2026-11-10T18:01:45.629Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.1-pro-preview-20260219",
      "model_slug": "google/gemini-3.1-pro-preview",
      "model_name": "Google: Gemini 3.1 Pro Preview",
      "provider_name": "Google AI Studio",
      "accuracy": 0.9511779999999999,
      "accuracy_std_dev": 0.003367000000000009,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.20247136363636364,
      "avg_duration_per_task_ms": 137578.30555555556,
      "avg_output_tokens_per_task": 16748.56818181818,
      "last_run_timestamp": "2026-10-01T19:12:42.155Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-fable-5.2-20261109",
      "model_slug": "anthropic/claude-fable-5.2",
      "model_name": "Anthropic: Claude Fable 5.2",
      "provider_name": "Azure",
      "accuracy": 0.949495,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.08434266666666666,
      "avg_duration_per_task_ms": 25783.79797979798,
      "avg_output_tokens_per_task": 1867.969696969697,
      "last_run_timestamp": "2026-11-10T18:03:17.074Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-astra-pro-20260903",
      "model_slug": "openai/gpt-6-astra-pro",
      "model_name": "OpenAI: GPT-6 Astra Pro",
      "provider_name": "Azure",
      "accuracy": 0.949495,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.35495546717171716,
      "avg_duration_per_task_ms": 72102.17676767676,
      "avg_output_tokens_per_task": 4622.60101010101,
      "last_run_timestamp": "2026-10-01T08:29:03.480Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.8-flash-20260902",
      "model_slug": "google/gemini-3.8-flash",
      "model_name": "Google: Gemini 3.8 Flash",
      "provider_name": "Google",
      "accuracy": 0.9482614885496183,
      "accuracy_std_dev": 0.0012429999999999941,
      "run_count": 2,
      "total_questions": 393,
      "avg_cost_per_task_usd": 0.07536056743002545,
      "avg_duration_per_task_ms": 132814.94147582698,
      "avg_output_tokens_per_task": 18975.51145038168,
      "last_run_timestamp": "2026-10-15T21:00:04.244Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "inclusionai/ling-3.1-flash-20261002",
      "model_slug": "inclusionai/ling-3.1-flash",
      "model_name": "inclusionAI: Ling 3.1 Flash",
      "provider_name": "DeepInfra",
      "accuracy": 0.947917,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 96,
      "avg_cost_per_task_usd": 0.009331666666666667,
      "avg_duration_per_task_ms": 232878.71875,
      "avg_output_tokens_per_task": 10278.75,
      "last_run_timestamp": "2026-10-14T18:00:04.673Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6.1-sol-20260929",
      "model_slug": "openai/gpt-6.1-sol",
      "model_name": "OpenAI: GPT-6.1 Sol",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.947811,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.021484267676767678,
      "avg_duration_per_task_ms": 55623.856060606064,
      "avg_output_tokens_per_task": 1890.328282828283,
      "last_run_timestamp": "2026-10-30T00:03:33.414Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-astra-20260903",
      "model_slug": "openai/gpt-6-astra",
      "model_name": "OpenAI: GPT-6 Astra",
      "provider_name": "OpenAI",
      "accuracy": 0.9472503333333333,
      "accuracy_std_dev": 0.0015872190215034077,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.11162585016835018,
      "avg_duration_per_task_ms": 48234.978114478115,
      "avg_output_tokens_per_task": 2076.191919191919,
      "last_run_timestamp": "2026-10-02T23:16:05.967Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-fable-5.2-20261109",
      "model_slug": "anthropic/claude-fable-5.2",
      "model_name": "Anthropic: Claude Fable 5.2",
      "provider_name": "Google",
      "accuracy": 0.9461279999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.08508327272727273,
      "avg_duration_per_task_ms": 22398.060606060608,
      "avg_output_tokens_per_task": 1886.4848484848485,
      "last_run_timestamp": "2026-11-10T18:01:46.740Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-astra-20260903",
      "model_slug": "openai/gpt-6-astra",
      "model_name": "OpenAI: GPT-6 Astra",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.9461276666666666,
      "accuracy_std_dev": 0.001374572014200141,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.1181814292929293,
      "avg_duration_per_task_ms": 36670.56565656565,
      "avg_output_tokens_per_task": 2059.811447811448,
      "last_run_timestamp": "2026-10-30T00:02:55.917Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.1-pro-preview-20260219",
      "model_slug": "google/gemini-3.1-pro-preview",
      "model_name": "Google: Gemini 3.1 Pro Preview",
      "provider_name": "auto-routing",
      "accuracy": 0.945286,
      "accuracy_std_dev": 0.0008420000000000094,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.20203182828282829,
      "avg_duration_per_task_ms": 133595.55303030304,
      "avg_output_tokens_per_task": 16712.858585858587,
      "last_run_timestamp": "2026-10-01T19:16:14.342Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.7-flash-20260813",
      "model_slug": "google/gemini-3.7-flash",
      "model_name": "Google: Gemini 3.7 Flash",
      "provider_name": "Google",
      "accuracy": 0.9450355212765958,
      "accuracy_std_dev": 0.00181198405683448,
      "run_count": 3,
      "total_questions": 564,
      "avg_cost_per_task_usd": 0.030139015957446804,
      "avg_duration_per_task_ms": 60727.07978723404,
      "avg_output_tokens_per_task": 7608.468085106383,
      "last_run_timestamp": "2026-10-15T21:01:11.190Z"
    },
    {
      "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": 0.941077,
      "accuracy_std_dev": 0.0033669999999999534,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.030745075757575756,
      "avg_duration_per_task_ms": 30416.38888888889,
      "avg_output_tokens_per_task": 8050.770202020202,
      "last_run_timestamp": "2026-10-01T16:16:03.631Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-sol-pro-20260709",
      "model_slug": "openai/gpt-5.6-sol-pro",
      "model_name": "OpenAI: GPT-5.6 Sol Pro",
      "provider_name": "auto-routing",
      "accuracy": 0.941077,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.15125092424242426,
      "avg_duration_per_task_ms": 97603.77272727272,
      "avg_output_tokens_per_task": 11559.585858585859,
      "last_run_timestamp": "2026-10-01T16:10:33.208Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3.1-pro-20261023",
      "model_slug": "moonshotai/kimi-k3.1-pro",
      "model_name": "MoonshotAI: Kimi K3.1 Pro",
      "provider_name": "Nebius",
      "accuracy": 0.939394,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.06875348484848484,
      "avg_duration_per_task_ms": 279532.6212121212,
      "avg_output_tokens_per_task": 6669.813131313132,
      "last_run_timestamp": "2026-11-04T21:00:08.522Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.5-20260423",
      "model_slug": "openai/gpt-5.5",
      "model_name": "OpenAI: GPT-5.5",
      "provider_name": "auto-routing",
      "accuracy": 0.939394,
      "accuracy_std_dev": 0.0033669999999999534,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.31252995454545457,
      "avg_duration_per_task_ms": 150832.55050505052,
      "avg_output_tokens_per_task": 10288.712121212122,
      "last_run_timestamp": "2026-10-01T19:12:11.505Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-sol-pro-20260709",
      "model_slug": "openai/gpt-5.6-sol-pro",
      "model_name": "OpenAI: GPT-5.6 Sol Pro",
      "provider_name": "Azure",
      "accuracy": 0.939394,
      "accuracy_std_dev": 0.0033669999999999534,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.32760574494949496,
      "avg_duration_per_task_ms": 78455.45202020202,
      "avg_output_tokens_per_task": 11977.780303030304,
      "last_run_timestamp": "2026-10-01T16:36:19.657Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-sol-pro-20260709",
      "model_slug": "openai/gpt-5.6-sol-pro",
      "model_name": "OpenAI: GPT-5.6 Sol Pro",
      "provider_name": "OpenAI",
      "accuracy": 0.9393935,
      "accuracy_std_dev": 0.005050499999999958,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.15067959090909092,
      "avg_duration_per_task_ms": 95544.60353535354,
      "avg_output_tokens_per_task": 11515.699494949495,
      "last_run_timestamp": "2026-10-02T23:21:48.863Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3.1-pro-20261023",
      "model_slug": "moonshotai/kimi-k3.1-pro",
      "model_name": "MoonshotAI: Kimi K3.1 Pro",
      "provider_name": "Sail Research",
      "accuracy": 0.9387350223978919,
      "accuracy_std_dev": 0.0014398627538762172,
      "run_count": 4,
      "total_questions": 759,
      "avg_cost_per_task_usd": 0.14429597496706192,
      "avg_duration_per_task_ms": 118270.40579710146,
      "avg_output_tokens_per_task": 6705.238471673254,
      "last_run_timestamp": "2026-11-06T22:04:38.207Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.7-flash-20260813",
      "model_slug": "google/gemini-3.7-flash",
      "model_name": "Google: Gemini 3.7 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.938528,
      "accuracy_std_dev": 0.004357499999999959,
      "run_count": 2,
      "total_questions": 385,
      "avg_cost_per_task_usd": 0.0293307974025974,
      "avg_duration_per_task_ms": 59248.22337662338,
      "avg_output_tokens_per_task": 7683.532467532467,
      "last_run_timestamp": "2026-10-01T16:17:32.289Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6.1-sol-20260929",
      "model_slug": "openai/gpt-6.1-sol",
      "model_name": "OpenAI: GPT-6.1 Sol",
      "provider_name": "Azure",
      "accuracy": 0.9382713333333333,
      "accuracy_std_dev": 0.004627357405210382,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.022108563973063967,
      "avg_duration_per_task_ms": 32928.50084175084,
      "avg_output_tokens_per_task": 1914.9974747474748,
      "last_run_timestamp": "2026-10-01T08:40:50.203Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.5-20260423",
      "model_slug": "openai/gpt-5.5",
      "model_name": "OpenAI: GPT-5.5",
      "provider_name": "Azure",
      "accuracy": 0.9368683333333334,
      "accuracy_std_dev": 0.0040948247690089186,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.3309691094276094,
      "avg_duration_per_task_ms": 165124.49915824915,
      "avg_output_tokens_per_task": 10209.419191919193,
      "last_run_timestamp": "2026-10-01T19:14:07.472Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3.1-pro-20261023",
      "model_slug": "moonshotai/kimi-k3.1-pro",
      "model_name": "MoonshotAI: Kimi K3.1 Pro",
      "provider_name": "River",
      "accuracy": 0.936027,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.08390751515151515,
      "avg_duration_per_task_ms": 94274.73232323233,
      "avg_output_tokens_per_task": 8182.020202020202,
      "last_run_timestamp": "2026-11-13T06:01:36.932Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.5-20260423",
      "model_slug": "openai/gpt-5.5",
      "model_name": "OpenAI: GPT-5.5",
      "provider_name": "OpenAI",
      "accuracy": 0.9360266666666668,
      "accuracy_std_dev": 0.0036369299629708854,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.3141816919191919,
      "avg_duration_per_task_ms": 142474.1195286195,
      "avg_output_tokens_per_task": 10342.412457912458,
      "last_run_timestamp": "2026-10-02T23:25:41.614Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Io Net",
      "accuracy": 0.9348024214876033,
      "accuracy_std_dev": 0.0015875000000000195,
      "run_count": 2,
      "total_questions": 363,
      "avg_cost_per_task_usd": 0.11736377685950414,
      "avg_duration_per_task_ms": 170976.24517906335,
      "avg_output_tokens_per_task": 7983.9201101928375,
      "last_run_timestamp": "2026-11-11T05:45:32.564Z"
    },
    {
      "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": 0.9335015000000001,
      "accuracy_std_dev": 0.000841499999999995,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.04714142045454545,
      "avg_duration_per_task_ms": 67600.7297979798,
      "avg_output_tokens_per_task": 11818.68813131313,
      "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-k3.1-pro-20261023",
      "model_slug": "moonshotai/kimi-k3.1-pro",
      "model_name": "MoonshotAI: Kimi K3.1 Pro",
      "provider_name": "Decart",
      "accuracy": 0.93266,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.04747170707070707,
      "avg_duration_per_task_ms": 97339.0303030303,
      "avg_output_tokens_per_task": 7164.717171717171,
      "last_run_timestamp": "2026-11-11T15:01:28.174Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3.1-pro-20261023",
      "model_slug": "moonshotai/kimi-k3.1-pro",
      "model_name": "MoonshotAI: Kimi K3.1 Pro",
      "provider_name": "DigitalOcean",
      "accuracy": 0.93266,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.06645131313131314,
      "avg_duration_per_task_ms": 52649.732323232325,
      "avg_output_tokens_per_task": 6433.60101010101,
      "last_run_timestamp": "2026-11-07T03:00:36.225Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3.1-pro-20261023",
      "model_slug": "moonshotai/kimi-k3.1-pro",
      "model_name": "MoonshotAI: Kimi K3.1 Pro",
      "provider_name": "Modal",
      "accuracy": 0.931818,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.07295092424242425,
      "avg_duration_per_task_ms": 39086.47474747475,
      "avg_output_tokens_per_task": 7118.949494949495,
      "last_run_timestamp": "2026-11-04T21:00:08.524Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3.1-pro-20261023",
      "model_slug": "moonshotai/kimi-k3.1-pro",
      "model_name": "MoonshotAI: Kimi K3.1 Pro",
      "provider_name": "Parasail",
      "accuracy": 0.930976,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.07401931818181819,
      "avg_duration_per_task_ms": 70607.26262626263,
      "avg_output_tokens_per_task": 7194.131313131313,
      "last_run_timestamp": "2026-11-04T21:00:08.540Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3.1-pro-20261023",
      "model_slug": "moonshotai/kimi-k3.1-pro",
      "model_name": "MoonshotAI: Kimi K3.1 Pro",
      "provider_name": "InferenceNet",
      "accuracy": 0.9306844829721364,
      "accuracy_std_dev": 0.007658821616932981,
      "run_count": 5,
      "total_questions": 969,
      "avg_cost_per_task_usd": 0.0836984004127967,
      "avg_duration_per_task_ms": 71158.75335397317,
      "avg_output_tokens_per_task": 7439.702786377709,
      "last_run_timestamp": "2026-11-05T00:00:09.926Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-sonnet-5.5-20260928",
      "model_slug": "anthropic/claude-sonnet-5.5",
      "model_name": "Anthropic: Claude Sonnet 5.5",
      "provider_name": "Google",
      "accuracy": 0.9301345000000001,
      "accuracy_std_dev": 0.006832688313053163,
      "run_count": 18,
      "total_questions": 3564,
      "avg_cost_per_task_usd": 0.026123152076318747,
      "avg_duration_per_task_ms": 18695.293490460157,
      "avg_output_tokens_per_task": 2221.0524691358023,
      "last_run_timestamp": "2026-10-31T08:07:18.284Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.6-flash-20260721",
      "model_slug": "google/gemini-3.6-flash",
      "model_name": "Google: Gemini 3.6 Flash",
      "provider_name": "Google AI Studio",
      "accuracy": 0.9301345,
      "accuracy_std_dev": 0.002525500000000014,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.044492378787878796,
      "avg_duration_per_task_ms": 69170.04040404041,
      "avg_output_tokens_per_task": 11715.780303030304,
      "last_run_timestamp": "2026-10-01T19:12:42.162Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-sonnet-5.5-20260928",
      "model_slug": "anthropic/claude-sonnet-5.5",
      "model_name": "Anthropic: Claude Sonnet 5.5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.929854,
      "accuracy_std_dev": 0.006348890821762996,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.025367877946127944,
      "avg_duration_per_task_ms": 20154.14898989899,
      "avg_output_tokens_per_task": 2304.1734006734005,
      "last_run_timestamp": "2026-10-31T08:33:04.486Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3.1-pro-20261023",
      "model_slug": "moonshotai/kimi-k3.1-pro",
      "model_name": "MoonshotAI: Kimi K3.1 Pro",
      "provider_name": "Moonshot AI",
      "accuracy": 0.9292930000000001,
      "accuracy_std_dev": 0.003367000000000009,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.10512464646464648,
      "avg_duration_per_task_ms": 151923.42676767678,
      "avg_output_tokens_per_task": 6756.267676767677,
      "last_run_timestamp": "2026-11-04T18:03:37.503Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3.1-pro-20261023",
      "model_slug": "moonshotai/kimi-k3.1-pro",
      "model_name": "MoonshotAI: Kimi K3.1 Pro",
      "provider_name": "AkashML",
      "accuracy": 0.9292929999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.08810357575757576,
      "avg_duration_per_task_ms": 85562.65151515152,
      "avg_output_tokens_per_task": 8602.494949494949,
      "last_run_timestamp": "2026-11-05T06:00:43.810Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3.1-pro-20261023",
      "model_slug": "moonshotai/kimi-k3.1-pro",
      "model_name": "MoonshotAI: Kimi K3.1 Pro",
      "provider_name": "Morph",
      "accuracy": 0.9292929999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.11162276262626263,
      "avg_duration_per_task_ms": 129833.19191919192,
      "avg_output_tokens_per_task": 9183.469696969696,
      "last_run_timestamp": "2026-11-05T06:00:06.218Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-sonnet-5.5-20260928",
      "model_slug": "anthropic/claude-sonnet-5.5",
      "model_name": "Anthropic: Claude Sonnet 5.5",
      "provider_name": "Anthropic",
      "accuracy": 0.9290121666666666,
      "accuracy_std_dev": 0.0044978590987515684,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.024844954545454545,
      "avg_duration_per_task_ms": 19502.118686868685,
      "avg_output_tokens_per_task": 2244.3215488215487,
      "last_run_timestamp": "2026-10-31T09:02:13.238Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-sol-20260709",
      "model_slug": "openai/gpt-5.6-sol",
      "model_name": "OpenAI: GPT-5.6 Sol",
      "provider_name": "Azure",
      "accuracy": 0.9287315,
      "accuracy_std_dev": 0.0045244150174359715,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.09770303535353536,
      "avg_duration_per_task_ms": 54704.919191919194,
      "avg_output_tokens_per_task": 3506.051346801347,
      "last_run_timestamp": "2026-10-01T19:42:48.407Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.5-20260423",
      "model_slug": "openai/gpt-5.5",
      "model_name": "OpenAI: GPT-5.5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.928451,
      "accuracy_std_dev": 0.0025249999999999995,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.3476244823232323,
      "avg_duration_per_task_ms": 112460.52777777778,
      "avg_output_tokens_per_task": 10405.60101010101,
      "last_run_timestamp": "2026-10-01T19:13:06.559Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "x-ai/grok-4.6-20260810",
      "model_slug": "x-ai/grok-4.6",
      "model_name": "SpaceXAI: Grok 4.6",
      "provider_name": "auto-routing",
      "accuracy": 0.928451,
      "accuracy_std_dev": 0.004209000000000018,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.12326839393939394,
      "avg_duration_per_task_ms": 361758.7878787879,
      "avg_output_tokens_per_task": 20322.222222222223,
      "last_run_timestamp": "2026-10-01T16:09:08.142Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-sonnet-5.5-20260928",
      "model_slug": "anthropic/claude-sonnet-5.5",
      "model_name": "Anthropic: Claude Sonnet 5.5",
      "provider_name": "auto-routing",
      "accuracy": 0.9282826,
      "accuracy_std_dev": 0.00626296053316642,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.025195250505050507,
      "avg_duration_per_task_ms": 19633.72121212121,
      "avg_output_tokens_per_task": 2281.29898989899,
      "last_run_timestamp": "2026-10-31T08:09:30.514Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Tenstorrent",
      "accuracy": 0.9281247875,
      "accuracy_std_dev": 0.003304499999999988,
      "run_count": 2,
      "total_questions": 160,
      "avg_cost_per_task_usd": 0.01232809375,
      "avg_duration_per_task_ms": 251386.81875,
      "avg_output_tokens_per_task": 10178.75625,
      "last_run_timestamp": "2026-11-04T18:00:19.969Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.6-flash-20260721",
      "model_slug": "google/gemini-3.6-flash",
      "model_name": "Google: Gemini 3.6 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.9276095000000001,
      "accuracy_std_dev": 0.005050500000000013,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.04436604545454545,
      "avg_duration_per_task_ms": 61670.28282828283,
      "avg_output_tokens_per_task": 11683.290404040405,
      "last_run_timestamp": "2026-10-01T19:11:04.522Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3.1-pro-20261023",
      "model_slug": "moonshotai/kimi-k3.1-pro",
      "model_name": "MoonshotAI: Kimi K3.1 Pro",
      "provider_name": "Fireworks",
      "accuracy": 0.927609,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.07610462626262626,
      "avg_duration_per_task_ms": 98167.92929292929,
      "avg_output_tokens_per_task": 7400.540404040404,
      "last_run_timestamp": "2026-11-05T15:00:03.242Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Parasail",
      "accuracy": 0.9275846571787361,
      "accuracy_std_dev": 0.010407425048021134,
      "run_count": 13,
      "total_questions": 2389,
      "avg_cost_per_task_usd": 0.13203736626203433,
      "avg_duration_per_task_ms": 128986.54792800335,
      "avg_output_tokens_per_task": 8711.713269150272,
      "last_run_timestamp": "2026-10-31T08:39:06.851Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3.1-pro-20261023",
      "model_slug": "moonshotai/kimi-k3.1-pro",
      "model_name": "MoonshotAI: Kimi K3.1 Pro",
      "provider_name": "PrimaLabs",
      "accuracy": 0.927326,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 172,
      "avg_cost_per_task_usd": 0.04753577906976744,
      "avg_duration_per_task_ms": 65225.773255813954,
      "avg_output_tokens_per_task": 5210.343023255814,
      "last_run_timestamp": "2026-11-06T15:43:35.998Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3.1-pro-20261023",
      "model_slug": "moonshotai/kimi-k3.1-pro",
      "model_name": "MoonshotAI: Kimi K3.1 Pro",
      "provider_name": "Phala",
      "accuracy": 0.9267675,
      "accuracy_std_dev": 0.002525500000000014,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.07679769191919193,
      "avg_duration_per_task_ms": 175516.4292929293,
      "avg_output_tokens_per_task": 7530.295454545455,
      "last_run_timestamp": "2026-11-05T18:00:48.183Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-sol-20260709",
      "model_slug": "openai/gpt-5.6-sol",
      "model_name": "OpenAI: GPT-5.6 Sol",
      "provider_name": "OpenAI",
      "accuracy": 0.9264870000000001,
      "accuracy_std_dev": 0.007815979699734826,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.03737200673400674,
      "avg_duration_per_task_ms": 60770.02693602694,
      "avg_output_tokens_per_task": 3580.875420875421,
      "last_run_timestamp": "2026-10-02T23:20:36.509Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Fireworks",
      "accuracy": 0.9264075634856123,
      "accuracy_std_dev": 0.007984700291764953,
      "run_count": 33,
      "total_questions": 6151,
      "avg_cost_per_task_usd": 0.1738332020809624,
      "avg_duration_per_task_ms": 167157.2952365469,
      "avg_output_tokens_per_task": 9377.01739554544,
      "last_run_timestamp": "2026-10-31T08:52:06.416Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-sonnet-5.5-20260928",
      "model_slug": "anthropic/claude-sonnet-5.5",
      "model_name": "Anthropic: Claude Sonnet 5.5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.9260940999999999,
      "accuracy_std_dev": 0.00286181765491794,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.026339350505050502,
      "avg_duration_per_task_ms": 19433.036868686868,
      "avg_output_tokens_per_task": 2294.320202020202,
      "last_run_timestamp": "2026-10-31T08:17:29.639Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Decart",
      "accuracy": 0.9259260000000001,
      "accuracy_std_dev": 0.004610436378478734,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.10788488484848485,
      "avg_duration_per_task_ms": 187765.80707070706,
      "avg_output_tokens_per_task": 9499.632323232323,
      "last_run_timestamp": "2026-10-24T08:37:09.424Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "KT",
      "accuracy": 0.925926,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.13830230303030303,
      "avg_duration_per_task_ms": 115013.06313131313,
      "avg_output_tokens_per_task": 9092.70202020202,
      "last_run_timestamp": "2026-11-11T01:44:59.803Z"
    },
    {
      "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": 0.9255047500000002,
      "accuracy_std_dev": 0.0055038172832589756,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.14877371590909091,
      "avg_duration_per_task_ms": 85367.24242424243,
      "avg_output_tokens_per_task": 15632.513888888889,
      "last_run_timestamp": "2026-10-01T19:13:26.156Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Relace",
      "accuracy": 0.9252448891402715,
      "accuracy_std_dev": 0.0069132275216017836,
      "run_count": 9,
      "total_questions": 1768,
      "avg_cost_per_task_usd": 0.10858752149321266,
      "avg_duration_per_task_ms": 188184.25339366516,
      "avg_output_tokens_per_task": 9458.231334841628,
      "last_run_timestamp": "2026-10-31T09:20:20.642Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Darkbloom",
      "accuracy": 0.925214,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 78,
      "avg_cost_per_task_usd": 0.0012695128205128204,
      "avg_duration_per_task_ms": 237789.78205128206,
      "avg_output_tokens_per_task": 6844.423076923077,
      "last_run_timestamp": "2026-10-30T01:38:27.650Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "x-ai/grok-4.6-20260810",
      "model_slug": "x-ai/grok-4.6",
      "model_name": "SpaceXAI: Grok 4.6",
      "provider_name": "xAI",
      "accuracy": 0.9246375695652174,
      "accuracy_std_dev": 0.006863146440316182,
      "run_count": 7,
      "total_questions": 1380,
      "avg_cost_per_task_usd": 0.12292555942028985,
      "avg_duration_per_task_ms": 340480.1579710145,
      "avg_output_tokens_per_task": 19662.684057971015,
      "last_run_timestamp": "2026-10-01T16:09:31.184Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-sol-20260922",
      "model_slug": "openai/gpt-6-sol",
      "model_name": "OpenAI: GPT-6 Sol",
      "provider_name": "Azure",
      "accuracy": 0.924599700084246,
      "accuracy_std_dev": 0.006819818440072705,
      "run_count": 6,
      "total_questions": 1187,
      "avg_cost_per_task_usd": 0.02751449789385004,
      "avg_duration_per_task_ms": 24957.139848357205,
      "avg_output_tokens_per_task": 2420.7127211457455,
      "last_run_timestamp": "2026-10-01T08:24:52.927Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Modal",
      "accuracy": 0.9243430301944727,
      "accuracy_std_dev": 0.011215611086765923,
      "run_count": 11,
      "total_questions": 1954,
      "avg_cost_per_task_usd": 0.14577470982599794,
      "avg_duration_per_task_ms": 113023.06345957011,
      "avg_output_tokens_per_task": 9570.704708290687,
      "last_run_timestamp": "2026-10-31T08:14:43.086Z"
    },
    {
      "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": 0.9243079168138753,
      "accuracy_std_dev": 0.02419971284017385,
      "run_count": 32,
      "total_questions": 6227,
      "avg_cost_per_task_usd": 0.1328945098763449,
      "avg_duration_per_task_ms": 158997.73036775333,
      "avg_output_tokens_per_task": 10041.62630480167,
      "last_run_timestamp": "2026-11-04T21:14:13.300Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3.1-pro-20261023",
      "model_slug": "moonshotai/kimi-k3.1-pro",
      "model_name": "MoonshotAI: Kimi K3.1 Pro",
      "provider_name": "Together",
      "accuracy": 0.9242420000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.09054113636363637,
      "avg_duration_per_task_ms": 151399.2323232323,
      "avg_output_tokens_per_task": 8895.065656565657,
      "last_run_timestamp": "2026-11-04T18:00:08.306Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-sol-20260709",
      "model_slug": "openai/gpt-5.6-sol",
      "model_name": "OpenAI: GPT-5.6 Sol",
      "provider_name": "auto-routing",
      "accuracy": 0.9242420000000001,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.03863523484848485,
      "avg_duration_per_task_ms": 63161.229797979795,
      "avg_output_tokens_per_task": 3708.2954545454545,
      "last_run_timestamp": "2026-10-01T19:12:28.573Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Makora",
      "accuracy": 0.9240912983587339,
      "accuracy_std_dev": 0.015687164131543977,
      "run_count": 10,
      "total_questions": 1706,
      "avg_cost_per_task_usd": 0.1119299531066823,
      "avg_duration_per_task_ms": 198230.14243845252,
      "avg_output_tokens_per_task": 8354.249120750294,
      "last_run_timestamp": "2026-10-31T08:14:43.089Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "River",
      "accuracy": 0.9236109999999998,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 192,
      "avg_cost_per_task_usd": 0.15688300520833334,
      "avg_duration_per_task_ms": 120831.05208333333,
      "avg_output_tokens_per_task": 10253.458333333334,
      "last_run_timestamp": "2026-11-14T00:02:16.722Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Near AI",
      "accuracy": 0.9234005000000001,
      "accuracy_std_dev": 0.005892500000000023,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.02620280303030303,
      "avg_duration_per_task_ms": 592141.2247474748,
      "avg_output_tokens_per_task": 21737.315656565657,
      "last_run_timestamp": "2026-10-31T10:06:11.748Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.5-flash-20260519",
      "model_slug": "google/gemini-3.5-flash",
      "model_name": "Google: Gemini 3.5 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.9234005000000001,
      "accuracy_std_dev": 0.000841499999999995,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.1436656515151515,
      "avg_duration_per_task_ms": 89621.648989899,
      "avg_output_tokens_per_task": 15839.800505050505,
      "last_run_timestamp": "2026-10-01T19:18:40.706Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "BaseTen",
      "accuracy": 0.9232600924690181,
      "accuracy_std_dev": 0.007367183577716027,
      "run_count": 11,
      "total_questions": 2098,
      "avg_cost_per_task_usd": 0.13289319971401337,
      "avg_duration_per_task_ms": 111823.89990467111,
      "avg_output_tokens_per_task": 8724.990943755958,
      "last_run_timestamp": "2026-10-31T08:14:19.970Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "DeepInfra",
      "accuracy": 0.9228069789473685,
      "accuracy_std_dev": 0.011126928806578514,
      "run_count": 3,
      "total_questions": 570,
      "avg_cost_per_task_usd": 0.00580749649122807,
      "avg_duration_per_task_ms": 218614.69122807018,
      "avg_output_tokens_per_task": 19521.97543859649,
      "last_run_timestamp": "2026-10-31T08:40:34.158Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "ParetoAlpha",
      "accuracy": 0.92256925,
      "accuracy_std_dev": 0.005618592528382903,
      "run_count": 5,
      "total_questions": 960,
      "avg_cost_per_task_usd": 0.148641165625,
      "avg_duration_per_task_ms": 385822.7,
      "avg_output_tokens_per_task": 9697.144791666668,
      "last_run_timestamp": "2026-10-31T08:38:36.234Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.5-flash-20260519",
      "model_slug": "google/gemini-3.5-flash",
      "model_name": "Google: Gemini 3.5 Flash",
      "provider_name": "Google AI Studio",
      "accuracy": 0.9225589999999999,
      "accuracy_std_dev": 0.003367000000000009,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.14588472727272728,
      "avg_duration_per_task_ms": 63933.833333333336,
      "avg_output_tokens_per_task": 16085.368686868687,
      "last_run_timestamp": "2026-10-01T19:12:44.398Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "Io Net",
      "accuracy": 0.9221647956015523,
      "accuracy_std_dev": 0.01297397580697607,
      "run_count": 5,
      "total_questions": 773,
      "avg_cost_per_task_usd": 0.013588241914618368,
      "avg_duration_per_task_ms": 479472.9379042691,
      "avg_output_tokens_per_task": 29097.7063389392,
      "last_run_timestamp": "2026-11-11T06:30:35.024Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-sol-20260709",
      "model_slug": "openai/gpt-5.6-sol",
      "model_name": "OpenAI: GPT-5.6 Sol",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.9219976666666666,
      "accuracy_std_dev": 0.0028614651647170097,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.08309612457912459,
      "avg_duration_per_task_ms": 49364.952861952865,
      "avg_output_tokens_per_task": 3748.658249158249,
      "last_run_timestamp": "2026-10-30T00:00:06.279Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Sail Research",
      "accuracy": 0.921964588235294,
      "accuracy_std_dev": 0.007325561540883652,
      "run_count": 17,
      "total_questions": 3366,
      "avg_cost_per_task_usd": 0.12449768538324421,
      "avg_duration_per_task_ms": 152742.1247771836,
      "avg_output_tokens_per_task": 11205.551396316103,
      "last_run_timestamp": "2026-11-06T22:04:48.119Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "DigitalOcean",
      "accuracy": 0.9208753333333334,
      "accuracy_std_dev": 0.003636621356271357,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.019656516835016836,
      "avg_duration_per_task_ms": 90437.9696969697,
      "avg_output_tokens_per_task": 20830.57239057239,
      "last_run_timestamp": "2026-10-31T08:09:30.885Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-sol-20260922",
      "model_slug": "openai/gpt-6-sol",
      "model_name": "OpenAI: GPT-6 Sol",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.9208750000000001,
      "accuracy_std_dev": 0.004761657064510223,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.026268675084175085,
      "avg_duration_per_task_ms": 25466.75420875421,
      "avg_output_tokens_per_task": 2308.39898989899,
      "last_run_timestamp": "2026-10-30T00:00:05.373Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-sol-20260922",
      "model_slug": "openai/gpt-6-sol",
      "model_name": "OpenAI: GPT-6 Sol",
      "provider_name": "OpenAI",
      "accuracy": 0.9208749999999999,
      "accuracy_std_dev": 0.002749143987983661,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.025590255892255893,
      "avg_duration_per_task_ms": 33502.22222222222,
      "avg_output_tokens_per_task": 2400.506734006734,
      "last_run_timestamp": "2026-10-02T23:17:42.236Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.9205386,
      "accuracy_std_dev": 0.004178401589124734,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.14947429494949496,
      "avg_duration_per_task_ms": 120588.33434343434,
      "avg_output_tokens_per_task": 9732.809090909092,
      "last_run_timestamp": "2026-10-31T08:09:30.327Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "auto-routing",
      "accuracy": 0.9202018,
      "accuracy_std_dev": 0.01007849159150319,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.13160375303030303,
      "avg_duration_per_task_ms": 237248.802020202,
      "avg_output_tokens_per_task": 10658.989393939393,
      "last_run_timestamp": "2026-10-31T08:21:51.280Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Alibaba",
      "accuracy": 0.9199569999999999,
      "accuracy_std_dev": 0.006554291127331654,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.17177113406795225,
      "avg_duration_per_task_ms": 299167.57943067036,
      "avg_output_tokens_per_task": 9764.863636363636,
      "last_run_timestamp": "2026-10-31T08:33:15.782Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3.1-pro-20261023",
      "model_slug": "moonshotai/kimi-k3.1-pro",
      "model_name": "MoonshotAI: Kimi K3.1 Pro",
      "provider_name": "Wafer",
      "accuracy": 0.9191920000000001,
      "accuracy_std_dev": 0.003367000000000009,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.07685085101010102,
      "avg_duration_per_task_ms": 125614,
      "avg_output_tokens_per_task": 7560.146464646465,
      "last_run_timestamp": "2026-11-05T21:03:19.219Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3.1-pro-20261023",
      "model_slug": "moonshotai/kimi-k3.1-pro",
      "model_name": "MoonshotAI: Kimi K3.1 Pro",
      "provider_name": "Relace",
      "accuracy": 0.9191919999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.062148696969696965,
      "avg_duration_per_task_ms": 92710.41414141415,
      "avg_output_tokens_per_task": 7696.424242424242,
      "last_run_timestamp": "2026-11-05T00:00:09.895Z"
    },
    {
      "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": "x-ai/grok-4.7-20260916",
      "model_slug": "x-ai/grok-4.7",
      "model_name": "SpaceXAI: Grok 4.7",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.9191919999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.21825907575757575,
      "avg_duration_per_task_ms": 803866.0202020202,
      "avg_output_tokens_per_task": 36106.61616161616,
      "last_run_timestamp": "2026-11-03T03:00:04.447Z"
    },
    {
      "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": "openai/gpt-6.1-terra-20261026",
      "model_slug": "openai/gpt-6.1-terra",
      "model_name": "OpenAI: GPT-6.1 Terra",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.9186305,
      "accuracy_std_dev": 0.0030220059425267055,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.012602135521885521,
      "avg_duration_per_task_ms": 20976.548821548822,
      "avg_output_tokens_per_task": 2132.777777777778,
      "last_run_timestamp": "2026-11-02T17:19:33.587Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "SiliconFlow",
      "accuracy": 0.9186303333333333,
      "accuracy_std_dev": 0.010408091158111372,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.025292042087542085,
      "avg_duration_per_task_ms": 160402.3451178451,
      "avg_output_tokens_per_task": 20957.713804713803,
      "last_run_timestamp": "2026-10-31T08:50:50.713Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "Venice",
      "accuracy": 0.9177144177922683,
      "accuracy_std_dev": 0.013751154902061724,
      "run_count": 11,
      "total_questions": 2147,
      "avg_cost_per_task_usd": 0.2460820894271076,
      "avg_duration_per_task_ms": 409070.95482068,
      "avg_output_tokens_per_task": 37154.30647414998,
      "last_run_timestamp": "2026-10-31T08:34:22.445Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "PrimaLabs",
      "accuracy": 0.9175825384615385,
      "accuracy_std_dev": 0.009135000000000004,
      "run_count": 2,
      "total_questions": 182,
      "avg_cost_per_task_usd": 0.03656871978021978,
      "avg_duration_per_task_ms": 44231.42857142857,
      "avg_output_tokens_per_task": 10074.60989010989,
      "last_run_timestamp": "2026-11-06T15:33:21.027Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-sonnet-5.5-20260928",
      "model_slug": "anthropic/claude-sonnet-5.5",
      "model_name": "Anthropic: Claude Sonnet 5.5",
      "provider_name": "Azure",
      "accuracy": 0.9175082499999999,
      "accuracy_std_dev": 0.003571399938516188,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.025937152356902356,
      "avg_duration_per_task_ms": 19174.19191919192,
      "avg_output_tokens_per_task": 2235.470117845118,
      "last_run_timestamp": "2026-10-31T08:20:35.966Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Moonshot AI",
      "accuracy": 0.9174755201839551,
      "accuracy_std_dev": 0.007025002493949755,
      "run_count": 10,
      "total_questions": 1957,
      "avg_cost_per_task_usd": 0.14208481297904957,
      "avg_duration_per_task_ms": 283983.31374552887,
      "avg_output_tokens_per_task": 9268.438426162493,
      "last_run_timestamp": "2026-10-24T08:44:42.687Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Cloudflare",
      "accuracy": 0.916877387341772,
      "accuracy_std_dev": 0.005308941032719149,
      "run_count": 6,
      "total_questions": 1185,
      "avg_cost_per_task_usd": 0.10198255443037975,
      "avg_duration_per_task_ms": 806625.7974683545,
      "avg_output_tokens_per_task": 39607.979746835445,
      "last_run_timestamp": "2026-10-31T16:24:40.574Z"
    },
    {
      "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": 0.9168302048835251,
      "accuracy_std_dev": 0.02695337897854574,
      "run_count": 18,
      "total_questions": 3563,
      "avg_cost_per_task_usd": 0.14734509009261856,
      "avg_duration_per_task_ms": 382336.6034240808,
      "avg_output_tokens_per_task": 9501.996070726915,
      "last_run_timestamp": "2026-10-31T09:12:10.294Z"
    },
    {
      "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": 0.9165900909090909,
      "accuracy_std_dev": 0.009424357296191685,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.0340643388429752,
      "avg_duration_per_task_ms": 223570.00459136823,
      "avg_output_tokens_per_task": 28227.440771349862,
      "last_run_timestamp": "2026-10-31T16:05:09.530Z"
    },
    {
      "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": 0.9164236554907678,
      "accuracy_std_dev": 0.014292148644484252,
      "run_count": 11,
      "total_questions": 2058,
      "avg_cost_per_task_usd": 0.1620518138969874,
      "avg_duration_per_task_ms": 492245.379494655,
      "avg_output_tokens_per_task": 11216.908163265307,
      "last_run_timestamp": "2026-10-31T08:37:30.440Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Baidu",
      "accuracy": 0.9159186669815778,
      "accuracy_std_dev": 0.007589099806154146,
      "run_count": 11,
      "total_questions": 2117,
      "avg_cost_per_task_usd": 0.09760643316013226,
      "avg_duration_per_task_ms": 645928.4369390648,
      "avg_output_tokens_per_task": 39652.34246575343,
      "last_run_timestamp": "2026-11-10T18:49:16.069Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "AkashML",
      "accuracy": 0.9158379972375691,
      "accuracy_std_dev": 0.012441503983078046,
      "run_count": 13,
      "total_questions": 1810,
      "avg_cost_per_task_usd": 0.11735693314917128,
      "avg_duration_per_task_ms": 179712.15801104973,
      "avg_output_tokens_per_task": 8118.493370165746,
      "last_run_timestamp": "2026-10-31T08:07:17.446Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "DekaLLM",
      "accuracy": 0.915825,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.014663378787878788,
      "avg_duration_per_task_ms": 96926.35858585859,
      "avg_output_tokens_per_task": 21419.76767676768,
      "last_run_timestamp": "2026-11-04T09:00:09.993Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3.1-pro-20261023",
      "model_slug": "moonshotai/kimi-k3.1-pro",
      "model_name": "MoonshotAI: Kimi K3.1 Pro",
      "provider_name": "DeepInfra",
      "accuracy": 0.915825,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.06562552525252525,
      "avg_duration_per_task_ms": 514778.64646464644,
      "avg_output_tokens_per_task": 6703.651515151515,
      "last_run_timestamp": "2026-11-05T06:03:33.391Z"
    },
    {
      "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": 0.9153655454545453,
      "accuracy_std_dev": 0.009766163789176628,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.1584063576675849,
      "avg_duration_per_task_ms": 611741.2534435262,
      "avg_output_tokens_per_task": 39713.49908172635,
      "last_run_timestamp": "2026-10-31T08:28:10.440Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "AtlasCloud",
      "accuracy": 0.9149830000000001,
      "accuracy_std_dev": 0.009335323856192691,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.15949898674242427,
      "avg_duration_per_task_ms": 579778.1357323233,
      "avg_output_tokens_per_task": 39960.98863636364,
      "last_run_timestamp": "2026-10-31T08:49:58.581Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Together",
      "accuracy": 0.9149829999999999,
      "accuracy_std_dev": 0.006208509031965745,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.035431615151515154,
      "avg_duration_per_task_ms": 289878.5762626263,
      "avg_output_tokens_per_task": 29385.477272727272,
      "last_run_timestamp": "2026-10-31T16:17:15.690Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "CoreWeave",
      "accuracy": 0.9149660000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 196,
      "avg_cost_per_task_usd": 0.022535153061224493,
      "avg_duration_per_task_ms": 106164.53571428571,
      "avg_output_tokens_per_task": 18639.61224489796,
      "last_run_timestamp": "2026-11-04T18:00:08.219Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "thinkingmachines/inkling-2-small-20261107",
      "model_slug": "thinkingmachines/inkling-2-small",
      "model_name": "Thinking Machines: Inkling 2 Small",
      "provider_name": "Together",
      "accuracy": 0.9141410000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.03261640909090909,
      "avg_duration_per_task_ms": 98347.65151515152,
      "avg_output_tokens_per_task": 32384.540404040403,
      "last_run_timestamp": "2026-11-13T00:01:54.553Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "thinkingmachines/inkling-2-small-20261107",
      "model_slug": "thinkingmachines/inkling-2-small",
      "model_name": "Thinking Machines: Inkling 2 Small",
      "provider_name": "Nebius",
      "accuracy": 0.9141410000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.030628545454545457,
      "avg_duration_per_task_ms": 320020.58585858584,
      "avg_output_tokens_per_task": 30417.116161616163,
      "last_run_timestamp": "2026-11-13T00:01:53.399Z"
    },
    {
      "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": 0.9140508968609865,
      "accuracy_std_dev": 0.01852693073581796,
      "run_count": 7,
      "total_questions": 892,
      "avg_cost_per_task_usd": 0.02261122085201794,
      "avg_duration_per_task_ms": 825583.0426008969,
      "avg_output_tokens_per_task": 20985.829596412557,
      "last_run_timestamp": "2026-10-31T16:03:21.470Z"
    },
    {
      "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": 0.9133203374352331,
      "accuracy_std_dev": 0.03022029270209009,
      "run_count": 8,
      "total_questions": 1544,
      "avg_cost_per_task_usd": 0.04059680051813471,
      "avg_duration_per_task_ms": 413773.9656735751,
      "avg_output_tokens_per_task": 12145.259715025906,
      "last_run_timestamp": "2026-10-31T16:08:56.804Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Baidu",
      "accuracy": 0.91311096,
      "accuracy_std_dev": 0.009099497963486788,
      "run_count": 4,
      "total_questions": 750,
      "avg_cost_per_task_usd": 0.01717951733333333,
      "avg_duration_per_task_ms": 130330.4,
      "avg_output_tokens_per_task": 20403.964,
      "last_run_timestamp": "2026-11-10T18:06:00.757Z"
    },
    {
      "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": "Wafer",
      "accuracy": 0.9129648851570964,
      "accuracy_std_dev": 0.008195106272424223,
      "run_count": 14,
      "total_questions": 2769,
      "avg_cost_per_task_usd": 0.0985225175153485,
      "avg_duration_per_task_ms": 283575.7085590466,
      "avg_output_tokens_per_task": 25018.89021307331,
      "last_run_timestamp": "2026-10-31T08:40:32.804Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Alibaba",
      "accuracy": 0.91245775,
      "accuracy_std_dev": 0.005709247427419856,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.027895760101010105,
      "avg_duration_per_task_ms": 177535.13888888888,
      "avg_output_tokens_per_task": 21414.281565656565,
      "last_run_timestamp": "2026-10-31T08:15:24.481Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Alibaba",
      "accuracy": 0.911926105263158,
      "accuracy_std_dev": 0.013578898854258508,
      "run_count": 19,
      "total_questions": 3762,
      "avg_cost_per_task_usd": 0.1411968139287613,
      "avg_duration_per_task_ms": 616719.149122807,
      "avg_output_tokens_per_task": 40116.86735778841,
      "last_run_timestamp": "2026-10-31T08:31:11.942Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "River",
      "accuracy": 0.911616,
      "accuracy_std_dev": 0.004209000000000018,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.009611237373737374,
      "avg_duration_per_task_ms": 119490.68686868687,
      "avg_output_tokens_per_task": 19858.734848484848,
      "last_run_timestamp": "2026-11-12T20:19:21.746Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Venice",
      "accuracy": 0.911616,
      "accuracy_std_dev": 0.004209000000000018,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.024379722222222223,
      "avg_duration_per_task_ms": 102982.82575757576,
      "avg_output_tokens_per_task": 20129.815656565657,
      "last_run_timestamp": "2026-10-31T08:34:49.804Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "GMICloud",
      "accuracy": 0.9114477999999999,
      "accuracy_std_dev": 0.0039265801609033885,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.018160035353535352,
      "avg_duration_per_task_ms": 129702.60808080809,
      "avg_output_tokens_per_task": 22106.812121212122,
      "last_run_timestamp": "2026-11-10T06:01:40.563Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6.1-terra-20261026",
      "model_slug": "openai/gpt-6.1-terra",
      "model_name": "OpenAI: GPT-6.1 Terra",
      "provider_name": "OpenAI",
      "accuracy": 0.910774,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.011462444444444444,
      "avg_duration_per_task_ms": 27557.444444444445,
      "avg_output_tokens_per_task": 2133.969696969697,
      "last_run_timestamp": "2026-10-28T00:00:04.598Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "thinkingmachines/inkling-2-small-20261107",
      "model_slug": "thinkingmachines/inkling-2-small",
      "model_name": "Thinking Machines: Inkling 2 Small",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.910774,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.01510519191919192,
      "avg_duration_per_task_ms": 254029.7676767677,
      "avg_output_tokens_per_task": 30029.237373737375,
      "last_run_timestamp": "2026-11-13T03:01:57.471Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "thinkingmachines/inkling-2-small-20261107",
      "model_slug": "thinkingmachines/inkling-2-small",
      "model_name": "Thinking Machines: Inkling 2 Small",
      "provider_name": "Cloudflare",
      "accuracy": 0.910774,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.030777131313131315,
      "avg_duration_per_task_ms": 94527.24747474748,
      "avg_output_tokens_per_task": 30575.68686868687,
      "last_run_timestamp": "2026-11-13T00:01:53.427Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "DekaLLM",
      "accuracy": 0.9106891772151898,
      "accuracy_std_dev": 0.008562817793226714,
      "run_count": 5,
      "total_questions": 948,
      "avg_cost_per_task_usd": 0.14278316033755276,
      "avg_duration_per_task_ms": 333972.91455696203,
      "avg_output_tokens_per_task": 37406.36181434599,
      "last_run_timestamp": "2026-10-31T08:31:11.854Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6.1-terra-20261026",
      "model_slug": "openai/gpt-6.1-terra",
      "model_name": "OpenAI: GPT-6.1 Terra",
      "provider_name": "Azure",
      "accuracy": 0.9096519999999999,
      "accuracy_std_dev": 0.0075707498968068145,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.012606855218855219,
      "avg_duration_per_task_ms": 20423.260942760942,
      "avg_output_tokens_per_task": 2135.06734006734,
      "last_run_timestamp": "2026-11-03T15:00:04.730Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta/muse-spark-open-20261015",
      "model_slug": "meta/muse-spark-open",
      "model_name": "Meta: Muse Spark Open",
      "provider_name": "Modal",
      "accuracy": 0.909632,
      "accuracy_std_dev": 0.0055275000000000185,
      "run_count": 2,
      "total_questions": 308,
      "avg_cost_per_task_usd": 0.05270355519480519,
      "avg_duration_per_task_ms": 55285.37662337662,
      "avg_output_tokens_per_task": 12143.461038961039,
      "last_run_timestamp": "2026-10-31T08:15:03.291Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-luna-pro-20260709",
      "model_slug": "openai/gpt-5.6-luna-pro",
      "model_name": "OpenAI: GPT-5.6 Luna Pro",
      "provider_name": "Azure",
      "accuracy": 0.9095115,
      "accuracy_std_dev": 0.002759842885020819,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.0416229191919192,
      "avg_duration_per_task_ms": 138307.48484848486,
      "avg_output_tokens_per_task": 27969.503787878788,
      "last_run_timestamp": "2026-10-01T16:11:23.394Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "NextBit",
      "accuracy": 0.9094272,
      "accuracy_std_dev": 0.00547072746168187,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.025930728282828285,
      "avg_duration_per_task_ms": 177886.11212121212,
      "avg_output_tokens_per_task": 28121.050505050505,
      "last_run_timestamp": "2026-10-03T09:34:35.533Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "StreamLake",
      "accuracy": 0.9090910000000001,
      "accuracy_std_dev": 0.003367000000000009,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.018701613636363638,
      "avg_duration_per_task_ms": 137003.63383838383,
      "avg_output_tokens_per_task": 22059.484848484848,
      "last_run_timestamp": "2026-10-31T08:36:23.829Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Reka",
      "accuracy": 0.9090909999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.1305062777777778,
      "avg_duration_per_task_ms": 540279.6767676767,
      "avg_output_tokens_per_task": 38448.40404040404,
      "last_run_timestamp": "2026-10-01T09:14:23.001Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": "thinkingmachines/inkling-2-small-20261107",
      "model_slug": "thinkingmachines/inkling-2-small",
      "model_name": "Thinking Machines: Inkling 2 Small",
      "provider_name": "Modal",
      "accuracy": 0.9090909999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.02941411111111111,
      "avg_duration_per_task_ms": 293246.38383838383,
      "avg_output_tokens_per_task": 29214.535353535353,
      "last_run_timestamp": "2026-11-13T06:01:37.481Z"
    },
    {
      "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": 0.909090857142857,
      "accuracy_std_dev": 0.0075824593339510055,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.038180986291486294,
      "avg_duration_per_task_ms": 147767.6544011544,
      "avg_output_tokens_per_task": 28692.146464646463,
      "last_run_timestamp": "2026-11-10T06:07:30.895Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "StepFun",
      "accuracy": 0.9090908333333333,
      "accuracy_std_dev": 0.007076201038614481,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.01709530134680135,
      "avg_duration_per_task_ms": 84455.94696969698,
      "avg_output_tokens_per_task": 21725.516835016835,
      "last_run_timestamp": "2026-11-02T17:24:31.225Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Phala",
      "accuracy": 0.9083779032258064,
      "accuracy_std_dev": 0.022205862232752862,
      "run_count": 12,
      "total_questions": 2232,
      "avg_cost_per_task_usd": 0.1474674775985663,
      "avg_duration_per_task_ms": 555743.3785842294,
      "avg_output_tokens_per_task": 37926.89829749104,
      "last_run_timestamp": "2026-10-31T08:44:11.309Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "InferenceNet",
      "accuracy": 0.908249,
      "accuracy_std_dev": 0.006574437713599543,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.009681809974747475,
      "avg_duration_per_task_ms": 167110.30050505052,
      "avg_output_tokens_per_task": 18361.854166666668,
      "last_run_timestamp": "2026-11-04T21:05:58.450Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Alibaba",
      "accuracy": 0.9081554444444445,
      "accuracy_std_dev": 0.008876832231665518,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.033613874298540965,
      "avg_duration_per_task_ms": 210737.89562289562,
      "avg_output_tokens_per_task": 27019.509539842875,
      "last_run_timestamp": "2026-10-31T08:40:16.353Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Novita",
      "accuracy": 0.9080193636363636,
      "accuracy_std_dev": 0.0188735967823871,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.037270338383838385,
      "avg_duration_per_task_ms": 248010.31542699726,
      "avg_output_tokens_per_task": 31082.94582185491,
      "last_run_timestamp": "2026-10-31T16:09:19.241Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Baidu",
      "accuracy": 0.9078681221719456,
      "accuracy_std_dev": 0.005280935005338756,
      "run_count": 7,
      "total_questions": 1326,
      "avg_cost_per_task_usd": 0.023841498491704374,
      "avg_duration_per_task_ms": 170556.5128205128,
      "avg_output_tokens_per_task": 27449.11161387632,
      "last_run_timestamp": "2026-11-10T18:05:49.922Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Mancer 2",
      "accuracy": 0.9078013574468085,
      "accuracy_std_dev": 0.012928926018815361,
      "run_count": 3,
      "total_questions": 470,
      "avg_cost_per_task_usd": 0.01807536170212766,
      "avg_duration_per_task_ms": 357786.7425531915,
      "avg_output_tokens_per_task": 17947.382978723403,
      "last_run_timestamp": "2026-10-31T09:05:48.501Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "StreamLake",
      "accuracy": 0.9076879166666667,
      "accuracy_std_dev": 0.026811471259078443,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.036969556397306404,
      "avg_duration_per_task_ms": 266792.257996633,
      "avg_output_tokens_per_task": 30839.63425925926,
      "last_run_timestamp": "2026-10-31T16:23:41.132Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Cohere",
      "accuracy": 0.9074074999999998,
      "accuracy_std_dev": 0.008417500000000022,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.014508287878787877,
      "avg_duration_per_task_ms": 90086.95707070707,
      "avg_output_tokens_per_task": 21790.61111111111,
      "last_run_timestamp": "2026-10-31T08:13:11.036Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Nebius",
      "accuracy": 0.9074072499999999,
      "accuracy_std_dev": 0.004610545920766882,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.025605553030303027,
      "avg_duration_per_task_ms": 96285.60858585859,
      "avg_output_tokens_per_task": 21160.45580808081,
      "last_run_timestamp": "2026-10-31T09:42:32.702Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "CoreWeave",
      "accuracy": 0.9073293664700928,
      "accuracy_std_dev": 0.008220232321331707,
      "run_count": 12,
      "total_questions": 2374,
      "avg_cost_per_task_usd": 0.030960309604043806,
      "avg_duration_per_task_ms": 400875.2552653749,
      "avg_output_tokens_per_task": 32028.098146588036,
      "last_run_timestamp": "2026-10-31T16:32:36.327Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "DeepInfra",
      "accuracy": 0.9071887349785409,
      "accuracy_std_dev": 0.014345789366918764,
      "run_count": 5,
      "total_questions": 932,
      "avg_cost_per_task_usd": 0.12532361909871245,
      "avg_duration_per_task_ms": 1379721.1190987125,
      "avg_output_tokens_per_task": 49889.66094420601,
      "last_run_timestamp": "2026-10-31T08:27:43.435Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "StreamLake",
      "accuracy": 0.9070474558880683,
      "accuracy_std_dev": 0.028977319190080154,
      "run_count": 13,
      "total_questions": 2573,
      "avg_cost_per_task_usd": 0.0985262642829382,
      "avg_duration_per_task_ms": 564503.555771473,
      "avg_output_tokens_per_task": 39384.385542168675,
      "last_run_timestamp": "2026-10-31T08:39:28.448Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Modal",
      "accuracy": 0.9070292653061225,
      "accuracy_std_dev": 0.005723067708842884,
      "run_count": 3,
      "total_questions": 588,
      "avg_cost_per_task_usd": 0.021270954081632654,
      "avg_duration_per_task_ms": 117754.6887755102,
      "avg_output_tokens_per_task": 21116.049319727892,
      "last_run_timestamp": "2026-10-31T08:51:31.997Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Phala",
      "accuracy": 0.9062928608176387,
      "accuracy_std_dev": 0.051292972174679034,
      "run_count": 11,
      "total_questions": 2177,
      "avg_cost_per_task_usd": 0.13471541892512628,
      "avg_duration_per_task_ms": 314455.0349104272,
      "avg_output_tokens_per_task": 10081.463481855764,
      "last_run_timestamp": "2026-10-31T09:37:23.390Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Fireworks",
      "accuracy": 0.9062632644787644,
      "accuracy_std_dev": 0.007463126815174401,
      "run_count": 8,
      "total_questions": 1554,
      "avg_cost_per_task_usd": 0.024727387387387386,
      "avg_duration_per_task_ms": 89200.08815958817,
      "avg_output_tokens_per_task": 20548.897683397685,
      "last_run_timestamp": "2026-10-31T08:46:47.079Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Venice",
      "accuracy": 0.9061779082526508,
      "accuracy_std_dev": 0.011623858653648528,
      "run_count": 11,
      "total_questions": 2169,
      "avg_cost_per_task_usd": 0.026847682342093126,
      "avg_duration_per_task_ms": 336788.01567542646,
      "avg_output_tokens_per_task": 22212.46196403873,
      "last_run_timestamp": "2026-10-31T16:12:28.227Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Morph",
      "accuracy": 0.9061447499999998,
      "accuracy_std_dev": 0.005102889738912664,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.007892248737373738,
      "avg_duration_per_task_ms": 123987.29040404041,
      "avg_output_tokens_per_task": 16306.166666666666,
      "last_run_timestamp": "2026-10-31T08:53:36.415Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "Novita",
      "accuracy": 0.9060041666666666,
      "accuracy_std_dev": 0.006179339377222198,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.032878036195286196,
      "avg_duration_per_task_ms": 809018.2592592592,
      "avg_output_tokens_per_task": 37448.21212121212,
      "last_run_timestamp": "2026-10-31T08:23:27.402Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "GMICloud",
      "accuracy": 0.9059826923076922,
      "accuracy_std_dev": 0.02531585573200691,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.03357483760683761,
      "avg_duration_per_task_ms": 222969.2094017094,
      "avg_output_tokens_per_task": 27670.721056721057,
      "last_run_timestamp": "2026-11-10T06:30:46.502Z"
    },
    {
      "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": 0.9059684472128378,
      "accuracy_std_dev": 0.029240642852196717,
      "run_count": 12,
      "total_questions": 2368,
      "avg_cost_per_task_usd": 0.15467415033783785,
      "avg_duration_per_task_ms": 833998.046875,
      "avg_output_tokens_per_task": 38756.0152027027,
      "last_run_timestamp": "2026-10-31T08:15:03.301Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Decart",
      "accuracy": 0.9059342499999998,
      "accuracy_std_dev": 0.010910172663874526,
      "run_count": 3,
      "total_questions": 528,
      "avg_cost_per_task_usd": 0.008819143939393941,
      "avg_duration_per_task_ms": 163195.54924242425,
      "avg_output_tokens_per_task": 22179.03787878788,
      "last_run_timestamp": "2026-10-31T08:49:09.016Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "SiliconFlow",
      "accuracy": 0.9057238888888889,
      "accuracy_std_dev": 0.009356618102764875,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.03432713860830528,
      "avg_duration_per_task_ms": 196757.96408529743,
      "avg_output_tokens_per_task": 28444.87317620651,
      "last_run_timestamp": "2026-10-31T08:29:56.453Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Relace",
      "accuracy": 0.9057234999999999,
      "accuracy_std_dev": 0.0029159075345422127,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.013535670454545456,
      "avg_duration_per_task_ms": 88017.25757575757,
      "avg_output_tokens_per_task": 21462.573232323233,
      "last_run_timestamp": "2026-11-01T12:00:08.725Z"
    },
    {
      "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": "DeepInfra",
      "accuracy": 0.9055437090526317,
      "accuracy_std_dev": 0.022030260791818957,
      "run_count": 12,
      "total_questions": 2375,
      "avg_cost_per_task_usd": 0.12441664378947369,
      "avg_duration_per_task_ms": 324658.0492631579,
      "avg_output_tokens_per_task": 8521.422736842105,
      "last_run_timestamp": "2026-10-31T08:25:00.194Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "auto-routing",
      "accuracy": 0.9054404524137931,
      "accuracy_std_dev": 0.015780138865932054,
      "run_count": 11,
      "total_questions": 2175,
      "avg_cost_per_task_usd": 0.10642273195402298,
      "avg_duration_per_task_ms": 537964.4225287356,
      "avg_output_tokens_per_task": 37458.18114942529,
      "last_run_timestamp": "2026-10-31T08:11:08.576Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Makora",
      "accuracy": 0.9054373465045593,
      "accuracy_std_dev": 0.0037554479359991244,
      "run_count": 5,
      "total_questions": 987,
      "avg_cost_per_task_usd": 0.026207787234042552,
      "avg_duration_per_task_ms": 127141.36068895644,
      "avg_output_tokens_per_task": 21741.29989868288,
      "last_run_timestamp": "2026-10-31T08:31:11.848Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "ModelRun",
      "accuracy": 0.9053869333333332,
      "accuracy_std_dev": 0.006878297322900644,
      "run_count": 15,
      "total_questions": 2970,
      "avg_cost_per_task_usd": 0.07875497138047138,
      "avg_duration_per_task_ms": 227261.18787878787,
      "avg_output_tokens_per_task": 31017.457912457914,
      "last_run_timestamp": "2026-10-31T16:10:09.223Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "InferenceNet",
      "accuracy": 0.905376,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 155,
      "avg_cost_per_task_usd": 0.09219292258064517,
      "avg_duration_per_task_ms": 892432.9419354838,
      "avg_output_tokens_per_task": 36790.02580645162,
      "last_run_timestamp": "2026-10-02T06:44:44.722Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "StreamLake",
      "accuracy": 0.9053496666666667,
      "accuracy_std_dev": 0.008612696635394351,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.02068693490460157,
      "avg_duration_per_task_ms": 208793.50953984287,
      "avg_output_tokens_per_task": 28854.00224466891,
      "last_run_timestamp": "2026-10-31T09:01:06.228Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Sail Research",
      "accuracy": 0.9053028750000003,
      "accuracy_std_dev": 0.00682546092651441,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.0031826717171717163,
      "avg_duration_per_task_ms": 242276.39330808082,
      "avg_output_tokens_per_task": 19735.85101010101,
      "last_run_timestamp": "2026-11-06T21:57:38.138Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-flash-20260826",
      "model_slug": "qwen/qwen3.8-flash",
      "model_name": "Qwen: Qwen3.8 Flash",
      "provider_name": "Alibaba",
      "accuracy": 0.9049724513274336,
      "accuracy_std_dev": 0.008906991898924113,
      "run_count": 4,
      "total_questions": 791,
      "avg_cost_per_task_usd": 0.020146505689001264,
      "avg_duration_per_task_ms": 644404.02528445,
      "avg_output_tokens_per_task": 42703.32616940582,
      "last_run_timestamp": "2026-10-28T03:59:43.480Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Wafer",
      "accuracy": 0.9048818749999998,
      "accuracy_std_dev": 0.008666358814945001,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.008252799242424242,
      "avg_duration_per_task_ms": 82042.59659090909,
      "avg_output_tokens_per_task": 16383.162878787878,
      "last_run_timestamp": "2026-10-31T08:56:01.471Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta/muse-spark-open-20261015",
      "model_slug": "meta/muse-spark-open",
      "model_name": "Meta: Muse Spark Open",
      "provider_name": "Makora",
      "accuracy": 0.9048634068157615,
      "accuracy_std_dev": 0.012266915351464668,
      "run_count": 5,
      "total_questions": 939,
      "avg_cost_per_task_usd": 0.06399653567625134,
      "avg_duration_per_task_ms": 75580.11395101172,
      "avg_output_tokens_per_task": 14917.831735889244,
      "last_run_timestamp": "2026-10-31T08:13:57.643Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Parasail",
      "accuracy": 0.9046013333333334,
      "accuracy_std_dev": 0.009152126540986068,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.02685950505050505,
      "avg_duration_per_task_ms": 136981.2441077441,
      "avg_output_tokens_per_task": 22236.84175084175,
      "last_run_timestamp": "2026-10-31T08:31:11.499Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "AtlasCloud",
      "accuracy": 0.9046013333333334,
      "accuracy_std_dev": 0.008285549120138134,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.01933321717171717,
      "avg_duration_per_task_ms": 139538.02861952863,
      "avg_output_tokens_per_task": 20900.63973063973,
      "last_run_timestamp": "2026-10-31T08:20:08.070Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "DeepInfra",
      "accuracy": 0.9044004522003033,
      "accuracy_std_dev": 0.00856689234728672,
      "run_count": 10,
      "total_questions": 1977,
      "avg_cost_per_task_usd": 0.014848743045017704,
      "avg_duration_per_task_ms": 370376.5558927668,
      "avg_output_tokens_per_task": 28855.582195245323,
      "last_run_timestamp": "2026-10-31T08:03:09.643Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Mistral",
      "accuracy": 0.904321,
      "accuracy_std_dev": 0.006024476685433623,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.025462516835016838,
      "avg_duration_per_task_ms": 81189.70454545454,
      "avg_output_tokens_per_task": 21063.152356902356,
      "last_run_timestamp": "2026-11-11T02:59:16.591Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Together",
      "accuracy": 0.9042355953038675,
      "accuracy_std_dev": 0.04114912403373615,
      "run_count": 11,
      "total_questions": 2172,
      "avg_cost_per_task_usd": 0.1376939341620626,
      "avg_duration_per_task_ms": 212484.72605893185,
      "avg_output_tokens_per_task": 9252.047421731124,
      "last_run_timestamp": "2026-10-31T08:40:21.641Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Relace",
      "accuracy": 0.9042316703786192,
      "accuracy_std_dev": 0.0031656583896560934,
      "run_count": 5,
      "total_questions": 898,
      "avg_cost_per_task_usd": 0.22350682405345215,
      "avg_duration_per_task_ms": 441723.66035634745,
      "avg_output_tokens_per_task": 42286.14922048998,
      "last_run_timestamp": "2026-10-31T08:44:55.396Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Alibaba",
      "accuracy": 0.9041981924485772,
      "accuracy_std_dev": 0.01113877499359892,
      "run_count": 18,
      "total_questions": 3549,
      "avg_cost_per_task_usd": 0.027435362073823615,
      "avg_duration_per_task_ms": 185052.6246830093,
      "avg_output_tokens_per_task": 25932.813750352212,
      "last_run_timestamp": "2026-10-31T08:28:13.141Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3-flash-preview-20251217",
      "model_slug": "google/gemini-3-flash-preview",
      "model_name": "Google: Gemini 3 Flash Preview",
      "provider_name": "Google",
      "accuracy": 0.9040405,
      "accuracy_std_dev": 0.011784499999999976,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.13174327272727274,
      "avg_duration_per_task_ms": 249918.9898989899,
      "avg_output_tokens_per_task": 43791.378787878784,
      "last_run_timestamp": "2026-10-01T19:18:35.588Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Novita",
      "accuracy": 0.9037997142857144,
      "accuracy_std_dev": 0.016282944378479363,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.12103858297258295,
      "avg_duration_per_task_ms": 424148.33405483403,
      "avg_output_tokens_per_task": 37524.782106782106,
      "last_run_timestamp": "2026-10-31T16:10:36.257Z"
    },
    {
      "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-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Novita",
      "accuracy": 0.903428,
      "accuracy_std_dev": 0.020115933056342978,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.12324611707988982,
      "avg_duration_per_task_ms": 520523.87649219466,
      "avg_output_tokens_per_task": 38776.62029384757,
      "last_run_timestamp": "2026-10-31T08:17:28.484Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.9033668,
      "accuracy_std_dev": 0.005061737504059227,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.026716864646464646,
      "avg_duration_per_task_ms": 106368.8292929293,
      "avg_output_tokens_per_task": 22085.938383838384,
      "last_run_timestamp": "2026-10-31T08:13:57.548Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Particle AI",
      "accuracy": 0.9031985,
      "accuracy_std_dev": 0.0075759444460740165,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.01504592297979798,
      "avg_duration_per_task_ms": 86737.82702020202,
      "avg_output_tokens_per_task": 17203.179292929293,
      "last_run_timestamp": "2026-11-03T01:49:06.293Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.4-20260305",
      "model_slug": "openai/gpt-5.4",
      "model_name": "OpenAI: GPT-5.4",
      "provider_name": "OpenAI",
      "accuracy": 0.9031985,
      "accuracy_std_dev": 0.004208500000000004,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.14226843434343436,
      "avg_duration_per_task_ms": 149341.0782828283,
      "avg_output_tokens_per_task": 9356.287878787878,
      "last_run_timestamp": "2026-10-01T19:14:55.028Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Together",
      "accuracy": 0.9029179999999999,
      "accuracy_std_dev": 0.0028613344206273234,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.023974107744107742,
      "avg_duration_per_task_ms": 50011.877104377105,
      "avg_output_tokens_per_task": 19809.430976430976,
      "last_run_timestamp": "2026-10-31T08:51:04.442Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "SiliconFlow",
      "accuracy": 0.9028377142857142,
      "accuracy_std_dev": 0.0030797912134375885,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.13983470995670993,
      "avg_duration_per_task_ms": 1849163.816017316,
      "avg_output_tokens_per_task": 58254.32756132756,
      "last_run_timestamp": "2026-10-31T16:13:09.866Z"
    },
    {
      "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": 0.9027775833333331,
      "accuracy_std_dev": 0.009146948712169298,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.004449537878787879,
      "avg_duration_per_task_ms": 380489.8497474748,
      "avg_output_tokens_per_task": 24420.810185185186,
      "last_run_timestamp": "2026-10-31T08:39:08.723Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Novita",
      "accuracy": 0.9027699539584503,
      "accuracy_std_dev": 0.006295075818094274,
      "run_count": 9,
      "total_questions": 1781,
      "avg_cost_per_task_usd": 0.028251830432341377,
      "avg_duration_per_task_ms": 172539.25996631107,
      "avg_output_tokens_per_task": 28841.043234138124,
      "last_run_timestamp": "2026-10-31T08:31:10.643Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "GMICloud",
      "accuracy": 0.9026157692307694,
      "accuracy_std_dev": 0.012508586614825496,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.022228931623931622,
      "avg_duration_per_task_ms": 201340.69813519812,
      "avg_output_tokens_per_task": 26432.28243978244,
      "last_run_timestamp": "2026-11-10T06:09:57.337Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.4-20260305",
      "model_slug": "openai/gpt-5.4",
      "model_name": "OpenAI: GPT-5.4",
      "provider_name": "auto-routing",
      "accuracy": 0.902357,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.1437863712121212,
      "avg_duration_per_task_ms": 149585.63636363635,
      "avg_output_tokens_per_task": 9456.126262626263,
      "last_run_timestamp": "2026-10-01T19:14:23.097Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "thinkingmachines/inkling-2-small-20261107",
      "model_slug": "thinkingmachines/inkling-2-small",
      "model_name": "Thinking Machines: Inkling 2 Small",
      "provider_name": "DeepInfra",
      "accuracy": 0.902357,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.027153818181818182,
      "avg_duration_per_task_ms": 429418.7121212121,
      "avg_output_tokens_per_task": 26931.89898989899,
      "last_run_timestamp": "2026-11-13T12:01:34.311Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta/muse-spark-open-20261015",
      "model_slug": "meta/muse-spark-open",
      "model_name": "Meta: Muse Spark Open",
      "provider_name": "Fireworks",
      "accuracy": 0.9023568,
      "accuracy_std_dev": 0.003838970664123373,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.0695745484848485,
      "avg_duration_per_task_ms": 344238.6484848485,
      "avg_output_tokens_per_task": 16158.315151515151,
      "last_run_timestamp": "2026-10-31T08:52:05.708Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-opus-5.5-20260921",
      "model_slug": "anthropic/claude-opus-5.5",
      "model_name": "Anthropic: Claude Opus 5.5",
      "provider_name": "auto-routing",
      "accuracy": 0.9020761666666667,
      "accuracy_std_dev": 0.005271695534223824,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.05407119865319866,
      "avg_duration_per_task_ms": 30482.292087542086,
      "avg_output_tokens_per_task": 2495.6936026936028,
      "last_run_timestamp": "2026-10-31T08:27:47.532Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Parasail",
      "accuracy": 0.9020202237762238,
      "accuracy_std_dev": 0.008897393508385223,
      "run_count": 12,
      "total_questions": 2145,
      "avg_cost_per_task_usd": 0.029685459207459214,
      "avg_duration_per_task_ms": 182816.28065268067,
      "avg_output_tokens_per_task": 23066.490442890445,
      "last_run_timestamp": "2026-11-06T06:01:51.918Z"
    },
    {
      "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": "PrimeIntellect",
      "accuracy": 0.9018877466467959,
      "accuracy_std_dev": 0.01830165892678099,
      "run_count": 6,
      "total_questions": 671,
      "avg_cost_per_task_usd": 0.0035507228017883754,
      "avg_duration_per_task_ms": 96273.04172876304,
      "avg_output_tokens_per_task": 15221.761549925484,
      "last_run_timestamp": "2026-11-02T17:22:52.620Z"
    },
    {
      "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": 0.9017445454545455,
      "accuracy_std_dev": 0.018612573415759153,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.03461236868686868,
      "avg_duration_per_task_ms": 310679.8080808081,
      "avg_output_tokens_per_task": 30520.936639118456,
      "last_run_timestamp": "2026-10-31T16:18:48.433Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Baidu",
      "accuracy": 0.9015653066016837,
      "accuracy_std_dev": 0.007303971162529,
      "run_count": 12,
      "total_questions": 2257,
      "avg_cost_per_task_usd": 0.013315422241914046,
      "avg_duration_per_task_ms": 151646.0602569783,
      "avg_output_tokens_per_task": 26073.0053167922,
      "last_run_timestamp": "2026-11-10T18:29:20.621Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta/muse-spark-open-20261015",
      "model_slug": "meta/muse-spark-open",
      "model_name": "Meta: Muse Spark Open",
      "provider_name": "auto-routing",
      "accuracy": 0.9015150000000001,
      "accuracy_std_dev": 0.004208999999999963,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.0736977297979798,
      "avg_duration_per_task_ms": 138589.0126262626,
      "avg_output_tokens_per_task": 17125.45707070707,
      "last_run_timestamp": "2026-10-31T08:09:26.224Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta/muse-spark-open-20261015",
      "model_slug": "meta/muse-spark-open",
      "model_name": "Meta: Muse Spark Open",
      "provider_name": "BaseTen",
      "accuracy": 0.901515,
      "accuracy_std_dev": 0.0025249999999999995,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.0720575808080808,
      "avg_duration_per_task_ms": 100990.45707070707,
      "avg_output_tokens_per_task": 16730.29797979798,
      "last_run_timestamp": "2026-10-31T08:17:04.260Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.4-20260305",
      "model_slug": "openai/gpt-5.4",
      "model_name": "OpenAI: GPT-5.4",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.901515,
      "accuracy_std_dev": 0.0025249999999999995,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.16319313383838385,
      "avg_duration_per_task_ms": 88415.74242424243,
      "avg_output_tokens_per_task": 9762.232323232323,
      "last_run_timestamp": "2026-10-01T19:12:25.243Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.8-opus-20260528",
      "model_slug": "anthropic/claude-opus-4.8",
      "model_name": "Anthropic: Claude Opus 4.8",
      "provider_name": "Azure",
      "accuracy": 0.9015149999999998,
      "accuracy_std_dev": 0.006314360868830512,
      "run_count": 22,
      "total_questions": 4356,
      "avg_cost_per_task_usd": 0.20703088957759413,
      "avg_duration_per_task_ms": 97686.47543617999,
      "avg_output_tokens_per_task": 7837.382690541782,
      "last_run_timestamp": "2026-10-31T08:46:46.997Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Novita",
      "accuracy": 0.9012853636363636,
      "accuracy_std_dev": 0.01164766215538338,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.03467139577594123,
      "avg_duration_per_task_ms": 250666.62396694216,
      "avg_output_tokens_per_task": 27670.98209366391,
      "last_run_timestamp": "2026-10-31T08:36:32.971Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "Xiaomi",
      "accuracy": 0.9012343333333332,
      "accuracy_std_dev": 0.0037221826780652403,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.03380553956228956,
      "avg_duration_per_task_ms": 826187.4343434343,
      "avg_output_tokens_per_task": 38504.42255892256,
      "last_run_timestamp": "2026-10-31T08:14:20.398Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-opus-5.5-20260921",
      "model_slug": "anthropic/claude-opus-5.5",
      "model_name": "Anthropic: Claude Opus 5.5",
      "provider_name": "Anthropic",
      "accuracy": 0.9011541428571429,
      "accuracy_std_dev": 0.004473489479416358,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.053932944444444436,
      "avg_duration_per_task_ms": 26662.532467532466,
      "avg_output_tokens_per_task": 2480.9805194805194,
      "last_run_timestamp": "2026-10-31T08:25:18.221Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Crusoe",
      "accuracy": 0.9010944999999999,
      "accuracy_std_dev": 0.0012624999999999997,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.00798360101010101,
      "avg_duration_per_task_ms": 317918.4797979798,
      "avg_output_tokens_per_task": 28057.032828282827,
      "last_run_timestamp": "2026-10-31T16:08:33.700Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "GMICloud",
      "accuracy": 0.900953777777778,
      "accuracy_std_dev": 0.018721426879248864,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.07315864141414141,
      "avg_duration_per_task_ms": 540347.3451178451,
      "avg_output_tokens_per_task": 36755.94332210999,
      "last_run_timestamp": "2026-11-10T07:51:59.421Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "auto-routing",
      "accuracy": 0.9006730000000001,
      "accuracy_std_dev": 0.006733999999999962,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.010972851010101009,
      "avg_duration_per_task_ms": 117479.04797979798,
      "avg_output_tokens_per_task": 19415.13888888889,
      "last_run_timestamp": "2026-10-31T08:11:46.712Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Cloudflare",
      "accuracy": 0.9005650372881355,
      "accuracy_std_dev": 0.0005644881654108368,
      "run_count": 3,
      "total_questions": 590,
      "avg_cost_per_task_usd": 0.013123520338983051,
      "avg_duration_per_task_ms": 181977.62711864407,
      "avg_output_tokens_per_task": 21777.491525423728,
      "last_run_timestamp": "2026-10-31T08:31:11.333Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "BaseTen",
      "accuracy": 0.9005043975240714,
      "accuracy_std_dev": 0.00897585252218416,
      "run_count": 4,
      "total_questions": 727,
      "avg_cost_per_task_usd": 0.027938022008253093,
      "avg_duration_per_task_ms": 73305.65749656121,
      "avg_output_tokens_per_task": 18908.090784044016,
      "last_run_timestamp": "2026-11-05T21:01:47.104Z"
    },
    {
      "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": 0.900418212121212,
      "accuracy_std_dev": 0.006022033356174897,
      "run_count": 33,
      "total_questions": 6534,
      "avg_cost_per_task_usd": 0.21616242209978576,
      "avg_duration_per_task_ms": 92587.19115396388,
      "avg_output_tokens_per_task": 7884.330272421182,
      "last_run_timestamp": "2026-10-31T08:23:44.152Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "reflection/beam-1.0-501b-a23b-20261016",
      "model_slug": "reflection/beam-1.0-501b-a23b",
      "model_name": "Reflection: Beam 1.0 501B A23B",
      "provider_name": "Nebius",
      "accuracy": 0.9002525,
      "accuracy_std_dev": 0.004014983343676526,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.014062290404040403,
      "avg_duration_per_task_ms": 87702.04797979798,
      "avg_output_tokens_per_task": 13895.555555555555,
      "last_run_timestamp": "2026-10-31T08:28:22.736Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "BaseTen",
      "accuracy": 0.9002466549491212,
      "accuracy_std_dev": 0.01549038271192753,
      "run_count": 11,
      "total_questions": 2162,
      "avg_cost_per_task_usd": 0.16790182839962997,
      "avg_duration_per_task_ms": 387266.13876040705,
      "avg_output_tokens_per_task": 42094.61147086031,
      "last_run_timestamp": "2026-10-31T08:09:30.320Z"
    },
    {
      "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": 0.9001631212121214,
      "accuracy_std_dev": 0.005552397340584662,
      "run_count": 33,
      "total_questions": 6534,
      "avg_cost_per_task_usd": 0.22268922956841136,
      "avg_duration_per_task_ms": 93998.23614937252,
      "avg_output_tokens_per_task": 8126.572849709213,
      "last_run_timestamp": "2026-10-31T08:31:12.840Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-opus-5.5-20260921",
      "model_slug": "anthropic/claude-opus-5.5",
      "model_name": "Anthropic: Claude Opus 5.5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.9000319047619048,
      "accuracy_std_dev": 0.005076517459315892,
      "run_count": 21,
      "total_questions": 4158,
      "avg_cost_per_task_usd": 0.05824434295334296,
      "avg_duration_per_task_ms": 35780.01707551708,
      "avg_output_tokens_per_task": 2517.5358345358345,
      "last_run_timestamp": "2026-10-31T09:01:04.740Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Wafer",
      "accuracy": 0.8999309479616755,
      "accuracy_std_dev": 0.123858969413793,
      "run_count": 27,
      "total_questions": 5323,
      "avg_cost_per_task_usd": 0.16280099473980839,
      "avg_duration_per_task_ms": 193274.2459139583,
      "avg_output_tokens_per_task": 10715.915085478115,
      "last_run_timestamp": "2026-10-31T09:24:16.791Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "Cohere",
      "accuracy": 0.8996625716694773,
      "accuracy_std_dev": 0.0020759481367959394,
      "run_count": 3,
      "total_questions": 593,
      "avg_cost_per_task_usd": 0.11324433389544686,
      "avg_duration_per_task_ms": 858665.3389544688,
      "avg_output_tokens_per_task": 45077.612141652615,
      "last_run_timestamp": "2026-10-31T08:28:18.045Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "DigitalOcean",
      "accuracy": 0.8995687234401349,
      "accuracy_std_dev": 0.007447984730328704,
      "run_count": 9,
      "total_questions": 1779,
      "avg_cost_per_task_usd": 0.0197627712197864,
      "avg_duration_per_task_ms": 382016.3063518831,
      "avg_output_tokens_per_task": 22859.64811691962,
      "last_run_timestamp": "2026-10-31T08:33:02.133Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.4-20260305",
      "model_slug": "openai/gpt-5.4",
      "model_name": "OpenAI: GPT-5.4",
      "provider_name": "Azure",
      "accuracy": 0.8995508333333332,
      "accuracy_std_dev": 0.0038470676025888335,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.15600541077441077,
      "avg_duration_per_task_ms": 138877.1792929293,
      "avg_output_tokens_per_task": 9623.925925925925,
      "last_run_timestamp": "2026-10-01T19:18:24.536Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "Novita",
      "accuracy": 0.8994372877637131,
      "accuracy_std_dev": 0.01977995405637288,
      "run_count": 6,
      "total_questions": 1185,
      "avg_cost_per_task_usd": 0.03544181097046414,
      "avg_duration_per_task_ms": 152645.21687763714,
      "avg_output_tokens_per_task": 26597.288607594935,
      "last_run_timestamp": "2026-10-03T09:28:39.969Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "NextBit",
      "accuracy": 0.8993638888888887,
      "accuracy_std_dev": 0.004029367042778795,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.030411617283950615,
      "avg_duration_per_task_ms": 237553.6509539843,
      "avg_output_tokens_per_task": 26633.1290684624,
      "last_run_timestamp": "2026-10-03T10:19:34.320Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-fable-5.1-20260831",
      "model_slug": "anthropic/claude-fable-5.1",
      "model_name": "Anthropic: Claude Fable 5.1",
      "provider_name": "auto-routing",
      "accuracy": 0.8993264,
      "accuracy_std_dev": 0.006511664598242117,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.13052546969696968,
      "avg_duration_per_task_ms": 30874.333333333332,
      "avg_output_tokens_per_task": 2403.130303030303,
      "last_run_timestamp": "2026-10-01T08:24:30.121Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-opus-5.5-20260921",
      "model_slug": "anthropic/claude-opus-5.5",
      "model_name": "Anthropic: Claude Opus 5.5",
      "provider_name": "Google",
      "accuracy": 0.8992302857142856,
      "accuracy_std_dev": 0.0048260090792136764,
      "run_count": 21,
      "total_questions": 4158,
      "avg_cost_per_task_usd": 0.05735562409812409,
      "avg_duration_per_task_ms": 25146.343674843676,
      "avg_output_tokens_per_task": 2479.3174603174602,
      "last_run_timestamp": "2026-10-31T09:06:56.339Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-opus-5.5-20260921",
      "model_slug": "anthropic/claude-opus-5.5",
      "model_name": "Anthropic: Claude Opus 5.5",
      "provider_name": "Azure",
      "accuracy": 0.8991100000000001,
      "accuracy_std_dev": 0.005600276243186554,
      "run_count": 14,
      "total_questions": 2772,
      "avg_cost_per_task_usd": 0.05768347799422799,
      "avg_duration_per_task_ms": 26053.75036075036,
      "avg_output_tokens_per_task": 2530.0901875901877,
      "last_run_timestamp": "2026-10-31T08:22:04.658Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-opus-5-20260723",
      "model_slug": "anthropic/claude-opus-5",
      "model_name": "Anthropic: Claude Opus 5",
      "provider_name": "Azure",
      "accuracy": 0.8989131363636366,
      "accuracy_std_dev": 0.008207886901663264,
      "run_count": 22,
      "total_questions": 4356,
      "avg_cost_per_task_usd": 0.1284000422405877,
      "avg_duration_per_task_ms": 58962.32116620753,
      "avg_output_tokens_per_task": 4769.593434343435,
      "last_run_timestamp": "2026-10-31T08:31:13.199Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "GMICloud",
      "accuracy": 0.8985305454545454,
      "accuracy_std_dev": 0.018596178744147863,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.023964589072543615,
      "avg_duration_per_task_ms": 168611.35950413224,
      "avg_output_tokens_per_task": 27930.285583103763,
      "last_run_timestamp": "2026-11-10T06:04:30.952Z"
    },
    {
      "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": 0.8981922631578948,
      "accuracy_std_dev": 0.013259923107827002,
      "run_count": 19,
      "total_questions": 3762,
      "avg_cost_per_task_usd": 0.007911754917597024,
      "avg_duration_per_task_ms": 156354.3657628921,
      "avg_output_tokens_per_task": 24268.396331738437,
      "last_run_timestamp": "2026-10-31T09:24:59.899Z"
    },
    {
      "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": 0.8981735454545454,
      "accuracy_std_dev": 0.007167908105686761,
      "run_count": 33,
      "total_questions": 6534,
      "avg_cost_per_task_usd": 0.2089634925007652,
      "avg_duration_per_task_ms": 71375.81910009183,
      "avg_output_tokens_per_task": 7612.6602387511475,
      "last_run_timestamp": "2026-10-31T08:51:35.523Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.8-opus-20260528",
      "model_slug": "anthropic/claude-opus-4.8",
      "model_name": "Anthropic: Claude Opus 4.8",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.8981480833333332,
      "accuracy_std_dev": 0.007606843601415013,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.19857798400673404,
      "avg_duration_per_task_ms": 95247.20033670033,
      "avg_output_tokens_per_task": 7731.943181818182,
      "last_run_timestamp": "2026-10-31T08:27:45.200Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "SiliconFlow",
      "accuracy": 0.898148,
      "accuracy_std_dev": 0.012012972737836369,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.03667170117845118,
      "avg_duration_per_task_ms": 157797.89814814815,
      "avg_output_tokens_per_task": 27519.53787878788,
      "last_run_timestamp": "2026-10-03T09:28:24.400Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8981478333333334,
      "accuracy_std_dev": 0.007386487717823356,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.12543263804713806,
      "avg_duration_per_task_ms": 965889.989057239,
      "avg_output_tokens_per_task": 49917.90404040404,
      "last_run_timestamp": "2026-10-31T08:45:01.505Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "auto-routing",
      "accuracy": 0.89814775,
      "accuracy_std_dev": 0.0055197238778311765,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.11915268055555557,
      "avg_duration_per_task_ms": 825703.6161616162,
      "avg_output_tokens_per_task": 47450.94823232323,
      "last_run_timestamp": "2026-10-31T08:07:17.850Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "GMICloud",
      "accuracy": 0.8980186153846152,
      "accuracy_std_dev": 0.03138066286630978,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.11786784654234653,
      "avg_duration_per_task_ms": 508070.64024864027,
      "avg_output_tokens_per_task": 36980.76418026418,
      "last_run_timestamp": "2026-11-10T08:27:28.930Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.7-opus-20260416",
      "model_slug": "anthropic/claude-opus-4.7",
      "model_name": "Anthropic: Claude Opus 4.7",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.8980075833333333,
      "accuracy_std_dev": 0.004784274352123693,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.23032522264309765,
      "avg_duration_per_task_ms": 110445.48653198653,
      "avg_output_tokens_per_task": 8998.862373737375,
      "last_run_timestamp": "2026-10-31T08:24:37.252Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Ionstream",
      "accuracy": 0.8979796,
      "accuracy_std_dev": 0.008689115757083655,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.014924614141414142,
      "avg_duration_per_task_ms": 134095.19090909092,
      "avg_output_tokens_per_task": 23160.81515151515,
      "last_run_timestamp": "2026-11-10T16:47:27.230Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "StreamLake",
      "accuracy": 0.8978673333333335,
      "accuracy_std_dev": 0.007570651065499923,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.03724634175084175,
      "avg_duration_per_task_ms": 745311.5892255892,
      "avg_output_tokens_per_task": 37370.69696969697,
      "last_run_timestamp": "2026-10-31T08:12:46.423Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-luna-pro-20260709",
      "model_slug": "openai/gpt-5.6-luna-pro",
      "model_name": "OpenAI: GPT-5.6 Luna Pro",
      "provider_name": "OpenAI",
      "accuracy": 0.8973063333333333,
      "accuracy_std_dev": 0.004955863351178606,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.03703221212121212,
      "avg_duration_per_task_ms": 137593.69696969696,
      "avg_output_tokens_per_task": 26012.949494949495,
      "last_run_timestamp": "2026-10-02T23:26:10.109Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Crusoe",
      "accuracy": 0.897306,
      "accuracy_std_dev": 0.004761657064510171,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.026144708754208754,
      "avg_duration_per_task_ms": 176560.02693602693,
      "avg_output_tokens_per_task": 21638.170033670034,
      "last_run_timestamp": "2026-10-31T08:57:38.005Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta/muse-spark-open-20261015",
      "model_slug": "meta/muse-spark-open",
      "model_name": "Meta: Muse Spark Open",
      "provider_name": "Together",
      "accuracy": 0.8973059999999999,
      "accuracy_std_dev": 0.0033669999999999534,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.07477493686868687,
      "avg_duration_per_task_ms": 142684.398989899,
      "avg_output_tokens_per_task": 17437.89393939394,
      "last_run_timestamp": "2026-10-31T08:49:18.270Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.8971531818181817,
      "accuracy_std_dev": 0.011450877590966798,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.20187068044077136,
      "avg_duration_per_task_ms": 96528.3347107438,
      "avg_output_tokens_per_task": 7855.07943067034,
      "last_run_timestamp": "2026-10-31T08:32:06.814Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta/muse-spark-open-20261015",
      "model_slug": "meta/muse-spark-open",
      "model_name": "Meta: Muse Spark Open",
      "provider_name": "Nebius",
      "accuracy": 0.8970256666666665,
      "accuracy_std_dev": 0.00667522615513675,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.06776180639730639,
      "avg_duration_per_task_ms": 222686.2255892256,
      "avg_output_tokens_per_task": 15778.796296296296,
      "last_run_timestamp": "2026-10-31T08:27:47.368Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8968690856353589,
      "accuracy_std_dev": 0.010670089513748447,
      "run_count": 11,
      "total_questions": 2172,
      "avg_cost_per_task_usd": 0.018374484806629836,
      "avg_duration_per_task_ms": 362424.26427255984,
      "avg_output_tokens_per_task": 28910.183241252304,
      "last_run_timestamp": "2026-10-31T09:16:30.257Z"
    },
    {
      "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": 0.8967014333929941,
      "accuracy_std_dev": 0.03454965314283198,
      "run_count": 20,
      "total_questions": 3911,
      "avg_cost_per_task_usd": 0.37683498056762965,
      "avg_duration_per_task_ms": 497556.26566095627,
      "avg_output_tokens_per_task": 57557.85144464331,
      "last_run_timestamp": "2026-10-31T16:18:08.626Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.8-opus-20260528",
      "model_slug": "anthropic/claude-opus-4.8",
      "model_name": "Anthropic: Claude Opus 4.8",
      "provider_name": "Anthropic",
      "accuracy": 0.896694090909091,
      "accuracy_std_dev": 0.006667655625269636,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.208940303030303,
      "avg_duration_per_task_ms": 100740.17171717172,
      "avg_output_tokens_per_task": 8143.670798898072,
      "last_run_timestamp": "2026-10-31T08:27:47.283Z"
    },
    {
      "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": 0.8963318181818182,
      "accuracy_std_dev": 0.010896439994787305,
      "run_count": 5,
      "total_questions": 836,
      "avg_cost_per_task_usd": 0.0957473456937799,
      "avg_duration_per_task_ms": 189939.80023923446,
      "avg_output_tokens_per_task": 25436.38875598086,
      "last_run_timestamp": "2026-09-26T09:19:13.012Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Inceptron",
      "accuracy": 0.8962463164155431,
      "accuracy_std_dev": 0.019589876018192762,
      "run_count": 7,
      "total_questions": 1261,
      "avg_cost_per_task_usd": 0.15202800396510704,
      "avg_duration_per_task_ms": 619590.4591593973,
      "avg_output_tokens_per_task": 49768.916732751786,
      "last_run_timestamp": "2026-10-31T16:34:03.928Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-opus-5-20260723",
      "model_slug": "anthropic/claude-opus-5",
      "model_name": "Anthropic: Claude Opus 5",
      "provider_name": "Anthropic",
      "accuracy": 0.8962348181818182,
      "accuracy_std_dev": 0.007295060767622528,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.12272349724517907,
      "avg_duration_per_task_ms": 58904.73461891644,
      "avg_output_tokens_per_task": 4673.36593204775,
      "last_run_timestamp": "2026-10-31T08:31:11.765Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Google",
      "accuracy": 0.896184,
      "accuracy_std_dev": 0.007570749896806769,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0237254696969697,
      "avg_duration_per_task_ms": 319850.3367003367,
      "avg_output_tokens_per_task": 19569.10606060606,
      "last_run_timestamp": "2026-10-31T08:20:35.720Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8961838333333332,
      "accuracy_std_dev": 0.011153184947458816,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.1947589175084175,
      "avg_duration_per_task_ms": 1704111.0387205386,
      "avg_output_tokens_per_task": 57112.030303030304,
      "last_run_timestamp": "2026-10-24T16:20:08.208Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-opus-5.5-20260921",
      "model_slug": "anthropic/claude-opus-5.5",
      "model_name": "Anthropic: Claude Opus 5.5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.8961037142857144,
      "accuracy_std_dev": 0.004737329618927418,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.054530834776334774,
      "avg_duration_per_task_ms": 30476.945887445887,
      "avg_output_tokens_per_task": 2521.178932178932,
      "last_run_timestamp": "2026-10-31T08:40:20.874Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-luna-pro-20260709",
      "model_slug": "openai/gpt-5.6-luna-pro",
      "model_name": "OpenAI: GPT-5.6 Luna Pro",
      "provider_name": "auto-routing",
      "accuracy": 0.895623,
      "accuracy_std_dev": 0.0033669999999999534,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.03841230555555556,
      "avg_duration_per_task_ms": 144059.0782828283,
      "avg_output_tokens_per_task": 27039.489898989897,
      "last_run_timestamp": "2026-10-01T16:12:34.647Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "GMICloud",
      "accuracy": 0.8956226666666667,
      "accuracy_std_dev": 0.0059917140749167225,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.11581659259259257,
      "avg_duration_per_task_ms": 1591244.2491582492,
      "avg_output_tokens_per_task": 46146.09090909091,
      "last_run_timestamp": "2026-11-10T07:04:59.972Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "AtlasCloud",
      "accuracy": 0.895412375,
      "accuracy_std_dev": 0.020051295631065218,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.02414115151515151,
      "avg_duration_per_task_ms": 210274.94823232322,
      "avg_output_tokens_per_task": 27810.135101010103,
      "last_run_timestamp": "2026-10-31T08:27:49.932Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Phala",
      "accuracy": 0.8952905937499999,
      "accuracy_std_dev": 0.016969367452854572,
      "run_count": 8,
      "total_questions": 1536,
      "avg_cost_per_task_usd": 0.032397834635416664,
      "avg_duration_per_task_ms": 236218.794921875,
      "avg_output_tokens_per_task": 32195.716796875,
      "last_run_timestamp": "2026-10-31T09:17:17.877Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.7-opus-20260416",
      "model_slug": "anthropic/claude-opus-4.7",
      "model_name": "Anthropic: Claude Opus 4.7",
      "provider_name": "auto-routing",
      "accuracy": 0.8951635454545456,
      "accuracy_std_dev": 0.01007786883462638,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.2307558640955005,
      "avg_duration_per_task_ms": 109981.68457300276,
      "avg_output_tokens_per_task": 9006.644168962352,
      "last_run_timestamp": "2026-10-31T08:13:10.867Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Cohere",
      "accuracy": 0.8951417142857143,
      "accuracy_std_dev": 0.009679824699036726,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.008688588023088023,
      "avg_duration_per_task_ms": 171107.75108225108,
      "avg_output_tokens_per_task": 30552.396103896102,
      "last_run_timestamp": "2026-10-31T09:09:17.234Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Novita",
      "accuracy": 0.8951417142857143,
      "accuracy_std_dev": 0.01785557157315559,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.2036656197691198,
      "avg_duration_per_task_ms": 957488.6248196248,
      "avg_output_tokens_per_task": 59748.26406926407,
      "last_run_timestamp": "2026-10-31T16:24:42.659Z"
    },
    {
      "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": 0.8950656277001271,
      "accuracy_std_dev": 0.016150328135211358,
      "run_count": 16,
      "total_questions": 3148,
      "avg_cost_per_task_usd": 0.007430722681067345,
      "avg_duration_per_task_ms": 254760.92153748413,
      "avg_output_tokens_per_task": 28431.650254129607,
      "last_run_timestamp": "2026-11-02T17:25:00.911Z"
    },
    {
      "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": 0.894573530232558,
      "accuracy_std_dev": 0.042752448470087655,
      "run_count": 12,
      "total_questions": 2365,
      "avg_cost_per_task_usd": 0.07040914122621564,
      "avg_duration_per_task_ms": 595191.9403805496,
      "avg_output_tokens_per_task": 39081.489217758986,
      "last_run_timestamp": "2026-10-31T16:51:33.999Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "ModelRun",
      "accuracy": 0.8943087542213883,
      "accuracy_std_dev": 0.03942204850926279,
      "run_count": 3,
      "total_questions": 533,
      "avg_cost_per_task_usd": 0.019450917448405252,
      "avg_duration_per_task_ms": 177927.2420262664,
      "avg_output_tokens_per_task": 28174.943714821762,
      "last_run_timestamp": "2026-10-28T05:00:14.622Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Sail Research",
      "accuracy": 0.8942835463539759,
      "accuracy_std_dev": 0.01749872529716121,
      "run_count": 14,
      "total_questions": 2729,
      "avg_cost_per_task_usd": 0.1169677999267131,
      "avg_duration_per_task_ms": 555475.2238915353,
      "avg_output_tokens_per_task": 43652.794796628805,
      "last_run_timestamp": "2026-11-06T22:44:36.218Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "Tencent",
      "accuracy": 0.894276,
      "accuracy_std_dev": 0.008869991905295046,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.11497575656565656,
      "avg_duration_per_task_ms": 654340.1272727273,
      "avg_output_tokens_per_task": 45820.55656565657,
      "last_run_timestamp": "2026-10-31T08:45:35.702Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Morph",
      "accuracy": 0.8941497500000002,
      "accuracy_std_dev": 0.012326495474687834,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.010706986111111112,
      "avg_duration_per_task_ms": 265829.2607323232,
      "avg_output_tokens_per_task": 20748.16351010101,
      "last_run_timestamp": "2026-10-31T08:33:46.211Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "reflection/beam-1.0-501b-a23b-20261016",
      "model_slug": "reflection/beam-1.0-501b-a23b",
      "model_name": "Reflection: Beam 1.0 501B A23B",
      "provider_name": "CoreWeave",
      "accuracy": 0.893939358490566,
      "accuracy_std_dev": 0.004245539018258487,
      "run_count": 3,
      "total_questions": 583,
      "avg_cost_per_task_usd": 0.023210332761578047,
      "avg_duration_per_task_ms": 116069.80960548884,
      "avg_output_tokens_per_task": 13628.193825042881,
      "last_run_timestamp": "2026-10-31T08:17:29.099Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Phala",
      "accuracy": 0.8939393333333333,
      "accuracy_std_dev": 0.012217692780917704,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.02448665488215488,
      "avg_duration_per_task_ms": 139886.7744107744,
      "avg_output_tokens_per_task": 20249.15824915825,
      "last_run_timestamp": "2026-10-31T09:02:19.878Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Moonshot AI",
      "accuracy": 0.8936585000000001,
      "accuracy_std_dev": 0.017069648744189203,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.23770161616161617,
      "avg_duration_per_task_ms": 925404.0513468013,
      "avg_output_tokens_per_task": 59269.02946127946,
      "last_run_timestamp": "2026-10-24T16:07:10.471Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "Makora",
      "accuracy": 0.8934709999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 194,
      "avg_cost_per_task_usd": 0.11686864948453608,
      "avg_duration_per_task_ms": 605687.7474226804,
      "avg_output_tokens_per_task": 46496.701030927834,
      "last_run_timestamp": "2026-10-13T03:00:07.016Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.8934035909090907,
      "accuracy_std_dev": 0.009447194358206847,
      "run_count": 22,
      "total_questions": 4356,
      "avg_cost_per_task_usd": 0.24392887970615243,
      "avg_duration_per_task_ms": 85202.9074839302,
      "avg_output_tokens_per_task": 9060.956611570247,
      "last_run_timestamp": "2026-10-31T08:12:47.061Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "reflection/beam-1.0-501b-a23b-20261016",
      "model_slug": "reflection/beam-1.0-501b-a23b",
      "model_name": "Reflection: Beam 1.0 501B A23B",
      "provider_name": "BaseTen",
      "accuracy": 0.8933780000000001,
      "accuracy_std_dev": 0.0028613344206273234,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.03135734848484849,
      "avg_duration_per_task_ms": 31339.205387205388,
      "avg_output_tokens_per_task": 13928.784511784512,
      "last_run_timestamp": "2026-10-31T08:28:17.755Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "reflection/beam-1.0-501b-a23b-20261016",
      "model_slug": "reflection/beam-1.0-501b-a23b",
      "model_name": "Reflection: Beam 1.0 501B A23B",
      "provider_name": "DigitalOcean",
      "accuracy": 0.8933779999999999,
      "accuracy_std_dev": 0.0034590473640392213,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.01648433501683502,
      "avg_duration_per_task_ms": 93266.93602693603,
      "avg_output_tokens_per_task": 13659.838383838383,
      "last_run_timestamp": "2026-10-31T08:40:23.037Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Parasail",
      "accuracy": 0.8931983140809443,
      "accuracy_std_dev": 0.01619141679512122,
      "run_count": 12,
      "total_questions": 2372,
      "avg_cost_per_task_usd": 0.1531594173693086,
      "avg_duration_per_task_ms": 667713.1500843171,
      "avg_output_tokens_per_task": 37442.99325463744,
      "last_run_timestamp": "2026-11-05T00:00:09.925Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.8931621743589744,
      "accuracy_std_dev": 0.0023230000000000195,
      "run_count": 2,
      "total_questions": 390,
      "avg_cost_per_task_usd": 0.020242015384615383,
      "avg_duration_per_task_ms": 141496.24102564104,
      "avg_output_tokens_per_task": 40323.36666666667,
      "last_run_timestamp": "2026-10-31T08:52:13.905Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "DigitalOcean",
      "accuracy": 0.8930975833333333,
      "accuracy_std_dev": 0.02298558731559965,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.08767985311447812,
      "avg_duration_per_task_ms": 506727.319023569,
      "avg_output_tokens_per_task": 39812.03493265993,
      "last_run_timestamp": "2026-10-31T16:16:10.979Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-terra-20260709",
      "model_slug": "openai/gpt-5.6-terra",
      "model_name": "OpenAI: GPT-5.6 Terra",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.8930975,
      "accuracy_std_dev": 0.010942499999999966,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.04334855050505051,
      "avg_duration_per_task_ms": 40520.42929292929,
      "avg_output_tokens_per_task": 3151.881313131313,
      "last_run_timestamp": "2026-10-01T19:11:43.444Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8930975,
      "accuracy_std_dev": 0.005892499999999967,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.11303334343434343,
      "avg_duration_per_task_ms": 814053.6843434343,
      "avg_output_tokens_per_task": 45048.43686868687,
      "last_run_timestamp": "2026-10-31T08:27:45.248Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Wafer",
      "accuracy": 0.8928929372738239,
      "accuracy_std_dev": 0.015585617718297187,
      "run_count": 17,
      "total_questions": 3316,
      "avg_cost_per_task_usd": 0.014895325693606755,
      "avg_duration_per_task_ms": 187175.30186972255,
      "avg_output_tokens_per_task": 21447.96381182147,
      "last_run_timestamp": "2026-10-31T08:57:07.688Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Cloudflare",
      "accuracy": 0.892885651478817,
      "accuracy_std_dev": 0.023028232143902604,
      "run_count": 7,
      "total_questions": 1251,
      "avg_cost_per_task_usd": 0.21098625099920065,
      "avg_duration_per_task_ms": 1517055.7913669064,
      "avg_output_tokens_per_task": 52600.646682653874,
      "last_run_timestamp": "2026-10-31T16:04:29.120Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Fireworks",
      "accuracy": 0.8928570861678005,
      "accuracy_std_dev": 0.015340394005370288,
      "run_count": 5,
      "total_questions": 882,
      "avg_cost_per_task_usd": 0.12683532766439912,
      "avg_duration_per_task_ms": 285532.1031746032,
      "avg_output_tokens_per_task": 28366.77097505669,
      "last_run_timestamp": "2026-09-19T09:14:09.704Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Parasail",
      "accuracy": 0.8928297457392905,
      "accuracy_std_dev": 0.010017422802463,
      "run_count": 11,
      "total_questions": 2171,
      "avg_cost_per_task_usd": 0.008073876554583142,
      "avg_duration_per_task_ms": 317611.8940580378,
      "avg_output_tokens_per_task": 28391.647167204053,
      "last_run_timestamp": "2026-10-31T08:16:11.302Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Novita",
      "accuracy": 0.892817,
      "accuracy_std_dev": 0.00915257617650169,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.007166784511784513,
      "avg_duration_per_task_ms": 121588.36531986532,
      "avg_output_tokens_per_task": 11764.878787878788,
      "last_run_timestamp": "2026-10-31T08:24:44.390Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "Novita",
      "accuracy": 0.892817,
      "accuracy_std_dev": 0.0067805750493597225,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.11382627104377104,
      "avg_duration_per_task_ms": 836209.1372053872,
      "avg_output_tokens_per_task": 45359.85101010101,
      "last_run_timestamp": "2026-10-31T09:36:23.668Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "InferenceNet",
      "accuracy": 0.8926684364569961,
      "accuracy_std_dev": 0.021409854620919034,
      "run_count": 12,
      "total_questions": 2337,
      "avg_cost_per_task_usd": 0.01926024176294395,
      "avg_duration_per_task_ms": 249786.02053915276,
      "avg_output_tokens_per_task": 30225.52802738554,
      "last_run_timestamp": "2026-11-04T21:06:11.704Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3-flash-preview-20251217",
      "model_slug": "google/gemini-3-flash-preview",
      "model_name": "Google: Gemini 3 Flash Preview",
      "provider_name": "Google AI Studio",
      "accuracy": 0.8922560000000002,
      "accuracy_std_dev": 0.010100999999999971,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.14220060606060608,
      "avg_duration_per_task_ms": 188559.2904040404,
      "avg_output_tokens_per_task": 47276.156565656565,
      "last_run_timestamp": "2026-10-01T19:12:30.168Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-terra-20260709",
      "model_slug": "openai/gpt-5.6-terra",
      "model_name": "OpenAI: GPT-5.6 Terra",
      "provider_name": "Azure",
      "accuracy": 0.8922558333333332,
      "accuracy_std_dev": 0.010423127464388435,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.04560339057239057,
      "avg_duration_per_task_ms": 43217.19528619529,
      "avg_output_tokens_per_task": 3431.844276094276,
      "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": "Anthropic",
      "accuracy": 0.8919497272727274,
      "accuracy_std_dev": 0.012304710177894609,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.20576803719008263,
      "avg_duration_per_task_ms": 98251.07162534435,
      "avg_output_tokens_per_task": 8002.224517906336,
      "last_run_timestamp": "2026-10-31T08:39:17.193Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "OpenInference",
      "accuracy": 0.8916946666666665,
      "accuracy_std_dev": 0.009944030347678707,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.011122356902356903,
      "avg_duration_per_task_ms": 302424.50673400675,
      "avg_output_tokens_per_task": 25444.905723905726,
      "last_run_timestamp": "2026-10-31T09:01:06.311Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8913374545454544,
      "accuracy_std_dev": 0.018875016904814263,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.0336423507805326,
      "avg_duration_per_task_ms": 191376.6694214876,
      "avg_output_tokens_per_task": 25231.550505050505,
      "last_run_timestamp": "2026-10-31T08:53:51.020Z"
    },
    {
      "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": 0.8912864848484849,
      "accuracy_std_dev": 0.009755318270195008,
      "run_count": 33,
      "total_questions": 6534,
      "avg_cost_per_task_usd": 0.13280288047138045,
      "avg_duration_per_task_ms": 59185.807315580045,
      "avg_output_tokens_per_task": 4745.987909397,
      "last_run_timestamp": "2026-10-31T08:46:09.151Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.8910532857142858,
      "accuracy_std_dev": 0.008331061289180467,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.02299151948051948,
      "avg_duration_per_task_ms": 246585.4624819625,
      "avg_output_tokens_per_task": 30738.57647907648,
      "last_run_timestamp": "2026-10-31T08:09:30.604Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.7-opus-20260416",
      "model_slug": "anthropic/claude-opus-4.7",
      "model_name": "Anthropic: Claude Opus 4.7",
      "provider_name": "Azure",
      "accuracy": 0.8904192727272727,
      "accuracy_std_dev": 0.008657537410634094,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.205998660697888,
      "avg_duration_per_task_ms": 99123.53581267217,
      "avg_output_tokens_per_task": 8011.966483011937,
      "last_run_timestamp": "2026-10-31T09:23:04.665Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-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": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "DeepInfra",
      "accuracy": 0.889876224111283,
      "accuracy_std_dev": 0.005998909101508626,
      "run_count": 7,
      "total_questions": 1294,
      "avg_cost_per_task_usd": 0.030587522411128287,
      "avg_duration_per_task_ms": 772887.7256568779,
      "avg_output_tokens_per_task": 36034.0115919629,
      "last_run_timestamp": "2026-10-31T08:09:29.858Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Ionstream",
      "accuracy": 0.8895201249999999,
      "accuracy_std_dev": 0.013120057749849085,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.10712794823232324,
      "avg_duration_per_task_ms": 473198.4646464646,
      "avg_output_tokens_per_task": 37189.18434343435,
      "last_run_timestamp": "2026-10-31T08:45:16.520Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-fable-5.1-20260831",
      "model_slug": "anthropic/claude-fable-5.1",
      "model_name": "Anthropic: Claude Fable 5.1",
      "provider_name": "Anthropic",
      "accuracy": 0.8894498333333334,
      "accuracy_std_dev": 0.02853141346198763,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.136280962962963,
      "avg_duration_per_task_ms": 40331.19781144781,
      "avg_output_tokens_per_task": 2516.2390572390573,
      "last_run_timestamp": "2026-10-01T08:24:17.840Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-luna-pro-20260922",
      "model_slug": "openai/gpt-6-luna-pro",
      "model_name": "OpenAI: GPT-6 Luna Pro",
      "provider_name": "Azure",
      "accuracy": 0.8894498333333334,
      "accuracy_std_dev": 0.009896244159893304,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.008765856902356902,
      "avg_duration_per_task_ms": 59664.589225589225,
      "avg_output_tokens_per_task": 12901.691919191919,
      "last_run_timestamp": "2026-10-01T08:29:53.989Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Cloudflare",
      "accuracy": 0.8893538140969164,
      "accuracy_std_dev": 0.014720288133832686,
      "run_count": 12,
      "total_questions": 2270,
      "avg_cost_per_task_usd": 0.037593757268722466,
      "avg_duration_per_task_ms": 339622.79823788546,
      "avg_output_tokens_per_task": 28197.38370044053,
      "last_run_timestamp": "2026-10-31T08:10:35.431Z"
    },
    {
      "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": 0.8893479090909092,
      "accuracy_std_dev": 0.0107458728521871,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.029853909090909094,
      "avg_duration_per_task_ms": 337392.91643709823,
      "avg_output_tokens_per_task": 27657.61432506887,
      "last_run_timestamp": "2026-10-31T09:11:16.308Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Reka",
      "accuracy": 0.8891708512988952,
      "accuracy_std_dev": 0.008107170901595804,
      "run_count": 17,
      "total_questions": 3349,
      "avg_cost_per_task_usd": 0.015164630934607346,
      "avg_duration_per_task_ms": 200146.0895789788,
      "avg_output_tokens_per_task": 27111.550313526426,
      "last_run_timestamp": "2026-10-31T09:08:32.372Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "GMICloud",
      "accuracy": 0.8890757777777778,
      "accuracy_std_dev": 0.019894746820749636,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.0056425325476992146,
      "avg_duration_per_task_ms": 237464.28787878787,
      "avg_output_tokens_per_task": 30270.649831649833,
      "last_run_timestamp": "2026-11-10T06:21:53.777Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Novita",
      "accuracy": 0.889041818181818,
      "accuracy_std_dev": 0.012149455384862324,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.11870665610651976,
      "avg_duration_per_task_ms": 546946.9178145088,
      "avg_output_tokens_per_task": 30914.909090909092,
      "last_run_timestamp": "2026-10-31T16:23:41.134Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-terra-20260709",
      "model_slug": "openai/gpt-5.6-terra",
      "model_name": "OpenAI: GPT-5.6 Terra",
      "provider_name": "OpenAI",
      "accuracy": 0.8888890000000002,
      "accuracy_std_dev": 0.002749143987983661,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0431288872053872,
      "avg_duration_per_task_ms": 52519.67676767677,
      "avg_output_tokens_per_task": 3463.8030303030305,
      "last_run_timestamp": "2026-10-02T23:21:06.556Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "amazon/nova-micro-v1",
      "model_slug": "amazon/nova-micro-v1",
      "model_name": "Amazon: Nova Micro 1.0",
      "provider_name": "Anthropic",
      "accuracy": 0.888889,
      "accuracy_std_dev": 0.006733999999999962,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.1979060606060606,
      "avg_duration_per_task_ms": 95702.44696969698,
      "avg_output_tokens_per_task": 7691.454545454545,
      "last_run_timestamp": "2026-10-01T19:11:18.427Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Nebius",
      "accuracy": 0.888889,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.13065590404040406,
      "avg_duration_per_task_ms": 663129.7929292929,
      "avg_output_tokens_per_task": 36983.707070707074,
      "last_run_timestamp": "2026-11-03T21:00:06.247Z"
    },
    {
      "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": 0.8887622653758542,
      "accuracy_std_dev": 0.006290540123073671,
      "run_count": 5,
      "total_questions": 878,
      "avg_cost_per_task_usd": 0.01762260478359909,
      "avg_duration_per_task_ms": 219383.70842824603,
      "avg_output_tokens_per_task": 26459.921412300682,
      "last_run_timestamp": "2026-09-19T08:54:58.652Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Alibaba",
      "accuracy": 0.8886065933593749,
      "accuracy_std_dev": 0.01380537387006623,
      "run_count": 13,
      "total_questions": 2560,
      "avg_cost_per_task_usd": 0.10022955078125,
      "avg_duration_per_task_ms": 460846.6203125,
      "avg_output_tokens_per_task": 34095.671484375,
      "last_run_timestamp": "2026-10-31T16:08:08.538Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Mancer 2",
      "accuracy": 0.8882767272727271,
      "accuracy_std_dev": 0.018777918676858473,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.013537202938475667,
      "avg_duration_per_task_ms": 349207.04545454547,
      "avg_output_tokens_per_task": 23738.730945821855,
      "last_run_timestamp": "2026-10-31T08:24:31.558Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Crusoe",
      "accuracy": 0.8881691813471503,
      "accuracy_std_dev": 0.0030120000000000147,
      "run_count": 2,
      "total_questions": 386,
      "avg_cost_per_task_usd": 0.1504010103626943,
      "avg_duration_per_task_ms": 651195.4274611399,
      "avg_output_tokens_per_task": 42839.41709844559,
      "last_run_timestamp": "2026-10-31T16:08:08.539Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8881671428571429,
      "accuracy_std_dev": 0.01839453545740899,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.008335702741702741,
      "avg_duration_per_task_ms": 417723.5966810967,
      "avg_output_tokens_per_task": 29344.838383838385,
      "last_run_timestamp": "2026-10-31T16:48:29.809Z"
    },
    {
      "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": 0.8880907953321364,
      "accuracy_std_dev": 0.02107123389837437,
      "run_count": 3,
      "total_questions": 557,
      "avg_cost_per_task_usd": 0.004236710951526032,
      "avg_duration_per_task_ms": 379493.5960502693,
      "avg_output_tokens_per_task": 16317.405745062837,
      "last_run_timestamp": "2026-09-05T08:24:32.148Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "auto-routing",
      "accuracy": 0.8876718896377422,
      "accuracy_std_dev": 0.0038308239938977078,
      "run_count": 6,
      "total_questions": 1187,
      "avg_cost_per_task_usd": 0.032938989890480205,
      "avg_duration_per_task_ms": 750389.2257792755,
      "avg_output_tokens_per_task": 38026.47177759057,
      "last_run_timestamp": "2026-10-31T08:03:09.435Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Nebius",
      "accuracy": 0.8875500348763478,
      "accuracy_std_dev": 0.013802209238356553,
      "run_count": 8,
      "total_questions": 1577,
      "avg_cost_per_task_usd": 0.1294895707038681,
      "avg_duration_per_task_ms": 1720933.2999365884,
      "avg_output_tokens_per_task": 32195.95688015219,
      "last_run_timestamp": "2026-10-31T16:35:08.639Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.8875475545108005,
      "accuracy_std_dev": 0.011882850320945717,
      "run_count": 20,
      "total_questions": 3935,
      "avg_cost_per_task_usd": 0.192012674205845,
      "avg_duration_per_task_ms": 351514.216772554,
      "avg_output_tokens_per_task": 31670.305209656926,
      "last_run_timestamp": "2026-10-24T16:13:46.475Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "NextBit",
      "accuracy": 0.887537924012158,
      "accuracy_std_dev": 0.020250680116973837,
      "run_count": 5,
      "total_questions": 987,
      "avg_cost_per_task_usd": 0.12033997061803446,
      "avg_duration_per_task_ms": 580755.4113475177,
      "avg_output_tokens_per_task": 34357.5663627153,
      "last_run_timestamp": "2026-10-03T16:44:56.199Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "DigitalOcean",
      "accuracy": 0.8874829878542512,
      "accuracy_std_dev": 0.0253227022698781,
      "run_count": 15,
      "total_questions": 2964,
      "avg_cost_per_task_usd": 0.14755026484480432,
      "avg_duration_per_task_ms": 305870.1987179487,
      "avg_output_tokens_per_task": 10905.781039136302,
      "last_run_timestamp": "2026-10-31T09:14:35.378Z"
    },
    {
      "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": "River",
      "accuracy": 0.8872055,
      "accuracy_std_dev": 0.011784499999999976,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.016706742424242426,
      "avg_duration_per_task_ms": 144417.60353535353,
      "avg_output_tokens_per_task": 34626.61363636364,
      "last_run_timestamp": "2026-11-12T20:24:33.604Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Phala",
      "accuracy": 0.8872054285714285,
      "accuracy_std_dev": 0.032307641642970675,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.26962912914862913,
      "avg_duration_per_task_ms": 1116880.2828282828,
      "avg_output_tokens_per_task": 58910.568542568544,
      "last_run_timestamp": "2026-10-31T16:10:09.220Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-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-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Ambient",
      "accuracy": 0.8872049999999999,
      "accuracy_std_dev": 0.006734000000000018,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.004174421717171718,
      "avg_duration_per_task_ms": 214515.8207070707,
      "avg_output_tokens_per_task": 22838.073232323233,
      "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-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Relace",
      "accuracy": 0.8871015424486148,
      "accuracy_std_dev": 0.007433711649565692,
      "run_count": 6,
      "total_questions": 1119,
      "avg_cost_per_task_usd": 0.100128890974084,
      "avg_duration_per_task_ms": 379563.4110813226,
      "avg_output_tokens_per_task": 36832.25379803396,
      "last_run_timestamp": "2026-10-31T16:50:08.743Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Crusoe",
      "accuracy": 0.8870104150464918,
      "accuracy_std_dev": 0.040746454673654994,
      "run_count": 6,
      "total_questions": 1183,
      "avg_cost_per_task_usd": 0.19607871766694843,
      "avg_duration_per_task_ms": 1537768.7954353339,
      "avg_output_tokens_per_task": 56311.583262890956,
      "last_run_timestamp": "2026-10-31T16:04:29.123Z"
    },
    {
      "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": 0.8869948749999998,
      "accuracy_std_dev": 0.02052275922261364,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.1362751912878788,
      "avg_duration_per_task_ms": 824252.7323232323,
      "avg_output_tokens_per_task": 63166.344696969696,
      "last_run_timestamp": "2026-11-10T18:52:31.004Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Fireworks",
      "accuracy": 0.8867051248554914,
      "accuracy_std_dev": 0.021208571037200977,
      "run_count": 5,
      "total_questions": 865,
      "avg_cost_per_task_usd": 0.16137210289017342,
      "avg_duration_per_task_ms": 430828.78150289017,
      "avg_output_tokens_per_task": 40500.787283237,
      "last_run_timestamp": "2026-09-19T08:53:57.162Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Cloudflare",
      "accuracy": 0.8866902897526502,
      "accuracy_std_dev": 0.024113592079662773,
      "run_count": 8,
      "total_questions": 1415,
      "avg_cost_per_task_usd": 0.035600583745583045,
      "avg_duration_per_task_ms": 349687.4614840989,
      "avg_output_tokens_per_task": 26740.03038869258,
      "last_run_timestamp": "2026-10-31T16:15:26.636Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Nebius",
      "accuracy": 0.8865318000000001,
      "accuracy_std_dev": 0.007110736653821441,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.030907244444444443,
      "avg_duration_per_task_ms": 199105.52626262626,
      "avg_output_tokens_per_task": 25534.58484848485,
      "last_run_timestamp": "2026-10-31T08:28:17.984Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "amazon/nova-micro-v1",
      "model_slug": "amazon/nova-micro-v1",
      "model_name": "Amazon: Nova Micro 1.0",
      "provider_name": "auto-routing",
      "accuracy": 0.8863635000000001,
      "accuracy_std_dev": 0.002525500000000014,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.22157512626262627,
      "avg_duration_per_task_ms": 106335.13383838384,
      "avg_output_tokens_per_task": 8638.217171717171,
      "last_run_timestamp": "2026-10-01T19:10:42.213Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "BaseTen",
      "accuracy": 0.886083,
      "accuracy_std_dev": 0.012320026406898109,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.022550329966329965,
      "avg_duration_per_task_ms": 149365.32154882155,
      "avg_output_tokens_per_task": 44803.96127946128,
      "last_run_timestamp": "2026-10-31T08:13:10.777Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Sail Research",
      "accuracy": 0.8856471629778674,
      "accuracy_std_dev": 0.009013068317726206,
      "run_count": 8,
      "total_questions": 1491,
      "avg_cost_per_task_usd": 0.010023977867203217,
      "avg_duration_per_task_ms": 270739.96177062375,
      "avg_output_tokens_per_task": 28220.264922870556,
      "last_run_timestamp": "2026-10-24T09:01:43.485Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-luna-20260709",
      "model_slug": "openai/gpt-5.6-luna",
      "model_name": "OpenAI: GPT-5.6 Luna",
      "provider_name": "Azure",
      "accuracy": 0.8856098161888702,
      "accuracy_std_dev": 0.01054693657998484,
      "run_count": 6,
      "total_questions": 1186,
      "avg_cost_per_task_usd": 0.010296545531197302,
      "avg_duration_per_task_ms": 75118.94940978078,
      "avg_output_tokens_per_task": 7923.973018549747,
      "last_run_timestamp": "2026-10-01T19:12:42.994Z"
    },
    {
      "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": 0.885566721955478,
      "accuracy_std_dev": 0.009041573880860321,
      "run_count": 12,
      "total_questions": 2291,
      "avg_cost_per_task_usd": 0.09590346355303361,
      "avg_duration_per_task_ms": 473697.19292885205,
      "avg_output_tokens_per_task": 28730.71977302488,
      "last_run_timestamp": "2026-10-31T16:12:28.231Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-flash-20260826",
      "model_slug": "qwen/qwen3.8-flash",
      "model_name": "Qwen: Qwen3.8 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.885522,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.02146120202020202,
      "avg_duration_per_task_ms": 674351.9393939395,
      "avg_output_tokens_per_task": 45464.17676767677,
      "last_run_timestamp": "2026-10-01T08:31:57.634Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "reflection/beam-1.0-501b-a23b-20261016",
      "model_slug": "reflection/beam-1.0-501b-a23b",
      "model_name": "Reflection: Beam 1.0 501B A23B",
      "provider_name": "Together",
      "accuracy": 0.8855216666666669,
      "accuracy_std_dev": 0.0023805928299947096,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.016851747474747474,
      "avg_duration_per_task_ms": 62951.132996632994,
      "avg_output_tokens_per_task": 13874.079124579124,
      "last_run_timestamp": "2026-10-31T08:03:10.178Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Inceptron",
      "accuracy": 0.8852179872935196,
      "accuracy_std_dev": 0.008384701841581625,
      "run_count": 12,
      "total_questions": 2361,
      "avg_cost_per_task_usd": 0.007895573909360442,
      "avg_duration_per_task_ms": 349145.3604404913,
      "avg_output_tokens_per_task": 20775.56925031766,
      "last_run_timestamp": "2026-10-31T08:07:18.580Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "DigitalOcean",
      "accuracy": 0.8851849999999999,
      "accuracy_std_dev": 0.05937202157919166,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.008028678282828283,
      "avg_duration_per_task_ms": 579179.9328282828,
      "avg_output_tokens_per_task": 32745.30505050505,
      "last_run_timestamp": "2026-10-31T08:07:18.761Z"
    },
    {
      "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": "google/gemini-3-flash-preview-20251217",
      "model_slug": "google/gemini-3-flash-preview",
      "model_name": "Google: Gemini 3 Flash Preview",
      "provider_name": "auto-routing",
      "accuracy": 0.88468,
      "accuracy_std_dev": 0.0008420000000000094,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.1319599772727273,
      "avg_duration_per_task_ms": 249775.19949494948,
      "avg_output_tokens_per_task": 43863.606060606064,
      "last_run_timestamp": "2026-10-01T19:15:31.426Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Nebius",
      "accuracy": 0.8846799999999999,
      "accuracy_std_dev": 0.021044000000000007,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.08896683838383838,
      "avg_duration_per_task_ms": 416558.4267676768,
      "avg_output_tokens_per_task": 22204.20202020202,
      "last_run_timestamp": "2026-10-31T08:57:00.457Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Parasail",
      "accuracy": 0.8844494345509893,
      "accuracy_std_dev": 0.01262210148905483,
      "run_count": 7,
      "total_questions": 1314,
      "avg_cost_per_task_usd": 0.18848695585996955,
      "avg_duration_per_task_ms": 741484.9490106545,
      "avg_output_tokens_per_task": 53820.38508371385,
      "last_run_timestamp": "2026-10-31T16:35:50.033Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "CoreWeave",
      "accuracy": 0.8840456666666667,
      "accuracy_std_dev": 0.005144110505120299,
      "run_count": 3,
      "total_questions": 585,
      "avg_cost_per_task_usd": 0.02067068717948718,
      "avg_duration_per_task_ms": 91684.32991452991,
      "avg_output_tokens_per_task": 41136.99487179487,
      "last_run_timestamp": "2026-10-31T08:35:13.435Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "DeepInfra",
      "accuracy": 0.8839239383667181,
      "accuracy_std_dev": 0.011423232412500406,
      "run_count": 8,
      "total_questions": 1298,
      "avg_cost_per_task_usd": 0.00878981278890601,
      "avg_duration_per_task_ms": 844366.6594761171,
      "avg_output_tokens_per_task": 29702.687981510015,
      "last_run_timestamp": "2026-10-31T08:34:49.723Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Together",
      "accuracy": 0.883838,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.11300773737373737,
      "avg_duration_per_task_ms": 418159.2121212121,
      "avg_output_tokens_per_task": 32149.79797979798,
      "last_run_timestamp": "2026-08-22T16:49:15.579Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Inferact vLLM",
      "accuracy": 0.883838,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.008843217171717172,
      "avg_duration_per_task_ms": 77278.5505050505,
      "avg_output_tokens_per_task": 29168.68181818182,
      "last_run_timestamp": "2026-11-13T21:02:01.371Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-terra-20260709",
      "model_slug": "openai/gpt-5.6-terra",
      "model_name": "OpenAI: GPT-5.6 Terra",
      "provider_name": "auto-routing",
      "accuracy": 0.883838,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.04134767424242424,
      "avg_duration_per_task_ms": 54936.305555555555,
      "avg_output_tokens_per_task": 3316.282828282828,
      "last_run_timestamp": "2026-10-01T19:13:07.806Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-luna-pro-20260922",
      "model_slug": "openai/gpt-6-luna-pro",
      "model_name": "OpenAI: GPT-6 Luna Pro",
      "provider_name": "auto-routing",
      "accuracy": 0.883838,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.008301368686868687,
      "avg_duration_per_task_ms": 72985.27777777778,
      "avg_output_tokens_per_task": 13098.969696969696,
      "last_run_timestamp": "2026-10-01T08:24:38.949Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "reflection/beam-1.0-501b-a23b-20261016",
      "model_slug": "reflection/beam-1.0-501b-a23b",
      "model_name": "Reflection: Beam 1.0 501B A23B",
      "provider_name": "auto-routing",
      "accuracy": 0.883838,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.014708931818181817,
      "avg_duration_per_task_ms": 71351.09343434343,
      "avg_output_tokens_per_task": 13550.636363636364,
      "last_run_timestamp": "2026-10-31T08:27:52.629Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "StreamLake",
      "accuracy": 0.8836312915601023,
      "accuracy_std_dev": 0.003620000000000012,
      "run_count": 2,
      "total_questions": 391,
      "avg_cost_per_task_usd": 0.006277416879795397,
      "avg_duration_per_task_ms": 104811.05115089515,
      "avg_output_tokens_per_task": 10905.734015345268,
      "last_run_timestamp": "2026-10-31T08:31:24.328Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "CoreWeave",
      "accuracy": 0.8833571428571427,
      "accuracy_std_dev": 0.012029934703878998,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.2045760606060606,
      "avg_duration_per_task_ms": 640185.823953824,
      "avg_output_tokens_per_task": 59868.3759018759,
      "last_run_timestamp": "2026-10-31T16:09:16.278Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "CoreWeave",
      "accuracy": 0.8831221012658229,
      "accuracy_std_dev": 0.01342904913771097,
      "run_count": 12,
      "total_questions": 2370,
      "avg_cost_per_task_usd": 0.1860454797468354,
      "avg_duration_per_task_ms": 320097.15611814347,
      "avg_output_tokens_per_task": 46697.94894514768,
      "last_run_timestamp": "2026-10-31T08:25:00.286Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Minimax",
      "accuracy": 0.8829965,
      "accuracy_std_dev": 0.002525500000000014,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.006819977272727273,
      "avg_duration_per_task_ms": 114306.47727272728,
      "avg_output_tokens_per_task": 11182.35606060606,
      "last_run_timestamp": "2026-10-31T08:07:19.817Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-opus-5-20260723",
      "model_slug": "anthropic/claude-opus-5",
      "model_name": "Anthropic: Claude Opus 5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.8829694193548384,
      "accuracy_std_dev": 0.009541162889069374,
      "run_count": 31,
      "total_questions": 6138,
      "avg_cost_per_task_usd": 0.1423835060280222,
      "avg_duration_per_task_ms": 63221.74942978169,
      "avg_output_tokens_per_task": 5118.9055066797,
      "last_run_timestamp": "2026-10-31T08:31:11.082Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Reka",
      "accuracy": 0.8829290000000001,
      "accuracy_std_dev": 0.012699110785799135,
      "run_count": 4,
      "total_questions": 783,
      "avg_cost_per_task_usd": 0.027103435504469985,
      "avg_duration_per_task_ms": 200899.3601532567,
      "avg_output_tokens_per_task": 30539.469987228607,
      "last_run_timestamp": "2026-10-31T08:15:03.390Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "Parasail",
      "accuracy": 0.882792129057187,
      "accuracy_std_dev": 0.012193080659094465,
      "run_count": 7,
      "total_questions": 1294,
      "avg_cost_per_task_usd": 0.06525332148377125,
      "avg_duration_per_task_ms": 302195.9992272025,
      "avg_output_tokens_per_task": 18028.40030911901,
      "last_run_timestamp": "2026-10-31T16:03:21.450Z"
    },
    {
      "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": 0.8826767013977129,
      "accuracy_std_dev": 0.009892996415464268,
      "run_count": 12,
      "total_questions": 2361,
      "avg_cost_per_task_usd": 0.10956175815332488,
      "avg_duration_per_task_ms": 529359.7882253282,
      "avg_output_tokens_per_task": 31450.83905124947,
      "last_run_timestamp": "2026-10-31T16:02:14.362Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Together",
      "accuracy": 0.882365125,
      "accuracy_std_dev": 0.0043279540904883765,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.029606650252525252,
      "avg_duration_per_task_ms": 78746.81123737374,
      "avg_output_tokens_per_task": 24519.371843434343,
      "last_run_timestamp": "2026-10-31T09:02:59.407Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.8821789256756756,
      "accuracy_std_dev": 0.01706691482241463,
      "run_count": 6,
      "total_questions": 1184,
      "avg_cost_per_task_usd": 0.04157157263513513,
      "avg_duration_per_task_ms": 164106.9873310811,
      "avg_output_tokens_per_task": 34476.61317567567,
      "last_run_timestamp": "2026-10-31T08:31:11.691Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "ModelRun",
      "accuracy": 0.8821550000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.026309151515151515,
      "avg_duration_per_task_ms": 202247.18686868687,
      "avg_output_tokens_per_task": 39119.5505050505,
      "last_run_timestamp": "2026-10-29T03:00:03.996Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Phala",
      "accuracy": 0.8821550000000001,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.129083601010101,
      "avg_duration_per_task_ms": 388751.6590909091,
      "avg_output_tokens_per_task": 32112.176767676767,
      "last_run_timestamp": "2026-10-17T16:07:32.047Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Novita",
      "accuracy": 0.8821546666666666,
      "accuracy_std_dev": 0.015120291935747206,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.007597324915824916,
      "avg_duration_per_task_ms": 210520.51515151514,
      "avg_output_tokens_per_task": 26812.25925925926,
      "last_run_timestamp": "2026-10-03T16:57:09.267Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Reka",
      "accuracy": 0.881818,
      "accuracy_std_dev": 0.03925680632450885,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.010723755555555554,
      "avg_duration_per_task_ms": 145330.13333333333,
      "avg_output_tokens_per_task": 16217.381818181819,
      "last_run_timestamp": "2026-10-31T09:01:13.864Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-opus-5-20260723",
      "model_slug": "anthropic/claude-opus-5",
      "model_name": "Anthropic: Claude Opus 5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.8816956363636362,
      "accuracy_std_dev": 0.019816803931249533,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.12889897842056933,
      "avg_duration_per_task_ms": 63188.49035812672,
      "avg_output_tokens_per_task": 4922.145546372819,
      "last_run_timestamp": "2026-10-31T08:31:10.506Z"
    },
    {
      "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": 0.8816457669902912,
      "accuracy_std_dev": 0.0330676479779465,
      "run_count": 11,
      "total_questions": 2163,
      "avg_cost_per_task_usd": 0.11147017290799817,
      "avg_duration_per_task_ms": 672376.8964401295,
      "avg_output_tokens_per_task": 27712.78640776699,
      "last_run_timestamp": "2026-10-31T16:08:08.536Z"
    },
    {
      "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-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Crusoe",
      "accuracy": 0.8810325,
      "accuracy_std_dev": 0.016020429392393534,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.03953295538720539,
      "avg_duration_per_task_ms": 203574.52356902356,
      "avg_output_tokens_per_task": 32800.2398989899,
      "last_run_timestamp": "2026-10-31T08:47:44.999Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-fable-5.1-20260831",
      "model_slug": "anthropic/claude-fable-5.1",
      "model_name": "Anthropic: Claude Fable 5.1",
      "provider_name": "Google",
      "accuracy": 0.8809522857142856,
      "accuracy_std_dev": 0.031683881867663896,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.12096548845598847,
      "avg_duration_per_task_ms": 28374.095238095237,
      "avg_output_tokens_per_task": 2176.287878787879,
      "last_run_timestamp": "2026-11-11T21:00:44.509Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "StreamLake",
      "accuracy": 0.8806113333333333,
      "accuracy_std_dev": 0.012070796855312507,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.006245051767676768,
      "avg_duration_per_task_ms": 409383.2984006734,
      "avg_output_tokens_per_task": 30642.196969696968,
      "last_run_timestamp": "2026-10-31T08:19:56.454Z"
    },
    {
      "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": 0.8802082288602942,
      "accuracy_std_dev": 0.029715678021281344,
      "run_count": 11,
      "total_questions": 2176,
      "avg_cost_per_task_usd": 0.17839703079044117,
      "avg_duration_per_task_ms": 356421.5234375,
      "avg_output_tokens_per_task": 40400.66819852941,
      "last_run_timestamp": "2026-10-31T16:32:36.327Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "auto-routing",
      "accuracy": 0.8801906666666667,
      "accuracy_std_dev": 0.011226983784119793,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.0227881835016835,
      "avg_duration_per_task_ms": 234700.68434343435,
      "avg_output_tokens_per_task": 25264.190235690236,
      "last_run_timestamp": "2026-10-03T09:30:18.263Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Modal",
      "accuracy": 0.8800631670886074,
      "accuracy_std_dev": 0.006743341790240229,
      "run_count": 8,
      "total_questions": 1580,
      "avg_cost_per_task_usd": 0.033778830379746834,
      "avg_duration_per_task_ms": 152045.89430379748,
      "avg_output_tokens_per_task": 27961.480379746834,
      "last_run_timestamp": "2026-10-31T09:42:46.430Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Parasail",
      "accuracy": 0.8799569142550911,
      "accuracy_std_dev": 0.006792606115475872,
      "run_count": 5,
      "total_questions": 933,
      "avg_cost_per_task_usd": 0.11584198606645231,
      "avg_duration_per_task_ms": 484616.05144694535,
      "avg_output_tokens_per_task": 33104.76848874598,
      "last_run_timestamp": "2026-10-17T16:07:16.258Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Wafer",
      "accuracy": 0.879872690530519,
      "accuracy_std_dev": 0.053065810239372686,
      "run_count": 18,
      "total_questions": 3506,
      "avg_cost_per_task_usd": 0.00976165202509983,
      "avg_duration_per_task_ms": 414485.4175698802,
      "avg_output_tokens_per_task": 26831.222760981174,
      "last_run_timestamp": "2026-10-31T08:47:06.377Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Venice",
      "accuracy": 0.87984,
      "accuracy_std_dev": 0.018052164925570553,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.036141941919191914,
      "avg_duration_per_task_ms": 139656.41224747474,
      "avg_output_tokens_per_task": 28081.721590909092,
      "last_run_timestamp": "2026-10-31T09:14:35.460Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "StreamLake",
      "accuracy": 0.8796295000000002,
      "accuracy_std_dev": 0.017716787991901916,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.02372089204545455,
      "avg_duration_per_task_ms": 758589.1104797979,
      "avg_output_tokens_per_task": 35952.905303030304,
      "last_run_timestamp": "2026-10-31T16:46:00.416Z"
    },
    {
      "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": "OpenInference",
      "accuracy": 0.879386,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 76,
      "avg_cost_per_task_usd": 0.03120878947368421,
      "avg_duration_per_task_ms": 274583.63157894736,
      "avg_output_tokens_per_task": 9741.526315789473,
      "last_run_timestamp": "2026-10-09T00:00:06.206Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "DigitalOcean",
      "accuracy": 0.8793102638453499,
      "accuracy_std_dev": 0.020773557275777312,
      "run_count": 10,
      "total_questions": 1914,
      "avg_cost_per_task_usd": 0.09333590334378267,
      "avg_duration_per_task_ms": 520471.578892372,
      "avg_output_tokens_per_task": 28040.05590386625,
      "last_run_timestamp": "2026-10-31T08:31:12.870Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-opus-5-20260723",
      "model_slug": "anthropic/claude-opus-5",
      "model_name": "Anthropic: Claude Opus 5",
      "provider_name": "auto-routing",
      "accuracy": 0.8792469090909092,
      "accuracy_std_dev": 0.01899374060269975,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.12746250872359963,
      "avg_duration_per_task_ms": 62359.71349862259,
      "avg_output_tokens_per_task": 4869.864095500459,
      "last_run_timestamp": "2026-10-31T08:12:45.681Z"
    },
    {
      "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": 0.8792467272727271,
      "accuracy_std_dev": 0.029423779211845618,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.1929496730945822,
      "avg_duration_per_task_ms": 46902.871441689625,
      "avg_output_tokens_per_task": 3649.5376492194673,
      "last_run_timestamp": "2026-10-31T08:26:18.906Z"
    },
    {
      "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": 0.8792337006896551,
      "accuracy_std_dev": 0.0059956127348350235,
      "run_count": 11,
      "total_questions": 2175,
      "avg_cost_per_task_usd": 0.11865828321839081,
      "avg_duration_per_task_ms": 795322.059770115,
      "avg_output_tokens_per_task": 31057.051954022987,
      "last_run_timestamp": "2026-10-31T16:19:28.502Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "CoreWeave",
      "accuracy": 0.8792084999999998,
      "accuracy_std_dev": 0.013918588167746535,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.009092756734006733,
      "avg_duration_per_task_ms": 214671.74326599325,
      "avg_output_tokens_per_task": 32045.307239057238,
      "last_run_timestamp": "2026-10-31T08:31:11.360Z"
    },
    {
      "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": 0.8789951917808219,
      "accuracy_std_dev": 0.028062210901257694,
      "run_count": 3,
      "total_questions": 511,
      "avg_cost_per_task_usd": 0.12092838356164383,
      "avg_duration_per_task_ms": 556595.4500978474,
      "avg_output_tokens_per_task": 50889.26614481409,
      "last_run_timestamp": "2026-10-03T17:01:24.374Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Baidu",
      "accuracy": 0.8788738633776092,
      "accuracy_std_dev": 0.01184394360000081,
      "run_count": 8,
      "total_questions": 1581,
      "avg_cost_per_task_usd": 0.05087817014547753,
      "avg_duration_per_task_ms": 522785.04807084124,
      "avg_output_tokens_per_task": 36266.31309297913,
      "last_run_timestamp": "2026-11-10T18:38:17.277Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-luna-20260709",
      "model_slug": "openai/gpt-5.6-luna",
      "model_name": "OpenAI: GPT-5.6 Luna",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.8787880000000001,
      "accuracy_std_dev": 0.007273551310513108,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.009556249158249158,
      "avg_duration_per_task_ms": 52070.99494949495,
      "avg_output_tokens_per_task": 7331.292929292929,
      "last_run_timestamp": "2026-10-30T00:00:06.145Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.7-plus-20260602",
      "model_slug": "qwen/qwen3.7-plus",
      "model_name": "Qwen: Qwen3.7 Plus",
      "provider_name": "auto-routing",
      "accuracy": 0.878788,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.043995565656565656,
      "avg_duration_per_task_ms": 588519.7272727273,
      "avg_output_tokens_per_task": 34166.82828282828,
      "last_run_timestamp": "2026-10-01T08:47:14.818Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8783067142857143,
      "accuracy_std_dev": 0.019872992683641802,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.12211109668109668,
      "avg_duration_per_task_ms": 631289.5389610389,
      "avg_output_tokens_per_task": 35901.76046176046,
      "last_run_timestamp": "2026-10-31T16:25:53.821Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "DeepInfra",
      "accuracy": 0.8782263333333332,
      "accuracy_std_dev": 0.004199386807089298,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.020176612794612796,
      "avg_duration_per_task_ms": 120195.73063973064,
      "avg_output_tokens_per_task": 42971.59090909091,
      "last_run_timestamp": "2026-10-31T08:09:30.418Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-luna-20260922",
      "model_slug": "openai/gpt-6-luna",
      "model_name": "OpenAI: GPT-6 Luna",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.8781846784586228,
      "accuracy_std_dev": 0.00428175396303899,
      "run_count": 8,
      "total_questions": 1583,
      "avg_cost_per_task_usd": 0.00286223562855338,
      "avg_duration_per_task_ms": 91558.78395451674,
      "avg_output_tokens_per_task": 5494.553379658875,
      "last_run_timestamp": "2026-11-03T03:00:05.640Z"
    },
    {
      "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": 0.8778809359190558,
      "accuracy_std_dev": 0.007512559071684942,
      "run_count": 6,
      "total_questions": 1186,
      "avg_cost_per_task_usd": 0.12372368465430017,
      "avg_duration_per_task_ms": 878302.9047217538,
      "avg_output_tokens_per_task": 43322.24536256324,
      "last_run_timestamp": "2026-10-24T16:19:29.598Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "GMICloud",
      "accuracy": 0.8777777999999999,
      "accuracy_std_dev": 0.016888770131658508,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.009364964646464646,
      "avg_duration_per_task_ms": 177367.14141414143,
      "avg_output_tokens_per_task": 18449.226262626264,
      "last_run_timestamp": "2026-11-10T06:10:46.202Z"
    },
    {
      "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": 0.8777164545454544,
      "accuracy_std_dev": 0.03678752360482025,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.18499438292011017,
      "avg_duration_per_task_ms": 44694.47474747475,
      "avg_output_tokens_per_task": 3489.084940312213,
      "last_run_timestamp": "2026-10-31T08:13:57.565Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Ionstream",
      "accuracy": 0.8774407999999999,
      "accuracy_std_dev": 0.00954717737134911,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.03921924343434343,
      "avg_duration_per_task_ms": 242470.04242424242,
      "avg_output_tokens_per_task": 37354.009090909094,
      "last_run_timestamp": "2026-10-21T11:23:15.101Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Reka",
      "accuracy": 0.8772070529695024,
      "accuracy_std_dev": 0.03893659390197737,
      "run_count": 4,
      "total_questions": 623,
      "avg_cost_per_task_usd": 0.09425666452648475,
      "avg_duration_per_task_ms": 461375.2568218299,
      "avg_output_tokens_per_task": 23922.329052969504,
      "last_run_timestamp": "2026-09-30T11:37:23.219Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-luna-20260922",
      "model_slug": "openai/gpt-6-luna",
      "model_name": "OpenAI: GPT-6 Luna",
      "provider_name": "Azure",
      "accuracy": 0.8771043333333335,
      "accuracy_std_dev": 0.007896528133018712,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.003143374579124579,
      "avg_duration_per_task_ms": 35450.33585858586,
      "avg_output_tokens_per_task": 5733.669191919192,
      "last_run_timestamp": "2026-10-01T08:25:15.234Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "GMICloud",
      "accuracy": 0.877104,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.07509321212121212,
      "avg_duration_per_task_ms": 263175.32828282827,
      "avg_output_tokens_per_task": 20738.459595959597,
      "last_run_timestamp": "2026-08-29T19:12:43.653Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Parasail",
      "accuracy": 0.876730111111111,
      "accuracy_std_dev": 0.012390967395149362,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.03705083277216611,
      "avg_duration_per_task_ms": 153849.39562289562,
      "avg_output_tokens_per_task": 30722.645342312007,
      "last_run_timestamp": "2026-11-05T00:00:09.938Z"
    },
    {
      "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": 0.8767260400808354,
      "accuracy_std_dev": 0.008082492995494786,
      "run_count": 15,
      "total_questions": 2969,
      "avg_cost_per_task_usd": 0.024447516335466487,
      "avg_duration_per_task_ms": 220112.43718423712,
      "avg_output_tokens_per_task": 34286.534523408554,
      "last_run_timestamp": "2026-11-02T17:21:00.007Z"
    },
    {
      "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": 0.876645090909091,
      "accuracy_std_dev": 0.013199311025094829,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.00596057116620753,
      "avg_duration_per_task_ms": 273115.2906336088,
      "avg_output_tokens_per_task": 18439.58402203857,
      "last_run_timestamp": "2026-10-31T08:36:26.194Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Z.AI",
      "accuracy": 0.8763216813793104,
      "accuracy_std_dev": 0.01096372812565248,
      "run_count": 11,
      "total_questions": 2175,
      "avg_cost_per_task_usd": 0.0099918183908046,
      "avg_duration_per_task_ms": 425649.0579310345,
      "avg_output_tokens_per_task": 22880.83264367816,
      "last_run_timestamp": "2026-10-31T08:32:56.980Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.8762625,
      "accuracy_std_dev": 0.009259500000000032,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.014013469696969696,
      "avg_duration_per_task_ms": 171317.10353535353,
      "avg_output_tokens_per_task": 26009.79292929293,
      "last_run_timestamp": "2026-10-31T08:09:25.873Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-luna-20260709",
      "model_slug": "openai/gpt-5.6-luna",
      "model_name": "OpenAI: GPT-5.6 Luna",
      "provider_name": "auto-routing",
      "accuracy": 0.8762625,
      "accuracy_std_dev": 0.005892500000000023,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.009382047979797978,
      "avg_duration_per_task_ms": 83118.84595959596,
      "avg_output_tokens_per_task": 7689.015151515152,
      "last_run_timestamp": "2026-10-01T19:14:22.294Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Decart",
      "accuracy": 0.8759833172804533,
      "accuracy_std_dev": 0.02383166285579381,
      "run_count": 6,
      "total_questions": 1059,
      "avg_cost_per_task_usd": 0.006660402266288952,
      "avg_duration_per_task_ms": 451840.41265344666,
      "avg_output_tokens_per_task": 36610.579792256845,
      "last_run_timestamp": "2026-10-31T08:15:03.294Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Novita",
      "accuracy": 0.8759819999999999,
      "accuracy_std_dev": 0.04446362050785635,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.019776964646464647,
      "avg_duration_per_task_ms": 142117.47643097644,
      "avg_output_tokens_per_task": 20178.936026936026,
      "last_run_timestamp": "2026-10-31T09:01:06.393Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "Venice",
      "accuracy": 0.8758864791489362,
      "accuracy_std_dev": 0.029346183036394313,
      "run_count": 6,
      "total_questions": 1175,
      "avg_cost_per_task_usd": 0.13275879148936168,
      "avg_duration_per_task_ms": 778330.9634042553,
      "avg_output_tokens_per_task": 39857.91574468085,
      "last_run_timestamp": "2026-10-31T16:02:14.362Z"
    },
    {
      "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": 0.8757797045454545,
      "accuracy_std_dev": 0.012408993441628312,
      "run_count": 8,
      "total_questions": 1496,
      "avg_cost_per_task_usd": 0.008467116978609626,
      "avg_duration_per_task_ms": 224989.5588235294,
      "avg_output_tokens_per_task": 29168.869652406418,
      "last_run_timestamp": "2026-10-28T03:34:12.001Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "Together",
      "accuracy": 0.8755709520547945,
      "accuracy_std_dev": 0.007533567739715949,
      "run_count": 3,
      "total_questions": 584,
      "avg_cost_per_task_usd": 0.031521625,
      "avg_duration_per_task_ms": 247947.07705479453,
      "avg_output_tokens_per_task": 62802.890410958906,
      "last_run_timestamp": "2026-10-31T08:14:43.030Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "Nebius",
      "accuracy": 0.875421,
      "accuracy_std_dev": 0.010101000000000027,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.05678013888888889,
      "avg_duration_per_task_ms": 453449.72222222225,
      "avg_output_tokens_per_task": 62821.7904040404,
      "last_run_timestamp": "2026-10-31T08:23:57.761Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Azure",
      "accuracy": 0.8754208571428571,
      "accuracy_std_dev": 0.009220961422271589,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.1371511341991342,
      "avg_duration_per_task_ms": 439856.0331890332,
      "avg_output_tokens_per_task": 35614.08585858586,
      "last_run_timestamp": "2026-10-31T16:30:53.502Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Together",
      "accuracy": 0.8754206666666668,
      "accuracy_std_dev": 0.004955863351178644,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.01787970538720539,
      "avg_duration_per_task_ms": 181946.53367003368,
      "avg_output_tokens_per_task": 29504.771043771045,
      "last_run_timestamp": "2026-10-31T08:09:31.246Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Cloudflare",
      "accuracy": 0.8754205,
      "accuracy_std_dev": 0.0016835000000000044,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.04156198232323233,
      "avg_duration_per_task_ms": 309570.8888888889,
      "avg_output_tokens_per_task": 39159.9494949495,
      "last_run_timestamp": "2026-10-17T08:26:18.265Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-5-fable-20260609",
      "model_slug": "anthropic/claude-fable-5",
      "model_name": "Anthropic: Claude Fable 5",
      "provider_name": "Azure",
      "accuracy": 0.8752677272727273,
      "accuracy_std_dev": 0.0335202236895631,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.19080393296602388,
      "avg_duration_per_task_ms": 47299.57070707071,
      "avg_output_tokens_per_task": 3603.7240587695132,
      "last_run_timestamp": "2026-10-31T08:09:25.606Z"
    },
    {
      "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": 0.8750631623672231,
      "accuracy_std_dev": 0.10207384869887097,
      "run_count": 10,
      "total_questions": 1977,
      "avg_cost_per_task_usd": 0.02678827718765807,
      "avg_duration_per_task_ms": 188464.65857359636,
      "avg_output_tokens_per_task": 22634.93171471927,
      "last_run_timestamp": "2026-10-31T16:04:29.094Z"
    },
    {
      "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": 0.8746991428571429,
      "accuracy_std_dev": 0.005681157161518922,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.08339443073593074,
      "avg_duration_per_task_ms": 446738.63275613275,
      "avg_output_tokens_per_task": 23715.052669552668,
      "last_run_timestamp": "2026-10-31T16:15:54.664Z"
    },
    {
      "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": 0.8746888260869566,
      "accuracy_std_dev": 0.03857164733172279,
      "run_count": 23,
      "total_questions": 4554,
      "avg_cost_per_task_usd": 0.19603994861660076,
      "avg_duration_per_task_ms": 42405.70619235837,
      "avg_output_tokens_per_task": 3518.768115942029,
      "last_run_timestamp": "2026-11-11T21:00:44.482Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Reka",
      "accuracy": 0.8746087667870037,
      "accuracy_std_dev": 0.007592482877484693,
      "run_count": 7,
      "total_questions": 1385,
      "avg_cost_per_task_usd": 0.2179415407942238,
      "avg_duration_per_task_ms": 1239989.4288808664,
      "avg_output_tokens_per_task": 48485.83321299639,
      "last_run_timestamp": "2026-10-24T16:18:35.554Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Makora",
      "accuracy": 0.8745317312013828,
      "accuracy_std_dev": 0.011186512454241005,
      "run_count": 6,
      "total_questions": 1157,
      "avg_cost_per_task_usd": 0.0043274416594641305,
      "avg_duration_per_task_ms": 185028.52031114954,
      "avg_output_tokens_per_task": 21870.51080380294,
      "last_run_timestamp": "2026-09-26T08:53:59.525Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "auto-routing",
      "accuracy": 0.8744105999999999,
      "accuracy_std_dev": 0.014149500063253112,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.09050255303030302,
      "avg_duration_per_task_ms": 196310.64444444445,
      "avg_output_tokens_per_task": 14775.43686868687,
      "last_run_timestamp": "2026-10-31T08:07:27.863Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "BaseTen",
      "accuracy": 0.8743250351006383,
      "accuracy_std_dev": 0.026296453097608783,
      "run_count": 21,
      "total_questions": 4074,
      "avg_cost_per_task_usd": 0.03975532817869414,
      "avg_duration_per_task_ms": 80921.18360333824,
      "avg_output_tokens_per_task": 25880.549828178693,
      "last_run_timestamp": "2026-10-31T09:03:20.938Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Together",
      "accuracy": 0.8742938723163841,
      "accuracy_std_dev": 0.008148155125586256,
      "run_count": 9,
      "total_questions": 1770,
      "avg_cost_per_task_usd": 0.009214897175141243,
      "avg_duration_per_task_ms": 375895.7553672316,
      "avg_output_tokens_per_task": 32530.198305084745,
      "last_run_timestamp": "2026-10-17T08:23:51.334Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Makora",
      "accuracy": 0.8740494840344654,
      "accuracy_std_dev": 0.0038017229265163467,
      "run_count": 10,
      "total_questions": 1973,
      "avg_cost_per_task_usd": 0.04434169640141917,
      "avg_duration_per_task_ms": 202197.7790167258,
      "avg_output_tokens_per_task": 38209.482513938165,
      "last_run_timestamp": "2026-10-31T08:10:37.345Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Together",
      "accuracy": 0.8737693295358652,
      "accuracy_std_dev": 0.014113612302391693,
      "run_count": 12,
      "total_questions": 2370,
      "avg_cost_per_task_usd": 0.09247935569620254,
      "avg_duration_per_task_ms": 203634.65358649788,
      "avg_output_tokens_per_task": 23372.77805907173,
      "last_run_timestamp": "2026-10-31T08:40:26.156Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Nebius",
      "accuracy": 0.873737,
      "accuracy_std_dev": 0.007273551310513108,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.01830383333333333,
      "avg_duration_per_task_ms": 381305.08585858584,
      "avg_output_tokens_per_task": 30305.045454545456,
      "last_run_timestamp": "2026-10-31T08:27:46.342Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-luna-20260922",
      "model_slug": "openai/gpt-6-luna",
      "model_name": "OpenAI: GPT-6 Luna",
      "provider_name": "auto-routing",
      "accuracy": 0.873737,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0029526111111111116,
      "avg_duration_per_task_ms": 49722.71717171717,
      "avg_output_tokens_per_task": 5746.707070707071,
      "last_run_timestamp": "2026-10-01T08:23:37.659Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.7-plus-20260602",
      "model_slug": "qwen/qwen3.7-plus",
      "model_name": "Qwen: Qwen3.7 Plus",
      "provider_name": "Alibaba",
      "accuracy": 0.873737,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.045554454545454544,
      "avg_duration_per_task_ms": 605480.6616161616,
      "avg_output_tokens_per_task": 35385.22727272727,
      "last_run_timestamp": "2026-10-01T08:40:06.225Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Decart",
      "accuracy": 0.8735040000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 195,
      "avg_cost_per_task_usd": 0.0022876666666666666,
      "avg_duration_per_task_ms": 331580.09743589745,
      "avg_output_tokens_per_task": 16588.984615384616,
      "last_run_timestamp": "2026-08-22T10:20:26.941Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "auto-routing",
      "accuracy": 0.8727751428571429,
      "accuracy_std_dev": 0.01839478573966448,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.047217443722943724,
      "avg_duration_per_task_ms": 671531.1060606061,
      "avg_output_tokens_per_task": 35962.7987012987,
      "last_run_timestamp": "2026-10-31T16:16:55.624Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "DigitalOcean",
      "accuracy": 0.8726149999999999,
      "accuracy_std_dev": 0.0020998715833752102,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.017651676767676767,
      "avg_duration_per_task_ms": 95564.90235690236,
      "avg_output_tokens_per_task": 29293.20202020202,
      "last_run_timestamp": "2026-10-31T08:33:05.096Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "Alibaba",
      "accuracy": 0.8723940682248895,
      "accuracy_std_dev": 0.00799901581348451,
      "run_count": 8,
      "total_questions": 1583,
      "avg_cost_per_task_usd": 0.0495316885660139,
      "avg_duration_per_task_ms": 303242.33101705625,
      "avg_output_tokens_per_task": 21030.309538850284,
      "last_run_timestamp": "2026-10-31T16:14:42.903Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Azure",
      "accuracy": 0.8722941428571429,
      "accuracy_std_dev": 0.007658356601565021,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.01657262121212121,
      "avg_duration_per_task_ms": 289911.9790764791,
      "avg_output_tokens_per_task": 29391.78354978355,
      "last_run_timestamp": "2026-10-31T16:04:48.120Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "OpenInference",
      "accuracy": 0.8722639999999998,
      "accuracy_std_dev": 0.01391911622553672,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.01943295707070707,
      "avg_duration_per_task_ms": 314686.08207070705,
      "avg_output_tokens_per_task": 35123.857954545456,
      "last_run_timestamp": "2026-10-31T08:09:31.159Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-sonnet-5-20260630",
      "model_slug": "anthropic/claude-sonnet-5",
      "model_name": "Anthropic: Claude Sonnet 5",
      "provider_name": "Azure",
      "accuracy": 0.8722068181818181,
      "accuracy_std_dev": 0.028274856431813575,
      "run_count": 22,
      "total_questions": 4356,
      "avg_cost_per_task_usd": 0.16051083654729112,
      "avg_duration_per_task_ms": 154457.4214876033,
      "avg_output_tokens_per_task": 15411.82598714417,
      "last_run_timestamp": "2026-10-31T08:57:30.925Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "Moonshot AI",
      "accuracy": 0.8720540000000001,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.12898378535353536,
      "avg_duration_per_task_ms": 597048.5454545454,
      "avg_output_tokens_per_task": 42966.707070707074,
      "last_run_timestamp": "2026-08-29T18:25:54.650Z"
    },
    {
      "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": 0.8720540000000001,
      "accuracy_std_dev": 0.003367000000000009,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.09837352525252525,
      "avg_duration_per_task_ms": 729156.0252525252,
      "avg_output_tokens_per_task": 28490.656565656565,
      "last_run_timestamp": "2026-08-29T11:31:24.094Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Morph",
      "accuracy": 0.871875,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 160,
      "avg_cost_per_task_usd": 0.04507655,
      "avg_duration_per_task_ms": 1504372.61875,
      "avg_output_tokens_per_task": 18693.5375,
      "last_run_timestamp": "2026-08-27T15:00:05.074Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-sonnet-5-20260630",
      "model_slug": "anthropic/claude-sonnet-5",
      "model_name": "Anthropic: Claude Sonnet 5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.8718365483870967,
      "accuracy_std_dev": 0.025673440848897774,
      "run_count": 31,
      "total_questions": 6138,
      "avg_cost_per_task_usd": 0.17548565819485176,
      "avg_duration_per_task_ms": 159833.02085369828,
      "avg_output_tokens_per_task": 16263.557673509287,
      "last_run_timestamp": "2026-10-31T08:33:01.888Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Wafer",
      "accuracy": 0.8714925000000001,
      "accuracy_std_dev": 0.007251768353305299,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.007901208754208754,
      "avg_duration_per_task_ms": 584978.2028619528,
      "avg_output_tokens_per_task": 44690.53282828283,
      "last_run_timestamp": "2026-10-31T16:48:05.771Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "DeepInfra",
      "accuracy": 0.8714859264426126,
      "accuracy_std_dev": 0.00752103815972237,
      "run_count": 8,
      "total_questions": 1577,
      "avg_cost_per_task_usd": 0.06945589727330374,
      "avg_duration_per_task_ms": 276858.71528218134,
      "avg_output_tokens_per_task": 11308.629676601142,
      "last_run_timestamp": "2026-10-10T08:22:14.215Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8713324285714285,
      "accuracy_std_dev": 0.009851679854683843,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.07396083838383838,
      "avg_duration_per_task_ms": 301937.24242424243,
      "avg_output_tokens_per_task": 21002.70779220779,
      "last_run_timestamp": "2026-10-31T16:09:19.242Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-5-fable-20260609",
      "model_slug": "anthropic/claude-fable-5",
      "model_name": "Anthropic: Claude Fable 5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.8712119166666668,
      "accuracy_std_dev": 0.038220154043075064,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.1817093308080808,
      "avg_duration_per_task_ms": 45516.91035353536,
      "avg_output_tokens_per_task": 3430.526515151515,
      "last_run_timestamp": "2026-10-31T08:20:29.734Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.8708191333333333,
      "accuracy_std_dev": 0.005011630059061523,
      "run_count": 15,
      "total_questions": 2970,
      "avg_cost_per_task_usd": 0.004255378451178452,
      "avg_duration_per_task_ms": 39486.23198653199,
      "avg_output_tokens_per_task": 7745.265656565656,
      "last_run_timestamp": "2026-10-31T08:49:45.152Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "DeepInfra",
      "accuracy": 0.8707621944670464,
      "accuracy_std_dev": 0.005695171399115766,
      "run_count": 7,
      "total_questions": 1229,
      "avg_cost_per_task_usd": 0.005226334418226199,
      "avg_duration_per_task_ms": 743725.4109031733,
      "avg_output_tokens_per_task": 28735.099267697315,
      "last_run_timestamp": "2026-10-31T16:22:14.129Z"
    },
    {
      "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": 0.870748394132653,
      "accuracy_std_dev": 0.011444362026233704,
      "run_count": 4,
      "total_questions": 784,
      "avg_cost_per_task_usd": 0.02509891071428572,
      "avg_duration_per_task_ms": 366798.57908163266,
      "avg_output_tokens_per_task": 20671.150510204083,
      "last_run_timestamp": "2026-09-12T08:54:29.103Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "Anthropic",
      "accuracy": 0.8703704,
      "accuracy_std_dev": 0.00583193054142453,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.004018268686868687,
      "avg_duration_per_task_ms": 37454.91717171717,
      "avg_output_tokens_per_task": 7796.7444444444445,
      "last_run_timestamp": "2026-10-31T08:14:20.331Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "thinkingmachines/inkling-small-20260730",
      "model_slug": "thinkingmachines/inkling-small",
      "model_name": "Thinking Machines: Inkling Small",
      "provider_name": "BaseTen",
      "accuracy": 0.8703701666666666,
      "accuracy_std_dev": 0.006941288747215641,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.024583196127946128,
      "avg_duration_per_task_ms": 150981.7457912458,
      "avg_output_tokens_per_task": 20283.477272727272,
      "last_run_timestamp": "2026-09-19T08:49:51.735Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Venice",
      "accuracy": 0.8703605355249205,
      "accuracy_std_dev": 0.03387763198941155,
      "run_count": 10,
      "total_questions": 1886,
      "avg_cost_per_task_usd": 0.014264545599151641,
      "avg_duration_per_task_ms": 348340.7725344645,
      "avg_output_tokens_per_task": 28651.488865323437,
      "last_run_timestamp": "2026-10-31T08:28:24.326Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Wafer",
      "accuracy": 0.870112333932584,
      "accuracy_std_dev": 0.07630745338957075,
      "run_count": 23,
      "total_questions": 4450,
      "avg_cost_per_task_usd": 0.21229808741573034,
      "avg_duration_per_task_ms": 491047.8352808989,
      "avg_output_tokens_per_task": 61310.42764044944,
      "last_run_timestamp": "2026-10-31T09:02:18.616Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "Together",
      "accuracy": 0.8700336,
      "accuracy_std_dev": 0.008343089106559986,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.07480730858585859,
      "avg_duration_per_task_ms": 106906.41464646465,
      "avg_output_tokens_per_task": 12161.163636363637,
      "last_run_timestamp": "2026-10-31T08:03:09.289Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "DeepInfra",
      "accuracy": 0.8699400591341078,
      "accuracy_std_dev": 0.01681928135474285,
      "run_count": 10,
      "total_questions": 1894,
      "avg_cost_per_task_usd": 0.07171590390707498,
      "avg_duration_per_task_ms": 408524.5543822598,
      "avg_output_tokens_per_task": 23628.092397043296,
      "last_run_timestamp": "2026-10-31T08:45:16.438Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "PrimaLabs",
      "accuracy": 0.8697916770833333,
      "accuracy_std_dev": 0.045953881606091436,
      "run_count": 3,
      "total_questions": 576,
      "avg_cost_per_task_usd": 0.32188484548611107,
      "avg_duration_per_task_ms": 172479.83506944444,
      "avg_output_tokens_per_task": 40101.135416666664,
      "last_run_timestamp": "2026-11-06T16:32:49.018Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Fireworks",
      "accuracy": 0.8697338870630417,
      "accuracy_std_dev": 0.011260986943785035,
      "run_count": 17,
      "total_questions": 3347,
      "avg_cost_per_task_usd": 0.04833613654018524,
      "avg_duration_per_task_ms": 200583.7881685091,
      "avg_output_tokens_per_task": 37254.27158649537,
      "last_run_timestamp": "2026-10-31T08:35:39.353Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "StreamLake",
      "accuracy": 0.8696521878754779,
      "accuracy_std_dev": 0.02976323739044529,
      "run_count": 10,
      "total_questions": 1831,
      "avg_cost_per_task_usd": 0.022690866193336975,
      "avg_duration_per_task_ms": 451948.2801747679,
      "avg_output_tokens_per_task": 21491.623702894594,
      "last_run_timestamp": "2026-10-17T08:18:04.346Z"
    },
    {
      "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": "Z.AI",
      "accuracy": 0.8695517762430939,
      "accuracy_std_dev": 0.027759107412948937,
      "run_count": 11,
      "total_questions": 2172,
      "avg_cost_per_task_usd": 0.10362995395948434,
      "avg_duration_per_task_ms": 356439.69521178637,
      "avg_output_tokens_per_task": 23309.89779005525,
      "last_run_timestamp": "2026-10-31T08:27:45.609Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.1-20251113",
      "model_slug": "openai/gpt-5.1",
      "model_name": "OpenAI: GPT-5.1",
      "provider_name": "OpenAI",
      "accuracy": 0.8695284999999999,
      "accuracy_std_dev": 0.005892500000000023,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.2202375707070707,
      "avg_duration_per_task_ms": 264875.98484848486,
      "avg_output_tokens_per_task": 21926.78787878788,
      "last_run_timestamp": "2026-10-01T16:12:49.000Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "ModelRun",
      "accuracy": 0.8695284999999998,
      "accuracy_std_dev": 0.01157224020015143,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.018078708333333332,
      "avg_duration_per_task_ms": 141813.27146464647,
      "avg_output_tokens_per_task": 29935.99873737374,
      "last_run_timestamp": "2026-10-31T08:24:37.341Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Parasail",
      "accuracy": 0.8694082857142857,
      "accuracy_std_dev": 0.006664936667029211,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.008169142135642135,
      "avg_duration_per_task_ms": 374429.6991341991,
      "avg_output_tokens_per_task": 28832.51875901876,
      "last_run_timestamp": "2026-10-31T16:18:48.433Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.8690236000000001,
      "accuracy_std_dev": 0.0056741455250989255,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.004051724242424242,
      "avg_duration_per_task_ms": 38666.534343434345,
      "avg_output_tokens_per_task": 7872.538383838384,
      "last_run_timestamp": "2026-10-31T08:35:55.398Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "GMICloud",
      "accuracy": 0.8687761417721519,
      "accuracy_std_dev": 0.03267208970972012,
      "run_count": 8,
      "total_questions": 1580,
      "avg_cost_per_task_usd": 0.03197368734177215,
      "avg_duration_per_task_ms": 624455.1689873418,
      "avg_output_tokens_per_task": 38087.244303797466,
      "last_run_timestamp": "2026-11-10T07:01:38.700Z"
    },
    {
      "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": 0.868687,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.005480646464646464,
      "avg_duration_per_task_ms": 745039.691919192,
      "avg_output_tokens_per_task": 30273.38888888889,
      "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": "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": "z-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": 0.8681285334935416,
      "accuracy_std_dev": 0.03709653553275389,
      "run_count": 17,
      "total_questions": 3329,
      "avg_cost_per_task_usd": 0.01566810363472514,
      "avg_duration_per_task_ms": 433897.7128266747,
      "avg_output_tokens_per_task": 40485.41213577651,
      "last_run_timestamp": "2026-11-04T21:50:43.543Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "auto-routing",
      "accuracy": 0.8681253333333332,
      "accuracy_std_dev": 0.00965467639137749,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.021851893939393942,
      "avg_duration_per_task_ms": 524499.1397306398,
      "avg_output_tokens_per_task": 46276.513468013465,
      "last_run_timestamp": "2026-10-31T08:03:09.464Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "SambaNova",
      "accuracy": 0.8678450000000001,
      "accuracy_std_dev": 0.012168917536083462,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.02358163005050505,
      "avg_duration_per_task_ms": 164224.13383838383,
      "avg_output_tokens_per_task": 28381.034090909092,
      "last_run_timestamp": "2026-11-11T19:09:33.977Z"
    },
    {
      "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": 0.8676650854638421,
      "accuracy_std_dev": 0.009102300097295443,
      "run_count": 7,
      "total_questions": 1369,
      "avg_cost_per_task_usd": 0.07595047918188458,
      "avg_duration_per_task_ms": 297193.17823228636,
      "avg_output_tokens_per_task": 20961.7319211103,
      "last_run_timestamp": "2026-10-31T16:05:36.103Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "Novita",
      "accuracy": 0.8676153636363637,
      "accuracy_std_dev": 0.004014278793432891,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.0702307419651056,
      "avg_duration_per_task_ms": 231760.75160697888,
      "avg_output_tokens_per_task": 11437.757575757576,
      "last_run_timestamp": "2026-10-31T08:17:29.729Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "Google",
      "accuracy": 0.8675645333333333,
      "accuracy_std_dev": 0.00826267276262484,
      "run_count": 15,
      "total_questions": 2970,
      "avg_cost_per_task_usd": 0.004299042424242424,
      "avg_duration_per_task_ms": 36217.24208754209,
      "avg_output_tokens_per_task": 7829.657575757576,
      "last_run_timestamp": "2026-10-31T08:38:41.553Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-luna-20260709",
      "model_slug": "openai/gpt-5.6-luna",
      "model_name": "OpenAI: GPT-5.6 Luna",
      "provider_name": "OpenAI",
      "accuracy": 0.8675643333333334,
      "accuracy_std_dev": 0.0028614651647170092,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.009448984848484849,
      "avg_duration_per_task_ms": 79377.02525252526,
      "avg_output_tokens_per_task": 7744.79797979798,
      "last_run_timestamp": "2026-10-02T23:19:10.742Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Parasail",
      "accuracy": 0.867521,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 195,
      "avg_cost_per_task_usd": 0.08994571282051282,
      "avg_duration_per_task_ms": 589733.9384615384,
      "avg_output_tokens_per_task": 26825.066666666666,
      "last_run_timestamp": "2026-08-22T09:51:24.721Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "auto-routing",
      "accuracy": 0.8674240000000001,
      "accuracy_std_dev": 0.0048169925783625705,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.004064253787878788,
      "avg_duration_per_task_ms": 38624.37373737374,
      "avg_output_tokens_per_task": 7893.444444444444,
      "last_run_timestamp": "2026-10-31T08:27:45.642Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "Azure",
      "accuracy": 0.8671716000000002,
      "accuracy_std_dev": 0.006408346388890055,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.0041782994949494945,
      "avg_duration_per_task_ms": 35977.948484848486,
      "avg_output_tokens_per_task": 7724.74494949495,
      "last_run_timestamp": "2026-10-31T09:34:38.477Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "auto-routing",
      "accuracy": 0.867147968231047,
      "accuracy_std_dev": 0.04215195953321502,
      "run_count": 7,
      "total_questions": 1385,
      "avg_cost_per_task_usd": 0.16483890613718413,
      "avg_duration_per_task_ms": 954961.5436823105,
      "avg_output_tokens_per_task": 56075.11624548736,
      "last_run_timestamp": "2026-10-31T16:11:41.349Z"
    },
    {
      "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": 0.8670032857142856,
      "accuracy_std_dev": 0.007740900274594989,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.02579789898989899,
      "avg_duration_per_task_ms": 254191.68181818182,
      "avg_output_tokens_per_task": 21252.31313131313,
      "last_run_timestamp": "2026-10-08T22:14:16.087Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.1-20251113",
      "model_slug": "openai/gpt-5.1",
      "model_name": "OpenAI: GPT-5.1",
      "provider_name": "auto-routing",
      "accuracy": 0.867003,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.22265377272727274,
      "avg_duration_per_task_ms": 265427.9646464646,
      "avg_output_tokens_per_task": 22169.171717171717,
      "last_run_timestamp": "2026-10-01T16:12:54.241Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "auto-routing",
      "accuracy": 0.8668349000000002,
      "accuracy_std_dev": 0.02737903360401897,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.008139828787878788,
      "avg_duration_per_task_ms": 416338.91818181815,
      "avg_output_tokens_per_task": 26892.52676767677,
      "last_run_timestamp": "2026-10-31T08:17:29.008Z"
    },
    {
      "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": 0.8665222857142858,
      "accuracy_std_dev": 0.007726115189302267,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.10016765367965368,
      "avg_duration_per_task_ms": 312867.60533910536,
      "avg_output_tokens_per_task": 22150.73520923521,
      "last_run_timestamp": "2026-10-31T16:15:29.279Z"
    },
    {
      "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": 0.866391,
      "accuracy_std_dev": 0.031399399620780255,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.1546900968778696,
      "avg_duration_per_task_ms": 152674.89577594123,
      "avg_output_tokens_per_task": 15252.516069788797,
      "last_run_timestamp": "2026-10-31T08:27:53.385Z"
    },
    {
      "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": 0.866331762553802,
      "accuracy_std_dev": 0.010391402837802988,
      "run_count": 8,
      "total_questions": 1394,
      "avg_cost_per_task_usd": 0.05799772740315638,
      "avg_duration_per_task_ms": 435430.42898134864,
      "avg_output_tokens_per_task": 19221.758967001435,
      "last_run_timestamp": "2026-10-31T16:05:14.702Z"
    },
    {
      "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": 0.86633,
      "accuracy_std_dev": 0.01754735666133223,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.10058035606060606,
      "avg_duration_per_task_ms": 457394.16717171716,
      "avg_output_tokens_per_task": 26933.21414141414,
      "last_run_timestamp": "2026-10-31T16:03:21.454Z"
    },
    {
      "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": 0.8661615,
      "accuracy_std_dev": 0.005892500000000023,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.09752832323232324,
      "avg_duration_per_task_ms": 1426009.4393939395,
      "avg_output_tokens_per_task": 43223.868686868685,
      "last_run_timestamp": "2026-08-29T18:36:10.818Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.6-opus-20260205",
      "model_slug": "anthropic/claude-opus-4.6",
      "model_name": "Anthropic: Claude Opus 4.6",
      "provider_name": "auto-routing",
      "accuracy": 0.8657789090909093,
      "accuracy_std_dev": 0.014068643001399471,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.7534191322314051,
      "avg_duration_per_task_ms": 505384.1258034894,
      "avg_output_tokens_per_task": 29965.463269054177,
      "last_run_timestamp": "2026-10-31T08:23:25.857Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "thinkingmachines/inkling-small-20260730",
      "model_slug": "thinkingmachines/inkling-small",
      "model_name": "Thinking Machines: Inkling Small",
      "provider_name": "auto-routing",
      "accuracy": 0.8656005,
      "accuracy_std_dev": 0.00803956708515244,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.02525451262626263,
      "avg_duration_per_task_ms": 233486.3627946128,
      "avg_output_tokens_per_task": 20811.0765993266,
      "last_run_timestamp": "2026-10-01T16:08:04.290Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Morph",
      "accuracy": 0.86532,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.02509249494949495,
      "avg_duration_per_task_ms": 884843.8888888889,
      "avg_output_tokens_per_task": 25370.131313131315,
      "last_run_timestamp": "2026-08-22T10:03:34.994Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Relace",
      "accuracy": 0.86532,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.01495640909090909,
      "avg_duration_per_task_ms": 120441.0101010101,
      "avg_output_tokens_per_task": 29795.70707070707,
      "last_run_timestamp": "2026-10-24T15:00:02.228Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8653197142857143,
      "accuracy_std_dev": 0.018961344600561016,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.16908653823953823,
      "avg_duration_per_task_ms": 1092542.5418470418,
      "avg_output_tokens_per_task": 53671.56421356421,
      "last_run_timestamp": "2026-10-31T16:05:14.701Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "CoreWeave",
      "accuracy": 0.8653195,
      "accuracy_std_dev": 0.005050500000000013,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.008214638888888888,
      "avg_duration_per_task_ms": 536604.0606060605,
      "avg_output_tokens_per_task": 28981.24494949495,
      "last_run_timestamp": "2026-08-29T18:49:16.416Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "NextBit",
      "accuracy": 0.8653195,
      "accuracy_std_dev": 0.0016835000000000044,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.008587179292929294,
      "avg_duration_per_task_ms": 240899.30303030304,
      "avg_output_tokens_per_task": 30387.573232323233,
      "last_run_timestamp": "2026-08-29T18:38:15.907Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Cohere",
      "accuracy": 0.8647586666666667,
      "accuracy_std_dev": 0.008943734131906087,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.05048964983164983,
      "avg_duration_per_task_ms": 169537.7323232323,
      "avg_output_tokens_per_task": 41902.75589225589,
      "last_run_timestamp": "2026-10-31T08:15:03.392Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.5-opus-20251124",
      "model_slug": "anthropic/claude-opus-4.5",
      "model_name": "Anthropic: Claude Opus 4.5",
      "provider_name": "auto-routing",
      "accuracy": 0.8647076363636366,
      "accuracy_std_dev": 0.01038443121629445,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.8432962465564738,
      "avg_duration_per_task_ms": 409882.468778696,
      "avg_output_tokens_per_task": 33540.75895316804,
      "last_run_timestamp": "2026-10-31T08:13:57.628Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "InferenceNet",
      "accuracy": 0.864685303613644,
      "accuracy_std_dev": 0.016464492614107477,
      "run_count": 15,
      "total_questions": 2961,
      "avg_cost_per_task_usd": 0.08223853157716987,
      "avg_duration_per_task_ms": 179300.8497129348,
      "avg_output_tokens_per_task": 24017.497804795676,
      "last_run_timestamp": "2026-11-04T21:43:38.767Z"
    },
    {
      "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": 0.8645034242424242,
      "accuracy_std_dev": 0.03391179125947315,
      "run_count": 33,
      "total_questions": 6534,
      "avg_cost_per_task_usd": 0.1657385367309458,
      "avg_duration_per_task_ms": 152502.7294153658,
      "avg_output_tokens_per_task": 15317.695592286502,
      "last_run_timestamp": "2026-10-31T08:51:08.192Z"
    },
    {
      "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": 0.8642484545454546,
      "accuracy_std_dev": 0.007188163178747238,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.7696710606060606,
      "avg_duration_per_task_ms": 439496.80716253445,
      "avg_output_tokens_per_task": 30614.35123966942,
      "last_run_timestamp": "2026-10-31T08:15:24.290Z"
    },
    {
      "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": 0.8640248461538463,
      "accuracy_std_dev": 0.04767222403655736,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.11144916588966587,
      "avg_duration_per_task_ms": 473283.12665112666,
      "avg_output_tokens_per_task": 28048.23271173271,
      "last_run_timestamp": "2026-11-10T06:16:52.306Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Phala",
      "accuracy": 0.863852179419525,
      "accuracy_std_dev": 0.02754414592685712,
      "run_count": 10,
      "total_questions": 1895,
      "avg_cost_per_task_usd": 0.012207302902374668,
      "avg_duration_per_task_ms": 440151.53984168865,
      "avg_output_tokens_per_task": 28225.067018469657,
      "last_run_timestamp": "2026-10-31T08:36:30.903Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Relace",
      "accuracy": 0.8636363333333335,
      "accuracy_std_dev": 0.008473463964374614,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.056405106902356894,
      "avg_duration_per_task_ms": 429778.9991582492,
      "avg_output_tokens_per_task": 55743.61616161616,
      "last_run_timestamp": "2026-10-31T16:10:09.223Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Morph",
      "accuracy": 0.86321525,
      "accuracy_std_dev": 0.03865413810560907,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.0056033421717171725,
      "avg_duration_per_task_ms": 440745.6477272727,
      "avg_output_tokens_per_task": 17145.257575757576,
      "last_run_timestamp": "2026-11-05T09:00:05.070Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "Fireworks",
      "accuracy": 0.86321525,
      "accuracy_std_dev": 0.010127171825712258,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.022006194444444443,
      "avg_duration_per_task_ms": 265798.13636363635,
      "avg_output_tokens_per_task": 33108.22601010101,
      "last_run_timestamp": "2026-09-19T08:51:58.825Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "Alibaba",
      "accuracy": 0.8630751666666665,
      "accuracy_std_dev": 0.009048483766478349,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.0646409191919192,
      "avg_duration_per_task_ms": 213655.1308922559,
      "avg_output_tokens_per_task": 10505.142255892257,
      "last_run_timestamp": "2026-10-31T08:13:57.551Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.5-opus-20251124",
      "model_slug": "anthropic/claude-opus-4.5",
      "model_name": "Anthropic: Claude Opus 4.5",
      "provider_name": "Anthropic",
      "accuracy": 0.8630240000000001,
      "accuracy_std_dev": 0.009717986304692007,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.8468043273645547,
      "avg_duration_per_task_ms": 463572.84573002753,
      "avg_output_tokens_per_task": 33681.08218549128,
      "last_run_timestamp": "2026-10-31T08:07:51.358Z"
    },
    {
      "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": 0.8630240000000001,
      "accuracy_std_dev": 0.019570785500099618,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.7589622681359046,
      "avg_duration_per_task_ms": 512637.1014692378,
      "avg_output_tokens_per_task": 30220.091827364555,
      "last_run_timestamp": "2026-10-31T08:30:25.667Z"
    },
    {
      "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": 0.8629476002066115,
      "accuracy_std_dev": 0.031092890107547094,
      "run_count": 10,
      "total_questions": 1936,
      "avg_cost_per_task_usd": 0.13384704958677685,
      "avg_duration_per_task_ms": 479346.4189049587,
      "avg_output_tokens_per_task": 43979.465392561986,
      "last_run_timestamp": "2026-10-31T16:16:35.161Z"
    },
    {
      "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": 0.862870909090909,
      "accuracy_std_dev": 0.016717786165173394,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.7582240289256198,
      "avg_duration_per_task_ms": 509993.8374655647,
      "avg_output_tokens_per_task": 30186.383379247014,
      "last_run_timestamp": "2026-10-31T08:09:30.224Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "DigitalOcean",
      "accuracy": 0.8627943333333332,
      "accuracy_std_dev": 0.06508388926010969,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.1701148493265993,
      "avg_duration_per_task_ms": 2045429.0942760943,
      "avg_output_tokens_per_task": 60250.10101010101,
      "last_run_timestamp": "2026-10-24T16:14:24.978Z"
    },
    {
      "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": 0.8627287304492512,
      "accuracy_std_dev": 0.07377561754013702,
      "run_count": 7,
      "total_questions": 1202,
      "avg_cost_per_task_usd": 0.045114597337770374,
      "avg_duration_per_task_ms": 123217.97254575707,
      "avg_output_tokens_per_task": 27990.334442595675,
      "last_run_timestamp": "2026-10-31T16:08:08.538Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8624118181818182,
      "accuracy_std_dev": 0.010959488075454655,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.0702179696969697,
      "avg_duration_per_task_ms": 231211.1772268136,
      "avg_output_tokens_per_task": 11433.751147842057,
      "last_run_timestamp": "2026-10-31T08:01:54.465Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Together",
      "accuracy": 0.862393,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 195,
      "avg_cost_per_task_usd": 0.11643089743589745,
      "avg_duration_per_task_ms": 352801.641025641,
      "avg_output_tokens_per_task": 29357.646153846155,
      "last_run_timestamp": "2026-08-22T10:19:36.962Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "DigitalOcean",
      "accuracy": 0.8622894,
      "accuracy_std_dev": 0.02509268697130698,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.0055417101010101,
      "avg_duration_per_task_ms": 936355.7888888889,
      "avg_output_tokens_per_task": 29101.413131313133,
      "last_run_timestamp": "2026-10-31T16:46:54.774Z"
    },
    {
      "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": 0.8622588181818182,
      "accuracy_std_dev": 0.01689329864617211,
      "run_count": 22,
      "total_questions": 4356,
      "avg_cost_per_task_usd": 0.8046746588613406,
      "avg_duration_per_task_ms": 489059.52777777775,
      "avg_output_tokens_per_task": 30498.26652892562,
      "last_run_timestamp": "2026-10-31T09:25:22.406Z"
    },
    {
      "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": 0.8622526105675148,
      "accuracy_std_dev": 0.011754560721269014,
      "run_count": 8,
      "total_questions": 1533,
      "avg_cost_per_task_usd": 0.0046349673842139595,
      "avg_duration_per_task_ms": 249077.7371167645,
      "avg_output_tokens_per_task": 28899.352250489235,
      "last_run_timestamp": "2026-11-10T18:11:20.521Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "DigitalOcean",
      "accuracy": 0.8622376508158508,
      "accuracy_std_dev": 0.05418494311168491,
      "run_count": 11,
      "total_questions": 2145,
      "avg_cost_per_task_usd": 0.07540476596736596,
      "avg_duration_per_task_ms": 485054.5156177156,
      "avg_output_tokens_per_task": 20286.058275058276,
      "last_run_timestamp": "2026-10-31T08:40:20.956Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.8620293636363636,
      "accuracy_std_dev": 0.007276086157877692,
      "run_count": 22,
      "total_questions": 4356,
      "avg_cost_per_task_usd": 0.8840942208448116,
      "avg_duration_per_task_ms": 412737.8966942149,
      "avg_output_tokens_per_task": 33491.24127640037,
      "last_run_timestamp": "2026-10-31T08:24:08.173Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "OpenInference",
      "accuracy": 0.8620252278481012,
      "accuracy_std_dev": 0.030307500000000043,
      "run_count": 2,
      "total_questions": 395,
      "avg_cost_per_task_usd": 0.02300040253164557,
      "avg_duration_per_task_ms": 74866.74430379747,
      "avg_output_tokens_per_task": 18412.721518987342,
      "last_run_timestamp": "2026-10-17T16:05:00.316Z"
    },
    {
      "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": 0.86199412283237,
      "accuracy_std_dev": 0.01125442664474918,
      "run_count": 7,
      "total_questions": 1384,
      "avg_cost_per_task_usd": 0.03778334826589595,
      "avg_duration_per_task_ms": 556168.3388728324,
      "avg_output_tokens_per_task": 35081.37789017341,
      "last_run_timestamp": "2026-10-31T16:11:17.542Z"
    },
    {
      "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": 0.8618590854662809,
      "accuracy_std_dev": 0.09124950603648056,
      "run_count": 23,
      "total_questions": 4493,
      "avg_cost_per_task_usd": 0.008348970843534385,
      "avg_duration_per_task_ms": 561394.6131760516,
      "avg_output_tokens_per_task": 28451.505230358336,
      "last_run_timestamp": "2026-11-06T22:28:22.713Z"
    },
    {
      "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": 0.8614718571428571,
      "accuracy_std_dev": 0.00972174233985086,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.014086771284271283,
      "avg_duration_per_task_ms": 839496.9516594517,
      "avg_output_tokens_per_task": 28569.064935064936,
      "last_run_timestamp": "2026-10-31T16:17:15.719Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "InferenceNet",
      "accuracy": 0.8614316774441879,
      "accuracy_std_dev": 0.016000304270056075,
      "run_count": 8,
      "total_questions": 1299,
      "avg_cost_per_task_usd": 0.013207368745188604,
      "avg_duration_per_task_ms": 402522.6505003849,
      "avg_output_tokens_per_task": 23550.163202463435,
      "last_run_timestamp": "2026-11-04T21:18:56.861Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "CoreWeave",
      "accuracy": 0.8607304246575341,
      "accuracy_std_dev": 0.01148253802030522,
      "run_count": 7,
      "total_questions": 1314,
      "avg_cost_per_task_usd": 0.02474931582952816,
      "avg_duration_per_task_ms": 139190.8691019787,
      "avg_output_tokens_per_task": 37895.824200913245,
      "last_run_timestamp": "2026-10-31T08:41:38.334Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Venice",
      "accuracy": 0.860546054237288,
      "accuracy_std_dev": 0.01522905267434664,
      "run_count": 9,
      "total_questions": 1770,
      "avg_cost_per_task_usd": 0.1431118,
      "avg_duration_per_task_ms": 301601.77118644066,
      "avg_output_tokens_per_task": 30086.04463276836,
      "last_run_timestamp": "2026-10-31T09:06:29.131Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "reflection/beam-1.0-501b-a23b-20261016",
      "model_slug": "reflection/beam-1.0-501b-a23b",
      "model_name": "Reflection: Beam 1.0 501B A23B",
      "provider_name": "Io Net",
      "accuracy": 0.860269,
      "accuracy_std_dev": 0.010100999999999971,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.00987439393939394,
      "avg_duration_per_task_ms": 74831.28787878787,
      "avg_output_tokens_per_task": 9997.058080808081,
      "last_run_timestamp": "2026-10-24T08:14:36.264Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.860028857142857,
      "accuracy_std_dev": 0.027595884472811062,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.08208907575757575,
      "avg_duration_per_task_ms": 365641.4213564214,
      "avg_output_tokens_per_task": 23745.34992784993,
      "last_run_timestamp": "2026-10-31T16:12:07.201Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Fireworks",
      "accuracy": 0.8597679673992673,
      "accuracy_std_dev": 0.009900199512162502,
      "run_count": 17,
      "total_questions": 2730,
      "avg_cost_per_task_usd": 0.17633374102564103,
      "avg_duration_per_task_ms": 246920.36886446885,
      "avg_output_tokens_per_task": 32290.727472527473,
      "last_run_timestamp": "2026-09-26T08:54:22.466Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "Io Net",
      "accuracy": 0.8596746149425287,
      "accuracy_std_dev": 0.000831499999999985,
      "run_count": 2,
      "total_questions": 348,
      "avg_cost_per_task_usd": 0.0899417959770115,
      "avg_duration_per_task_ms": 1024745.4281609196,
      "avg_output_tokens_per_task": 35825.26724137931,
      "last_run_timestamp": "2026-10-27T13:57:40.548Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Novita",
      "accuracy": 0.8596166717241379,
      "accuracy_std_dev": 0.025707588739385667,
      "run_count": 11,
      "total_questions": 2175,
      "avg_cost_per_task_usd": 0.08281105701149424,
      "avg_duration_per_task_ms": 496103.58022988506,
      "avg_output_tokens_per_task": 41921.770574712646,
      "last_run_timestamp": "2026-10-31T16:30:10.027Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "DigitalOcean",
      "accuracy": 0.859429902017291,
      "accuracy_std_dev": 0.05098775585122179,
      "run_count": 11,
      "total_questions": 2082,
      "avg_cost_per_task_usd": 0.011825536023054757,
      "avg_duration_per_task_ms": 442610.81796349667,
      "avg_output_tokens_per_task": 25201.657060518733,
      "last_run_timestamp": "2026-10-31T08:27:46.434Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "DigitalOcean",
      "accuracy": 0.8593071428571429,
      "accuracy_std_dev": 0.07000414295174368,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.07085715151515151,
      "avg_duration_per_task_ms": 1036918.955988456,
      "avg_output_tokens_per_task": 34951.957431457435,
      "last_run_timestamp": "2026-10-31T16:05:14.716Z"
    },
    {
      "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": 0.8592592,
      "accuracy_std_dev": 0.008288623417673138,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.00619469898989899,
      "avg_duration_per_task_ms": 555649.1151515151,
      "avg_output_tokens_per_task": 32304.478787878787,
      "last_run_timestamp": "2026-10-31T16:16:55.696Z"
    },
    {
      "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": 0.859217,
      "accuracy_std_dev": 0.016791584253726632,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.09269683143939395,
      "avg_duration_per_task_ms": 184470.00315656565,
      "avg_output_tokens_per_task": 25619.8952020202,
      "last_run_timestamp": "2026-10-31T16:04:29.122Z"
    },
    {
      "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": 0.8588918181818183,
      "accuracy_std_dev": 0.009103185045188405,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.8537164830119376,
      "avg_duration_per_task_ms": 482176.7851239669,
      "avg_output_tokens_per_task": 33957.568411386594,
      "last_run_timestamp": "2026-10-31T08:48:00.920Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "DekaLLM",
      "accuracy": 0.8588278767123287,
      "accuracy_std_dev": 0.014585318196353787,
      "run_count": 3,
      "total_questions": 438,
      "avg_cost_per_task_usd": 0.005061143835616439,
      "avg_duration_per_task_ms": 414471.07305936073,
      "avg_output_tokens_per_task": 13714.753424657534,
      "last_run_timestamp": "2026-10-31T16:45:39.994Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Chutes",
      "accuracy": 0.8588262857142857,
      "accuracy_std_dev": 0.023627943787656442,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.13331038239538243,
      "avg_duration_per_task_ms": 1309203.440836941,
      "avg_output_tokens_per_task": 52545.645021645025,
      "last_run_timestamp": "2026-10-31T16:08:31.067Z"
    },
    {
      "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": 0.858586,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.009637575757575758,
      "avg_duration_per_task_ms": 172308.71717171717,
      "avg_output_tokens_per_task": 16318.166666666666,
      "last_run_timestamp": "2026-08-22T09:56:05.550Z"
    },
    {
      "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": "z-ai/glm-5.3-flashx-20260918",
      "model_slug": "z-ai/glm-5.3-flashx",
      "model_name": "Z.ai: GLM 5.3 FlashX",
      "provider_name": "Novita",
      "accuracy": 0.858586,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.019546777777777778,
      "avg_duration_per_task_ms": 118523.70707070707,
      "avg_output_tokens_per_task": 15401.424242424242,
      "last_run_timestamp": "2026-11-12T21:02:44.111Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.5-opus-20251124",
      "model_slug": "anthropic/claude-opus-4.5",
      "model_name": "Anthropic: Claude Opus 4.5",
      "provider_name": "Azure",
      "accuracy": 0.8584327272727273,
      "accuracy_std_dev": 0.01770708058217958,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.8375554407713499,
      "avg_duration_per_task_ms": 457074.21533516987,
      "avg_output_tokens_per_task": 33325.96694214876,
      "last_run_timestamp": "2026-10-31T08:40:44.982Z"
    },
    {
      "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": 0.8584054285714288,
      "accuracy_std_dev": 0.00621956903323561,
      "run_count": 14,
      "total_questions": 2772,
      "avg_cost_per_task_usd": 0.06424625072150071,
      "avg_duration_per_task_ms": 592730.4098124098,
      "avg_output_tokens_per_task": 35466.90873015873,
      "last_run_timestamp": "2026-10-31T16:08:56.835Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.5-opus-20251124",
      "model_slug": "anthropic/claude-opus-4.5",
      "model_name": "Anthropic: Claude Opus 4.5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.8581648333333333,
      "accuracy_std_dev": 0.025909089369798823,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.8336664709595959,
      "avg_duration_per_task_ms": 439138.7053872054,
      "avg_output_tokens_per_task": 33233.371632996634,
      "last_run_timestamp": "2026-10-31T08:35:35.377Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.8579124,
      "accuracy_std_dev": 0.017384854495796048,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.02057177575757576,
      "avg_duration_per_task_ms": 445221.4888888889,
      "avg_output_tokens_per_task": 40903.725252525255,
      "last_run_timestamp": "2026-10-31T08:25:16.899Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "Xiaomi",
      "accuracy": 0.8572763333333334,
      "accuracy_std_dev": 0.011384368786483787,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.030025383277216613,
      "avg_duration_per_task_ms": 811489.1139169473,
      "avg_output_tokens_per_task": 34239.19135802469,
      "last_run_timestamp": "2026-10-17T08:19:33.830Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "BaseTen",
      "accuracy": 0.857222111631923,
      "accuracy_std_dev": 0.02112164754592157,
      "run_count": 41,
      "total_questions": 8107,
      "avg_cost_per_task_usd": 0.29757759652152466,
      "avg_duration_per_task_ms": 345556.9242629826,
      "avg_output_tokens_per_task": 53533.547551498705,
      "last_run_timestamp": "2026-10-31T16:26:37.892Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Wafer",
      "accuracy": 0.8569025,
      "accuracy_std_dev": 0.005050499999999958,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.018898722222222223,
      "avg_duration_per_task_ms": 164757.63383838383,
      "avg_output_tokens_per_task": 31337.035353535353,
      "last_run_timestamp": "2026-10-24T08:17:56.393Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "GMICloud",
      "accuracy": 0.8569022222222222,
      "accuracy_std_dev": 0.02084630876346526,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.029755960157126823,
      "avg_duration_per_task_ms": 618931.7239057239,
      "avg_output_tokens_per_task": 34230.03367003367,
      "last_run_timestamp": "2026-11-10T07:06:13.165Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5-2025-08-07",
      "model_slug": "openai/gpt-5",
      "model_name": "OpenAI: GPT-5",
      "provider_name": "auto-routing",
      "accuracy": 0.8569019999999999,
      "accuracy_std_dev": 0.0033669999999999534,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.20989276262626264,
      "avg_duration_per_task_ms": 263850.72222222225,
      "avg_output_tokens_per_task": 20893.070707070707,
      "last_run_timestamp": "2026-10-01T16:31:36.833Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Near AI",
      "accuracy": 0.8568247201068804,
      "accuracy_std_dev": 0.015059304005690374,
      "run_count": 9,
      "total_questions": 1497,
      "avg_cost_per_task_usd": 0.004656902471609886,
      "avg_duration_per_task_ms": 413401.8376753507,
      "avg_output_tokens_per_task": 9906.163660654642,
      "last_run_timestamp": "2026-10-31T09:01:06.018Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "DigitalOcean",
      "accuracy": 0.8565399329113924,
      "accuracy_std_dev": 0.05737653598692915,
      "run_count": 8,
      "total_questions": 1580,
      "avg_cost_per_task_usd": 0.07123199113924049,
      "avg_duration_per_task_ms": 1681972.4069620252,
      "avg_output_tokens_per_task": 20457.24683544304,
      "last_run_timestamp": "2026-10-31T16:13:09.868Z"
    },
    {
      "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": "Baidu",
      "accuracy": 0.8561370454545454,
      "accuracy_std_dev": 0.0107404034056508,
      "run_count": 22,
      "total_questions": 4356,
      "avg_cost_per_task_usd": 0.15750592470156105,
      "avg_duration_per_task_ms": 443794.3243801653,
      "avg_output_tokens_per_task": 45571.12006427915,
      "last_run_timestamp": "2026-11-10T19:22:00.700Z"
    },
    {
      "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": 0.8560681935807422,
      "accuracy_std_dev": 0.011718959590718295,
      "run_count": 6,
      "total_questions": 997,
      "avg_cost_per_task_usd": 0.09006469408224674,
      "avg_duration_per_task_ms": 1284430.8184553662,
      "avg_output_tokens_per_task": 40831.585757271816,
      "last_run_timestamp": "2026-10-31T16:04:23.355Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "GMICloud",
      "accuracy": 0.8556465066354411,
      "accuracy_std_dev": 0.020339777161859043,
      "run_count": 13,
      "total_questions": 2562,
      "avg_cost_per_task_usd": 0.08425942115534739,
      "avg_duration_per_task_ms": 311656.0573770492,
      "avg_output_tokens_per_task": 25603.65261514442,
      "last_run_timestamp": "2026-11-10T06:48:32.349Z"
    },
    {
      "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": 0.8552316718446601,
      "accuracy_std_dev": 0.0035618211195257714,
      "run_count": 8,
      "total_questions": 1545,
      "avg_cost_per_task_usd": 0.0868813637540453,
      "avg_duration_per_task_ms": 692640.7055016181,
      "avg_output_tokens_per_task": 43299.1715210356,
      "last_run_timestamp": "2026-10-31T16:11:17.543Z"
    },
    {
      "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": 0.8550984285714284,
      "accuracy_std_dev": 0.023860950363405426,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.04428129942279942,
      "avg_duration_per_task_ms": 655320.3124098124,
      "avg_output_tokens_per_task": 36753.11616161616,
      "last_run_timestamp": "2026-10-31T16:02:14.360Z"
    },
    {
      "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": 0.8549381666666667,
      "accuracy_std_dev": 0.008098201002211657,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.015701257575757577,
      "avg_duration_per_task_ms": 717198.1986531987,
      "avg_output_tokens_per_task": 19145.29797979798,
      "last_run_timestamp": "2026-10-31T16:05:36.294Z"
    },
    {
      "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": 0.854737857142857,
      "accuracy_std_dev": 0.020612976983503603,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.04354637373737373,
      "avg_duration_per_task_ms": 599777.3571428572,
      "avg_output_tokens_per_task": 37914.28932178932,
      "last_run_timestamp": "2026-10-31T16:03:21.441Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Inceptron",
      "accuracy": 0.8545531474226805,
      "accuracy_std_dev": 0.018117679996070137,
      "run_count": 10,
      "total_questions": 1940,
      "avg_cost_per_task_usd": 0.08358154742268041,
      "avg_duration_per_task_ms": 325785.14793814434,
      "avg_output_tokens_per_task": 28959.811340206186,
      "last_run_timestamp": "2026-10-31T08:56:36.753Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Parasail",
      "accuracy": 0.85452,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 118,
      "avg_cost_per_task_usd": 0.11718616949152542,
      "avg_duration_per_task_ms": 289492.906779661,
      "avg_output_tokens_per_task": 26545.86440677966,
      "last_run_timestamp": "2026-08-22T17:29:49.714Z"
    },
    {
      "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": 0.8540377841483979,
      "accuracy_std_dev": 0.008466319992902585,
      "run_count": 9,
      "total_questions": 1779,
      "avg_cost_per_task_usd": 0.03240532996065205,
      "avg_duration_per_task_ms": 782505.4761101742,
      "avg_output_tokens_per_task": 33540.41765036537,
      "last_run_timestamp": "2026-10-17T08:19:39.242Z"
    },
    {
      "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": 0.8539581570952205,
      "accuracy_std_dev": 0.020366121093319547,
      "run_count": 14,
      "total_questions": 2699,
      "avg_cost_per_task_usd": 0.22091287662097076,
      "avg_duration_per_task_ms": 240855.18228973693,
      "avg_output_tokens_per_task": 44279.295294553536,
      "last_run_timestamp": "2026-10-24T16:29:51.093Z"
    },
    {
      "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": 0.8538717999999998,
      "accuracy_std_dev": 0.04946631113151657,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.005526387878787878,
      "avg_duration_per_task_ms": 190074.4404040404,
      "avg_output_tokens_per_task": 19384.131313131315,
      "last_run_timestamp": "2026-11-12T15:01:54.124Z"
    },
    {
      "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": 0.8537664327538884,
      "accuracy_std_dev": 0.01580166085394366,
      "run_count": 12,
      "total_questions": 2186,
      "avg_cost_per_task_usd": 0.04007263220494053,
      "avg_duration_per_task_ms": 147595.21180237876,
      "avg_output_tokens_per_task": 19176.248856358645,
      "last_run_timestamp": "2026-10-31T16:17:39.770Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "auto-routing",
      "accuracy": 0.8537037,
      "accuracy_std_dev": 0.012630753865466617,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.10325877878787879,
      "avg_duration_per_task_ms": 455294.6515151515,
      "avg_output_tokens_per_task": 31795.09494949495,
      "last_run_timestamp": "2026-10-31T08:16:11.390Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.6-opus-20260205",
      "model_slug": "anthropic/claude-opus-4.6",
      "model_name": "Anthropic: Claude Opus 4.6",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.8536755833333335,
      "accuracy_std_dev": 0.03297161670148616,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.7094076557239056,
      "avg_duration_per_task_ms": 485859.81018518517,
      "avg_output_tokens_per_task": 28361.97053872054,
      "last_run_timestamp": "2026-10-31T08:09:42.551Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "BaseTen",
      "accuracy": 0.853571253359394,
      "accuracy_std_dev": 0.031222921105250087,
      "run_count": 21,
      "total_questions": 4093,
      "avg_cost_per_task_usd": 0.003661041534326899,
      "avg_duration_per_task_ms": 73059.1441485463,
      "avg_output_tokens_per_task": 13652.101881260689,
      "last_run_timestamp": "2026-10-31T08:12:47.827Z"
    },
    {
      "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": 0.8533128573596358,
      "accuracy_std_dev": 0.012686455748159126,
      "run_count": 10,
      "total_questions": 1977,
      "avg_cost_per_task_usd": 0.08956905058168944,
      "avg_duration_per_task_ms": 291296.215477997,
      "avg_output_tokens_per_task": 29754.587253414265,
      "last_run_timestamp": "2026-11-10T18:31:05.774Z"
    },
    {
      "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": 0.8533057272727274,
      "accuracy_std_dev": 0.008106430938580997,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.09884094352617079,
      "avg_duration_per_task_ms": 553983.2718089991,
      "avg_output_tokens_per_task": 30406.843893480258,
      "last_run_timestamp": "2026-10-31T16:01:53.623Z"
    },
    {
      "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": 0.852923,
      "accuracy_std_dev": 0.04905195955496394,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.16991116437098255,
      "avg_duration_per_task_ms": 168270.75114784206,
      "avg_output_tokens_per_task": 16771.511019283746,
      "last_run_timestamp": "2026-10-31T08:12:45.468Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "DekaLLM",
      "accuracy": 0.8523164263322883,
      "accuracy_std_dev": 0.010034068893524688,
      "run_count": 5,
      "total_questions": 957,
      "avg_cost_per_task_usd": 0.08246928213166145,
      "avg_duration_per_task_ms": 272317.5339602926,
      "avg_output_tokens_per_task": 27908.60710553814,
      "last_run_timestamp": "2026-10-31T08:44:39.028Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Venice",
      "accuracy": 0.8521062280219781,
      "accuracy_std_dev": 0.006055500000000047,
      "run_count": 2,
      "total_questions": 364,
      "avg_cost_per_task_usd": 0.005501717032967033,
      "avg_duration_per_task_ms": 539158.467032967,
      "avg_output_tokens_per_task": 15491.851648351649,
      "last_run_timestamp": "2026-10-17T08:17:58.236Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Tenstorrent",
      "accuracy": 0.851852,
      "accuracy_std_dev": 0.005323694440893428,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.011784135101010102,
      "avg_duration_per_task_ms": 222866.32323232322,
      "avg_output_tokens_per_task": 29178.76893939394,
      "last_run_timestamp": "2026-11-04T18:07:25.864Z"
    },
    {
      "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": 0.8518518181818181,
      "accuracy_std_dev": 0.019763599225105198,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.1265775752984389,
      "avg_duration_per_task_ms": 369781.93434343435,
      "avg_output_tokens_per_task": 40176.337006427915,
      "last_run_timestamp": "2026-10-31T16:08:08.543Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5-2025-08-07",
      "model_slug": "openai/gpt-5",
      "model_name": "OpenAI: GPT-5",
      "provider_name": "Azure",
      "accuracy": 0.8518517499999999,
      "accuracy_std_dev": 0.0020620621201845155,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.21063902146464644,
      "avg_duration_per_task_ms": 251655.29545454544,
      "avg_output_tokens_per_task": 19971.159090909092,
      "last_run_timestamp": "2026-10-01T16:27:58.636Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "auto-routing",
      "accuracy": 0.8517752727272727,
      "accuracy_std_dev": 0.014241886054177125,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.1367359797979798,
      "avg_duration_per_task_ms": 439341.9412304867,
      "avg_output_tokens_per_task": 49949.342975206615,
      "last_run_timestamp": "2026-10-31T16:09:19.244Z"
    },
    {
      "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": 0.8517346009049774,
      "accuracy_std_dev": 0.012647581720936017,
      "run_count": 7,
      "total_questions": 1105,
      "avg_cost_per_task_usd": 0.00618648778280543,
      "avg_duration_per_task_ms": 347941.31402714935,
      "avg_output_tokens_per_task": 15241.635294117646,
      "last_run_timestamp": "2026-10-31T16:08:33.701Z"
    },
    {
      "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": 0.8517072126779972,
      "accuracy_std_dev": 0.008685929974427657,
      "run_count": 11,
      "total_questions": 2177,
      "avg_cost_per_task_usd": 0.03815233899862195,
      "avg_duration_per_task_ms": 381837.27468994027,
      "avg_output_tokens_per_task": 38692.63114377584,
      "last_run_timestamp": "2026-10-31T16:26:59.494Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Parasail",
      "accuracy": 0.8512906666666668,
      "accuracy_std_dev": 0.00555530025191155,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.017986696969696972,
      "avg_duration_per_task_ms": 129036.91750841751,
      "avg_output_tokens_per_task": 29778.104377104377,
      "last_run_timestamp": "2026-10-31T08:56:30.449Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-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": 0.8507849505644222,
      "accuracy_std_dev": 0.03243006205932389,
      "run_count": 13,
      "total_questions": 2569,
      "avg_cost_per_task_usd": 0.008747444920202414,
      "avg_duration_per_task_ms": 480929.24172829895,
      "avg_output_tokens_per_task": 31305.17399766446,
      "last_run_timestamp": "2026-11-10T08:10:11.354Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "BaseTen",
      "accuracy": 0.8502042599087539,
      "accuracy_std_dev": 0.01590746948888509,
      "run_count": 37,
      "total_questions": 7014,
      "avg_cost_per_task_usd": 0.16371538508696884,
      "avg_duration_per_task_ms": 188635.26946107784,
      "avg_output_tokens_per_task": 29341.152266894784,
      "last_run_timestamp": "2026-10-31T09:24:22.125Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Parasail",
      "accuracy": 0.850060552,
      "accuracy_std_dev": 0.01012123337849893,
      "run_count": 7,
      "total_questions": 1375,
      "avg_cost_per_task_usd": 0.0072516559999999996,
      "avg_duration_per_task_ms": 627511.1978181818,
      "avg_output_tokens_per_task": 17945.874909090908,
      "last_run_timestamp": "2026-10-31T16:12:28.235Z"
    },
    {
      "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": "River",
      "accuracy": 0.8498725725190839,
      "accuracy_std_dev": 0.009880999999999973,
      "run_count": 2,
      "total_questions": 393,
      "avg_cost_per_task_usd": 0.10012308905852417,
      "avg_duration_per_task_ms": 342870.14503816795,
      "avg_output_tokens_per_task": 28230.23409669211,
      "last_run_timestamp": "2026-11-12T21:00:32.103Z"
    },
    {
      "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": 0.8497887502109704,
      "accuracy_std_dev": 0.08999992173407093,
      "run_count": 6,
      "total_questions": 1185,
      "avg_cost_per_task_usd": 0.08140508607594936,
      "avg_duration_per_task_ms": 452408.74599156115,
      "avg_output_tokens_per_task": 20689.04641350211,
      "last_run_timestamp": "2026-10-31T16:16:30.410Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "StreamLake",
      "accuracy": 0.8497475000000001,
      "accuracy_std_dev": 0.008615114711946663,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.16274348863636365,
      "avg_duration_per_task_ms": 1045148.2323232323,
      "avg_output_tokens_per_task": 53462.20012626263,
      "last_run_timestamp": "2026-10-31T16:39:29.347Z"
    },
    {
      "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": "NextBit",
      "accuracy": 0.8495076506329113,
      "accuracy_std_dev": 0.005079115528864803,
      "run_count": 6,
      "total_questions": 1185,
      "avg_cost_per_task_usd": 0.007176422784810127,
      "avg_duration_per_task_ms": 1312452.7088607594,
      "avg_output_tokens_per_task": 17778.172151898732,
      "last_run_timestamp": "2026-11-02T17:22:59.212Z"
    },
    {
      "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": 0.8494951090909092,
      "accuracy_std_dev": 0.012579000000000007,
      "run_count": 2,
      "total_questions": 165,
      "avg_cost_per_task_usd": 0.005448739393939394,
      "avg_duration_per_task_ms": 236133.1696969697,
      "avg_output_tokens_per_task": 25659.842424242423,
      "last_run_timestamp": "2026-10-31T16:10:36.259Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Crusoe",
      "accuracy": 0.8493265000000001,
      "accuracy_std_dev": 0.000841499999999995,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.2185174267676768,
      "avg_duration_per_task_ms": 797991.7474747475,
      "avg_output_tokens_per_task": 49726.7398989899,
      "last_run_timestamp": "2026-08-29T18:53:23.077Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8492499090909091,
      "accuracy_std_dev": 0.017323134780636528,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.05284465794306703,
      "avg_duration_per_task_ms": 274246.5909090909,
      "avg_output_tokens_per_task": 45771.48347107438,
      "last_run_timestamp": "2026-10-31T16:53:23.847Z"
    },
    {
      "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": "SambaNova",
      "accuracy": 0.8491455076923077,
      "accuracy_std_dev": 0.009587999999999985,
      "run_count": 2,
      "total_questions": 390,
      "avg_cost_per_task_usd": 0.14742170256410256,
      "avg_duration_per_task_ms": 263938.93333333335,
      "avg_output_tokens_per_task": 33257.10769230769,
      "last_run_timestamp": "2026-11-11T19:11:45.316Z"
    },
    {
      "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": "StreamLake",
      "accuracy": 0.8491083950617284,
      "accuracy_std_dev": 0.03011880589939707,
      "run_count": 6,
      "total_questions": 486,
      "avg_cost_per_task_usd": 0.026545792181069956,
      "avg_duration_per_task_ms": 90656.26131687243,
      "avg_output_tokens_per_task": 8757.810699588477,
      "last_run_timestamp": "2026-11-02T17:24:29.725Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Fireworks",
      "accuracy": 0.8490562452830188,
      "accuracy_std_dev": 0.004459999999999964,
      "run_count": 2,
      "total_questions": 265,
      "avg_cost_per_task_usd": 0.017200528301886793,
      "avg_duration_per_task_ms": 165535.98490566039,
      "avg_output_tokens_per_task": 28515.603773584906,
      "last_run_timestamp": "2026-10-31T08:20:07.444Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.8489438181818183,
      "accuracy_std_dev": 0.010179603259086118,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.054512841597796145,
      "avg_duration_per_task_ms": 250309.3691460055,
      "avg_output_tokens_per_task": 23083.604683195594,
      "last_run_timestamp": "2026-10-31T16:08:08.536Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-sonnet-5-20260630",
      "model_slug": "anthropic/claude-sonnet-5",
      "model_name": "Anthropic: Claude Sonnet 5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.84890575,
      "accuracy_std_dev": 0.05035112340045156,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.1629483771043771,
      "avg_duration_per_task_ms": 162891.44065656565,
      "avg_output_tokens_per_task": 16117.493265993266,
      "last_run_timestamp": "2026-10-31T08:45:37.246Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "Venice",
      "accuracy": 0.8488202684365781,
      "accuracy_std_dev": 0.035164132211494975,
      "run_count": 4,
      "total_questions": 678,
      "avg_cost_per_task_usd": 0.027723622418879056,
      "avg_duration_per_task_ms": 605422.2300884955,
      "avg_output_tokens_per_task": 13768.485250737464,
      "last_run_timestamp": "2026-10-17T08:18:07.707Z"
    },
    {
      "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": 0.8487092666666665,
      "accuracy_std_dev": 0.01899568044746548,
      "run_count": 30,
      "total_questions": 5940,
      "avg_cost_per_task_usd": 0.24548661565656565,
      "avg_duration_per_task_ms": 502671.36026936024,
      "avg_output_tokens_per_task": 48693.465656565655,
      "last_run_timestamp": "2026-10-31T17:01:17.198Z"
    },
    {
      "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": 0.8485243124999999,
      "accuracy_std_dev": 0.026025999999999994,
      "run_count": 2,
      "total_questions": 384,
      "avg_cost_per_task_usd": 0.007128966145833333,
      "avg_duration_per_task_ms": 721849.125,
      "avg_output_tokens_per_task": 18412.1875,
      "last_run_timestamp": "2026-08-29T18:08:44.008Z"
    },
    {
      "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-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Sail Research",
      "accuracy": 0.8484845,
      "accuracy_std_dev": 0.005050500000000013,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.09501090909090909,
      "avg_duration_per_task_ms": 411875.595959596,
      "avg_output_tokens_per_task": 30402.136363636364,
      "last_run_timestamp": "2026-08-29T11:19:17.684Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Minimax",
      "accuracy": 0.848064,
      "accuracy_std_dev": 0.003834285135458768,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.05505487626262626,
      "avg_duration_per_task_ms": 405186.76767676766,
      "avg_output_tokens_per_task": 28115.5202020202,
      "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": "ModelRun",
      "accuracy": 0.8479167041666666,
      "accuracy_std_dev": 0.00828816902819917,
      "run_count": 5,
      "total_questions": 960,
      "avg_cost_per_task_usd": 0.16138468541666665,
      "avg_duration_per_task_ms": 277988.090625,
      "avg_output_tokens_per_task": 24274.81875,
      "last_run_timestamp": "2026-10-28T05:08:57.249Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8477888928571429,
      "accuracy_std_dev": 0.029215151613674717,
      "run_count": 10,
      "total_questions": 1960,
      "avg_cost_per_task_usd": 0.1592395025510204,
      "avg_duration_per_task_ms": 438119.2229591837,
      "avg_output_tokens_per_task": 36131.57244897959,
      "last_run_timestamp": "2026-10-31T08:13:34.182Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8475227142857141,
      "accuracy_std_dev": 0.03525429692033209,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.01139066955266955,
      "avg_duration_per_task_ms": 480834.867965368,
      "avg_output_tokens_per_task": 40336.09018759019,
      "last_run_timestamp": "2026-10-31T16:04:29.124Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Alibaba",
      "accuracy": 0.8475027499999999,
      "accuracy_std_dev": 0.011176715424824044,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.0940345744949495,
      "avg_duration_per_task_ms": 518383.0500841751,
      "avg_output_tokens_per_task": 34690.1148989899,
      "last_run_timestamp": "2026-10-31T16:05:36.297Z"
    },
    {
      "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": 0.8471073636363637,
      "accuracy_std_dev": 0.010373179378243821,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.019450327364554636,
      "avg_duration_per_task_ms": 145903.90633608814,
      "avg_output_tokens_per_task": 18981.601469237834,
      "last_run_timestamp": "2026-10-31T16:05:14.635Z"
    },
    {
      "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": 0.8469436303724929,
      "accuracy_std_dev": 0.009207938453420692,
      "run_count": 11,
      "total_questions": 2094,
      "avg_cost_per_task_usd": 0.06545764565425023,
      "avg_duration_per_task_ms": 187651.79751671443,
      "avg_output_tokens_per_task": 20359.965616045843,
      "last_run_timestamp": "2026-10-31T16:08:31.065Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Inceptron",
      "accuracy": 0.8469361045003815,
      "accuracy_std_dev": 0.04231320024956013,
      "run_count": 8,
      "total_questions": 1311,
      "avg_cost_per_task_usd": 0.02445259496567506,
      "avg_duration_per_task_ms": 437957.52021357743,
      "avg_output_tokens_per_task": 38826.28070175439,
      "last_run_timestamp": "2026-10-31T08:55:46.529Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Nebius",
      "accuracy": 0.8468582586644126,
      "accuracy_std_dev": 0.013598361278927057,
      "run_count": 6,
      "total_questions": 1183,
      "avg_cost_per_task_usd": 0.26757311665257816,
      "avg_duration_per_task_ms": 498008.1420118343,
      "avg_output_tokens_per_task": 60614.17836010144,
      "last_run_timestamp": "2026-10-31T16:45:45.621Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Crusoe",
      "accuracy": 0.8466142222222222,
      "accuracy_std_dev": 0.010836379374616662,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.14184849326599328,
      "avg_duration_per_task_ms": 144926.00224466893,
      "avg_output_tokens_per_task": 32057.991582491584,
      "last_run_timestamp": "2026-10-31T08:40:49.272Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Phala",
      "accuracy": 0.8465387902641119,
      "accuracy_std_dev": 0.025032600265254102,
      "run_count": 10,
      "total_questions": 1931,
      "avg_cost_per_task_usd": 0.10684122164681513,
      "avg_duration_per_task_ms": 489966.58933195233,
      "avg_output_tokens_per_task": 34902.65044018643,
      "last_run_timestamp": "2026-10-31T09:15:06.724Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "KT",
      "accuracy": 0.8459595,
      "accuracy_std_dev": 0.002525500000000014,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.006682772727272727,
      "avg_duration_per_task_ms": 172002.25757575757,
      "avg_output_tokens_per_task": 19423.946969696968,
      "last_run_timestamp": "2026-11-11T01:47:13.395Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Novita",
      "accuracy": 0.8459293152073732,
      "accuracy_std_dev": 0.017246738162321286,
      "run_count": 11,
      "total_questions": 2170,
      "avg_cost_per_task_usd": 0.010452895391705068,
      "avg_duration_per_task_ms": 454110.96359447006,
      "avg_output_tokens_per_task": 34047.99170506912,
      "last_run_timestamp": "2026-10-31T08:27:51.327Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Fireworks",
      "accuracy": 0.8456636089524402,
      "accuracy_std_dev": 0.018418318454142355,
      "run_count": 17,
      "total_questions": 3217,
      "avg_cost_per_task_usd": 0.14255490332608023,
      "avg_duration_per_task_ms": 266408.3546782717,
      "avg_output_tokens_per_task": 28803.044140503574,
      "last_run_timestamp": "2026-10-31T08:33:10.900Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-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": 0.8451602030567684,
      "accuracy_std_dev": 0.02474124360172708,
      "run_count": 5,
      "total_questions": 916,
      "avg_cost_per_task_usd": 0.028169002183406112,
      "avg_duration_per_task_ms": 430146.16484716156,
      "avg_output_tokens_per_task": 38733.88864628821,
      "last_run_timestamp": "2026-11-11T15:01:28.360Z"
    },
    {
      "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": 0.8451178333333333,
      "accuracy_std_dev": 0.04928392495671267,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.09621632239057239,
      "avg_duration_per_task_ms": 1254499.6717171718,
      "avg_output_tokens_per_task": 37856.02356902357,
      "last_run_timestamp": "2026-10-31T16:01:53.650Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8446468343465044,
      "accuracy_std_dev": 0.01783018725336332,
      "run_count": 10,
      "total_questions": 1974,
      "avg_cost_per_task_usd": 0.0660811013171226,
      "avg_duration_per_task_ms": 245353.75430597772,
      "avg_output_tokens_per_task": 20336.286220871327,
      "last_run_timestamp": "2026-10-31T08:32:53.731Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "StreamLake",
      "accuracy": 0.8444862499999999,
      "accuracy_std_dev": 0.011931925502093117,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.13937785227272728,
      "avg_duration_per_task_ms": 872540.9444444445,
      "avg_output_tokens_per_task": 55302.43308080808,
      "last_run_timestamp": "2026-10-31T16:10:36.259Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Modal",
      "accuracy": 0.844322,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 182,
      "avg_cost_per_task_usd": 0.0016337307692307693,
      "avg_duration_per_task_ms": 52884.66483516483,
      "avg_output_tokens_per_task": 15097.615384615385,
      "last_run_timestamp": "2026-11-04T09:00:10.020Z"
    },
    {
      "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": 0.844303929113924,
      "accuracy_std_dev": 0.0093105,
      "run_count": 2,
      "total_questions": 395,
      "avg_cost_per_task_usd": 0.008944994936708861,
      "avg_duration_per_task_ms": 480112.0911392405,
      "avg_output_tokens_per_task": 19303.701265822783,
      "last_run_timestamp": "2026-08-29T19:08:46.030Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.5-flash-lite-20260721",
      "model_slug": "google/gemini-3.5-flash-lite",
      "model_name": "Google: Gemini 3.5 Flash Lite",
      "provider_name": "auto-routing",
      "accuracy": 0.844276,
      "accuracy_std_dev": 0.009259000000000017,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.036391522727272724,
      "avg_duration_per_task_ms": 53072.40404040404,
      "avg_output_tokens_per_task": 14468.015151515152,
      "last_run_timestamp": "2026-10-01T19:12:26.032Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Novita",
      "accuracy": 0.844032090487239,
      "accuracy_std_dev": 0.01125640617422735,
      "run_count": 7,
      "total_questions": 1293,
      "avg_cost_per_task_usd": 0.007000730858468677,
      "avg_duration_per_task_ms": 651373.6117556071,
      "avg_output_tokens_per_task": 17260.92807424594,
      "last_run_timestamp": "2026-10-31T16:34:27.227Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.8439077604166666,
      "accuracy_std_dev": 0.009070260174817954,
      "run_count": 12,
      "total_questions": 2112,
      "avg_cost_per_task_usd": 0.12450989630681818,
      "avg_duration_per_task_ms": 153233.97348484848,
      "avg_output_tokens_per_task": 31747.522727272728,
      "last_run_timestamp": "2026-10-31T08:45:40.383Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Cohere",
      "accuracy": 0.843434,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.11786523232323233,
      "avg_duration_per_task_ms": 258487.5,
      "avg_output_tokens_per_task": 33493.90909090909,
      "last_run_timestamp": "2026-11-13T00:01:54.511Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "thinkingmachines/inkling-20260715",
      "model_slug": "thinkingmachines/inkling",
      "model_name": "Thinking Machines: Inkling",
      "provider_name": "DeepInfra",
      "accuracy": 0.8434174840336134,
      "accuracy_std_dev": 0.0058964576706039785,
      "run_count": 7,
      "total_questions": 1190,
      "avg_cost_per_task_usd": 0.07875789159663865,
      "avg_duration_per_task_ms": 236163.5705882353,
      "avg_output_tokens_per_task": 19324.579831932773,
      "last_run_timestamp": "2026-10-08T22:17:54.321Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Relace",
      "accuracy": 0.8433784397283531,
      "accuracy_std_dev": 0.003767029717016947,
      "run_count": 6,
      "total_questions": 1178,
      "avg_cost_per_task_usd": 0.1209133650254669,
      "avg_duration_per_task_ms": 239041.5050933786,
      "avg_output_tokens_per_task": 32072.49235993209,
      "last_run_timestamp": "2026-10-31T09:12:37.593Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "DeepInfra",
      "accuracy": 0.8432478707692308,
      "accuracy_std_dev": 0.040161275624730285,
      "run_count": 6,
      "total_questions": 975,
      "avg_cost_per_task_usd": 0.045744206153846154,
      "avg_duration_per_task_ms": 765827.1415384616,
      "avg_output_tokens_per_task": 33817.81538461539,
      "last_run_timestamp": "2026-09-26T09:56:00.145Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Venice",
      "accuracy": 0.8429531428571428,
      "accuracy_std_dev": 0.006641024607448381,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.005697821789321789,
      "avg_duration_per_task_ms": 491338.5937950938,
      "avg_output_tokens_per_task": 26035.01370851371,
      "last_run_timestamp": "2026-10-31T16:45:40.019Z"
    },
    {
      "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": 0.8428716426914153,
      "accuracy_std_dev": 0.019938122355458276,
      "run_count": 7,
      "total_questions": 1293,
      "avg_cost_per_task_usd": 0.006783645784996132,
      "avg_duration_per_task_ms": 813203.3232791957,
      "avg_output_tokens_per_task": 18142.77648878577,
      "last_run_timestamp": "2026-10-31T16:04:29.121Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "Darkbloom",
      "accuracy": 0.842743368078176,
      "accuracy_std_dev": 0.03194672742801677,
      "run_count": 5,
      "total_questions": 921,
      "avg_cost_per_task_usd": 0.011946617806731813,
      "avg_duration_per_task_ms": 1590074.9272529858,
      "avg_output_tokens_per_task": 34007.04125950054,
      "last_run_timestamp": "2026-11-13T03:05:40.266Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "StreamLake",
      "accuracy": 0.8426947486187845,
      "accuracy_std_dev": 0.027181441039645704,
      "run_count": 11,
      "total_questions": 2172,
      "avg_cost_per_task_usd": 0.01345692311233886,
      "avg_duration_per_task_ms": 682984.7329650092,
      "avg_output_tokens_per_task": 42267.71869244936,
      "last_run_timestamp": "2026-10-31T08:57:00.536Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8425925,
      "accuracy_std_dev": 0.005892500000000023,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.08320908333333334,
      "avg_duration_per_task_ms": 696703.1691919192,
      "avg_output_tokens_per_task": 44847.378787878784,
      "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": "Together",
      "accuracy": 0.8425437868421053,
      "accuracy_std_dev": 0.00573439306339873,
      "run_count": 6,
      "total_questions": 1140,
      "avg_cost_per_task_usd": 0.08756267807017544,
      "avg_duration_per_task_ms": 282086.36140350875,
      "avg_output_tokens_per_task": 21480.616666666665,
      "last_run_timestamp": "2026-10-01T16:06:12.801Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Together",
      "accuracy": 0.842516,
      "accuracy_std_dev": 0.010871595667777411,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.2414962369146006,
      "avg_duration_per_task_ms": 281047.5523415978,
      "avg_output_tokens_per_task": 54665.6129476584,
      "last_run_timestamp": "2026-10-31T16:57:18.670Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Decart",
      "accuracy": 0.8422337612565445,
      "accuracy_std_dev": 0.023212455418589382,
      "run_count": 10,
      "total_questions": 1910,
      "avg_cost_per_task_usd": 0.10300291204188483,
      "avg_duration_per_task_ms": 203070.60157068062,
      "avg_output_tokens_per_task": 30819.346073298428,
      "last_run_timestamp": "2026-10-31T08:52:06.333Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-122b-a10b-20260224",
      "model_slug": "qwen/qwen3.5-122b-a10b",
      "model_name": "Qwen: Qwen3.5-122B-A10B",
      "provider_name": "DeepInfra",
      "accuracy": 0.8420896313446128,
      "accuracy_std_dev": 0.007016167749483269,
      "run_count": 6,
      "total_questions": 1123,
      "avg_cost_per_task_usd": 0.05385980587711488,
      "avg_duration_per_task_ms": 265038.4915405165,
      "avg_output_tokens_per_task": 22350.218165627783,
      "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-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Together",
      "accuracy": 0.842080958563536,
      "accuracy_std_dev": 0.01924201603413352,
      "run_count": 11,
      "total_questions": 2172,
      "avg_cost_per_task_usd": 0.02133896546961326,
      "avg_duration_per_task_ms": 337269.2900552486,
      "avg_output_tokens_per_task": 42459.25598526704,
      "last_run_timestamp": "2026-10-31T08:44:51.872Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Mistral",
      "accuracy": 0.8420313611111111,
      "accuracy_std_dev": 0.011106999512701637,
      "run_count": 36,
      "total_questions": 7128,
      "avg_cost_per_task_usd": 0.15133402413019081,
      "avg_duration_per_task_ms": 168852.7923681257,
      "avg_output_tokens_per_task": 33127.37401795735,
      "last_run_timestamp": "2026-11-11T03:10:59.273Z"
    },
    {
      "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": 0.841751,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.15673991919191918,
      "avg_duration_per_task_ms": 1499085.4595959596,
      "avg_output_tokens_per_task": 52100.893939393936,
      "last_run_timestamp": "2026-08-29T18:25:52.753Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.4-mini-20260317",
      "model_slug": "openai/gpt-5.4-mini",
      "model_name": "OpenAI: GPT-5.4 Mini",
      "provider_name": "OpenAI",
      "accuracy": 0.841751,
      "accuracy_std_dev": 0.004761657064510223,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.06159810101010101,
      "avg_duration_per_task_ms": 79946.62794612795,
      "avg_output_tokens_per_task": 13559.319865319865,
      "last_run_timestamp": "2026-10-02T23:24:30.140Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Makora",
      "accuracy": 0.8416364037363778,
      "accuracy_std_dev": 0.015589563439410502,
      "run_count": 11,
      "total_questions": 1927,
      "avg_cost_per_task_usd": 0.1375403404255319,
      "avg_duration_per_task_ms": 272709.34509600414,
      "avg_output_tokens_per_task": 32850.36170212766,
      "last_run_timestamp": "2026-10-31T09:42:32.959Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8414446363636365,
      "accuracy_std_dev": 0.019030622147536685,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.13825922910927457,
      "avg_duration_per_task_ms": 553721.6345270891,
      "avg_output_tokens_per_task": 44693.3769513315,
      "last_run_timestamp": "2026-10-31T16:18:28.926Z"
    },
    {
      "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": "PrimeIntellect",
      "accuracy": 0.841228,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 190,
      "avg_cost_per_task_usd": 0.0038195157894736842,
      "avg_duration_per_task_ms": 400775.6,
      "avg_output_tokens_per_task": 17024.936842105264,
      "last_run_timestamp": "2026-11-01T18:00:07.350Z"
    },
    {
      "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": 0.8412182282003711,
      "accuracy_std_dev": 0.02921932152413157,
      "run_count": 11,
      "total_questions": 2156,
      "avg_cost_per_task_usd": 0.06684410575139148,
      "avg_duration_per_task_ms": 1133131.4666048237,
      "avg_output_tokens_per_task": 33323.142393320966,
      "last_run_timestamp": "2026-10-31T16:05:36.103Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Alibaba",
      "accuracy": 0.8409620247933886,
      "accuracy_std_dev": 0.0474612402328205,
      "run_count": 8,
      "total_questions": 1573,
      "avg_cost_per_task_usd": 0.19201579465988555,
      "avg_duration_per_task_ms": 598063.229497775,
      "avg_output_tokens_per_task": 45700.98919262556,
      "last_run_timestamp": "2026-10-31T16:25:21.093Z"
    },
    {
      "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": 0.8409089999999999,
      "accuracy_std_dev": 0.045212357351171156,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.0038926938131313127,
      "avg_duration_per_task_ms": 564281.6047979798,
      "avg_output_tokens_per_task": 43083.319444444445,
      "last_run_timestamp": "2026-10-31T16:18:02.140Z"
    },
    {
      "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 AI Studio",
      "accuracy": 0.8409089999999999,
      "accuracy_std_dev": 0.010943000000000036,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.03708839646464646,
      "avg_duration_per_task_ms": 47515.843434343435,
      "avg_output_tokens_per_task": 14746.045454545454,
      "last_run_timestamp": "2026-10-01T19:13:29.958Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Novita",
      "accuracy": 0.8409089999999999,
      "accuracy_std_dev": 0.0008420000000000094,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.025631659090909094,
      "avg_duration_per_task_ms": 236135.89141414143,
      "avg_output_tokens_per_task": 21303.184343434343,
      "last_run_timestamp": "2026-08-29T18:12:43.943Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Fireworks",
      "accuracy": 0.840891889212828,
      "accuracy_std_dev": 0.01539818649719325,
      "run_count": 16,
      "total_questions": 3087,
      "avg_cost_per_task_usd": 0.022577127632005183,
      "avg_duration_per_task_ms": 250189.5425979916,
      "avg_output_tokens_per_task": 38449.29964366699,
      "last_run_timestamp": "2026-10-31T08:54:15.638Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Together",
      "accuracy": 0.8408359820854386,
      "accuracy_std_dev": 0.011558364030339736,
      "run_count": 11,
      "total_questions": 2177,
      "avg_cost_per_task_usd": 0.14478948920532844,
      "avg_duration_per_task_ms": 203069.29581993568,
      "avg_output_tokens_per_task": 32673.29122645843,
      "last_run_timestamp": "2026-10-31T09:21:48.528Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.8407407,
      "accuracy_std_dev": 0.024165205569371837,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.013536882323232325,
      "avg_duration_per_task_ms": 528967.1176767676,
      "avg_output_tokens_per_task": 40146.823737373736,
      "last_run_timestamp": "2026-10-31T08:45:25.206Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Cloudflare",
      "accuracy": 0.8406767200863932,
      "accuracy_std_dev": 0.03362626844282443,
      "run_count": 12,
      "total_questions": 2315,
      "avg_cost_per_task_usd": 0.030011868250539957,
      "avg_duration_per_task_ms": 286135.6241900648,
      "avg_output_tokens_per_task": 9270.458315334774,
      "last_run_timestamp": "2026-10-31T08:16:12.311Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8406685714285715,
      "accuracy_std_dev": 0.01850043854968342,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.02123720418470418,
      "avg_duration_per_task_ms": 436896.4862914863,
      "avg_output_tokens_per_task": 42244.7886002886,
      "last_run_timestamp": "2026-10-31T08:09:30.600Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "SambaNova",
      "accuracy": 0.8406374175298806,
      "accuracy_std_dev": 0.019535919283063564,
      "run_count": 8,
      "total_questions": 1255,
      "avg_cost_per_task_usd": 0.019024535458167333,
      "avg_duration_per_task_ms": 105363.14422310756,
      "avg_output_tokens_per_task": 16342.26374501992,
      "last_run_timestamp": "2026-11-11T18:58:39.610Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "Venice",
      "accuracy": 0.8404879999999999,
      "accuracy_std_dev": 0.020003728864889173,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.037026556818181816,
      "avg_duration_per_task_ms": 113840.68308080808,
      "avg_output_tokens_per_task": 34919.68308080808,
      "last_run_timestamp": "2026-10-31T09:18:03.023Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-122b-a10b-20260224",
      "model_slug": "qwen/qwen3.5-122b-a10b",
      "model_name": "Qwen: Qwen3.5-122B-A10B",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8403734545454546,
      "accuracy_std_dev": 0.034018740226698145,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.052804829201101924,
      "avg_duration_per_task_ms": 204681.84710743802,
      "avg_output_tokens_per_task": 21900.736455463728,
      "last_run_timestamp": "2026-10-31T16:28:22.255Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Sail Research",
      "accuracy": 0.8403390039479797,
      "accuracy_std_dev": 0.012076015641029829,
      "run_count": 22,
      "total_questions": 4306,
      "avg_cost_per_task_usd": 0.10892927728750582,
      "avg_duration_per_task_ms": 343834.3037621923,
      "avg_output_tokens_per_task": 34310.84417092429,
      "last_run_timestamp": "2026-11-06T23:05:09.500Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Reka",
      "accuracy": 0.8401781603650587,
      "accuracy_std_dev": 0.014388231606073072,
      "run_count": 16,
      "total_questions": 3068,
      "avg_cost_per_task_usd": 0.07506749576271186,
      "avg_duration_per_task_ms": 388300.5521512386,
      "avg_output_tokens_per_task": 22304.916558018253,
      "last_run_timestamp": "2026-10-31T08:43:48.577Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "Novita",
      "accuracy": 0.8400672857142857,
      "accuracy_std_dev": 0.010723216824312801,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.09338691269841272,
      "avg_duration_per_task_ms": 1164264.9783549784,
      "avg_output_tokens_per_task": 47107.84848484849,
      "last_run_timestamp": "2026-10-31T16:33:40.856Z"
    },
    {
      "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": 0.8398570000000002,
      "accuracy_std_dev": 0.010883944758220736,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.21990282449494947,
      "avg_duration_per_task_ms": 317333.1452020202,
      "avg_output_tokens_per_task": 49811.11111111111,
      "last_run_timestamp": "2026-09-12T08:53:53.147Z"
    },
    {
      "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": 0.8398265714285714,
      "accuracy_std_dev": 0.008557108588054464,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.03983322727272727,
      "avg_duration_per_task_ms": 52955.70346320346,
      "avg_output_tokens_per_task": 14785.20634920635,
      "last_run_timestamp": "2026-10-01T19:13:57.987Z"
    },
    {
      "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": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8394550909090909,
      "accuracy_std_dev": 0.011310932454565176,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.0878581368227732,
      "avg_duration_per_task_ms": 712685.4439853076,
      "avg_output_tokens_per_task": 50105.391184573,
      "last_run_timestamp": "2026-10-31T16:46:10.481Z"
    },
    {
      "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": 0.8392255000000001,
      "accuracy_std_dev": 0.002525500000000014,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.028252058080808082,
      "avg_duration_per_task_ms": 430220.3787878788,
      "avg_output_tokens_per_task": 23440.575757575756,
      "last_run_timestamp": "2026-08-29T18:16:26.161Z"
    },
    {
      "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": 0.8391051428571429,
      "accuracy_std_dev": 0.00943165809432062,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.150706683982684,
      "avg_duration_per_task_ms": 316554.5598845599,
      "avg_output_tokens_per_task": 62067.65873015873,
      "last_run_timestamp": "2026-10-31T16:45:48.945Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Mistral",
      "accuracy": 0.8389995344074279,
      "accuracy_std_dev": 0.011455015921374732,
      "run_count": 37,
      "total_questions": 7324,
      "avg_cost_per_task_usd": 0.26465632960131075,
      "avg_duration_per_task_ms": 285335.27007099945,
      "avg_output_tokens_per_task": 57439.03536318951,
      "last_run_timestamp": "2026-11-11T03:18:20.413Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "StreamLake",
      "accuracy": 0.8387440275454993,
      "accuracy_std_dev": 0.019690853286246784,
      "run_count": 11,
      "total_questions": 2033,
      "avg_cost_per_task_usd": 0.07400631874077719,
      "avg_duration_per_task_ms": 403244.14658140676,
      "avg_output_tokens_per_task": 24425.756025577964,
      "last_run_timestamp": "2026-10-31T16:18:04.837Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Novita",
      "accuracy": 0.8386898181818183,
      "accuracy_std_dev": 0.022320469186247312,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.1030642401285583,
      "avg_duration_per_task_ms": 384608.90220385673,
      "avg_output_tokens_per_task": 35835.99357208448,
      "last_run_timestamp": "2026-10-31T08:36:48.545Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Cloudflare",
      "accuracy": 0.8385089339622642,
      "accuracy_std_dev": 0.013106640774000717,
      "run_count": 13,
      "total_questions": 2544,
      "avg_cost_per_task_usd": 0.3152681855345912,
      "avg_duration_per_task_ms": 413001.90290880506,
      "avg_output_tokens_per_task": 53140.55856918239,
      "last_run_timestamp": "2026-10-31T16:56:08.871Z"
    },
    {
      "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": 0.8383837500000001,
      "accuracy_std_dev": 0.0064105770167357185,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.06462981691919192,
      "avg_duration_per_task_ms": 108153.94696969698,
      "avg_output_tokens_per_task": 13551.515151515152,
      "last_run_timestamp": "2026-10-01T19:14:05.661Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Nebius",
      "accuracy": 0.8383837142857142,
      "accuracy_std_dev": 0.012139811903865117,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.1527254054834055,
      "avg_duration_per_task_ms": 220532.05050505052,
      "avg_output_tokens_per_task": 34470.08658008658,
      "last_run_timestamp": "2026-10-31T08:45:52.318Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "auto-routing",
      "accuracy": 0.8383140372413792,
      "accuracy_std_dev": 0.03002615781240096,
      "run_count": 11,
      "total_questions": 2175,
      "avg_cost_per_task_usd": 0.08971148505747126,
      "avg_duration_per_task_ms": 771909.1783908046,
      "avg_output_tokens_per_task": 31778.30390804598,
      "last_run_timestamp": "2026-10-31T16:01:53.624Z"
    },
    {
      "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": 0.8380034220665499,
      "accuracy_std_dev": 0.010074705190062243,
      "run_count": 3,
      "total_questions": 571,
      "avg_cost_per_task_usd": 0.01617205078809107,
      "avg_duration_per_task_ms": 843918.8598949212,
      "avg_output_tokens_per_task": 17113.875656742555,
      "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": "InferenceNet",
      "accuracy": 0.8378226666666666,
      "accuracy_std_dev": 0.00975199857579064,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.15828378956228956,
      "avg_duration_per_task_ms": 446608.7861952862,
      "avg_output_tokens_per_task": 56298.93097643097,
      "last_run_timestamp": "2026-10-10T08:30:28.864Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8375420000000001,
      "accuracy_std_dev": 0.007576000000000027,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.028651648989898987,
      "avg_duration_per_task_ms": 499930.6313131313,
      "avg_output_tokens_per_task": 23687.20202020202,
      "last_run_timestamp": "2026-08-29T18:16:39.840Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.4-mini-20260317",
      "model_slug": "openai/gpt-5.4-mini",
      "model_name": "OpenAI: GPT-5.4 Mini",
      "provider_name": "auto-routing",
      "accuracy": 0.837542,
      "accuracy_std_dev": 0.005892000000000008,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.06186689898989899,
      "avg_duration_per_task_ms": 75089.61363636363,
      "avg_output_tokens_per_task": 13620.507575757576,
      "last_run_timestamp": "2026-10-01T19:12:44.401Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "DeepInfra",
      "accuracy": 0.8374978740239459,
      "accuracy_std_dev": 0.020695537575457914,
      "run_count": 11,
      "total_questions": 1921,
      "avg_cost_per_task_usd": 0.09704594429984383,
      "avg_duration_per_task_ms": 615903.0510150963,
      "avg_output_tokens_per_task": 30251.17542946382,
      "last_run_timestamp": "2026-10-31T16:13:36.316Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Modal",
      "accuracy": 0.8371593636363637,
      "accuracy_std_dev": 0.013027195241792026,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.141865238292011,
      "avg_duration_per_task_ms": 181084.47015610652,
      "avg_output_tokens_per_task": 32015.05968778696,
      "last_run_timestamp": "2026-10-31T08:16:11.213Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.8369407142857143,
      "accuracy_std_dev": 0.020626950634783515,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.10078755411255412,
      "avg_duration_per_task_ms": 1217533.0151515151,
      "avg_output_tokens_per_task": 46038.86075036075,
      "last_run_timestamp": "2026-10-31T16:08:56.903Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "Reka",
      "accuracy": 0.836139,
      "accuracy_std_dev": 0.0007933738084913017,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.05002690909090909,
      "avg_duration_per_task_ms": 537584.335016835,
      "avg_output_tokens_per_task": 49932.06734006734,
      "last_run_timestamp": "2026-09-30T11:44:08.104Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Reka",
      "accuracy": 0.8361277244542866,
      "accuracy_std_dev": 0.026614847390527968,
      "run_count": 16,
      "total_questions": 3161,
      "avg_cost_per_task_usd": 0.02298093546346093,
      "avg_duration_per_task_ms": 307170.6646630813,
      "avg_output_tokens_per_task": 18491.486871243276,
      "last_run_timestamp": "2026-10-31T08:31:11.445Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "inclusionai/ling-3.0-flash-vl-20260910",
      "model_slug": "inclusionai/ling-3.0-flash-vl",
      "model_name": "inclusionAI: Ling 3.0 Flash VL",
      "provider_name": "Novita",
      "accuracy": 0.8358585,
      "accuracy_std_dev": 0.01430950000000003,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.0011451691919191918,
      "avg_duration_per_task_ms": 116299.33333333333,
      "avg_output_tokens_per_task": 18355.219696969696,
      "last_run_timestamp": "2026-10-01T08:23:48.135Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "DigitalOcean",
      "accuracy": 0.8356902,
      "accuracy_std_dev": 0.013324017868496005,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.0493619898989899,
      "avg_duration_per_task_ms": 1019058.498989899,
      "avg_output_tokens_per_task": 29486.257575757576,
      "last_run_timestamp": "2026-10-17T08:20:47.476Z"
    },
    {
      "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": 0.835577888888889,
      "accuracy_std_dev": 0.03515262553634887,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.10956455723905723,
      "avg_duration_per_task_ms": 1079749.1066217732,
      "avg_output_tokens_per_task": 34161.65993265993,
      "last_run_timestamp": "2026-10-24T16:17:57.470Z"
    },
    {
      "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": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "AkashML",
      "accuracy": 0.8350828124309392,
      "accuracy_std_dev": 0.015312607695726205,
      "run_count": 19,
      "total_questions": 3620,
      "avg_cost_per_task_usd": 0.04118064751381217,
      "avg_duration_per_task_ms": 454296.9718232044,
      "avg_output_tokens_per_task": 45405.74917127072,
      "last_run_timestamp": "2026-10-31T16:27:38.037Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "AkashML",
      "accuracy": 0.8348581735005084,
      "accuracy_std_dev": 0.011926219294758369,
      "run_count": 18,
      "total_questions": 2951,
      "avg_cost_per_task_usd": 0.1190305462555066,
      "avg_duration_per_task_ms": 291312.74483226024,
      "avg_output_tokens_per_task": 28021.837004405286,
      "last_run_timestamp": "2026-10-31T08:09:30.690Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.8347106363636364,
      "accuracy_std_dev": 0.07009507473200917,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.21278814876033061,
      "avg_duration_per_task_ms": 570294.2180899908,
      "avg_output_tokens_per_task": 48358.50367309458,
      "last_run_timestamp": "2026-10-31T16:18:48.433Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Mara",
      "accuracy": 0.8346120933503837,
      "accuracy_std_dev": 0.029860364128538167,
      "run_count": 6,
      "total_questions": 782,
      "avg_cost_per_task_usd": 0.09530405242966751,
      "avg_duration_per_task_ms": 330065.4283887468,
      "avg_output_tokens_per_task": 39472.66240409207,
      "last_run_timestamp": "2026-10-31T16:08:31.063Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Parasail",
      "accuracy": 0.8345096889411766,
      "accuracy_std_dev": 0.010890152035457172,
      "run_count": 11,
      "total_questions": 2125,
      "avg_cost_per_task_usd": 0.2468675487058824,
      "avg_duration_per_task_ms": 311483.9990588235,
      "avg_output_tokens_per_task": 55935.44047058823,
      "last_run_timestamp": "2026-10-31T16:28:00.196Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Modal",
      "accuracy": 0.8344576188870152,
      "accuracy_std_dev": 0.007843600547934322,
      "run_count": 3,
      "total_questions": 593,
      "avg_cost_per_task_usd": 0.033416617200674535,
      "avg_duration_per_task_ms": 208747.37942664418,
      "avg_output_tokens_per_task": 14517.00168634064,
      "last_run_timestamp": "2026-10-31T08:47:23.899Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "DekaLLM",
      "accuracy": 0.8341543388067869,
      "accuracy_std_dev": 0.008301927538228698,
      "run_count": 10,
      "total_questions": 1827,
      "avg_cost_per_task_usd": 0.0431522140120416,
      "avg_duration_per_task_ms": 94743.94636015326,
      "avg_output_tokens_per_task": 15061.821018062397,
      "last_run_timestamp": "2026-10-31T08:22:53.415Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Alibaba",
      "accuracy": 0.8341254705882353,
      "accuracy_std_dev": 0.010649185432078013,
      "run_count": 17,
      "total_questions": 3366,
      "avg_cost_per_task_usd": 0.25025737195484254,
      "avg_duration_per_task_ms": 384709.65567439096,
      "avg_output_tokens_per_task": 40253.002970885325,
      "last_run_timestamp": "2026-10-31T08:41:02.403Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "thinkingmachines/inkling-20260715",
      "model_slug": "thinkingmachines/inkling",
      "model_name": "Thinking Machines: Inkling",
      "provider_name": "auto-routing",
      "accuracy": 0.8340364430379746,
      "accuracy_std_dev": 0.01069546220626094,
      "run_count": 6,
      "total_questions": 1185,
      "avg_cost_per_task_usd": 0.0954248565400844,
      "avg_duration_per_task_ms": 293770.05400843883,
      "avg_output_tokens_per_task": 23403.932489451476,
      "last_run_timestamp": "2026-10-01T16:11:05.832Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8338943333333333,
      "accuracy_std_dev": 0.011690806112307073,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.03773407239057239,
      "avg_duration_per_task_ms": 276396.1228956229,
      "avg_output_tokens_per_task": 28290.345117845118,
      "last_run_timestamp": "2026-09-12T09:56:44.384Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "Novita",
      "accuracy": 0.833814142857143,
      "accuracy_std_dev": 0.020613070322551417,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.10298939610389608,
      "avg_duration_per_task_ms": 1389091.401875902,
      "avg_output_tokens_per_task": 47060.21284271284,
      "last_run_timestamp": "2026-10-31T16:08:33.697Z"
    },
    {
      "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": 0.8334736666666667,
      "accuracy_std_dev": 0.007416946298099319,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.08273329461279462,
      "avg_duration_per_task_ms": 719974.9494949495,
      "avg_output_tokens_per_task": 60693.068181818184,
      "last_run_timestamp": "2026-10-31T16:05:36.105Z"
    },
    {
      "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": 0.8333335000000001,
      "accuracy_std_dev": 0.0016835000000000044,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.02674798484848485,
      "avg_duration_per_task_ms": 520849.86363636365,
      "avg_output_tokens_per_task": 22257.11111111111,
      "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-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "Google",
      "accuracy": 0.8333332,
      "accuracy_std_dev": 0.00946378413532349,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.10132244444444444,
      "avg_duration_per_task_ms": 359966.904040404,
      "avg_output_tokens_per_task": 46075.3101010101,
      "last_run_timestamp": "2026-10-17T16:06:16.399Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-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-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Z.AI",
      "accuracy": 0.8330927142857143,
      "accuracy_std_dev": 0.013865607871011393,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.23029656782106786,
      "avg_duration_per_task_ms": 1228281.2157287158,
      "avg_output_tokens_per_task": 52325.608225108226,
      "last_run_timestamp": "2026-10-31T16:30:07.021Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Cloudflare",
      "accuracy": 0.8330725023474179,
      "accuracy_std_dev": 0.010168177287347969,
      "run_count": 7,
      "total_questions": 1278,
      "avg_cost_per_task_usd": 0.029710525039123633,
      "avg_duration_per_task_ms": 655555.5469483568,
      "avg_output_tokens_per_task": 42323.18544600939,
      "last_run_timestamp": "2026-10-03T09:40:53.626Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Friendli",
      "accuracy": 0.8329060884237507,
      "accuracy_std_dev": 0.015886988525511157,
      "run_count": 12,
      "total_questions": 2341,
      "avg_cost_per_task_usd": 0.2567825809483127,
      "avg_duration_per_task_ms": 427385.35027765913,
      "avg_output_tokens_per_task": 58144.50192225545,
      "last_run_timestamp": "2026-10-31T16:06:46.433Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "Crusoe",
      "accuracy": 0.832772,
      "accuracy_std_dev": 0.0698509274145066,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.024015700336700335,
      "avg_duration_per_task_ms": 613738.0185185185,
      "avg_output_tokens_per_task": 59848.74074074074,
      "last_run_timestamp": "2026-10-31T08:32:40.133Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Parasail",
      "accuracy": 0.8327310906791908,
      "accuracy_std_dev": 0.020914023298509846,
      "run_count": 14,
      "total_questions": 2768,
      "avg_cost_per_task_usd": 0.02508389053468208,
      "avg_duration_per_task_ms": 393221.2268786127,
      "avg_output_tokens_per_task": 46919.84176300578,
      "last_run_timestamp": "2026-10-31T08:25:10.696Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Alibaba",
      "accuracy": 0.8326857692307692,
      "accuracy_std_dev": 0.011020358329744859,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.03860720590520591,
      "avg_duration_per_task_ms": 230762.9052059052,
      "avg_output_tokens_per_task": 14476.024864024865,
      "last_run_timestamp": "2026-10-31T08:25:00.370Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Relace",
      "accuracy": 0.8325688873598369,
      "accuracy_std_dev": 0.019237670701256954,
      "run_count": 10,
      "total_questions": 1962,
      "avg_cost_per_task_usd": 0.013576182466870541,
      "avg_duration_per_task_ms": 438297.6029561672,
      "avg_output_tokens_per_task": 39848.997961264016,
      "last_run_timestamp": "2026-10-31T08:26:26.747Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8325147922396857,
      "accuracy_std_dev": 0.029981271466154875,
      "run_count": 11,
      "total_questions": 2036,
      "avg_cost_per_task_usd": 0.011545964145383104,
      "avg_duration_per_task_ms": 399504.8664047151,
      "avg_output_tokens_per_task": 22873.502946954814,
      "last_run_timestamp": "2026-10-31T08:40:04.563Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8323711428571429,
      "accuracy_std_dev": 0.01158903863174871,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.008056092352092353,
      "avg_duration_per_task_ms": 525780.0266955267,
      "avg_output_tokens_per_task": 20733.657287157286,
      "last_run_timestamp": "2026-10-31T16:01:53.649Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Friendli",
      "accuracy": 0.8323670282608697,
      "accuracy_std_dev": 0.013729350448377906,
      "run_count": 8,
      "total_questions": 1380,
      "avg_cost_per_task_usd": 0.20936515,
      "avg_duration_per_task_ms": 475047.03623188403,
      "avg_output_tokens_per_task": 47461.076811594205,
      "last_run_timestamp": "2026-10-31T16:35:08.642Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "CoreWeave",
      "accuracy": 0.8320348961038961,
      "accuracy_std_dev": 0.011673500000000003,
      "run_count": 2,
      "total_questions": 385,
      "avg_cost_per_task_usd": 0.02834737922077922,
      "avg_duration_per_task_ms": 116343.6909090909,
      "avg_output_tokens_per_task": 22533.716883116882,
      "last_run_timestamp": "2026-08-29T18:29:13.946Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "BaseTen",
      "accuracy": 0.8318478292335116,
      "accuracy_std_dev": 0.010811745465322452,
      "run_count": 17,
      "total_questions": 2805,
      "avg_cost_per_task_usd": 0.020570678431372547,
      "avg_duration_per_task_ms": 255956.05668449198,
      "avg_output_tokens_per_task": 40944.02673796791,
      "last_run_timestamp": "2026-10-31T08:20:07.217Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Parasail",
      "accuracy": 0.8317821610051186,
      "accuracy_std_dev": 0.013001907940092663,
      "run_count": 11,
      "total_questions": 2149,
      "avg_cost_per_task_usd": 0.15365638017682642,
      "avg_duration_per_task_ms": 272025.46998604,
      "avg_output_tokens_per_task": 34707.94276407632,
      "last_run_timestamp": "2026-11-05T00:00:09.943Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Parasail",
      "accuracy": 0.831769061433447,
      "accuracy_std_dev": 0.006801018573632123,
      "run_count": 12,
      "total_questions": 2344,
      "avg_cost_per_task_usd": 0.03448967320819113,
      "avg_duration_per_task_ms": 208520.74530716724,
      "avg_output_tokens_per_task": 14194.120733788395,
      "last_run_timestamp": "2026-10-31T08:37:28.112Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "tencent/hy4-20261104",
      "model_slug": "tencent/hy4",
      "model_name": "Tencent: Hy4",
      "provider_name": "Tencent",
      "accuracy": 0.83165,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.06418498484848485,
      "avg_duration_per_task_ms": 493220.3232323232,
      "avg_output_tokens_per_task": 51169.5101010101,
      "last_run_timestamp": "2026-11-10T19:46:03.094Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Cerebras",
      "accuracy": 0.8316496,
      "accuracy_std_dev": 0.007062535765573171,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.022192562626262628,
      "avg_duration_per_task_ms": 11557.427272727273,
      "avg_output_tokens_per_task": 14346.162626262627,
      "last_run_timestamp": "2026-10-31T08:15:46.337Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Ionstream",
      "accuracy": 0.8316495,
      "accuracy_std_dev": 0.055555500000000035,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.04943669696969697,
      "avg_duration_per_task_ms": 1128190.707070707,
      "avg_output_tokens_per_task": 25885.7702020202,
      "last_run_timestamp": "2026-08-29T19:26:41.546Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.831532,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 185,
      "avg_cost_per_task_usd": 0.01692656216216216,
      "avg_duration_per_task_ms": 560352.7027027027,
      "avg_output_tokens_per_task": 13578.437837837839,
      "last_run_timestamp": "2026-10-31T18:41:15.616Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "DeepInfra",
      "accuracy": 0.8314150997442455,
      "accuracy_std_dev": 0.012107134010057875,
      "run_count": 8,
      "total_questions": 1564,
      "avg_cost_per_task_usd": 0.02770090409207161,
      "avg_duration_per_task_ms": 583377.9558823529,
      "avg_output_tokens_per_task": 14708.68989769821,
      "last_run_timestamp": "2026-10-31T08:48:50.088Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Crusoe",
      "accuracy": 0.8313539815914488,
      "accuracy_std_dev": 0.013525888763386546,
      "run_count": 9,
      "total_questions": 1684,
      "avg_cost_per_task_usd": 0.02395000890736342,
      "avg_duration_per_task_ms": 287524.2090261283,
      "avg_output_tokens_per_task": 47679.917458432305,
      "last_run_timestamp": "2026-10-31T08:55:40.953Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.8310883333333333,
      "accuracy_std_dev": 0.009654676391377534,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.20155932323232326,
      "avg_duration_per_task_ms": 266665.86195286195,
      "avg_output_tokens_per_task": 37910.23063973064,
      "last_run_timestamp": "2026-10-31T08:09:31.329Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Wafer",
      "accuracy": 0.8309762,
      "accuracy_std_dev": 0.004338052761320463,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.04722357777777777,
      "avg_duration_per_task_ms": 258256.24646464645,
      "avg_output_tokens_per_task": 15298.06111111111,
      "last_run_timestamp": "2026-10-31T08:20:09.369Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "CoreWeave",
      "accuracy": 0.830421644785601,
      "accuracy_std_dev": 0.015773715151716793,
      "run_count": 12,
      "total_questions": 1889,
      "avg_cost_per_task_usd": 0.04062480518793012,
      "avg_duration_per_task_ms": 302854.55849655904,
      "avg_output_tokens_per_task": 33135.881948120696,
      "last_run_timestamp": "2026-10-31T16:19:08.399Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Cloudflare",
      "accuracy": 0.8302074978687126,
      "accuracy_std_dev": 0.018798024809619108,
      "run_count": 12,
      "total_questions": 2346,
      "avg_cost_per_task_usd": 0.14303728772378516,
      "avg_duration_per_task_ms": 291115.02642796247,
      "avg_output_tokens_per_task": 32292.15643648764,
      "last_run_timestamp": "2026-10-31T08:14:43.196Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Phala",
      "accuracy": 0.82996625,
      "accuracy_std_dev": 0.034767651864736286,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.20791771717171717,
      "avg_duration_per_task_ms": 1562678.1717171718,
      "avg_output_tokens_per_task": 49349.89141414141,
      "last_run_timestamp": "2026-10-31T16:31:54.109Z"
    },
    {
      "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": 0.8294050000000001,
      "accuracy_std_dev": 0.004827531943619518,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.02358655555555556,
      "avg_duration_per_task_ms": 376720.68518518517,
      "avg_output_tokens_per_task": 21014.917508417508,
      "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-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Morph",
      "accuracy": 0.8288813468715698,
      "accuracy_std_dev": 0.048221988340541345,
      "run_count": 14,
      "total_questions": 2733,
      "avg_cost_per_task_usd": 0.09670949213318698,
      "avg_duration_per_task_ms": 542922.2052689352,
      "avg_output_tokens_per_task": 29030.10171972192,
      "last_run_timestamp": "2026-10-31T08:36:47.227Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8285234285714286,
      "accuracy_std_dev": 0.008462147918097126,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.048541420634920646,
      "avg_duration_per_task_ms": 445627.215007215,
      "avg_output_tokens_per_task": 26857.45165945166,
      "last_run_timestamp": "2026-10-31T16:10:31.145Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "Alibaba",
      "accuracy": 0.8284931249999998,
      "accuracy_std_dev": 0.006861031326220224,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.029469156565656567,
      "avg_duration_per_task_ms": 133597.33396464647,
      "avg_output_tokens_per_task": 22565.453914141413,
      "last_run_timestamp": "2026-10-31T16:08:08.538Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Decart",
      "accuracy": 0.82828275,
      "accuracy_std_dev": 0.004610363346581274,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.0223316047979798,
      "avg_duration_per_task_ms": 615131.3409090909,
      "avg_output_tokens_per_task": 56159.15404040404,
      "last_run_timestamp": "2026-10-17T08:18:07.611Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "Venice",
      "accuracy": 0.8281987,
      "accuracy_std_dev": 0.027781391164050814,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.038244685353535354,
      "avg_duration_per_task_ms": 378160.9792929293,
      "avg_output_tokens_per_task": 38488.014646464646,
      "last_run_timestamp": "2026-10-24T16:01:48.224Z"
    },
    {
      "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": 0.8281297272727273,
      "accuracy_std_dev": 0.007612402500834456,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.06470958723599633,
      "avg_duration_per_task_ms": 560112.3668503214,
      "avg_output_tokens_per_task": 50830.36317722681,
      "last_run_timestamp": "2026-10-31T16:08:56.901Z"
    },
    {
      "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": 0.8280023333333335,
      "accuracy_std_dev": 0.027257023221344042,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.1507835429292929,
      "avg_duration_per_task_ms": 1669561.2037037036,
      "avg_output_tokens_per_task": 49318.73569023569,
      "last_run_timestamp": "2026-10-31T16:08:56.833Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Nebius",
      "accuracy": 0.8275762754844144,
      "accuracy_std_dev": 0.06091378546606307,
      "run_count": 6,
      "total_questions": 1187,
      "avg_cost_per_task_usd": 0.20493896293176075,
      "avg_duration_per_task_ms": 1744742.5619208089,
      "avg_output_tokens_per_task": 46988.62679022746,
      "last_run_timestamp": "2026-10-31T16:22:14.129Z"
    },
    {
      "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": 0.8275058461538461,
      "accuracy_std_dev": 0.01065993423883712,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.035353606837606834,
      "avg_duration_per_task_ms": 305384.14801864803,
      "avg_output_tokens_per_task": 15605.497668997668,
      "last_run_timestamp": "2026-11-10T17:02:13.232Z"
    },
    {
      "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": 0.8270801428571428,
      "accuracy_std_dev": 0.019019147618188608,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.09306387085137086,
      "avg_duration_per_task_ms": 1107119.2698412698,
      "avg_output_tokens_per_task": 45248.218614718615,
      "last_run_timestamp": "2026-10-31T16:08:08.535Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Novita",
      "accuracy": 0.8270583636363636,
      "accuracy_std_dev": 0.00941702237510474,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.04288225068870524,
      "avg_duration_per_task_ms": 229942.66437098256,
      "avg_output_tokens_per_task": 14178.259871441689,
      "last_run_timestamp": "2026-10-31T08:36:47.145Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "BaseTen",
      "accuracy": 0.8269358,
      "accuracy_std_dev": 0.006767606146932625,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.04578402121212121,
      "avg_duration_per_task_ms": 60291.652525252524,
      "avg_output_tokens_per_task": 11285.754545454545,
      "last_run_timestamp": "2026-09-26T09:19:03.901Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.1-terminus",
      "model_slug": "deepseek/deepseek-v3.1-terminus",
      "model_name": "DeepSeek: DeepSeek V3.1 Terminus",
      "provider_name": "Novita",
      "accuracy": 0.8265989999999999,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.01892201515151515,
      "avg_duration_per_task_ms": 495658.10101010103,
      "avg_output_tokens_per_task": 18768.85606060606,
      "last_run_timestamp": "2026-08-29T18:35:57.732Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Nebius",
      "accuracy": 0.8265989999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.051792560606060606,
      "avg_duration_per_task_ms": 150332.1313131313,
      "avg_output_tokens_per_task": 17173.429292929293,
      "last_run_timestamp": "2026-11-04T18:02:01.782Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Friendli",
      "accuracy": 0.826459,
      "accuracy_std_dev": 0.009410002267091476,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.14383780639730645,
      "avg_duration_per_task_ms": 198190.3564814815,
      "avg_output_tokens_per_task": 35703.72769360269,
      "last_run_timestamp": "2026-10-31T08:32:01.077Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "CoreWeave",
      "accuracy": 0.8264159143968872,
      "accuracy_std_dev": 0.0040531378816294926,
      "run_count": 4,
      "total_questions": 771,
      "avg_cost_per_task_usd": 0.13822659403372245,
      "avg_duration_per_task_ms": 272716.0363164721,
      "avg_output_tokens_per_task": 34382.090791180286,
      "last_run_timestamp": "2026-10-31T08:38:15.143Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "CoreWeave",
      "accuracy": 0.8264111325878595,
      "accuracy_std_dev": 0.01425677691801343,
      "run_count": 10,
      "total_questions": 1878,
      "avg_cost_per_task_usd": 0.02454355484558041,
      "avg_duration_per_task_ms": 264967.17305644305,
      "avg_output_tokens_per_task": 50263.019169329076,
      "last_run_timestamp": "2026-10-31T08:17:29.105Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Io Net",
      "accuracy": 0.8263836795903439,
      "accuracy_std_dev": 0.029166866420929173,
      "run_count": 7,
      "total_questions": 1367,
      "avg_cost_per_task_usd": 0.07172207900512069,
      "avg_duration_per_task_ms": 248028.12801755668,
      "avg_output_tokens_per_task": 21803.532553035846,
      "last_run_timestamp": "2026-10-31T09:17:12.929Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "GMICloud",
      "accuracy": 0.8262311079545455,
      "accuracy_std_dev": 0.0007570000000000077,
      "run_count": 2,
      "total_questions": 352,
      "avg_cost_per_task_usd": 0.03419934943181818,
      "avg_duration_per_task_ms": 187950.64204545456,
      "avg_output_tokens_per_task": 10629.338068181818,
      "last_run_timestamp": "2026-11-10T06:16:01.718Z"
    },
    {
      "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": 0.8261184761715648,
      "accuracy_std_dev": 0.021573637693110777,
      "run_count": 13,
      "total_questions": 2518,
      "avg_cost_per_task_usd": 0.022734466243050037,
      "avg_duration_per_task_ms": 240016.77958697377,
      "avg_output_tokens_per_task": 45243.0552025417,
      "last_run_timestamp": "2026-11-12T09:43:53.858Z"
    },
    {
      "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": 0.8260636363636363,
      "accuracy_std_dev": 0.058329593420450306,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.051863643709825526,
      "avg_duration_per_task_ms": 534077.3172635445,
      "avg_output_tokens_per_task": 24831.462809917357,
      "last_run_timestamp": "2026-10-31T16:01:53.625Z"
    },
    {
      "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": 0.8260381111111113,
      "accuracy_std_dev": 0.019244088918963866,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.012581548821548822,
      "avg_duration_per_task_ms": 794792.8148148148,
      "avg_output_tokens_per_task": 37224.931537598204,
      "last_run_timestamp": "2026-10-17T08:04:21.262Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Venice",
      "accuracy": 0.82575775,
      "accuracy_std_dev": 0.010065865980505628,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.14380135795454543,
      "avg_duration_per_task_ms": 418612.6521464646,
      "avg_output_tokens_per_task": 28784.420454545456,
      "last_run_timestamp": "2026-10-31T08:40:11.483Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-2.5-pro",
      "model_slug": "google/gemini-2.5-pro",
      "model_name": "Google: Gemini 2.5 Pro",
      "provider_name": "auto-routing",
      "accuracy": 0.8257574999999999,
      "accuracy_std_dev": 0.015993499999999994,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.2609003181818182,
      "avg_duration_per_task_ms": 231557.84848484848,
      "avg_output_tokens_per_task": 25997.717171717173,
      "last_run_timestamp": "2026-10-01T19:12:42.150Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "DeepInfra",
      "accuracy": 0.8254768888888889,
      "accuracy_std_dev": 0.05394208057505227,
      "run_count": 18,
      "total_questions": 3564,
      "avg_cost_per_task_usd": 0.007830105218855219,
      "avg_duration_per_task_ms": 541491.9023569024,
      "avg_output_tokens_per_task": 17689.749438832772,
      "last_run_timestamp": "2026-10-31T16:45:45.611Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Modal",
      "accuracy": 0.825389385046729,
      "accuracy_std_dev": 0.014184335849948446,
      "run_count": 11,
      "total_questions": 2140,
      "avg_cost_per_task_usd": 0.03383954813084113,
      "avg_duration_per_task_ms": 407090.0654205608,
      "avg_output_tokens_per_task": 47991.5476635514,
      "last_run_timestamp": "2026-10-31T08:48:01.125Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "CoreWeave",
      "accuracy": 0.824495,
      "accuracy_std_dev": 0.023148000000000002,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.11508162373737373,
      "avg_duration_per_task_ms": 562224.2297979798,
      "avg_output_tokens_per_task": 32139.14393939394,
      "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-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "Baidu",
      "accuracy": 0.8244826858605246,
      "accuracy_std_dev": 0.009259257906785777,
      "run_count": 8,
      "total_questions": 1563,
      "avg_cost_per_task_usd": 0.008255877799104288,
      "avg_duration_per_task_ms": 352541.7984644914,
      "avg_output_tokens_per_task": 19223.356365962893,
      "last_run_timestamp": "2026-11-10T18:18:26.718Z"
    },
    {
      "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": 0.8244347142857142,
      "accuracy_std_dev": 0.01399044581863215,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.1977735064935065,
      "avg_duration_per_task_ms": 958897.5014430014,
      "avg_output_tokens_per_task": 52647.930735930735,
      "last_run_timestamp": "2026-10-31T16:50:02.353Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-2.5-pro",
      "model_slug": "google/gemini-2.5-pro",
      "model_name": "Google: Gemini 2.5 Pro",
      "provider_name": "Google",
      "accuracy": 0.8243546666666667,
      "accuracy_std_dev": 0.03111308863734931,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.25956171212121215,
      "avg_duration_per_task_ms": 233681.0909090909,
      "avg_output_tokens_per_task": 25863.887205387204,
      "last_run_timestamp": "2026-10-01T19:17:25.275Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "NextBit",
      "accuracy": 0.8243546666666666,
      "accuracy_std_dev": 0.03801456116045827,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.029119777356902358,
      "avg_duration_per_task_ms": 574036.2205387205,
      "avg_output_tokens_per_task": 53199.25168350169,
      "last_run_timestamp": "2026-11-01T11:18:30.093Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "AkashML",
      "accuracy": 0.8242303292875989,
      "accuracy_std_dev": 0.016434035763013298,
      "run_count": 20,
      "total_questions": 3790,
      "avg_cost_per_task_usd": 0.030382969920844324,
      "avg_duration_per_task_ms": 266171.0530343008,
      "avg_output_tokens_per_task": 14053.306068601583,
      "last_run_timestamp": "2026-10-31T08:10:37.346Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "Xiaomi",
      "accuracy": 0.8241675555555555,
      "accuracy_std_dev": 0.024242370268744616,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.011208364197530863,
      "avg_duration_per_task_ms": 802349.7974186308,
      "avg_output_tokens_per_task": 39790.19079685747,
      "last_run_timestamp": "2026-10-17T08:02:36.531Z"
    },
    {
      "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": 0.8240802188112345,
      "accuracy_std_dev": 0.011135789059800642,
      "run_count": 8,
      "total_questions": 1531,
      "avg_cost_per_task_usd": 0.02478115610711953,
      "avg_duration_per_task_ms": 443159.53821032005,
      "avg_output_tokens_per_task": 24668.269105160027,
      "last_run_timestamp": "2026-10-31T16:42:08.154Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Featherless",
      "accuracy": 0.822917,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 176,
      "avg_cost_per_task_usd": 0.036827073863636366,
      "avg_duration_per_task_ms": 130682.76704545454,
      "avg_output_tokens_per_task": 8202.113636363636,
      "last_run_timestamp": "2026-11-14T00:00:42.343Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "DigitalOcean",
      "accuracy": 0.8225597382029736,
      "accuracy_std_dev": 0.013482485406421735,
      "run_count": 8,
      "total_questions": 1547,
      "avg_cost_per_task_usd": 0.018713002585649646,
      "avg_duration_per_task_ms": 601976.5061409179,
      "avg_output_tokens_per_task": 20229.648351648353,
      "last_run_timestamp": "2026-10-31T16:13:07.320Z"
    },
    {
      "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": 0.8225107142857142,
      "accuracy_std_dev": 0.013731671521541366,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.007629560606060605,
      "avg_duration_per_task_ms": 439138.09451659454,
      "avg_output_tokens_per_task": 18615.266233766233,
      "last_run_timestamp": "2026-10-31T16:13:09.863Z"
    },
    {
      "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": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "StreamLake",
      "accuracy": 0.8223905000000001,
      "accuracy_std_dev": 0.019360500000000003,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.023091813131313133,
      "avg_duration_per_task_ms": 454744.9166666667,
      "avg_output_tokens_per_task": 21871.75,
      "last_run_timestamp": "2026-08-29T18:09:03.379Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Chutes",
      "accuracy": 0.8223905,
      "accuracy_std_dev": 0.019262412208322532,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.03192426473063974,
      "avg_duration_per_task_ms": 260830.34259259258,
      "avg_output_tokens_per_task": 14583.287878787878,
      "last_run_timestamp": "2026-10-31T08:17:29.187Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "auto-routing",
      "accuracy": 0.8220747500000002,
      "accuracy_std_dev": 0.06699828629105002,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.029949901515151516,
      "avg_duration_per_task_ms": 747877.6811868687,
      "avg_output_tokens_per_task": 30535.987373737375,
      "last_run_timestamp": "2026-10-17T08:18:04.911Z"
    },
    {
      "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": 0.8220297142857144,
      "accuracy_std_dev": 0.020075599465977787,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.006425138528138528,
      "avg_duration_per_task_ms": 406726.4509379509,
      "avg_output_tokens_per_task": 18776.16810966811,
      "last_run_timestamp": "2026-10-31T16:12:28.233Z"
    },
    {
      "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": 0.8216246926454446,
      "accuracy_std_dev": 0.030556920617104074,
      "run_count": 5,
      "total_questions": 911,
      "avg_cost_per_task_usd": 0.02863867069154775,
      "avg_duration_per_task_ms": 543580.5345773875,
      "avg_output_tokens_per_task": 16890.06256860593,
      "last_run_timestamp": "2026-10-17T16:03:44.233Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Cohere",
      "accuracy": 0.8215486666666668,
      "accuracy_std_dev": 0.015611992961253275,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.02833054377104377,
      "avg_duration_per_task_ms": 228888.5606060606,
      "avg_output_tokens_per_task": 56421.28451178451,
      "last_run_timestamp": "2026-10-31T08:03:09.999Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-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": 0.821448952348303,
      "accuracy_std_dev": 0.023090870602325352,
      "run_count": 15,
      "total_questions": 2917,
      "avg_cost_per_task_usd": 0.02527841960918752,
      "avg_duration_per_task_ms": 834456.179293795,
      "avg_output_tokens_per_task": 51002.6424408639,
      "last_run_timestamp": "2026-11-12T02:16:49.980Z"
    },
    {
      "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": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Alibaba",
      "accuracy": 0.8213131235765501,
      "accuracy_std_dev": 0.034016985407619854,
      "run_count": 12,
      "total_questions": 2371,
      "avg_cost_per_task_usd": 0.08167362969211302,
      "avg_duration_per_task_ms": 336290.14972585405,
      "avg_output_tokens_per_task": 20283.436524673132,
      "last_run_timestamp": "2026-10-31T16:02:14.359Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta/muse-glimmer-30b-20260810",
      "model_slug": "meta/muse-glimmer-30b",
      "model_name": "Meta: Muse Glimmer 30B",
      "provider_name": "Parasail",
      "accuracy": 0.8212070710659898,
      "accuracy_std_dev": 0.011946281848340899,
      "run_count": 3,
      "total_questions": 591,
      "avg_cost_per_task_usd": 0.01807649915397631,
      "avg_duration_per_task_ms": 302940.0592216582,
      "avg_output_tokens_per_task": 16413.382402707277,
      "last_run_timestamp": "2026-08-29T09:59:35.603Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Io Net",
      "accuracy": 0.8210525603238865,
      "accuracy_std_dev": 0.07599126782917703,
      "run_count": 7,
      "total_questions": 1235,
      "avg_cost_per_task_usd": 0.007044106882591094,
      "avg_duration_per_task_ms": 525854.9595141701,
      "avg_output_tokens_per_task": 14201.662348178137,
      "last_run_timestamp": "2026-10-03T10:22:02.091Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta/muse-glimmer-30b-20260810",
      "model_slug": "meta/muse-glimmer-30b",
      "model_name": "Meta: Muse Glimmer 30B",
      "provider_name": "Phala",
      "accuracy": 0.820736340116279,
      "accuracy_std_dev": 0.02413812698210204,
      "run_count": 7,
      "total_questions": 1376,
      "avg_cost_per_task_usd": 0.01966482122093023,
      "avg_duration_per_task_ms": 283517.4375,
      "avg_output_tokens_per_task": 17874.627180232557,
      "last_run_timestamp": "2026-10-03T09:25:02.390Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.1-flash-lite-20260507",
      "model_slug": "google/gemini-3.1-flash-lite",
      "model_name": "Google: Gemini 3.1 Flash Lite",
      "provider_name": "auto-routing",
      "accuracy": 0.8207070000000001,
      "accuracy_std_dev": 0.024411000000000016,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.04063828030303031,
      "avg_duration_per_task_ms": 83296.42424242424,
      "avg_output_tokens_per_task": 26969.131313131315,
      "last_run_timestamp": "2026-10-01T19:23:01.327Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "StreamLake",
      "accuracy": 0.8207070000000001,
      "accuracy_std_dev": 0.007575999999999972,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.07039754797979798,
      "avg_duration_per_task_ms": 879593.0631313132,
      "avg_output_tokens_per_task": 40130.17171717172,
      "last_run_timestamp": "2026-08-30T00:37:59.094Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "CoreWeave",
      "accuracy": 0.8203833076923077,
      "accuracy_std_dev": 0.018076297388844354,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.042996285159285164,
      "avg_duration_per_task_ms": 241748.8038073038,
      "avg_output_tokens_per_task": 14318.661227661227,
      "last_run_timestamp": "2026-10-31T08:09:30.241Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "thinkingmachines/inkling-20260715",
      "model_slug": "thinkingmachines/inkling",
      "model_name": "Thinking Machines: Inkling",
      "provider_name": "BaseTen",
      "accuracy": 0.8202743410372041,
      "accuracy_std_dev": 0.011484132116496031,
      "run_count": 9,
      "total_questions": 1774,
      "avg_cost_per_task_usd": 0.09063781341600902,
      "avg_duration_per_task_ms": 168767.6234498309,
      "avg_output_tokens_per_task": 22236.227170236754,
      "last_run_timestamp": "2026-09-19T10:14:07.966Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Mancer 2",
      "accuracy": 0.8200275393258427,
      "accuracy_std_dev": 0.011794769394173345,
      "run_count": 9,
      "total_questions": 1691,
      "avg_cost_per_task_usd": 0.03818077646363098,
      "avg_duration_per_task_ms": 215799.96629213484,
      "avg_output_tokens_per_task": 14692.68894145476,
      "last_run_timestamp": "2026-10-31T08:55:11.136Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "BaseTen",
      "accuracy": 0.819865,
      "accuracy_std_dev": 0.006733999999999962,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.08986796717171718,
      "avg_duration_per_task_ms": 195255.59343434343,
      "avg_output_tokens_per_task": 25486.853535353537,
      "last_run_timestamp": "2026-08-29T18:58:14.477Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "StreamLake",
      "accuracy": 0.819865,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.13854771717171718,
      "avg_duration_per_task_ms": 540824.1212121212,
      "avg_output_tokens_per_task": 31240.651515151516,
      "last_run_timestamp": "2026-09-07T06:00:57.184Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "auto-routing",
      "accuracy": 0.8194060909090909,
      "accuracy_std_dev": 0.013884218840464785,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.033594433884297524,
      "avg_duration_per_task_ms": 211994.46005509642,
      "avg_output_tokens_per_task": 13897.44214876033,
      "last_run_timestamp": "2026-10-31T08:09:30.951Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Tenstorrent",
      "accuracy": 0.8193976666666667,
      "accuracy_std_dev": 0.013990614068010025,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.017960378226711565,
      "avg_duration_per_task_ms": 372150.59315375984,
      "avg_output_tokens_per_task": 13044.02581369248,
      "last_run_timestamp": "2026-11-08T03:44:37.321Z"
    },
    {
      "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": 0.8193842857142857,
      "accuracy_std_dev": 0.05010024752352396,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.20815405194805192,
      "avg_duration_per_task_ms": 879947.1565656565,
      "avg_output_tokens_per_task": 51956.655122655124,
      "last_run_timestamp": "2026-10-31T16:08:08.537Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "SambaNova",
      "accuracy": 0.8192149207920791,
      "accuracy_std_dev": 0.018823182524040232,
      "run_count": 6,
      "total_questions": 909,
      "avg_cost_per_task_usd": 0.11224348184818483,
      "avg_duration_per_task_ms": 401079.6094609461,
      "avg_output_tokens_per_task": 46532.80198019802,
      "last_run_timestamp": "2026-10-17T16:06:46.637Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Darkbloom",
      "accuracy": 0.8189473014736841,
      "accuracy_std_dev": 0.010171864904671552,
      "run_count": 12,
      "total_questions": 2375,
      "avg_cost_per_task_usd": 0.03239228378947368,
      "avg_duration_per_task_ms": 542473.8867368421,
      "avg_output_tokens_per_task": 15887.374736842105,
      "last_run_timestamp": "2026-11-13T01:53:10.915Z"
    },
    {
      "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": 0.8186628571428571,
      "accuracy_std_dev": 0.011047136143797251,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.024918836219336223,
      "avg_duration_per_task_ms": 190477.32106782106,
      "avg_output_tokens_per_task": 16519.824675324675,
      "last_run_timestamp": "2026-10-03T10:10:35.343Z"
    },
    {
      "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": 0.818182,
      "accuracy_std_dev": 0.003367000000000009,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.10403051010101011,
      "avg_duration_per_task_ms": 527188.4621212122,
      "avg_output_tokens_per_task": 38544.431818181816,
      "last_run_timestamp": "2026-08-29T18:26:14.942Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "KT",
      "accuracy": 0.818182,
      "accuracy_std_dev": 0.006734000000000018,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.023744409090909087,
      "avg_duration_per_task_ms": 279872.4292929293,
      "avg_output_tokens_per_task": 47254.60101010101,
      "last_run_timestamp": "2026-11-11T02:01:08.052Z"
    },
    {
      "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": 0.8179302505415164,
      "accuracy_std_dev": 0.004936247761122178,
      "run_count": 7,
      "total_questions": 1385,
      "avg_cost_per_task_usd": 0.02662889386281589,
      "avg_duration_per_task_ms": 314143.65198555955,
      "avg_output_tokens_per_task": 26506.814440433212,
      "last_run_timestamp": "2026-10-31T16:04:48.178Z"
    },
    {
      "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": "Tenstorrent",
      "accuracy": 0.817884202247191,
      "accuracy_std_dev": 0.008045999999999998,
      "run_count": 2,
      "total_questions": 356,
      "avg_cost_per_task_usd": 0.006599825842696629,
      "avg_duration_per_task_ms": 342879.2528089888,
      "avg_output_tokens_per_task": 26299.71629213483,
      "last_run_timestamp": "2026-11-08T04:17:28.377Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Near AI",
      "accuracy": 0.8178727706422019,
      "accuracy_std_dev": 0.03174962566960436,
      "run_count": 5,
      "total_questions": 981,
      "avg_cost_per_task_usd": 0.023857419979612642,
      "avg_duration_per_task_ms": 166236.06523955148,
      "avg_output_tokens_per_task": 14523.781855249745,
      "last_run_timestamp": "2026-10-31T08:07:51.399Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "StreamLake",
      "accuracy": 0.8177376144306652,
      "accuracy_std_dev": 0.03151916867911054,
      "run_count": 9,
      "total_questions": 1774,
      "avg_cost_per_task_usd": 0.013423886696730551,
      "avg_duration_per_task_ms": 855848.4718151072,
      "avg_output_tokens_per_task": 39679.556933483655,
      "last_run_timestamp": "2026-10-17T08:19:03.202Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Reka",
      "accuracy": 0.8172424786585366,
      "accuracy_std_dev": 0.032381379178472315,
      "run_count": 8,
      "total_questions": 984,
      "avg_cost_per_task_usd": 0.0567359319105691,
      "avg_duration_per_task_ms": 328244.918699187,
      "avg_output_tokens_per_task": 21710.94918699187,
      "last_run_timestamp": "2026-10-31T16:26:17.224Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "auto-routing",
      "accuracy": 0.8172197142857143,
      "accuracy_std_dev": 0.01956805149591904,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.009476494949494948,
      "avg_duration_per_task_ms": 577700.4018759019,
      "avg_output_tokens_per_task": 19548.448051948053,
      "last_run_timestamp": "2026-10-31T16:08:08.535Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "OpenInference",
      "accuracy": 0.8171817336142323,
      "accuracy_std_dev": 0.04982830145510509,
      "run_count": 11,
      "total_questions": 2136,
      "avg_cost_per_task_usd": 0.018343408239700373,
      "avg_duration_per_task_ms": 793559.7439138577,
      "avg_output_tokens_per_task": 48776.84925093633,
      "last_run_timestamp": "2026-10-31T08:31:53.887Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "auto-routing",
      "accuracy": 0.8164981428571428,
      "accuracy_std_dev": 0.013316748910714663,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.026554869408369406,
      "avg_duration_per_task_ms": 277649.69264069264,
      "avg_output_tokens_per_task": 25393.416305916307,
      "last_run_timestamp": "2026-10-31T16:05:36.303Z"
    },
    {
      "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": 0.8161921818181818,
      "accuracy_std_dev": 0.005314989759986385,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.34637480073461896,
      "avg_duration_per_task_ms": 358931.4338842975,
      "avg_output_tokens_per_task": 22900.560146923784,
      "last_run_timestamp": "2026-10-31T08:17:29.306Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.5-sonnet-20250929",
      "model_slug": "anthropic/claude-sonnet-4.5",
      "model_name": "Anthropic: Claude Sonnet 4.5",
      "provider_name": "auto-routing",
      "accuracy": 0.8158860909090909,
      "accuracy_std_dev": 0.011310770084341126,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.31360446280991733,
      "avg_duration_per_task_ms": 289460.9269972452,
      "avg_output_tokens_per_task": 20716.85261707989,
      "last_run_timestamp": "2026-10-31T08:15:24.404Z"
    },
    {
      "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": 0.8157767142857143,
      "accuracy_std_dev": 0.008190962872306891,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.02609219480519481,
      "avg_duration_per_task_ms": 163166.5974025974,
      "avg_output_tokens_per_task": 23394.820346320346,
      "last_run_timestamp": "2026-10-31T16:09:44.821Z"
    },
    {
      "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": 0.8156564999999999,
      "accuracy_std_dev": 0.015993499999999994,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.018819113636363637,
      "avg_duration_per_task_ms": 509395.56313131313,
      "avg_output_tokens_per_task": 18621.05808080808,
      "last_run_timestamp": "2026-08-30T00:16:38.394Z"
    },
    {
      "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": 0.8154460000000001,
      "accuracy_std_dev": 0.020060118637485684,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.12119656250000001,
      "avg_duration_per_task_ms": 800383.8623737374,
      "avg_output_tokens_per_task": 63007.16098484849,
      "last_run_timestamp": "2026-10-31T16:10:09.226Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "Friendli",
      "accuracy": 0.8153198,
      "accuracy_std_dev": 0.009214558967199644,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.030977021212121213,
      "avg_duration_per_task_ms": 329320.2121212121,
      "avg_output_tokens_per_task": 20456.146464646463,
      "last_run_timestamp": "2026-10-17T16:09:25.569Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Relace",
      "accuracy": 0.8152402988505746,
      "accuracy_std_dev": 0.029604949040447602,
      "run_count": 4,
      "total_questions": 783,
      "avg_cost_per_task_usd": 0.28207722477650066,
      "avg_duration_per_task_ms": 913648.006385696,
      "avg_output_tokens_per_task": 58474.00127713921,
      "last_run_timestamp": "2026-10-31T16:12:07.198Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.8149678181818182,
      "accuracy_std_dev": 0.009372109986726867,
      "run_count": 22,
      "total_questions": 4356,
      "avg_cost_per_task_usd": 0.3310624931129476,
      "avg_duration_per_task_ms": 290282.1565656566,
      "avg_output_tokens_per_task": 20829.40679522498,
      "last_run_timestamp": "2026-10-31T08:27:57.306Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Tenstorrent",
      "accuracy": 0.8148578328496808,
      "accuracy_std_dev": 0.00794054106748679,
      "run_count": 9,
      "total_questions": 1723,
      "avg_cost_per_task_usd": 0.04277770748694139,
      "avg_duration_per_task_ms": 184421.18630295995,
      "avg_output_tokens_per_task": 16933.951247823563,
      "last_run_timestamp": "2026-11-08T04:02:09.197Z"
    },
    {
      "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": 0.8143555909090909,
      "accuracy_std_dev": 0.010013659936938356,
      "run_count": 22,
      "total_questions": 4356,
      "avg_cost_per_task_usd": 0.36341748691460063,
      "avg_duration_per_task_ms": 357684.4742883379,
      "avg_output_tokens_per_task": 22888.384756657484,
      "last_run_timestamp": "2026-10-31T08:33:07.626Z"
    },
    {
      "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": 0.8140932857142855,
      "accuracy_std_dev": 0.003468669103119897,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.012047585137085136,
      "avg_duration_per_task_ms": 883223.4725829726,
      "avg_output_tokens_per_task": 18603.608946608947,
      "last_run_timestamp": "2026-10-31T16:13:09.867Z"
    },
    {
      "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": 0.813973,
      "accuracy_std_dev": 0.0008419999999999539,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.006049762626262625,
      "avg_duration_per_task_ms": 483167.9494949495,
      "avg_output_tokens_per_task": 18365.43181818182,
      "last_run_timestamp": "2026-08-29T19:09:30.030Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.1-flash-lite-20260507",
      "model_slug": "google/gemini-3.1-flash-lite",
      "model_name": "Google: Gemini 3.1 Flash Lite",
      "provider_name": "Google AI Studio",
      "accuracy": 0.813973,
      "accuracy_std_dev": 0.01262599999999997,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.04186610353535354,
      "avg_duration_per_task_ms": 76135.87878787878,
      "avg_output_tokens_per_task": 27786.689393939392,
      "last_run_timestamp": "2026-10-01T19:16:58.595Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/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": 0.8133718571428571,
      "accuracy_std_dev": 0.016208212586997554,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.04426409668109668,
      "avg_duration_per_task_ms": 84524.13997113997,
      "avg_output_tokens_per_task": 27404.910533910534,
      "last_run_timestamp": "2026-10-01T19:25:58.919Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "inclusionai/ling-3.1-flash-20261002",
      "model_slug": "inclusionai/ling-3.1-flash",
      "model_name": "inclusionAI: Ling 3.1 Flash",
      "provider_name": "Novita",
      "accuracy": 0.813051,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 189,
      "avg_cost_per_task_usd": 0.00766168253968254,
      "avg_duration_per_task_ms": 356938.3703703704,
      "avg_output_tokens_per_task": 33904.37566137566,
      "last_run_timestamp": "2026-10-14T18:00:03.925Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.5-sonnet-20250929",
      "model_slug": "anthropic/claude-sonnet-4.5",
      "model_name": "Anthropic: Claude Sonnet 4.5",
      "provider_name": "Anthropic",
      "accuracy": 0.8128251818181819,
      "accuracy_std_dev": 0.015105822103463543,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.3406421253443526,
      "avg_duration_per_task_ms": 353395.67309458216,
      "avg_output_tokens_per_task": 22530.53305785124,
      "last_run_timestamp": "2026-10-31T08:46:10.648Z"
    },
    {
      "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": 0.8120124868421054,
      "accuracy_std_dev": 0.08527436462791073,
      "run_count": 7,
      "total_questions": 1368,
      "avg_cost_per_task_usd": 0.0753633048245614,
      "avg_duration_per_task_ms": 666389.4437134503,
      "avg_output_tokens_per_task": 27234.816520467837,
      "last_run_timestamp": "2026-10-10T16:29:40.552Z"
    },
    {
      "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": 0.8117845999999999,
      "accuracy_std_dev": 0.01594972366657178,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.026062873737373735,
      "avg_duration_per_task_ms": 323333.02424242423,
      "avg_output_tokens_per_task": 17452.424242424244,
      "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-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "River",
      "accuracy": 0.811448,
      "accuracy_std_dev": 0.01683499999999999,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.020069919191919194,
      "avg_duration_per_task_ms": 287040.1212121212,
      "avg_output_tokens_per_task": 49971.14141414141,
      "last_run_timestamp": "2026-11-12T20:40:55.525Z"
    },
    {
      "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": 0.8108866666666666,
      "accuracy_std_dev": 0.07631444718764303,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.09796533164983165,
      "avg_duration_per_task_ms": 791756.9158249159,
      "avg_output_tokens_per_task": 39080.49158249158,
      "last_run_timestamp": "2026-10-03T17:58:22.995Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Mistral",
      "accuracy": 0.8107880872313526,
      "accuracy_std_dev": 0.020350576011120062,
      "run_count": 8,
      "total_questions": 1582,
      "avg_cost_per_task_usd": 0.025940944374209858,
      "avg_duration_per_task_ms": 221408.600505689,
      "avg_output_tokens_per_task": 51645.53539823009,
      "last_run_timestamp": "2026-11-11T03:11:48.726Z"
    },
    {
      "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": 0.8107262857142856,
      "accuracy_std_dev": 0.02121111496843299,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.29609035353535357,
      "avg_duration_per_task_ms": 1026681.7662337662,
      "avg_output_tokens_per_task": 67260.45959595959,
      "last_run_timestamp": "2026-10-31T16:12:28.234Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "ModelRun",
      "accuracy": 0.810606,
      "accuracy_std_dev": 0.010546997511140292,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.044063582702020196,
      "avg_duration_per_task_ms": 62743.707702020205,
      "avg_output_tokens_per_task": 15119.022095959595,
      "last_run_timestamp": "2026-10-31T08:43:59.854Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "DeepInfra",
      "accuracy": 0.8099128055555557,
      "accuracy_std_dev": 0.02251018815558855,
      "run_count": 5,
      "total_questions": 612,
      "avg_cost_per_task_usd": 0.06893759803921569,
      "avg_duration_per_task_ms": 1733704.9052287582,
      "avg_output_tokens_per_task": 39280.651960784315,
      "last_run_timestamp": "2026-10-31T16:05:14.633Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "auto-routing",
      "accuracy": 0.8098844285714284,
      "accuracy_std_dev": 0.01753250978575957,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.23733018614718615,
      "avg_duration_per_task_ms": 1080648.3585858585,
      "avg_output_tokens_per_task": 63860.35858585859,
      "last_run_timestamp": "2026-10-31T16:12:28.230Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5-mini-2025-08-07",
      "model_slug": "openai/gpt-5-mini",
      "model_name": "OpenAI: GPT-5 Mini",
      "provider_name": "auto-routing",
      "accuracy": 0.8097645,
      "accuracy_std_dev": 0.021885500000000002,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.044968992424242425,
      "avg_duration_per_task_ms": 208495.61616161617,
      "avg_output_tokens_per_task": 22388.290404040403,
      "last_run_timestamp": "2026-10-01T19:19:28.473Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Chutes",
      "accuracy": 0.8097639999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.06257432828282827,
      "avg_duration_per_task_ms": 366659.297979798,
      "avg_output_tokens_per_task": 47164.93434343435,
      "last_run_timestamp": "2026-11-01T18:00:07.322Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.8096608426103648,
      "accuracy_std_dev": 0.012817931849127972,
      "run_count": 3,
      "total_questions": 521,
      "avg_cost_per_task_usd": 0.0079749443378119,
      "avg_duration_per_task_ms": 289711.4932821497,
      "avg_output_tokens_per_task": 31380.54894433781,
      "last_run_timestamp": "2026-10-31T16:04:23.357Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Morph",
      "accuracy": 0.8081742222222223,
      "accuracy_std_dev": 0.08442583700078994,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.11942196632996631,
      "avg_duration_per_task_ms": 649773.9371492704,
      "avg_output_tokens_per_task": 32692.530303030304,
      "last_run_timestamp": "2026-10-31T17:00:44.290Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "upstage/solar-pro4-20260810",
      "model_slug": "upstage/solar-pro4",
      "model_name": "Upstage: Solar Pro 4",
      "provider_name": "KT",
      "accuracy": 0.808081,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.05294089393939394,
      "avg_duration_per_task_ms": 409688.16161616164,
      "avg_output_tokens_per_task": 43927.25252525252,
      "last_run_timestamp": "2026-11-11T02:02:05.405Z"
    },
    {
      "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": 0.8078402857142857,
      "accuracy_std_dev": 0.013630118652814278,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.02220728354978355,
      "avg_duration_per_task_ms": 246897.44444444444,
      "avg_output_tokens_per_task": 17520.37012987013,
      "last_run_timestamp": "2026-10-03T09:13:36.277Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "DekaLLM",
      "accuracy": 0.807532574733096,
      "accuracy_std_dev": 0.023704551349195916,
      "run_count": 6,
      "total_questions": 1124,
      "avg_cost_per_task_usd": 0.039211227758007124,
      "avg_duration_per_task_ms": 252786.0587188612,
      "avg_output_tokens_per_task": 39158.109430604985,
      "last_run_timestamp": "2026-10-31T16:09:44.822Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Particle AI",
      "accuracy": 0.80723875,
      "accuracy_std_dev": 0.007668707465244688,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.02855746717171717,
      "avg_duration_per_task_ms": 256654.82196969696,
      "avg_output_tokens_per_task": 56910.10353535353,
      "last_run_timestamp": "2026-11-03T01:57:50.570Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5-mini-2025-08-07",
      "model_slug": "openai/gpt-5-mini",
      "model_name": "OpenAI: GPT-5 Mini",
      "provider_name": "Azure",
      "accuracy": 0.8068179999999999,
      "accuracy_std_dev": 0.027363478050131,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.046258910353535354,
      "avg_duration_per_task_ms": 196125.59974747474,
      "avg_output_tokens_per_task": 21946.10606060606,
      "last_run_timestamp": "2026-10-01T19:34:33.810Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8063975,
      "accuracy_std_dev": 0.0016835000000000044,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.17857601262626266,
      "avg_duration_per_task_ms": 685061.2348484849,
      "avg_output_tokens_per_task": 56591.982323232325,
      "last_run_timestamp": "2026-08-29T19:16:33.097Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.8063972857142858,
      "accuracy_std_dev": 0.03755804913811719,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.18206293145743146,
      "avg_duration_per_task_ms": 561423.8109668109,
      "avg_output_tokens_per_task": 56651.43217893218,
      "last_run_timestamp": "2026-10-31T16:15:54.668Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5-mini-2025-08-07",
      "model_slug": "openai/gpt-5-mini",
      "model_name": "OpenAI: GPT-5 Mini",
      "provider_name": "OpenAI",
      "accuracy": 0.806397,
      "accuracy_std_dev": 0.003367000000000009,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.04417995707070707,
      "avg_duration_per_task_ms": 210130.46464646465,
      "avg_output_tokens_per_task": 21993.772727272728,
      "last_run_timestamp": "2026-10-01T19:26:25.385Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "Parasail",
      "accuracy": 0.8047882407261246,
      "accuracy_std_dev": 0.03134006389099038,
      "run_count": 9,
      "total_questions": 1267,
      "avg_cost_per_task_usd": 0.007205067876874505,
      "avg_duration_per_task_ms": 327813.7805840568,
      "avg_output_tokens_per_task": 28563.743488555643,
      "last_run_timestamp": "2026-10-31T16:04:29.125Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.5-sonnet-20250929",
      "model_slug": "anthropic/claude-sonnet-4.5",
      "model_name": "Anthropic: Claude Sonnet 4.5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.8038719166666666,
      "accuracy_std_dev": 0.01983021612530708,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.3387011611952862,
      "avg_duration_per_task_ms": 355211.85016835015,
      "avg_output_tokens_per_task": 22440.643097643097,
      "last_run_timestamp": "2026-10-31T08:53:23.022Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.80303,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.05439790909090909,
      "avg_duration_per_task_ms": 437250.6313131313,
      "avg_output_tokens_per_task": 23570.0202020202,
      "last_run_timestamp": "2026-08-22T17:18:01.115Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Makora",
      "accuracy": 0.8028182609489051,
      "accuracy_std_dev": 0.07654548992882891,
      "run_count": 9,
      "total_questions": 1644,
      "avg_cost_per_task_usd": 0.019102108272506083,
      "avg_duration_per_task_ms": 309069.2311435523,
      "avg_output_tokens_per_task": 40420.16605839416,
      "last_run_timestamp": "2026-10-31T08:33:53.031Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "Venice",
      "accuracy": 0.8016327222511724,
      "accuracy_std_dev": 0.1200400557122913,
      "run_count": 10,
      "total_questions": 1919,
      "avg_cost_per_task_usd": 0.10716173944762898,
      "avg_duration_per_task_ms": 645521.1886399166,
      "avg_output_tokens_per_task": 30270.389265242313,
      "last_run_timestamp": "2026-10-31T16:40:50.323Z"
    },
    {
      "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": 0.8007854444444445,
      "accuracy_std_dev": 0.13709711112858092,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.009282588103254769,
      "avg_duration_per_task_ms": 461356.3843995511,
      "avg_output_tokens_per_task": 29453.966891133558,
      "last_run_timestamp": "2026-11-10T07:09:23.085Z"
    },
    {
      "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": 0.8001147113402061,
      "accuracy_std_dev": 0.1242203500272354,
      "run_count": 3,
      "total_questions": 582,
      "avg_cost_per_task_usd": 0.11411094845360827,
      "avg_duration_per_task_ms": 1146503.6958762887,
      "avg_output_tokens_per_task": 38030.19243986254,
      "last_run_timestamp": "2026-10-24T16:13:05.627Z"
    },
    {
      "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": 0.7996630000000001,
      "accuracy_std_dev": 0.006734000000000018,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.01377076515151515,
      "avg_duration_per_task_ms": 290656.38383838383,
      "avg_output_tokens_per_task": 14276.108585858587,
      "last_run_timestamp": "2026-08-29T19:12:57.906Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "auto-routing",
      "accuracy": 0.7980561818181818,
      "accuracy_std_dev": 0.04308950805605,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.0430184435261708,
      "avg_duration_per_task_ms": 516278.7502295684,
      "avg_output_tokens_per_task": 42641.152892561986,
      "last_run_timestamp": "2026-10-31T16:13:09.867Z"
    },
    {
      "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": 0.7979796666666666,
      "accuracy_std_dev": 0.004956316261992276,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.014001683501683501,
      "avg_duration_per_task_ms": 340373.8703703704,
      "avg_output_tokens_per_task": 14266.20707070707,
      "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": 0.79586,
      "accuracy_std_dev": 0.006604009430641372,
      "run_count": 5,
      "total_questions": 934,
      "avg_cost_per_task_usd": 0.05025589079229122,
      "avg_duration_per_task_ms": 1087344.4850107066,
      "avg_output_tokens_per_task": 22883.693790149893,
      "last_run_timestamp": "2026-10-17T16:06:10.469Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Venice",
      "accuracy": 0.7955151838886546,
      "accuracy_std_dev": 0.021167952217358864,
      "run_count": 12,
      "total_questions": 2371,
      "avg_cost_per_task_usd": 0.0645802066638549,
      "avg_duration_per_task_ms": 194462.45592576973,
      "avg_output_tokens_per_task": 20212.99072121468,
      "last_run_timestamp": "2026-10-31T08:22:42.983Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.1-terminus",
      "model_slug": "deepseek/deepseek-v3.1-terminus",
      "model_name": "DeepSeek: DeepSeek V3.1 Terminus",
      "provider_name": "StreamLake",
      "accuracy": 0.7954545,
      "accuracy_std_dev": 0.010942500000000022,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.014605156565656566,
      "avg_duration_per_task_ms": 301530.7196969697,
      "avg_output_tokens_per_task": 13980.005050505051,
      "last_run_timestamp": "2026-08-29T16:10:33.070Z"
    },
    {
      "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": 0.7950563915254236,
      "accuracy_std_dev": 0.05895769598336308,
      "run_count": 6,
      "total_questions": 1180,
      "avg_cost_per_task_usd": 0.007454267796610169,
      "avg_duration_per_task_ms": 818243.9161016949,
      "avg_output_tokens_per_task": 18528.625423728812,
      "last_run_timestamp": "2026-10-24T16:21:54.634Z"
    },
    {
      "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": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7943721428571427,
      "accuracy_std_dev": 0.008070114451809616,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.017872657287157288,
      "avg_duration_per_task_ms": 60674.55339105339,
      "avg_output_tokens_per_task": 5803.906204906205,
      "last_run_timestamp": "2026-10-31T16:10:36.258Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "Venice",
      "accuracy": 0.7942385555555556,
      "accuracy_std_dev": 0.012314380393228746,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.004576078563411897,
      "avg_duration_per_task_ms": 256060.55387205386,
      "avg_output_tokens_per_task": 29957.45061728395,
      "last_run_timestamp": "2026-10-17T16:03:34.733Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "Phala",
      "accuracy": 0.7934233709949409,
      "accuracy_std_dev": 0.05668542998401107,
      "run_count": 6,
      "total_questions": 1186,
      "avg_cost_per_task_usd": 0.01885696121416526,
      "avg_duration_per_task_ms": 1103043.2150084318,
      "avg_output_tokens_per_task": 18169.33305227656,
      "last_run_timestamp": "2026-10-31T16:45:39.993Z"
    },
    {
      "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": 0.7932098333333334,
      "accuracy_std_dev": 0.006693092768336313,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.05432222053872054,
      "avg_duration_per_task_ms": 460720.4478114478,
      "avg_output_tokens_per_task": 23581.273569023568,
      "last_run_timestamp": "2026-10-17T16:03:34.700Z"
    },
    {
      "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": 0.7923028091603053,
      "accuracy_std_dev": 0.012838036519947284,
      "run_count": 8,
      "total_questions": 1572,
      "avg_cost_per_task_usd": 0.01830340776081425,
      "avg_duration_per_task_ms": 189913.7779898219,
      "avg_output_tokens_per_task": 15102.418575063613,
      "last_run_timestamp": "2026-10-08T22:29:11.504Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7920875,
      "accuracy_std_dev": 0.007575500000000013,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.02811962121212121,
      "avg_duration_per_task_ms": 464497.5707070707,
      "avg_output_tokens_per_task": 23480.530303030304,
      "last_run_timestamp": "2026-08-29T18:09:57.600Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "GMICloud",
      "accuracy": 0.7913752307692307,
      "accuracy_std_dev": 0.03764112974915735,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.10873279254079253,
      "avg_duration_per_task_ms": 370663.50815850816,
      "avg_output_tokens_per_task": 27165.075757575756,
      "last_run_timestamp": "2026-11-10T06:47:58.589Z"
    },
    {
      "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": 0.791077650901999,
      "accuracy_std_dev": 0.011802764782558223,
      "run_count": 11,
      "total_questions": 2051,
      "avg_cost_per_task_usd": 0.00883669673330083,
      "avg_duration_per_task_ms": 380017.00048756704,
      "avg_output_tokens_per_task": 34825.446123842026,
      "last_run_timestamp": "2026-10-31T16:15:26.662Z"
    },
    {
      "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": 0.7909730112634672,
      "accuracy_std_dev": 0.015262253874258334,
      "run_count": 11,
      "total_questions": 2042,
      "avg_cost_per_task_usd": 0.004672809010773752,
      "avg_duration_per_task_ms": 1542673.8849167484,
      "avg_output_tokens_per_task": 30652.18511263467,
      "last_run_timestamp": "2026-10-31T16:08:08.537Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "auto-routing",
      "accuracy": 0.7905242857142856,
      "accuracy_std_dev": 0.03382434521682478,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.16089077777777777,
      "avg_duration_per_task_ms": 885338.5844155845,
      "avg_output_tokens_per_task": 66850.18398268399,
      "last_run_timestamp": "2026-10-31T16:04:51.396Z"
    },
    {
      "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": 0.7888407142857142,
      "accuracy_std_dev": 0.009727651246601654,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.0210191240981241,
      "avg_duration_per_task_ms": 240196.7041847042,
      "avg_output_tokens_per_task": 12493.777056277057,
      "last_run_timestamp": "2026-10-31T16:10:36.260Z"
    },
    {
      "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": 0.7887204999999999,
      "accuracy_std_dev": 0.022727499999999956,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.02925468181818182,
      "avg_duration_per_task_ms": 425922.92171717173,
      "avg_output_tokens_per_task": 25932.09595959596,
      "last_run_timestamp": "2026-08-29T18:05:36.579Z"
    },
    {
      "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": 0.7886696422660725,
      "accuracy_std_dev": 0.030494803180214167,
      "run_count": 8,
      "total_questions": 1571,
      "avg_cost_per_task_usd": 0.011359753660089117,
      "avg_duration_per_task_ms": 640851.9363462763,
      "avg_output_tokens_per_task": 37620.8707829408,
      "last_run_timestamp": "2026-10-31T16:03:21.448Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "Io Net",
      "accuracy": 0.7884532538560409,
      "accuracy_std_dev": 0.014149184070446418,
      "run_count": 8,
      "total_questions": 1556,
      "avg_cost_per_task_usd": 0.011693410025706941,
      "avg_duration_per_task_ms": 858091.7879177377,
      "avg_output_tokens_per_task": 38791.87210796915,
      "last_run_timestamp": "2026-11-11T08:00:18.448Z"
    },
    {
      "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": 0.7883595714285714,
      "accuracy_std_dev": 0.03112971003424745,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.03924310822510823,
      "avg_duration_per_task_ms": 198693.47186147186,
      "avg_output_tokens_per_task": 21702.200577200576,
      "last_run_timestamp": "2026-10-31T16:08:31.063Z"
    },
    {
      "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": 0.7878784999999998,
      "accuracy_std_dev": 0.0016835000000000044,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.006171542929292929,
      "avg_duration_per_task_ms": 348997.62626262626,
      "avg_output_tokens_per_task": 14953.704545454546,
      "last_run_timestamp": "2026-08-29T19:20:50.265Z"
    },
    {
      "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": 0.7878303017902814,
      "accuracy_std_dev": 0.15859528241072463,
      "run_count": 8,
      "total_questions": 1564,
      "avg_cost_per_task_usd": 0.01594299360613811,
      "avg_duration_per_task_ms": 555279.5345268542,
      "avg_output_tokens_per_task": 26007.55051150895,
      "last_run_timestamp": "2026-10-17T08:10:43.825Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.7872267663934426,
      "accuracy_std_dev": 0.013963506030602306,
      "run_count": 3,
      "total_questions": 488,
      "avg_cost_per_task_usd": 0.008632290983606556,
      "avg_duration_per_task_ms": 586832.4200819673,
      "avg_output_tokens_per_task": 37091.375,
      "last_run_timestamp": "2026-10-31T16:20:33.127Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-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": 0.7871492666666666,
      "accuracy_std_dev": 0.11453375331110427,
      "run_count": 15,
      "total_questions": 2970,
      "avg_cost_per_task_usd": 0.01689483905723906,
      "avg_duration_per_task_ms": 609264.4754208755,
      "avg_output_tokens_per_task": 30856.628282828282,
      "last_run_timestamp": "2026-11-06T22:10:48.831Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4-sonnet-20250522",
      "model_slug": "anthropic/claude-sonnet-4",
      "model_name": "Anthropic: Claude Sonnet 4",
      "provider_name": "auto-routing",
      "accuracy": 0.786195,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.2643429696969697,
      "avg_duration_per_task_ms": 180778.88888888888,
      "avg_output_tokens_per_task": 17428.555555555555,
      "last_run_timestamp": "2026-09-01T13:19:59.945Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.786195,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.057962348484848485,
      "avg_duration_per_task_ms": 638736.3484848485,
      "avg_output_tokens_per_task": 23055.60101010101,
      "last_run_timestamp": "2026-08-22T17:09:41.940Z"
    },
    {
      "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": 0.7852381866666666,
      "accuracy_std_dev": 0.022517162489606328,
      "run_count": 6,
      "total_questions": 1050,
      "avg_cost_per_task_usd": 0.039451490476190475,
      "avg_duration_per_task_ms": 759087.4819047619,
      "avg_output_tokens_per_task": 18229.481904761906,
      "last_run_timestamp": "2026-10-17T16:06:15.029Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Nebius",
      "accuracy": 0.78493275,
      "accuracy_std_dev": 0.010934561842501992,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.03407076388888889,
      "avg_duration_per_task_ms": 483005.77398989897,
      "avg_output_tokens_per_task": 67906.58964646465,
      "last_run_timestamp": "2026-10-31T08:31:13.285Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "Darkbloom",
      "accuracy": 0.782641111111111,
      "accuracy_std_dev": 0.020988606207731148,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.020732816498316497,
      "avg_duration_per_task_ms": 463791.91694725025,
      "avg_output_tokens_per_task": 27564.918630751963,
      "last_run_timestamp": "2026-11-13T02:10:04.930Z"
    },
    {
      "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": 0.7824160174311926,
      "accuracy_std_dev": 0.017840425618521553,
      "run_count": 8,
      "total_questions": 1090,
      "avg_cost_per_task_usd": 0.04672766880733945,
      "avg_duration_per_task_ms": 971010.7899082568,
      "avg_output_tokens_per_task": 23260.208256880735,
      "last_run_timestamp": "2026-10-31T16:04:29.123Z"
    },
    {
      "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": 0.7823447220338983,
      "accuracy_std_dev": 0.020068261404654758,
      "run_count": 6,
      "total_questions": 1180,
      "avg_cost_per_task_usd": 0.022374091525423728,
      "avg_duration_per_task_ms": 880666.9830508474,
      "avg_output_tokens_per_task": 37830.4686440678,
      "last_run_timestamp": "2026-10-24T16:37:21.981Z"
    },
    {
      "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": 0.7818664285714286,
      "accuracy_std_dev": 0.016143829133821997,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.3387973888888889,
      "avg_duration_per_task_ms": 1030179.549062049,
      "avg_output_tokens_per_task": 85385.09307359307,
      "last_run_timestamp": "2026-10-31T16:04:02.366Z"
    },
    {
      "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": 0.781081,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 185,
      "avg_cost_per_task_usd": 0.04895784324324324,
      "avg_duration_per_task_ms": 686648.6648648649,
      "avg_output_tokens_per_task": 21222.45945945946,
      "last_run_timestamp": "2026-08-22T16:06:14.894Z"
    },
    {
      "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": "PrimeIntellect",
      "accuracy": 0.77933,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 179,
      "avg_cost_per_task_usd": 0.00797877653631285,
      "avg_duration_per_task_ms": 499606.7541899441,
      "avg_output_tokens_per_task": 38804.5530726257,
      "last_run_timestamp": "2026-11-01T18:00:07.325Z"
    },
    {
      "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": 0.7792742222222223,
      "accuracy_std_dev": 0.04102605862613494,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.2320418187429854,
      "avg_duration_per_task_ms": 897509.8860830527,
      "avg_output_tokens_per_task": 77846.84624017957,
      "last_run_timestamp": "2026-10-17T16:06:46.266Z"
    },
    {
      "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": 0.7787130000000001,
      "accuracy_std_dev": 0.07769734491388151,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.1402097396184063,
      "avg_duration_per_task_ms": 939428.9349046015,
      "avg_output_tokens_per_task": 73044.72109988777,
      "last_run_timestamp": "2026-11-10T07:06:06.465Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.4-nano-20260317",
      "model_slug": "openai/gpt-5.4-nano",
      "model_name": "OpenAI: GPT-5.4 Nano",
      "provider_name": "auto-routing",
      "accuracy": 0.776936,
      "accuracy_std_dev": 0.004208999999999963,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.013052893939393939,
      "avg_duration_per_task_ms": 64108.87373737374,
      "avg_output_tokens_per_task": 10319.823232323231,
      "last_run_timestamp": "2026-10-01T19:12:45.924Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.4-nano-20260317",
      "model_slug": "openai/gpt-5.4-nano",
      "model_name": "OpenAI: GPT-5.4 Nano",
      "provider_name": "OpenAI",
      "accuracy": 0.7760940000000002,
      "accuracy_std_dev": 0.012598160421267844,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.013424281144781143,
      "avg_duration_per_task_ms": 64995.4595959596,
      "avg_output_tokens_per_task": 10615.195286195287,
      "last_run_timestamp": "2026-10-02T23:20:38.343Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "GMICloud",
      "accuracy": 0.7752525,
      "accuracy_std_dev": 0.02576620138055538,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.011368148148148148,
      "avg_duration_per_task_ms": 910071.1422558923,
      "avg_output_tokens_per_task": 43361.55303030303,
      "last_run_timestamp": "2026-10-31T09:00:40.586Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "DeepInfra",
      "accuracy": 0.7749944977228367,
      "accuracy_std_dev": 0.02795561944851158,
      "run_count": 8,
      "total_questions": 1537,
      "avg_cost_per_task_usd": 0.007373610930383865,
      "avg_duration_per_task_ms": 852404.3311646064,
      "avg_output_tokens_per_task": 18993.68770331815,
      "last_run_timestamp": "2026-10-31T16:04:29.109Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Cerebras",
      "accuracy": 0.7748314999999999,
      "accuracy_std_dev": 0.0012624999999999997,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.004441212121212121,
      "avg_duration_per_task_ms": 4186.169191919192,
      "avg_output_tokens_per_task": 2455.989898989899,
      "last_run_timestamp": "2026-08-29T18:58:30.919Z"
    },
    {
      "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": 0.7746466307789741,
      "accuracy_std_dev": 0.006675073159711417,
      "run_count": 8,
      "total_questions": 1579,
      "avg_cost_per_task_usd": 0.011284899936668778,
      "avg_duration_per_task_ms": 924382.4306523115,
      "avg_output_tokens_per_task": 11730.89170360988,
      "last_run_timestamp": "2026-10-31T16:02:14.358Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "DekaLLM",
      "accuracy": 0.7744107142857142,
      "accuracy_std_dev": 0.007947545348170354,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.005179176046176046,
      "avg_duration_per_task_ms": 801836.0988455989,
      "avg_output_tokens_per_task": 28610.891053391053,
      "last_run_timestamp": "2026-10-31T16:17:18.501Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "DeepInfra",
      "accuracy": 0.7741524627118646,
      "accuracy_std_dev": 0.024920321732780762,
      "run_count": 7,
      "total_questions": 1180,
      "avg_cost_per_task_usd": 0.010626239830508475,
      "avg_duration_per_task_ms": 1250036.8076271187,
      "avg_output_tokens_per_task": 40411.0593220339,
      "last_run_timestamp": "2026-10-31T08:17:43.227Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "auto-routing",
      "accuracy": 0.7739515454545453,
      "accuracy_std_dev": 0.019647060545183753,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.005980178604224059,
      "avg_duration_per_task_ms": 705416.41184573,
      "avg_output_tokens_per_task": 37567.65151515151,
      "last_run_timestamp": "2026-10-31T16:03:21.447Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.4-nano-20260317",
      "model_slug": "openai/gpt-5.4-nano",
      "model_name": "OpenAI: GPT-5.4 Nano",
      "provider_name": "Azure",
      "accuracy": 0.7735690000000001,
      "accuracy_std_dev": 0.005892142861472351,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.014036280303030302,
      "avg_duration_per_task_ms": 123381.52777777778,
      "avg_output_tokens_per_task": 10576.208333333334,
      "last_run_timestamp": "2026-10-01T19:20:04.986Z"
    },
    {
      "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": 0.773333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 175,
      "avg_cost_per_task_usd": 0.13656869142857142,
      "avg_duration_per_task_ms": 1521793.1542857143,
      "avg_output_tokens_per_task": 38871.68571428571,
      "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": "Venice",
      "accuracy": 0.7732982293193716,
      "accuracy_std_dev": 0.013698592432801259,
      "run_count": 5,
      "total_questions": 955,
      "avg_cost_per_task_usd": 0.00232973612565445,
      "avg_duration_per_task_ms": 792784.167539267,
      "avg_output_tokens_per_task": 15348.367539267016,
      "last_run_timestamp": "2026-10-31T16:10:36.260Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Parasail",
      "accuracy": 0.7729887726692208,
      "accuracy_std_dev": 0.01457930786894563,
      "run_count": 8,
      "total_questions": 1566,
      "avg_cost_per_task_usd": 0.02178225223499361,
      "avg_duration_per_task_ms": 240176.16922094507,
      "avg_output_tokens_per_task": 29760.224137931036,
      "last_run_timestamp": "2026-11-05T00:00:07.969Z"
    },
    {
      "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": 0.7722016904681711,
      "accuracy_std_dev": 0.014666151853692065,
      "run_count": 15,
      "total_questions": 2969,
      "avg_cost_per_task_usd": 0.006039911754799596,
      "avg_duration_per_task_ms": 437079.74840013473,
      "avg_output_tokens_per_task": 29140.058268777368,
      "last_run_timestamp": "2026-10-31T16:02:14.375Z"
    },
    {
      "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": 0.772166,
      "accuracy_std_dev": 0.04805772781423054,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.01390386111111111,
      "avg_duration_per_task_ms": 894648.4250841751,
      "avg_output_tokens_per_task": 40593.27777777778,
      "last_run_timestamp": "2026-10-17T08:18:06.210Z"
    },
    {
      "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": 0.7718855,
      "accuracy_std_dev": 0.0025254999999999583,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.017657098484848484,
      "avg_duration_per_task_ms": 452144.0101010101,
      "avg_output_tokens_per_task": 17493.924242424244,
      "last_run_timestamp": "2026-08-29T18:21:38.666Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Io Net",
      "accuracy": 0.771605,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 162,
      "avg_cost_per_task_usd": 0.0022807283950617282,
      "avg_duration_per_task_ms": 290757.45061728393,
      "avg_output_tokens_per_task": 15097.716049382716,
      "last_run_timestamp": "2026-11-11T15:03:02.397Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "CoreWeave",
      "accuracy": 0.7714860192771085,
      "accuracy_std_dev": 0.016470310033005864,
      "run_count": 6,
      "total_questions": 830,
      "avg_cost_per_task_usd": 0.009284253012048193,
      "avg_duration_per_task_ms": 720483.7698795181,
      "avg_output_tokens_per_task": 30820.519277108433,
      "last_run_timestamp": "2026-10-31T16:24:40.574Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Xiaomi",
      "accuracy": 0.7713804000000001,
      "accuracy_std_dev": 0.01747606822600553,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.014499158585858585,
      "avg_duration_per_task_ms": 1080917.518181818,
      "avg_output_tokens_per_task": 51478.64141414141,
      "last_run_timestamp": "2026-10-31T08:09:31.070Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "CoreWeave",
      "accuracy": 0.7712842857142856,
      "accuracy_std_dev": 0.011713027432421244,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.00498178354978355,
      "avg_duration_per_task_ms": 549029.411976912,
      "avg_output_tokens_per_task": 29264.81168831169,
      "last_run_timestamp": "2026-10-31T16:20:12.041Z"
    },
    {
      "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": 0.7710438,
      "accuracy_std_dev": 0.0072995281874926425,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.020946333333333338,
      "avg_duration_per_task_ms": 767702.9585858586,
      "avg_output_tokens_per_task": 52149.88787878788,
      "last_run_timestamp": "2026-10-31T16:23:19.027Z"
    },
    {
      "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": 0.7706229166666667,
      "accuracy_std_dev": 0.010605727544259059,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.014841917929292927,
      "avg_duration_per_task_ms": 138155.14436026936,
      "avg_output_tokens_per_task": 29486.722643097644,
      "last_run_timestamp": "2026-10-31T16:51:23.396Z"
    },
    {
      "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": 0.7705627142857142,
      "accuracy_std_dev": 0.03293717088949932,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.06274982683982684,
      "avg_duration_per_task_ms": 1017904.803030303,
      "avg_output_tokens_per_task": 30923.48124098124,
      "last_run_timestamp": "2026-10-31T16:08:31.065Z"
    },
    {
      "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": 0.7678882984293194,
      "accuracy_std_dev": 0.015081000000000011,
      "run_count": 2,
      "total_questions": 191,
      "avg_cost_per_task_usd": 0.038533612565445026,
      "avg_duration_per_task_ms": 148852.86910994764,
      "avg_output_tokens_per_task": 40034.97905759162,
      "last_run_timestamp": "2026-08-29T19:44:46.543Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "SambaNova",
      "accuracy": 0.7673434936548225,
      "accuracy_std_dev": 0.011201352155404033,
      "run_count": 8,
      "total_questions": 1576,
      "avg_cost_per_task_usd": 0.025760851522842635,
      "avg_duration_per_task_ms": 78397.31725888325,
      "avg_output_tokens_per_task": 26980.187817258884,
      "last_run_timestamp": "2026-11-11T18:52:31.116Z"
    },
    {
      "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": 0.7671418075356415,
      "accuracy_std_dev": 0.022561888094930423,
      "run_count": 7,
      "total_questions": 982,
      "avg_cost_per_task_usd": 0.013307315682281062,
      "avg_duration_per_task_ms": 53274.17718940937,
      "avg_output_tokens_per_task": 17643.751527494907,
      "last_run_timestamp": "2026-10-31T16:39:49.876Z"
    },
    {
      "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": 0.7670034,
      "accuracy_std_dev": 0.04110764542320566,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.052571260606060606,
      "avg_duration_per_task_ms": 891054.0545454546,
      "avg_output_tokens_per_task": 23291.960606060606,
      "last_run_timestamp": "2026-10-17T16:03:24.588Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "DekaLLM",
      "accuracy": 0.7669798106936416,
      "accuracy_std_dev": 0.011686708686296891,
      "run_count": 7,
      "total_questions": 1384,
      "avg_cost_per_task_usd": 0.019905706647398845,
      "avg_duration_per_task_ms": 1154492.3439306358,
      "avg_output_tokens_per_task": 60181.387283236996,
      "last_run_timestamp": "2026-10-31T16:03:21.445Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4-sonnet-20250522",
      "model_slug": "anthropic/claude-sonnet-4",
      "model_name": "Anthropic: Claude Sonnet 4",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7669488305084746,
      "accuracy_std_dev": 0.016365499999999977,
      "run_count": 2,
      "total_questions": 354,
      "avg_cost_per_task_usd": 0.21194037853107345,
      "avg_duration_per_task_ms": 159764.62146892655,
      "avg_output_tokens_per_task": 13982.706214689266,
      "last_run_timestamp": "2026-09-01T13:33:46.786Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-2.5-flash",
      "model_slug": "google/gemini-2.5-flash",
      "model_name": "Google: Gemini 2.5 Flash",
      "provider_name": "Google AI Studio",
      "accuracy": 0.7668349999999999,
      "accuracy_std_dev": 0.004208999999999963,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.057140755050505046,
      "avg_duration_per_task_ms": 101024.92676767676,
      "avg_output_tokens_per_task": 22766.989898989897,
      "last_run_timestamp": "2026-10-01T19:11:42.323Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Nebius",
      "accuracy": 0.76599325,
      "accuracy_std_dev": 0.01423513635998966,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.018594656565656564,
      "avg_duration_per_task_ms": 110236.10416666667,
      "avg_output_tokens_per_task": 30754.41792929293,
      "last_run_timestamp": "2026-10-31T16:52:10.938Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Reka",
      "accuracy": 0.7653769053571428,
      "accuracy_std_dev": 0.011846747415441054,
      "run_count": 17,
      "total_questions": 3360,
      "avg_cost_per_task_usd": 0.016902593452380953,
      "avg_duration_per_task_ms": 110624.59017857142,
      "avg_output_tokens_per_task": 5988.1654761904765,
      "last_run_timestamp": "2026-10-31T09:06:38.813Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Near AI",
      "accuracy": 0.76536225,
      "accuracy_std_dev": 0.00536920838555365,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.004458252525252525,
      "avg_duration_per_task_ms": 487032.047979798,
      "avg_output_tokens_per_task": 29595.234848484848,
      "last_run_timestamp": "2026-10-31T16:16:55.698Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Tenstorrent",
      "accuracy": 0.7653508070175438,
      "accuracy_std_dev": 0.03268549999999998,
      "run_count": 2,
      "total_questions": 228,
      "avg_cost_per_task_usd": 0.014143986842105261,
      "avg_duration_per_task_ms": 648386.7105263158,
      "avg_output_tokens_per_task": 35206.833333333336,
      "last_run_timestamp": "2026-11-08T04:33:18.539Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "OpenInference",
      "accuracy": 0.7653198,
      "accuracy_std_dev": 0.07983844813722271,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.004923061616161616,
      "avg_duration_per_task_ms": 448741.63535353536,
      "avg_output_tokens_per_task": 16983.609090909093,
      "last_run_timestamp": "2026-10-17T08:26:03.926Z"
    },
    {
      "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": 0.7650311428571429,
      "accuracy_std_dev": 0.012201722845900242,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.021599636363636363,
      "avg_duration_per_task_ms": 1135053.7258297259,
      "avg_output_tokens_per_task": 53902.65151515151,
      "last_run_timestamp": "2026-10-31T16:51:21.148Z"
    },
    {
      "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": 0.7649719016949154,
      "accuracy_std_dev": 0.008301844132480422,
      "run_count": 3,
      "total_questions": 590,
      "avg_cost_per_task_usd": 0.0026733169491525424,
      "avg_duration_per_task_ms": 102640.67457627118,
      "avg_output_tokens_per_task": 31256.29491525424,
      "last_run_timestamp": "2026-10-01T16:04:09.046Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.76431,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0026081363636363633,
      "avg_duration_per_task_ms": 302574.7474747475,
      "avg_output_tokens_per_task": 27415.863636363636,
      "last_run_timestamp": "2026-11-01T06:00:07.957Z"
    },
    {
      "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": 0.7643099999999999,
      "accuracy_std_dev": 0.01683499999999999,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.057833444444444444,
      "avg_duration_per_task_ms": 573166.5858585859,
      "avg_output_tokens_per_task": 26406.381313131315,
      "last_run_timestamp": "2026-08-29T18:26:43.144Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Sail Research",
      "accuracy": 0.7640448089887641,
      "accuracy_std_dev": 0.007632810273199963,
      "run_count": 3,
      "total_questions": 534,
      "avg_cost_per_task_usd": 0.012456207865168538,
      "avg_duration_per_task_ms": 247061.07865168538,
      "avg_output_tokens_per_task": 31050.842696629214,
      "last_run_timestamp": "2026-11-06T21:57:37.505Z"
    },
    {
      "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": 0.7635882857142857,
      "accuracy_std_dev": 0.013965487794808272,
      "run_count": 14,
      "total_questions": 2772,
      "avg_cost_per_task_usd": 0.01827745598845599,
      "avg_duration_per_task_ms": 144604.34848484848,
      "avg_output_tokens_per_task": 30221.237734487735,
      "last_run_timestamp": "2026-10-31T16:08:33.703Z"
    },
    {
      "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": 0.7633477142857142,
      "accuracy_std_dev": 0.011623763131918104,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.01745055844155844,
      "avg_duration_per_task_ms": 63367.67532467532,
      "avg_output_tokens_per_task": 28845.276334776336,
      "last_run_timestamp": "2026-10-31T16:23:59.761Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Crusoe",
      "accuracy": 0.7632998,
      "accuracy_std_dev": 0.007030464092789305,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.007469111111111111,
      "avg_duration_per_task_ms": 123178.52121212121,
      "avg_output_tokens_per_task": 29681.70909090909,
      "last_run_timestamp": "2026-10-31T16:38:46.120Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4-sonnet-20250522",
      "model_slug": "anthropic/claude-sonnet-4",
      "model_name": "Anthropic: Claude Sonnet 4",
      "provider_name": "Google",
      "accuracy": 0.7629069999999999,
      "accuracy_std_dev": 0.00819966182229484,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.2718667558922559,
      "avg_duration_per_task_ms": 264601.48148148146,
      "avg_output_tokens_per_task": 17934.003367003366,
      "last_run_timestamp": "2026-09-01T13:23:02.059Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-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": "inclusionai/ling-3.0-flash-20260723",
      "model_slug": "inclusionai/ling-3.0-flash",
      "model_name": "inclusionAI: Ling 3.0 Flash",
      "provider_name": "DeepInfra",
      "accuracy": 0.7614677541284404,
      "accuracy_std_dev": 0.10796943486314385,
      "run_count": 3,
      "total_questions": 545,
      "avg_cost_per_task_usd": 0.0038067559633027524,
      "avg_duration_per_task_ms": 346287.8605504587,
      "avg_output_tokens_per_task": 20923.759633027523,
      "last_run_timestamp": "2026-10-08T22:27:01.555Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Together",
      "accuracy": 0.7602212857142857,
      "accuracy_std_dev": 0.01112536459388809,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.017620217893217896,
      "avg_duration_per_task_ms": 231549.2683982684,
      "avg_output_tokens_per_task": 29213.04834054834,
      "last_run_timestamp": "2026-10-31T16:22:37.563Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "mistralai/mistral-small-2603",
      "model_slug": "mistralai/mistral-small-2603",
      "model_name": "Mistral: Mistral Small 4",
      "provider_name": "auto-routing",
      "accuracy": 0.7575759999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.014556641414141414,
      "avg_duration_per_task_ms": 173179.91919191918,
      "avg_output_tokens_per_task": 23980.964646464647,
      "last_run_timestamp": "2026-09-01T13:30:26.223Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Phala",
      "accuracy": 0.7574711434482759,
      "accuracy_std_dev": 0.05835625158448523,
      "run_count": 11,
      "total_questions": 2175,
      "avg_cost_per_task_usd": 0.014400611954022986,
      "avg_duration_per_task_ms": 105760.68091954023,
      "avg_output_tokens_per_task": 6123.435402298851,
      "last_run_timestamp": "2026-10-31T08:19:36.628Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Novita",
      "accuracy": 0.7563452081218275,
      "accuracy_std_dev": 0.0358022938821275,
      "run_count": 6,
      "total_questions": 1182,
      "avg_cost_per_task_usd": 0.014391308798646361,
      "avg_duration_per_task_ms": 1040556.6945854484,
      "avg_output_tokens_per_task": 51071.41201353638,
      "last_run_timestamp": "2026-10-31T08:41:02.325Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "mistralai/mistral-small-2603",
      "model_slug": "mistralai/mistral-small-2603",
      "model_name": "Mistral: Mistral Small 4",
      "provider_name": "Venice",
      "accuracy": 0.755892,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.018171858585858586,
      "avg_duration_per_task_ms": 285760.0353535354,
      "avg_output_tokens_per_task": 24027.954545454544,
      "last_run_timestamp": "2026-09-01T12:59:55.618Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-coder-next-2025-02-03",
      "model_slug": "qwen/qwen3-coder-next",
      "model_name": "Qwen: Qwen3 Coder Next",
      "provider_name": "Alibaba",
      "accuracy": 0.75477,
      "accuracy_std_dev": 0.008054024625407281,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.022580063973063974,
      "avg_duration_per_task_ms": 93134.21212121213,
      "avg_output_tokens_per_task": 14890.888888888889,
      "last_run_timestamp": "2026-10-03T16:56:57.458Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "DeepInfra",
      "accuracy": 0.7539042317380352,
      "accuracy_std_dev": 0.03122595132646972,
      "run_count": 12,
      "total_questions": 1985,
      "avg_cost_per_task_usd": 0.045778590428211585,
      "avg_duration_per_task_ms": 263292.8277078086,
      "avg_output_tokens_per_task": 17254.131486146096,
      "last_run_timestamp": "2026-10-31T08:39:08.315Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "Baidu",
      "accuracy": 0.753787875,
      "accuracy_std_dev": 0.02753108760309653,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.2093788106060606,
      "avg_duration_per_task_ms": 1100289.5448232323,
      "avg_output_tokens_per_task": 94261.211489899,
      "last_run_timestamp": "2026-11-10T19:04:50.024Z"
    },
    {
      "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": 0.7534605555555556,
      "accuracy_std_dev": 0.03816866389275402,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.350709655443322,
      "avg_duration_per_task_ms": 957253.5684624017,
      "avg_output_tokens_per_task": 84600.8125701459,
      "last_run_timestamp": "2026-11-10T07:04:16.019Z"
    },
    {
      "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": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "DeepInfra",
      "accuracy": 0.7526884764267989,
      "accuracy_std_dev": 0.036890165142246543,
      "run_count": 3,
      "total_questions": 403,
      "avg_cost_per_task_usd": 0.056140801488833755,
      "avg_duration_per_task_ms": 723763.2431761787,
      "avg_output_tokens_per_task": 50423.94789081886,
      "last_run_timestamp": "2026-08-29T19:07:57.500Z"
    },
    {
      "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": 0.75252525,
      "accuracy_std_dev": 0.009888238426914091,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.01365957702020202,
      "avg_duration_per_task_ms": 93762.47853535354,
      "avg_output_tokens_per_task": 15014.820707070707,
      "last_run_timestamp": "2026-10-03T16:54:39.591Z"
    },
    {
      "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": 0.7525252,
      "accuracy_std_dev": 0.014481875091299456,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.015641921717171717,
      "avg_duration_per_task_ms": 1362988.2676767677,
      "avg_output_tokens_per_task": 56157.93787878788,
      "last_run_timestamp": "2026-11-01T13:06:18.223Z"
    },
    {
      "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": "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": 0.7514741137971698,
      "accuracy_std_dev": 0.03193872589510217,
      "run_count": 9,
      "total_questions": 1696,
      "avg_cost_per_task_usd": 0.014983457547169812,
      "avg_duration_per_task_ms": 382125.23054245283,
      "avg_output_tokens_per_task": 49724.931603773584,
      "last_run_timestamp": "2026-11-11T06:34:22.186Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "Makora",
      "accuracy": 0.7514029999999999,
      "accuracy_std_dev": 0.013279676389129373,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.015289352693602695,
      "avg_duration_per_task_ms": 350945.7154882155,
      "avg_output_tokens_per_task": 62534.38636363636,
      "last_run_timestamp": "2026-10-31T16:38:47.420Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.7510417714285714,
      "accuracy_std_dev": 0.04292803699474532,
      "run_count": 6,
      "total_questions": 1120,
      "avg_cost_per_task_usd": 0.018983601785714288,
      "avg_duration_per_task_ms": 779779.53125,
      "avg_output_tokens_per_task": 47360.87946428572,
      "last_run_timestamp": "2026-10-31T16:47:37.740Z"
    },
    {
      "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": 0.750842,
      "accuracy_std_dev": 0.010996575951934644,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.014766501683501684,
      "avg_duration_per_task_ms": 106033.30808080808,
      "avg_output_tokens_per_task": 14469.075757575758,
      "last_run_timestamp": "2026-10-03T16:43:33.673Z"
    },
    {
      "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": 0.7495937585365854,
      "accuracy_std_dev": 0.031913877008090064,
      "run_count": 4,
      "total_questions": 410,
      "avg_cost_per_task_usd": 0.047968058536585374,
      "avg_duration_per_task_ms": 440215.96341463417,
      "avg_output_tokens_per_task": 14034.10731707317,
      "last_run_timestamp": "2026-09-19T08:53:38.031Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "DeepInfra",
      "accuracy": 0.7492805738953785,
      "accuracy_std_dev": 0.12024394573944976,
      "run_count": 11,
      "total_questions": 1969,
      "avg_cost_per_task_usd": 0.0343216658202133,
      "avg_duration_per_task_ms": 1417972.940578974,
      "avg_output_tokens_per_task": 36042.608938547484,
      "last_run_timestamp": "2026-10-31T16:04:48.182Z"
    },
    {
      "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": 0.749158,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.021431318181818183,
      "avg_duration_per_task_ms": 113907.66161616161,
      "avg_output_tokens_per_task": 12443.959595959595,
      "last_run_timestamp": "2026-08-22T17:12:08.919Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-thinking-2507",
      "model_slug": "qwen/qwen3-235b-a22b-thinking-2507",
      "model_name": "Qwen: Qwen3 235B A22B Thinking 2507",
      "provider_name": "Venice",
      "accuracy": 0.749158,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.07988080808080808,
      "avg_duration_per_task_ms": 853640.3434343435,
      "avg_output_tokens_per_task": 22758.171717171717,
      "last_run_timestamp": "2026-08-22T17:38:59.682Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "inclusionai/ling-3.0-flash-20260723",
      "model_slug": "inclusionai/ling-3.0-flash",
      "model_name": "inclusionAI: Ling 3.0 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.7487311624365482,
      "accuracy_std_dev": 0.019042500000000018,
      "run_count": 2,
      "total_questions": 394,
      "avg_cost_per_task_usd": 0.00260698730964467,
      "avg_duration_per_task_ms": 166191.14467005077,
      "avg_output_tokens_per_task": 30980.81725888325,
      "last_run_timestamp": "2026-10-01T16:04:24.420Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "Wafer",
      "accuracy": 0.7481013645569621,
      "accuracy_std_dev": 0.004003499999999993,
      "run_count": 2,
      "total_questions": 395,
      "avg_cost_per_task_usd": 0.010130237974683544,
      "avg_duration_per_task_ms": 298779.6734177215,
      "avg_output_tokens_per_task": 60721.07341772152,
      "last_run_timestamp": "2026-10-10T08:39:12.933Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.7453702499999999,
      "accuracy_std_dev": 0.0095511571910162,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.006491083333333334,
      "avg_duration_per_task_ms": 157741.5366161616,
      "avg_output_tokens_per_task": 7540.814393939394,
      "last_run_timestamp": "2026-10-03T16:58:32.590Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-coder-next-2025-02-03",
      "model_slug": "qwen/qwen3-coder-next",
      "model_name": "Qwen: Qwen3 Coder Next",
      "provider_name": "Parasail",
      "accuracy": 0.7438658599640935,
      "accuracy_std_dev": 0.0027091388955820363,
      "run_count": 3,
      "total_questions": 557,
      "avg_cost_per_task_usd": 0.008930811490125673,
      "avg_duration_per_task_ms": 129877.56014362657,
      "avg_output_tokens_per_task": 11161.86355475763,
      "last_run_timestamp": "2026-10-03T17:03:40.665Z"
    },
    {
      "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": 0.7431237709174842,
      "accuracy_std_dev": 0.10658009891156979,
      "run_count": 9,
      "total_questions": 1733,
      "avg_cost_per_task_usd": 0.005974010963646855,
      "avg_duration_per_task_ms": 574816.0046162724,
      "avg_output_tokens_per_task": 26107.00807847663,
      "last_run_timestamp": "2026-10-27T02:37:18.076Z"
    },
    {
      "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": 0.7418633333333333,
      "accuracy_std_dev": 0.005722799567422317,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.023266708754208755,
      "avg_duration_per_task_ms": 95383.41582491582,
      "avg_output_tokens_per_task": 15402.81313131313,
      "last_run_timestamp": "2026-10-03T16:58:08.462Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "DigitalOcean",
      "accuracy": 0.741453,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 156,
      "avg_cost_per_task_usd": 0.09924664102564103,
      "avg_duration_per_task_ms": 1908511.7756410257,
      "avg_output_tokens_per_task": 29914.98717948718,
      "last_run_timestamp": "2026-08-22T17:09:42.218Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Nebius",
      "accuracy": 0.7414141999999999,
      "accuracy_std_dev": 0.0027351667883330276,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.00532230808080808,
      "avg_duration_per_task_ms": 145939.18484848485,
      "avg_output_tokens_per_task": 8599.7,
      "last_run_timestamp": "2026-10-17T16:06:15.944Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Alibaba",
      "accuracy": 0.7407406666666667,
      "accuracy_std_dev": 0.008584264726165454,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.004939484848484848,
      "avg_duration_per_task_ms": 108598.20370370371,
      "avg_output_tokens_per_task": 8056.897306397306,
      "last_run_timestamp": "2026-10-03T16:59:36.480Z"
    },
    {
      "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": 0.7401770526570047,
      "accuracy_std_dev": 0.0865423410071625,
      "run_count": 12,
      "total_questions": 2070,
      "avg_cost_per_task_usd": 0.0039563893719806766,
      "avg_duration_per_task_ms": 552542.9028985507,
      "avg_output_tokens_per_task": 24944.29468599034,
      "last_run_timestamp": "2026-11-04T18:42:44.798Z"
    },
    {
      "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": 0.7392041135371179,
      "accuracy_std_dev": 0.011064834624399307,
      "run_count": 7,
      "total_questions": 1374,
      "avg_cost_per_task_usd": 0.006584693595342066,
      "avg_duration_per_task_ms": 301209.6688500728,
      "avg_output_tokens_per_task": 26146.835516739447,
      "last_run_timestamp": "2026-10-31T16:16:13.798Z"
    },
    {
      "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": 0.7388167142857144,
      "accuracy_std_dev": 0.023177894935318268,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.006051409090909092,
      "avg_duration_per_task_ms": 406464.2871572872,
      "avg_output_tokens_per_task": 23256.089466089466,
      "last_run_timestamp": "2026-10-31T16:08:31.072Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-2.5-flash",
      "model_slug": "google/gemini-2.5-flash",
      "model_name": "Google: Gemini 2.5 Flash",
      "provider_name": "Google",
      "accuracy": 0.7384959999999999,
      "accuracy_std_dev": 0.008170584087648423,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.06071549915824916,
      "avg_duration_per_task_ms": 125902.70959595959,
      "avg_output_tokens_per_task": 24199.07239057239,
      "last_run_timestamp": "2026-10-01T19:15:31.968Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Baidu",
      "accuracy": 0.7380049999999998,
      "accuracy_std_dev": 0.023295947045569962,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.3327117329545455,
      "avg_duration_per_task_ms": 1075619.288510101,
      "avg_output_tokens_per_task": 99143.67108585859,
      "last_run_timestamp": "2026-11-10T19:38:16.555Z"
    },
    {
      "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": 0.7369396718907987,
      "accuracy_std_dev": 0.01294608625956121,
      "run_count": 10,
      "total_questions": 1978,
      "avg_cost_per_task_usd": 0.006695506572295248,
      "avg_duration_per_task_ms": 136027.5783619818,
      "avg_output_tokens_per_task": 6995.519716885743,
      "last_run_timestamp": "2026-10-17T16:03:44.090Z"
    },
    {
      "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": 0.7365419711229947,
      "accuracy_std_dev": 0.1044922112691581,
      "run_count": 7,
      "total_questions": 935,
      "avg_cost_per_task_usd": 0.09201655614973261,
      "avg_duration_per_task_ms": 1267131.9229946523,
      "avg_output_tokens_per_task": 26199.03101604278,
      "last_run_timestamp": "2026-10-31T16:05:14.699Z"
    },
    {
      "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": 0.7361712857142858,
      "accuracy_std_dev": 0.00933947044559173,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.0070149227994227995,
      "avg_duration_per_task_ms": 250938.70995670996,
      "avg_output_tokens_per_task": 8674.670995670996,
      "last_run_timestamp": "2026-10-31T16:02:14.362Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7353578611713666,
      "accuracy_std_dev": 0.08764063669791727,
      "run_count": 7,
      "total_questions": 1383,
      "avg_cost_per_task_usd": 0.004913945770065076,
      "avg_duration_per_task_ms": 616099.819956616,
      "avg_output_tokens_per_task": 32251.255242227042,
      "last_run_timestamp": "2026-10-17T16:03:21.011Z"
    },
    {
      "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": 0.7352091428571429,
      "accuracy_std_dev": 0.007011626557339723,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.11408521356421357,
      "avg_duration_per_task_ms": 767202.2121212122,
      "avg_output_tokens_per_task": 44535.16594516594,
      "last_run_timestamp": "2026-10-31T16:04:48.154Z"
    },
    {
      "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": 0.733364333488806,
      "accuracy_std_dev": 0.03217972751863435,
      "run_count": 11,
      "total_questions": 2144,
      "avg_cost_per_task_usd": 0.010386112406716419,
      "avg_duration_per_task_ms": 251820.99860074627,
      "avg_output_tokens_per_task": 48631.178638059704,
      "last_run_timestamp": "2026-10-31T08:46:43.680Z"
    },
    {
      "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": 0.7328843333333334,
      "accuracy_std_dev": 0.0007938452130121063,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0046818838383838384,
      "avg_duration_per_task_ms": 164385.80303030304,
      "avg_output_tokens_per_task": 7946.1447811447815,
      "last_run_timestamp": "2026-10-03T16:57:10.991Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Relace",
      "accuracy": 0.732440575,
      "accuracy_std_dev": 0.057246596407316816,
      "run_count": 3,
      "total_questions": 560,
      "avg_cost_per_task_usd": 0.014524144642857144,
      "avg_duration_per_task_ms": 428029.5660714286,
      "avg_output_tokens_per_task": 58878.205357142855,
      "last_run_timestamp": "2026-10-31T08:03:10.090Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.7323230000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0787700707070707,
      "avg_duration_per_task_ms": 1116199.0808080807,
      "avg_output_tokens_per_task": 26175.146464646463,
      "last_run_timestamp": "2026-08-22T17:45:06.645Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.732155,
      "accuracy_std_dev": 0.02424727451075689,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.016461675757575758,
      "avg_duration_per_task_ms": 673272.2666666667,
      "avg_output_tokens_per_task": 60883.92525252525,
      "last_run_timestamp": "2026-10-31T08:31:11.846Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Venice",
      "accuracy": 0.7320426666666667,
      "accuracy_std_dev": 0.006532295121590524,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.006095060606060606,
      "avg_duration_per_task_ms": 411941.8905723906,
      "avg_output_tokens_per_task": 7966.672558922559,
      "last_run_timestamp": "2026-10-31T16:08:56.903Z"
    },
    {
      "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": 0.7315579999999999,
      "accuracy_std_dev": 0.007776888066572624,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.19043861111111113,
      "avg_duration_per_task_ms": 238176.99081726355,
      "avg_output_tokens_per_task": 37896.631313131315,
      "last_run_timestamp": "2026-10-31T08:20:09.279Z"
    },
    {
      "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": 0.7294072124711316,
      "accuracy_std_dev": 0.019970689066868837,
      "run_count": 6,
      "total_questions": 866,
      "avg_cost_per_task_usd": 0.016059699769053117,
      "avg_duration_per_task_ms": 62647.899538106234,
      "avg_output_tokens_per_task": 9232.347575057736,
      "last_run_timestamp": "2026-10-31T16:30:53.500Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "DeepInfra",
      "accuracy": 0.7291284439132034,
      "accuracy_std_dev": 0.048248946497751195,
      "run_count": 14,
      "total_questions": 2719,
      "avg_cost_per_task_usd": 0.009029743655755791,
      "avg_duration_per_task_ms": 527982.3622655388,
      "avg_output_tokens_per_task": 26831.921662375873,
      "last_run_timestamp": "2026-10-31T16:52:55.644Z"
    },
    {
      "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": 0.7289052424242423,
      "accuracy_std_dev": 0.011972354458274426,
      "run_count": 33,
      "total_questions": 6534,
      "avg_cost_per_task_usd": 0.2043336472298745,
      "avg_duration_per_task_ms": 238207.2739516376,
      "avg_output_tokens_per_task": 38129.07560453015,
      "last_run_timestamp": "2026-10-31T08:33:11.343Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-2.5-flash",
      "model_slug": "google/gemini-2.5-flash",
      "model_name": "Google: Gemini 2.5 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.7281145,
      "accuracy_std_dev": 0.004208499999999948,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.060785368686868686,
      "avg_duration_per_task_ms": 121296.46212121213,
      "avg_output_tokens_per_task": 24226.65404040404,
      "last_run_timestamp": "2026-10-01T19:12:41.089Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.5-haiku-20251001",
      "model_slug": "anthropic/claude-haiku-4.5",
      "model_name": "Anthropic: Claude Haiku 4.5",
      "provider_name": "Azure",
      "accuracy": 0.7278849090909092,
      "accuracy_std_dev": 0.012272399739063737,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.18833512167125802,
      "avg_duration_per_task_ms": 344775.9109274564,
      "avg_output_tokens_per_task": 37475.9334251607,
      "last_run_timestamp": "2026-10-31T08:16:11.695Z"
    },
    {
      "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": 0.7272728571428572,
      "accuracy_std_dev": 0.016272465871865285,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.005508002886002886,
      "avg_duration_per_task_ms": 323368.19624819624,
      "avg_output_tokens_per_task": 8110.7409812409815,
      "last_run_timestamp": "2026-10-31T16:22:55.771Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "Phala",
      "accuracy": 0.7271808509532062,
      "accuracy_std_dev": 0.03540222199908109,
      "run_count": 6,
      "total_questions": 1154,
      "avg_cost_per_task_usd": 0.011152869150779895,
      "avg_duration_per_task_ms": 504501.5831889082,
      "avg_output_tokens_per_task": 55506.62998266898,
      "last_run_timestamp": "2026-10-10T08:24:27.616Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Cohere",
      "accuracy": 0.7267116666666668,
      "accuracy_std_dev": 0.01049864519301841,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.021460314814814814,
      "avg_duration_per_task_ms": 244130.04882154881,
      "avg_output_tokens_per_task": 76264.67003367003,
      "last_run_timestamp": "2026-10-31T08:19:36.711Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Tenstorrent",
      "accuracy": 0.7264309999999999,
      "accuracy_std_dev": 0.021044000000000007,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.1303771111111111,
      "avg_duration_per_task_ms": 318087.7601010101,
      "avg_output_tokens_per_task": 43267.416666666664,
      "last_run_timestamp": "2026-10-04T05:57:29.940Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.5-haiku-20251001",
      "model_slug": "anthropic/claude-haiku-4.5",
      "model_name": "Anthropic: Claude Haiku 4.5",
      "provider_name": "Anthropic",
      "accuracy": 0.7252831818181819,
      "accuracy_std_dev": 0.010667148505559918,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.1864189655647383,
      "avg_duration_per_task_ms": 310620.87190082646,
      "avg_output_tokens_per_task": 37109.52800734619,
      "last_run_timestamp": "2026-10-31T08:09:30.774Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-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": 0.7251266559865094,
      "accuracy_std_dev": 0.005701732738582379,
      "run_count": 3,
      "total_questions": 593,
      "avg_cost_per_task_usd": 0.01145135581787521,
      "avg_duration_per_task_ms": 98413.20741989883,
      "avg_output_tokens_per_task": 10340.510961214166,
      "last_run_timestamp": "2026-10-03T16:43:17.105Z"
    },
    {
      "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": "anthropic/claude-4.5-haiku-20251001",
      "model_slug": "anthropic/claude-haiku-4.5",
      "model_name": "Anthropic: Claude Haiku 4.5",
      "provider_name": "Google",
      "accuracy": 0.7226813939393939,
      "accuracy_std_dev": 0.014086885758036204,
      "run_count": 33,
      "total_questions": 6534,
      "avg_cost_per_task_usd": 0.19769731894704623,
      "avg_duration_per_task_ms": 272979.47413529234,
      "avg_output_tokens_per_task": 36868.905111723296,
      "last_run_timestamp": "2026-10-31T08:29:35.324Z"
    },
    {
      "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": 0.72106475,
      "accuracy_std_dev": 0.011038884066675383,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.004405270833333332,
      "avg_duration_per_task_ms": 708002.5321969697,
      "avg_output_tokens_per_task": 7885.837121212121,
      "last_run_timestamp": "2026-10-31T16:02:18.363Z"
    },
    {
      "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": 0.720845,
      "accuracy_std_dev": 0.13313462084603755,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.6171295247933883,
      "avg_duration_per_task_ms": 640700.6698806244,
      "avg_output_tokens_per_task": 40986.32323232323,
      "last_run_timestamp": "2026-10-31T08:09:30.979Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.6-sonnet-20260217",
      "model_slug": "anthropic/claude-sonnet-4.6",
      "model_name": "Anthropic: Claude Sonnet 4.6",
      "provider_name": "Azure",
      "accuracy": 0.7205387272727274,
      "accuracy_std_dev": 0.14017884451338383,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.6223622332415059,
      "avg_duration_per_task_ms": 640560.214876033,
      "avg_output_tokens_per_task": 41343.55968778696,
      "last_run_timestamp": "2026-10-31T08:28:17.671Z"
    },
    {
      "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": 0.7205387272727272,
      "accuracy_std_dev": 0.13030117338555663,
      "run_count": 33,
      "total_questions": 6534,
      "avg_cost_per_task_usd": 0.6590140376492195,
      "avg_duration_per_task_ms": 429144.30731558,
      "avg_output_tokens_per_task": 41017.19651056015,
      "last_run_timestamp": "2026-10-31T08:31:32.893Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.6-sonnet-20260217",
      "model_slug": "anthropic/claude-sonnet-4.6",
      "model_name": "Anthropic: Claude Sonnet 4.6",
      "provider_name": "Google",
      "accuracy": 0.719695823610473,
      "accuracy_std_dev": 0.13461887126264796,
      "run_count": 33,
      "total_questions": 6531,
      "avg_cost_per_task_usd": 0.6638264376052672,
      "avg_duration_per_task_ms": 604267.6686571735,
      "avg_output_tokens_per_task": 41332.92175777063,
      "last_run_timestamp": "2026-10-31T08:51:32.081Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Makora",
      "accuracy": 0.7183742857142856,
      "accuracy_std_dev": 0.012394494661678109,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.01898589466089466,
      "avg_duration_per_task_ms": 501749.6746031746,
      "avg_output_tokens_per_task": 67465.82683982684,
      "last_run_timestamp": "2026-10-31T08:06:04.074Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-next-80b-a3b-instruct-2509",
      "model_slug": "qwen/qwen3-next-80b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Next 80B A3B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.7177330000000001,
      "accuracy_std_dev": 0.0028613344206273234,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.011208954545454548,
      "avg_duration_per_task_ms": 85332.55555555556,
      "avg_output_tokens_per_task": 10503.252525252525,
      "last_run_timestamp": "2026-10-03T16:59:33.949Z"
    },
    {
      "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": 0.7171206858771375,
      "accuracy_std_dev": 0.09050148781226955,
      "run_count": 8,
      "total_questions": 1579,
      "avg_cost_per_task_usd": 0.011264650411652945,
      "avg_duration_per_task_ms": 777628.9563014567,
      "avg_output_tokens_per_task": 32969.615579480684,
      "last_run_timestamp": "2026-10-31T16:08:08.535Z"
    },
    {
      "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": 0.7166907142857143,
      "accuracy_std_dev": 0.03789838215813547,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.3448133593073593,
      "avg_duration_per_task_ms": 1259875.4646464647,
      "avg_output_tokens_per_task": 107708.73232323233,
      "last_run_timestamp": "2026-10-31T16:18:48.427Z"
    },
    {
      "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": 0.716486229602888,
      "accuracy_std_dev": 0.037837545924002784,
      "run_count": 7,
      "total_questions": 1385,
      "avg_cost_per_task_usd": 0.33939834729241875,
      "avg_duration_per_task_ms": 1241884.499638989,
      "avg_output_tokens_per_task": 105962.96534296029,
      "last_run_timestamp": "2026-10-31T16:18:28.925Z"
    },
    {
      "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": 0.7158777222222221,
      "accuracy_std_dev": 0.12243858819244856,
      "run_count": 5,
      "total_questions": 972,
      "avg_cost_per_task_usd": 0.03800503395061729,
      "avg_duration_per_task_ms": 541259.3065843622,
      "avg_output_tokens_per_task": 34594.609053497945,
      "last_run_timestamp": "2026-09-19T17:18:00.087Z"
    },
    {
      "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": 0.7150290909090908,
      "accuracy_std_dev": 0.14315254393475418,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.6259749265381084,
      "avg_duration_per_task_ms": 641693.616620753,
      "avg_output_tokens_per_task": 41568.741965105604,
      "last_run_timestamp": "2026-10-31T08:22:47.508Z"
    },
    {
      "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": 0.7149273333333334,
      "accuracy_std_dev": 0.0015872190215034077,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.01138610101010101,
      "avg_duration_per_task_ms": 186927.4898989899,
      "avg_output_tokens_per_task": 5929.171717171717,
      "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": "Alibaba",
      "accuracy": 0.7132433333333333,
      "accuracy_std_dev": 0.008837261503932612,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.00784910101010101,
      "avg_duration_per_task_ms": 79021.32323232324,
      "avg_output_tokens_per_task": 9968.289562289563,
      "last_run_timestamp": "2026-10-03T16:40:12.993Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Darkbloom",
      "accuracy": 0.713213,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 111,
      "avg_cost_per_task_usd": 0.007665369369369369,
      "avg_duration_per_task_ms": 1401051.936936937,
      "avg_output_tokens_per_task": 27183.8018018018,
      "last_run_timestamp": "2026-10-03T09:34:00.571Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_slug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_name": "Qwen: Qwen3 VL 235B A22B Instruct",
      "provider_name": "Alibaba",
      "accuracy": 0.7115600000000001,
      "accuracy_std_dev": 0.0034590473640392213,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.006004388888888889,
      "avg_duration_per_task_ms": 110119.68181818182,
      "avg_output_tokens_per_task": 5570.340067340067,
      "last_run_timestamp": "2026-10-03T17:02:24.058Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_slug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_name": "Qwen: Qwen3 VL 235B A22B Instruct",
      "provider_name": "Novita",
      "accuracy": 0.7115600000000001,
      "accuracy_std_dev": 0.00975207530050231,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.008760998316498316,
      "avg_duration_per_task_ms": 118710.20370370371,
      "avg_output_tokens_per_task": 5678.178451178451,
      "last_run_timestamp": "2026-10-03T16:45:33.688Z"
    },
    {
      "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": 0.7110037019328587,
      "accuracy_std_dev": 0.031890632320636075,
      "run_count": 10,
      "total_questions": 1966,
      "avg_cost_per_task_usd": 0.0072013123092573746,
      "avg_duration_per_task_ms": 309646.26246185147,
      "avg_output_tokens_per_task": 40164.048830111904,
      "last_run_timestamp": "2026-10-31T08:15:24.314Z"
    },
    {
      "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": 0.7104376666666667,
      "accuracy_std_dev": 0.004123715988711595,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.015477676767676767,
      "avg_duration_per_task_ms": 105803.74579124579,
      "avg_output_tokens_per_task": 10237.208754208754,
      "last_run_timestamp": "2026-10-03T16:46:22.205Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "AkashML",
      "accuracy": 0.71001675,
      "accuracy_std_dev": 0.018948683824675006,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.015486457070707072,
      "avg_duration_per_task_ms": 1105788.4797979798,
      "avg_output_tokens_per_task": 64321.568181818184,
      "last_run_timestamp": "2026-10-31T16:02:14.361Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-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": "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": 0.7061267289719626,
      "accuracy_std_dev": 0.026110787988875415,
      "run_count": 10,
      "total_questions": 1926,
      "avg_cost_per_task_usd": 0.00886934890965732,
      "avg_duration_per_task_ms": 374407.95742471446,
      "avg_output_tokens_per_task": 45095.30893042575,
      "last_run_timestamp": "2026-10-31T08:03:09.796Z"
    },
    {
      "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": 0.7048043337547408,
      "accuracy_std_dev": 0.0036369767922685555,
      "run_count": 4,
      "total_questions": 791,
      "avg_cost_per_task_usd": 0.010453273072060683,
      "avg_duration_per_task_ms": 132200.4804045512,
      "avg_output_tokens_per_task": 9436.553729456384,
      "last_run_timestamp": "2026-10-08T22:40:37.311Z"
    },
    {
      "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": "anthropic/claude-4.6-sonnet-20260217",
      "model_slug": "anthropic/claude-sonnet-4.6",
      "model_name": "Anthropic: Claude Sonnet 4.6",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.703844,
      "accuracy_std_dev": 0.13064484273849722,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.611304741161616,
      "avg_duration_per_task_ms": 637806.0361952862,
      "avg_output_tokens_per_task": 40663.66961279461,
      "last_run_timestamp": "2026-10-31T08:38:33.095Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-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": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "Parasail",
      "accuracy": 0.7001132585034013,
      "accuracy_std_dev": 0.0646565,
      "run_count": 2,
      "total_questions": 294,
      "avg_cost_per_task_usd": 0.006708948979591837,
      "avg_duration_per_task_ms": 418040.18707482994,
      "avg_output_tokens_per_task": 32070.63945578231,
      "last_run_timestamp": "2026-08-29T12:30:09.617Z"
    },
    {
      "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": 0.6952379824175824,
      "accuracy_std_dev": 0.04288600202768226,
      "run_count": 7,
      "total_questions": 1365,
      "avg_cost_per_task_usd": 0.031070087912087915,
      "avg_duration_per_task_ms": 755530.3274725275,
      "avg_output_tokens_per_task": 18572.407326007327,
      "last_run_timestamp": "2026-10-31T16:23:19.099Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-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": 0.6947253333333333,
      "accuracy_std_dev": 0.03660940543388032,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.010800447811447811,
      "avg_duration_per_task_ms": 53436.71548821549,
      "avg_output_tokens_per_task": 8968.276094276094,
      "last_run_timestamp": "2026-10-03T16:36:16.895Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-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": 0.694725,
      "accuracy_std_dev": 0.007570749896806768,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.007473897306397308,
      "avg_duration_per_task_ms": 291565.11447811447,
      "avg_output_tokens_per_task": 5943.813131313132,
      "last_run_timestamp": "2026-10-03T17:02:48.284Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "Darkbloom",
      "accuracy": 0.6944443846153847,
      "accuracy_std_dev": 0.1332283983619215,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.02886994483294483,
      "avg_duration_per_task_ms": 752020.932012432,
      "avg_output_tokens_per_task": 41085.64568764569,
      "last_run_timestamp": "2026-11-13T02:11:08.485Z"
    },
    {
      "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": 0.6927611428571429,
      "accuracy_std_dev": 0.04357394862080053,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.014586953823953825,
      "avg_duration_per_task_ms": 929278.9754689755,
      "avg_output_tokens_per_task": 50665.61544011544,
      "last_run_timestamp": "2026-10-31T16:11:46.431Z"
    },
    {
      "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": 0.6917989999999999,
      "accuracy_std_dev": 0.053895225463061976,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.14352880086580086,
      "avg_duration_per_task_ms": 700927.1991341992,
      "avg_output_tokens_per_task": 43119.867965367965,
      "last_run_timestamp": "2026-10-31T16:56:31.817Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_slug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_name": "Qwen: Qwen3 VL 235B A22B Instruct",
      "provider_name": "DeepInfra",
      "accuracy": 0.69149825,
      "accuracy_std_dev": 0.011926428811152977,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.005578758838383838,
      "avg_duration_per_task_ms": 386937.1830808081,
      "avg_output_tokens_per_task": 6154.853535353535,
      "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-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Friendli",
      "accuracy": 0.690446125,
      "accuracy_std_dev": 0.011976425410337372,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.00649373611111111,
      "avg_duration_per_task_ms": 110586.78093434343,
      "avg_output_tokens_per_task": 15958.085227272728,
      "last_run_timestamp": "2026-10-31T16:08:56.904Z"
    },
    {
      "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": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "Darkbloom",
      "accuracy": 0.6884587777777779,
      "accuracy_std_dev": 0.0671979156138009,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.007667775533108866,
      "avg_duration_per_task_ms": 1347693.569023569,
      "avg_output_tokens_per_task": 59908.588664422,
      "last_run_timestamp": "2026-11-13T01:44:30.123Z"
    },
    {
      "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": "DekaLLM",
      "accuracy": 0.6866834096080909,
      "accuracy_std_dev": 0.0076575271261354015,
      "run_count": 4,
      "total_questions": 791,
      "avg_cost_per_task_usd": 0.007103366624525917,
      "avg_duration_per_task_ms": 1621893.5347661187,
      "avg_output_tokens_per_task": 50543.86852085967,
      "last_run_timestamp": "2026-10-31T16:36:31.734Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "Venice",
      "accuracy": 0.6866618266276519,
      "accuracy_std_dev": 0.06204446895442602,
      "run_count": 7,
      "total_questions": 1367,
      "avg_cost_per_task_usd": 0.051509942209217265,
      "avg_duration_per_task_ms": 1178589.2787125092,
      "avg_output_tokens_per_task": 24138.589612289685,
      "last_run_timestamp": "2026-10-31T16:02:14.362Z"
    },
    {
      "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": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "SambaNova",
      "accuracy": 0.6821764194977844,
      "accuracy_std_dev": 0.028397019030270698,
      "run_count": 6,
      "total_questions": 677,
      "avg_cost_per_task_usd": 0.008931531757754801,
      "avg_duration_per_task_ms": 55086.47267355982,
      "avg_output_tokens_per_task": 5789.069423929099,
      "last_run_timestamp": "2026-10-31T16:13:09.867Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "SiliconFlow",
      "accuracy": 0.6821534336283186,
      "accuracy_std_dev": 0.02131712035248777,
      "run_count": 7,
      "total_questions": 1356,
      "avg_cost_per_task_usd": 0.0023845553097345137,
      "avg_duration_per_task_ms": 215586.99410029498,
      "avg_output_tokens_per_task": 4060.5553097345132,
      "last_run_timestamp": "2026-10-31T16:18:48.428Z"
    },
    {
      "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": 0.6809765,
      "accuracy_std_dev": 0.000841499999999995,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.00570934090909091,
      "avg_duration_per_task_ms": 525980.6186868687,
      "avg_output_tokens_per_task": 16193.580808080807,
      "last_run_timestamp": "2026-08-29T18:40:21.422Z"
    },
    {
      "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": 0.6804673184713376,
      "accuracy_std_dev": 0.00007650000000003487,
      "run_count": 2,
      "total_questions": 314,
      "avg_cost_per_task_usd": 0.03821310191082802,
      "avg_duration_per_task_ms": 669073.7388535031,
      "avg_output_tokens_per_task": 40207.68152866242,
      "last_run_timestamp": "2026-08-29T18:51:07.939Z"
    },
    {
      "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": 0.6803991483087598,
      "accuracy_std_dev": 0.010339641763888907,
      "run_count": 6,
      "total_questions": 1153,
      "avg_cost_per_task_usd": 0.018036671292281006,
      "avg_duration_per_task_ms": 929552.2064180399,
      "avg_output_tokens_per_task": 89989.56808326105,
      "last_run_timestamp": "2026-09-26T16:34:21.219Z"
    },
    {
      "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": 0.679672686416185,
      "accuracy_std_dev": 0.07384322715989433,
      "run_count": 7,
      "total_questions": 1384,
      "avg_cost_per_task_usd": 0.010168833092485549,
      "avg_duration_per_task_ms": 385418.2059248555,
      "avg_output_tokens_per_task": 28139.643063583815,
      "last_run_timestamp": "2026-10-31T16:12:07.129Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "AkashML",
      "accuracy": 0.6785352626262626,
      "accuracy_std_dev": 0.014156861524116758,
      "run_count": 13,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.005069473232323233,
      "avg_duration_per_task_ms": 756595.5121212121,
      "avg_output_tokens_per_task": 50528.773232323234,
      "last_run_timestamp": "2026-10-31T16:04:29.122Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Mancer 2",
      "accuracy": 0.6767676666666665,
      "accuracy_std_dev": 0.1409255711923457,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.007406489898989899,
      "avg_duration_per_task_ms": 706859.9983164984,
      "avg_output_tokens_per_task": 26149.993265993267,
      "last_run_timestamp": "2026-10-31T16:45:39.993Z"
    },
    {
      "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": 0.6734009999999999,
      "accuracy_std_dev": 0.063973,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.00633028787878788,
      "avg_duration_per_task_ms": 117057.46717171717,
      "avg_output_tokens_per_task": 7022.232323232323,
      "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-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "CoreWeave",
      "accuracy": 0.6709957142857143,
      "accuracy_std_dev": 0.0163185813959275,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.007971292207792208,
      "avg_duration_per_task_ms": 515085.1370851371,
      "avg_output_tokens_per_task": 61096.35209235209,
      "last_run_timestamp": "2026-10-31T16:08:31.061Z"
    },
    {
      "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": 0.6708364177407126,
      "accuracy_std_dev": 0.019300239418446996,
      "run_count": 8,
      "total_questions": 1319,
      "avg_cost_per_task_usd": 0.006608144806671721,
      "avg_duration_per_task_ms": 469922.3040181956,
      "avg_output_tokens_per_task": 47033.74753601213,
      "last_run_timestamp": "2026-10-31T16:03:21.483Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-chat-v3-0324",
      "model_slug": "deepseek/deepseek-chat-v3-0324",
      "model_name": "DeepSeek: DeepSeek V3 0324",
      "provider_name": "SiliconFlow",
      "accuracy": 0.6691920000000001,
      "accuracy_std_dev": 0.007576000000000027,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.002490050505050505,
      "avg_duration_per_task_ms": 88307.96464646465,
      "avg_output_tokens_per_task": 2307.4141414141413,
      "last_run_timestamp": "2026-08-29T20:54:16.286Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-chat-v3",
      "model_slug": "deepseek/deepseek-chat",
      "model_name": "DeepSeek: DeepSeek V3",
      "provider_name": "StreamLake",
      "accuracy": 0.6689113333333334,
      "accuracy_std_dev": 0.0067804647005611455,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0026409175084175086,
      "avg_duration_per_task_ms": 47935.90572390572,
      "avg_output_tokens_per_task": 2383.6902356902356,
      "last_run_timestamp": "2026-10-02T15:53:31.869Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 VL 30B A3B Instruct",
      "provider_name": "Alibaba",
      "accuracy": 0.6683501428571427,
      "accuracy_std_dev": 0.007142653957110113,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.005125032467532468,
      "avg_duration_per_task_ms": 102657.88744588745,
      "avg_output_tokens_per_task": 9652.720057720058,
      "last_run_timestamp": "2026-10-03T09:23:08.256Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.6-20251208",
      "model_slug": "z-ai/glm-4.6v",
      "model_name": "Z.ai: GLM 4.6V",
      "provider_name": "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": "Amazon Bedrock",
      "accuracy": 0.667508642857143,
      "accuracy_std_dev": 0.01141368504163279,
      "run_count": 14,
      "total_questions": 2772,
      "avg_cost_per_task_usd": 0.009206428571428572,
      "avg_duration_per_task_ms": 323973.83044733043,
      "avg_output_tokens_per_task": 60939.117965367965,
      "last_run_timestamp": "2026-10-31T16:08:08.537Z"
    },
    {
      "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": 0.666667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 196,
      "avg_cost_per_task_usd": 0.05274626530612245,
      "avg_duration_per_task_ms": 2852978.0561224488,
      "avg_output_tokens_per_task": 28832.622448979593,
      "last_run_timestamp": "2026-08-22T18:19:05.094Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Phala",
      "accuracy": 0.6666668592964823,
      "accuracy_std_dev": 0.0020275000000000154,
      "run_count": 2,
      "total_questions": 199,
      "avg_cost_per_task_usd": 0.004585065326633166,
      "avg_duration_per_task_ms": 615862.0050251256,
      "avg_output_tokens_per_task": 30806.96984924623,
      "last_run_timestamp": "2026-08-29T17:24:11.557Z"
    },
    {
      "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": 0.6666665714285714,
      "accuracy_std_dev": 0.017285925272620274,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.00577388023088023,
      "avg_duration_per_task_ms": 421454.3968253968,
      "avg_output_tokens_per_task": 15830.38961038961,
      "last_run_timestamp": "2026-10-31T16:17:15.718Z"
    },
    {
      "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": 0.6652235714285715,
      "accuracy_std_dev": 0.018626001977704986,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.008422984848484848,
      "avg_duration_per_task_ms": 801001.3318903319,
      "avg_output_tokens_per_task": 56678.3354978355,
      "last_run_timestamp": "2026-10-31T16:08:56.902Z"
    },
    {
      "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": 0.663861,
      "accuracy_std_dev": 0.007815979699734826,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.002078132996632997,
      "avg_duration_per_task_ms": 241271.14814814815,
      "avg_output_tokens_per_task": 3140.030303030303,
      "last_run_timestamp": "2026-10-05T20:44:48.611Z"
    },
    {
      "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": "auto-routing",
      "accuracy": 0.6632995,
      "accuracy_std_dev": 0.005050500000000013,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.0029102499999999996,
      "avg_duration_per_task_ms": 81353.34090909091,
      "avg_output_tokens_per_task": 2988.2853535353534,
      "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-4.6-20251208",
      "model_slug": "z-ai/glm-4.6v",
      "model_name": "Z.ai: GLM 4.6V",
      "provider_name": "Z.AI",
      "accuracy": 0.659933,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.014984520202020202,
      "avg_duration_per_task_ms": 348987.1767676768,
      "avg_output_tokens_per_task": 16461.863636363636,
      "last_run_timestamp": "2026-10-01T08:28:22.621Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4.1-2025-04-14",
      "model_slug": "openai/gpt-4.1",
      "model_name": "OpenAI: GPT-4.1",
      "provider_name": "OpenAI",
      "accuracy": 0.6599326666666666,
      "accuracy_std_dev": 0.013255807590972711,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.015927558922558923,
      "avg_duration_per_task_ms": 14095.8367003367,
      "avg_output_tokens_per_task": 1794.7289562289561,
      "last_run_timestamp": "2026-10-02T23:17:48.072Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 VL 30B A3B Instruct",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.658249,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0066776414141414145,
      "avg_duration_per_task_ms": 101359.9898989899,
      "avg_output_tokens_per_task": 9307.358585858587,
      "last_run_timestamp": "2026-10-04T03:07:35.168Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "ibm-granite/granite-4.2-8b-20260831",
      "model_slug": "ibm-granite/granite-4.2-8b",
      "model_name": "IBM: Granite 4.2 8B",
      "provider_name": "auto-routing",
      "accuracy": 0.65736,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 197,
      "avg_cost_per_task_usd": 0.014864365482233502,
      "avg_duration_per_task_ms": 1717906.416243655,
      "avg_output_tokens_per_task": 64415.350253807104,
      "last_run_timestamp": "2026-10-01T08:48:38.304Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.657127,
      "accuracy_std_dev": 0.05754067397936871,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.01756317676767677,
      "avg_duration_per_task_ms": 295248.13804713805,
      "avg_output_tokens_per_task": 9861.76430976431,
      "last_run_timestamp": "2026-10-03T17:36:48.437Z"
    },
    {
      "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": 0.6523570000000002,
      "accuracy_std_dev": 0.01094299999999998,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.003023964646464647,
      "avg_duration_per_task_ms": 45756.330808080806,
      "avg_output_tokens_per_task": 2857.7727272727275,
      "last_run_timestamp": "2026-08-29T19:51:51.198Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4.1-mini-2025-04-14",
      "model_slug": "openai/gpt-4.1-mini",
      "model_name": "OpenAI: GPT-4.1 Mini",
      "provider_name": "auto-routing",
      "accuracy": 0.652357,
      "accuracy_std_dev": 0.01262599999999997,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.004161952020202021,
      "avg_duration_per_task_ms": 27160.11111111111,
      "avg_output_tokens_per_task": 2408.8232323232323,
      "last_run_timestamp": "2026-10-01T19:12:31.553Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "ibm-granite/granite-4.2-8b-20260831",
      "model_slug": "ibm-granite/granite-4.2-8b",
      "model_name": "IBM: Granite 4.2 8B",
      "provider_name": "DeepInfra",
      "accuracy": 0.6523569999999999,
      "accuracy_std_dev": 0.0025249999999999995,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.01652235606060606,
      "avg_duration_per_task_ms": 1917916.494949495,
      "avg_output_tokens_per_task": 65966.15909090909,
      "last_run_timestamp": "2026-10-09T00:15:37.020Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "auto-routing",
      "accuracy": 0.6522367142857142,
      "accuracy_std_dev": 0.024854328583708306,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.0057348217893217895,
      "avg_duration_per_task_ms": 642012.7799422799,
      "avg_output_tokens_per_task": 41761.8759018759,
      "last_run_timestamp": "2026-10-31T16:01:53.622Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-r1-0528",
      "model_slug": "deepseek/deepseek-r1-0528",
      "model_name": "DeepSeek: R1 0528",
      "provider_name": "Novita",
      "accuracy": 0.651515,
      "accuracy_std_dev": 0.1905456270048375,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.058297676767676776,
      "avg_duration_per_task_ms": 1098090.5791245792,
      "avg_output_tokens_per_task": 23174.20202020202,
      "last_run_timestamp": "2026-10-17T16:03:44.041Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4.1-mini-2025-04-14",
      "model_slug": "openai/gpt-4.1-mini",
      "model_name": "OpenAI: GPT-4.1 Mini",
      "provider_name": "OpenAI",
      "accuracy": 0.6492706666666667,
      "accuracy_std_dev": 0.010498645193018418,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0041127255892255896,
      "avg_duration_per_task_ms": 24949.132996632998,
      "avg_output_tokens_per_task": 2375.9343434343436,
      "last_run_timestamp": "2026-10-02T23:15:53.868Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.6492706666666667,
      "accuracy_std_dev": 0.0055553002519115975,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.002898063973063973,
      "avg_duration_per_task_ms": 166642.234006734,
      "avg_output_tokens_per_task": 10315.60606060606,
      "last_run_timestamp": "2026-10-03T17:16:11.892Z"
    },
    {
      "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": "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": 0.6473065,
      "accuracy_std_dev": 0.012626499999999985,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.002362989898989899,
      "avg_duration_per_task_ms": 57564.656565656565,
      "avg_output_tokens_per_task": 2763.429292929293,
      "last_run_timestamp": "2026-08-29T16:11:23.388Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "Alibaba",
      "accuracy": 0.646465,
      "accuracy_std_dev": 0.007273551310513057,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.004810424242424242,
      "avg_duration_per_task_ms": 80194.98484848485,
      "avg_output_tokens_per_task": 9051.29797979798,
      "last_run_timestamp": "2026-10-03T16:51:25.541Z"
    },
    {
      "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": 0.6450988763882182,
      "accuracy_std_dev": 0.03520097627520112,
      "run_count": 11,
      "total_questions": 2071,
      "avg_cost_per_task_usd": 0.005406431675519074,
      "avg_duration_per_task_ms": 792878.7054563013,
      "avg_output_tokens_per_task": 36344.80685659102,
      "last_run_timestamp": "2026-11-13T01:39:57.890Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-4-maverick-17b-128e-instruct",
      "model_slug": "meta-llama/llama-4-maverick",
      "model_name": "Meta: Llama 4 Maverick",
      "provider_name": "Novita",
      "accuracy": 0.644781,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.002468472222222222,
      "avg_duration_per_task_ms": 55306.53787878788,
      "avg_output_tokens_per_task": 2662.2853535353534,
      "last_run_timestamp": "2026-08-29T17:03:33.297Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-chat-v3-0324",
      "model_slug": "deepseek/deepseek-chat-v3-0324",
      "model_name": "DeepSeek: DeepSeek V3 0324",
      "provider_name": "Crusoe",
      "accuracy": 0.6443605,
      "accuracy_std_dev": 0.007996499999999962,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.0038927146464646464,
      "avg_duration_per_task_ms": 63456.618686868685,
      "avg_output_tokens_per_task": 2359.770202020202,
      "last_run_timestamp": "2026-08-29T16:10:45.295Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-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": 0.6418952857142858,
      "accuracy_std_dev": 0.013932428582414525,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.007236988455988457,
      "avg_duration_per_task_ms": 236970.17243867245,
      "avg_output_tokens_per_task": 11526.8658008658,
      "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": "Ambient",
      "accuracy": 0.641414,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.00284089898989899,
      "avg_duration_per_task_ms": 209352.64646464647,
      "avg_output_tokens_per_task": 9266.040404040405,
      "last_run_timestamp": "2026-10-09T15:00:06.492Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Darkbloom",
      "accuracy": 0.6409465555555555,
      "accuracy_std_dev": 0.012642544119063567,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.0015052222222222225,
      "avg_duration_per_task_ms": 370215.3058361392,
      "avg_output_tokens_per_task": 16243.532547699215,
      "last_run_timestamp": "2026-11-13T01:55:29.605Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4.1-mini-2025-04-14",
      "model_slug": "openai/gpt-4.1-mini",
      "model_name": "OpenAI: GPT-4.1 Mini",
      "provider_name": "Azure",
      "accuracy": 0.6401515,
      "accuracy_std_dev": 0.002759842885020819,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.004124544191919192,
      "avg_duration_per_task_ms": 34476.35353535353,
      "avg_output_tokens_per_task": 2259.632575757576,
      "last_run_timestamp": "2026-10-01T19:13:29.560Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4.1-2025-04-14",
      "model_slug": "openai/gpt-4.1",
      "model_name": "OpenAI: GPT-4.1",
      "provider_name": "auto-routing",
      "accuracy": 0.6372055000000001,
      "accuracy_std_dev": 0.004208500000000004,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.015519383838383837,
      "avg_duration_per_task_ms": 14506.310606060606,
      "avg_output_tokens_per_task": 1746.2525252525252,
      "last_run_timestamp": "2026-10-01T19:12:45.033Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4.1-2025-04-14",
      "model_slug": "openai/gpt-4.1",
      "model_name": "OpenAI: GPT-4.1",
      "provider_name": "Azure",
      "accuracy": 0.6367847499999999,
      "accuracy_std_dev": 0.012277609526593496,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.015237281565656567,
      "avg_duration_per_task_ms": 18840.242424242424,
      "avg_output_tokens_per_task": 1618.909090909091,
      "last_run_timestamp": "2026-10-01T19:14:40.644Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "auto-routing",
      "accuracy": 0.6366041428571428,
      "accuracy_std_dev": 0.13875937770258326,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.014381060606060606,
      "avg_duration_per_task_ms": 413704.7886002886,
      "avg_output_tokens_per_task": 12007.351370851371,
      "last_run_timestamp": "2026-10-31T16:01:53.629Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 VL 30B A3B Instruct",
      "provider_name": "Novita",
      "accuracy": 0.6356422857142857,
      "accuracy_std_dev": 0.03719902777421659,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.006775251803751804,
      "avg_duration_per_task_ms": 102294.06782106782,
      "avg_output_tokens_per_task": 9446.849206349207,
      "last_run_timestamp": "2026-10-03T10:32:36.729Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-32b-04-28",
      "model_slug": "qwen/qwen3-32b",
      "model_name": "Qwen: Qwen3 32B",
      "provider_name": "Groq",
      "accuracy": 0.633429739884393,
      "accuracy_std_dev": 0.015053999999999956,
      "run_count": 2,
      "total_questions": 346,
      "avg_cost_per_task_usd": 0.007234326589595376,
      "avg_duration_per_task_ms": 26683.387283236993,
      "avg_output_tokens_per_task": 11969.693641618496,
      "last_run_timestamp": "2026-08-29T16:23:45.123Z"
    },
    {
      "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": 0.633117,
      "accuracy_std_dev": 0.010839761238804363,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.01982476406926407,
      "avg_duration_per_task_ms": 532402.84992785,
      "avg_output_tokens_per_task": 99141.9025974026,
      "last_run_timestamp": "2026-10-03T17:08:02.072Z"
    },
    {
      "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": 0.6322752857142858,
      "accuracy_std_dev": 0.02198430561439089,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.018390331168831167,
      "avg_duration_per_task_ms": 674719.9083694083,
      "avg_output_tokens_per_task": 89392.03174603175,
      "last_run_timestamp": "2026-10-03T17:40:51.448Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "inclusionai/ling-3.0-flash-fin-20260827",
      "model_slug": "inclusionai/ling-3.0-flash-fin",
      "model_name": "inclusionAI: Ling 3.0 Flash Fin",
      "provider_name": "Novita",
      "accuracy": 0.632155,
      "accuracy_std_dev": 0.005892000000000008,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.005966434343434343,
      "avg_duration_per_task_ms": 235473.01515151514,
      "avg_output_tokens_per_task": 48249.11363636364,
      "last_run_timestamp": "2026-10-01T08:24:52.614Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "SiliconFlow",
      "accuracy": 0.6318743333333333,
      "accuracy_std_dev": 0.008054489114083432,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0027108804713804713,
      "avg_duration_per_task_ms": 284802.3080808081,
      "avg_output_tokens_per_task": 8805.387205387206,
      "last_run_timestamp": "2026-10-03T17:03:50.406Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Phala",
      "accuracy": 0.6308699561200923,
      "accuracy_std_dev": 0.2332223523022349,
      "run_count": 7,
      "total_questions": 1299,
      "avg_cost_per_task_usd": 0.009968328714395689,
      "avg_duration_per_task_ms": 169608.67359507314,
      "avg_output_tokens_per_task": 16432.014626635875,
      "last_run_timestamp": "2026-10-31T16:54:17.849Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "DekaLLM",
      "accuracy": 0.6306148156028369,
      "accuracy_std_dev": 0.05281625698168657,
      "run_count": 3,
      "total_questions": 282,
      "avg_cost_per_task_usd": 0.0037584255319148938,
      "avg_duration_per_task_ms": 208452.59574468085,
      "avg_output_tokens_per_task": 12308.468085106382,
      "last_run_timestamp": "2026-10-23T09:29:14.171Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "SiliconFlow",
      "accuracy": 0.6294955496017379,
      "accuracy_std_dev": 0.007954455460524567,
      "run_count": 7,
      "total_questions": 1381,
      "avg_cost_per_task_usd": 0.0018547262853005066,
      "avg_duration_per_task_ms": 225698.48443157133,
      "avg_output_tokens_per_task": 10120.918175235336,
      "last_run_timestamp": "2026-10-31T16:10:09.221Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "StreamLake",
      "accuracy": 0.62878825,
      "accuracy_std_dev": 0.0068900206957236525,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.001630263888888889,
      "avg_duration_per_task_ms": 77290.25,
      "avg_output_tokens_per_task": 8242.138888888889,
      "last_run_timestamp": "2026-10-03T16:47:32.559Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3-nano-30b-a3b",
      "model_slug": "nvidia/nemotron-3-nano-30b-a3b",
      "model_name": "NVIDIA: Nemotron 3 Nano 30B A3B",
      "provider_name": "Nebius",
      "accuracy": 0.627104375,
      "accuracy_std_dev": 0.008333089267155095,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.024497946969696968,
      "avg_duration_per_task_ms": 1060409.1414141415,
      "avg_output_tokens_per_task": 101874.34027777778,
      "last_run_timestamp": "2026-10-23T09:03:34.804Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "AtlasCloud",
      "accuracy": 0.6253323876329787,
      "accuracy_std_dev": 0.21157174135765242,
      "run_count": 8,
      "total_questions": 1504,
      "avg_cost_per_task_usd": 0.010399158909574468,
      "avg_duration_per_task_ms": 276119.94281914894,
      "avg_output_tokens_per_task": 11912.803191489362,
      "last_run_timestamp": "2026-10-17T08:17:57.529Z"
    },
    {
      "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": 0.6245795,
      "accuracy_std_dev": 0.011784500000000031,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.004032424242424243,
      "avg_duration_per_task_ms": 154426.69696969696,
      "avg_output_tokens_per_task": 13170.60101010101,
      "last_run_timestamp": "2026-08-29T18:20:19.448Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Groq",
      "accuracy": 0.621446169117647,
      "accuracy_std_dev": 0.008868533772011368,
      "run_count": 7,
      "total_questions": 1360,
      "avg_cost_per_task_usd": 0.017286922794117646,
      "avg_duration_per_task_ms": 67123.2,
      "avg_output_tokens_per_task": 57389.13382352941,
      "last_run_timestamp": "2026-10-31T16:52:47.514Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-chat-v3",
      "model_slug": "deepseek/deepseek-chat",
      "model_name": "DeepSeek: DeepSeek V3",
      "provider_name": "auto-routing",
      "accuracy": 0.6212125000000001,
      "accuracy_std_dev": 0.011784500000000031,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.0023548611111111114,
      "avg_duration_per_task_ms": 64938.90151515151,
      "avg_output_tokens_per_task": 2165.318181818182,
      "last_run_timestamp": "2026-10-01T19:21:21.945Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.621212,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.00552190909090909,
      "avg_duration_per_task_ms": 174055.64141414143,
      "avg_output_tokens_per_task": 10415.944444444445,
      "last_run_timestamp": "2026-10-04T03:09:40.101Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.6191077500000001,
      "accuracy_std_dev": 0.015215861565731303,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.004869287878787879,
      "avg_duration_per_task_ms": 286467.2689393939,
      "avg_output_tokens_per_task": 15960.142676767677,
      "last_run_timestamp": "2026-10-23T08:08:11.470Z"
    },
    {
      "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": 0.6173469387755102,
      "accuracy_std_dev": 0.01386353555194346,
      "run_count": 3,
      "total_questions": 588,
      "avg_cost_per_task_usd": 0.017528746598639455,
      "avg_duration_per_task_ms": 899129.887755102,
      "avg_output_tokens_per_task": 18516.460884353743,
      "last_run_timestamp": "2026-10-03T11:06:38.068Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Fireworks",
      "accuracy": 0.61687,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 164,
      "avg_cost_per_task_usd": 0.010753804878048782,
      "avg_duration_per_task_ms": 340169.43292682926,
      "avg_output_tokens_per_task": 37200.54878048781,
      "last_run_timestamp": "2026-08-22T17:46:40.285Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-32b-04-28",
      "model_slug": "qwen/qwen3-32b",
      "model_name": "Qwen: Qwen3 32B",
      "provider_name": "Nebius",
      "accuracy": 0.6136364999999999,
      "accuracy_std_dev": 0.007575500000000013,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.005658078282828283,
      "avg_duration_per_task_ms": 686384.2954545454,
      "avg_output_tokens_per_task": 18940.444444444445,
      "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": 0.6125139999999999,
      "accuracy_std_dev": 0.004677949194536715,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.004050520202020202,
      "avg_duration_per_task_ms": 21444.56228956229,
      "avg_output_tokens_per_task": 2151.1430976430975,
      "last_run_timestamp": "2026-10-03T17:03:50.598Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-14b-04-28",
      "model_slug": "qwen/qwen3-14b",
      "model_name": "Qwen: Qwen3 14B",
      "provider_name": "DeepInfra",
      "accuracy": 0.6123375670628184,
      "accuracy_std_dev": 0.008273645159312946,
      "run_count": 3,
      "total_questions": 589,
      "avg_cost_per_task_usd": 0.003321100169779287,
      "avg_duration_per_task_ms": 426798.44821731746,
      "avg_output_tokens_per_task": 13430.818336162989,
      "last_run_timestamp": "2026-10-08T22:16:42.056Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-32b-04-28",
      "model_slug": "qwen/qwen3-32b",
      "model_name": "Qwen: Qwen3 32B",
      "provider_name": "auto-routing",
      "accuracy": 0.6094280000000001,
      "accuracy_std_dev": 0.010100999999999971,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.008297136363636364,
      "avg_duration_per_task_ms": 338157.7095959596,
      "avg_output_tokens_per_task": 17281.333333333332,
      "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": 0.609428,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.00778269696969697,
      "avg_duration_per_task_ms": 302947.1666666667,
      "avg_output_tokens_per_task": 15369.68686868687,
      "last_run_timestamp": "2026-10-08T22:57:53.933Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "DeepInfra",
      "accuracy": 0.607744,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.04297312626262627,
      "avg_duration_per_task_ms": 320571.8686868687,
      "avg_output_tokens_per_task": 21377.484848484848,
      "last_run_timestamp": "2026-08-22T10:29:25.897Z"
    },
    {
      "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": 0.6074393716319267,
      "accuracy_std_dev": 0.07076264023225816,
      "run_count": 10,
      "total_questions": 1967,
      "avg_cost_per_task_usd": 0.012005192679206914,
      "avg_duration_per_task_ms": 1240549.4712760549,
      "avg_output_tokens_per_task": 56702.37315709202,
      "last_run_timestamp": "2026-11-13T01:55:21.496Z"
    },
    {
      "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": 0.6060606666666667,
      "accuracy_std_dev": 0.013538129913527765,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0036622003367003365,
      "avg_duration_per_task_ms": 32739.0202020202,
      "avg_output_tokens_per_task": 2163.530303030303,
      "last_run_timestamp": "2026-10-03T17:03:29.436Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "Mara",
      "accuracy": 0.6059481003717473,
      "accuracy_std_dev": 0.024067533945591992,
      "run_count": 3,
      "total_questions": 269,
      "avg_cost_per_task_usd": 0.03276318587360595,
      "avg_duration_per_task_ms": 68520.27137546468,
      "avg_output_tokens_per_task": 7032.501858736059,
      "last_run_timestamp": "2026-10-31T16:08:33.699Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "auto-routing",
      "accuracy": 0.6043773333333333,
      "accuracy_std_dev": 0.004955863351178606,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0035131498316498316,
      "avg_duration_per_task_ms": 35423.020202020205,
      "avg_output_tokens_per_task": 2122.909090909091,
      "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-vl-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 VL 30B A3B Instruct",
      "provider_name": "DeepInfra",
      "accuracy": 0.6030365565974547,
      "accuracy_std_dev": 0.012321947684289796,
      "run_count": 8,
      "total_questions": 1493,
      "avg_cost_per_task_usd": 0.008523111185532485,
      "avg_duration_per_task_ms": 567385.2880107167,
      "avg_output_tokens_per_task": 14003.128600133958,
      "last_run_timestamp": "2026-10-08T22:50:36.237Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "Alibaba",
      "accuracy": 0.6021326666666665,
      "accuracy_std_dev": 0.003459479954495376,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.011048464646464645,
      "avg_duration_per_task_ms": 31195.183501683503,
      "avg_output_tokens_per_task": 2103.8636363636365,
      "last_run_timestamp": "2026-10-03T17:43:44.972Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.599327,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.012440297979797978,
      "avg_duration_per_task_ms": 230792.08080808082,
      "avg_output_tokens_per_task": 41242.71717171717,
      "last_run_timestamp": "2026-10-04T03:37:16.537Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "DeepInfra",
      "accuracy": 0.5985173543758967,
      "accuracy_std_dev": 0.021740745111426138,
      "run_count": 4,
      "total_questions": 697,
      "avg_cost_per_task_usd": 0.002190916786226686,
      "avg_duration_per_task_ms": 51577.241032998565,
      "avg_output_tokens_per_task": 2056.6857962697272,
      "last_run_timestamp": "2026-10-08T23:15:07.689Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-14b-04-28",
      "model_slug": "qwen/qwen3-14b",
      "model_name": "Qwen: Qwen3 14B",
      "provider_name": "NextBit",
      "accuracy": 0.5978837142857143,
      "accuracy_std_dev": 0.01426892240114143,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.0030797525252525253,
      "avg_duration_per_task_ms": 643402.1818181818,
      "avg_output_tokens_per_task": 13628.22077922078,
      "last_run_timestamp": "2026-11-01T11:08:51.026Z"
    },
    {
      "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": 0.5976433333333334,
      "accuracy_std_dev": 0.0036366213562713575,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.003224691919191919,
      "avg_duration_per_task_ms": 76581.48821548822,
      "avg_output_tokens_per_task": 2072.1902356902356,
      "last_run_timestamp": "2026-10-03T16:59:44.383Z"
    },
    {
      "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": 0.5965210000000001,
      "accuracy_std_dev": 0.03456524546419422,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.030341809764309766,
      "avg_duration_per_task_ms": 1499106.207070707,
      "avg_output_tokens_per_task": 75742.45791245792,
      "last_run_timestamp": "2026-10-13T21:28:41.390Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-2.5-flash-lite",
      "model_slug": "google/gemini-2.5-flash-lite",
      "model_name": "Google: Gemini 2.5 Flash Lite",
      "provider_name": "Google AI Studio",
      "accuracy": 0.5963805,
      "accuracy_std_dev": 0.0012624999999999997,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.01735105808080808,
      "avg_duration_per_task_ms": 112232.29292929293,
      "avg_output_tokens_per_task": 43198.833333333336,
      "last_run_timestamp": "2026-10-01T19:14:48.696Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.595620394160584,
      "accuracy_std_dev": 0.009567520605183589,
      "run_count": 7,
      "total_questions": 1370,
      "avg_cost_per_task_usd": 0.009091688321167881,
      "avg_duration_per_task_ms": 479820.2708029197,
      "avg_output_tokens_per_task": 45264.08613138686,
      "last_run_timestamp": "2026-10-03T17:04:03.802Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-14b-04-28",
      "model_slug": "qwen/qwen3-14b",
      "model_name": "Qwen: Qwen3 14B",
      "provider_name": "auto-routing",
      "accuracy": 0.591751,
      "accuracy_std_dev": 0.0025249999999999995,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.0029208080808080807,
      "avg_duration_per_task_ms": 521963.51515151514,
      "avg_output_tokens_per_task": 11688.588383838383,
      "last_run_timestamp": "2026-10-01T16:32:22.405Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Cerebras",
      "accuracy": 0.5897068571428572,
      "accuracy_std_dev": 0.008664658157111118,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.010310333333333333,
      "avg_duration_per_task_ms": 10820.968253968254,
      "avg_output_tokens_per_task": 13292.640692640693,
      "last_run_timestamp": "2026-10-31T16:21:16.340Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "DeepInfra",
      "accuracy": 0.5880285,
      "accuracy_std_dev": 0.0387325,
      "run_count": 2,
      "total_questions": 142,
      "avg_cost_per_task_usd": 0.03859155633802817,
      "avg_duration_per_task_ms": 448942.5140845071,
      "avg_output_tokens_per_task": 18452.429577464787,
      "last_run_timestamp": "2026-08-29T19:07:49.385Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "SambaNova",
      "accuracy": 0.5873616193353474,
      "accuracy_std_dev": 0.048784153240032144,
      "run_count": 7,
      "total_questions": 662,
      "avg_cost_per_task_usd": 0.03112247280966767,
      "avg_duration_per_task_ms": 62861.41993957704,
      "avg_output_tokens_per_task": 6675.285498489426,
      "last_run_timestamp": "2026-10-31T16:41:30.138Z"
    },
    {
      "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": 0.5833335000000001,
      "accuracy_std_dev": 0.007575500000000013,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.025034136363636363,
      "avg_duration_per_task_ms": 1106648.0227272727,
      "avg_output_tokens_per_task": 68373.12626262626,
      "last_run_timestamp": "2026-08-30T00:32:50.853Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Novita",
      "accuracy": 0.5824916666666666,
      "accuracy_std_dev": 0.004956316261992276,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.006444644781144781,
      "avg_duration_per_task_ms": 663316.5858585859,
      "avg_output_tokens_per_task": 42705.53198653198,
      "last_run_timestamp": "2026-10-03T16:12:07.492Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "Fireworks",
      "accuracy": 0.5810130708860759,
      "accuracy_std_dev": 0.012019500000000016,
      "run_count": 2,
      "total_questions": 395,
      "avg_cost_per_task_usd": 0.01962629113924051,
      "avg_duration_per_task_ms": 376708.9949367089,
      "avg_output_tokens_per_task": 97953.7417721519,
      "last_run_timestamp": "2026-10-31T08:09:25.783Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Google",
      "accuracy": 0.5774414999999999,
      "accuracy_std_dev": 0.005050500000000013,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.010289136363636365,
      "avg_duration_per_task_ms": 331432.74242424243,
      "avg_output_tokens_per_task": 40883.856060606064,
      "last_run_timestamp": "2026-10-17T16:06:05.579Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-chat-v3-0324",
      "model_slug": "deepseek/deepseek-chat-v3-0324",
      "model_name": "DeepSeek: DeepSeek V3 0324",
      "provider_name": "GMICloud",
      "accuracy": 0.5709940405679513,
      "accuracy_std_dev": 0.154429513944712,
      "run_count": 5,
      "total_questions": 986,
      "avg_cost_per_task_usd": 0.0030763367139959434,
      "avg_duration_per_task_ms": 63553.28093306288,
      "avg_output_tokens_per_task": 2517.3924949290063,
      "last_run_timestamp": "2026-11-10T06:14:22.398Z"
    },
    {
      "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": 0.5614105961383749,
      "accuracy_std_dev": 0.06778815603610641,
      "run_count": 7,
      "total_questions": 1243,
      "avg_cost_per_task_usd": 0.06303039259855191,
      "avg_duration_per_task_ms": 382701.67095736123,
      "avg_output_tokens_per_task": 23894.755430410296,
      "last_run_timestamp": "2026-10-31T16:21:16.340Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "Mancer 2",
      "accuracy": 0.5608581453590192,
      "accuracy_std_dev": 0.2662797085149478,
      "run_count": 6,
      "total_questions": 1142,
      "avg_cost_per_task_usd": 0.09908757443082312,
      "avg_duration_per_task_ms": 1949562.3021015762,
      "avg_output_tokens_per_task": 42074.91681260946,
      "last_run_timestamp": "2026-10-31T16:02:18.348Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-2.5-flash-lite",
      "model_slug": "google/gemini-2.5-flash-lite",
      "model_name": "Google: Gemini 2.5 Flash Lite",
      "provider_name": "Google",
      "accuracy": 0.56060625,
      "accuracy_std_dev": 0.009297342103391676,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.017428513888888886,
      "avg_duration_per_task_ms": 127446.60858585859,
      "avg_output_tokens_per_task": 43389.579545454544,
      "last_run_timestamp": "2026-10-01T19:27:01.395Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-32b-04-28",
      "model_slug": "qwen/qwen3-32b",
      "model_name": "Qwen: Qwen3 32B",
      "provider_name": "DeepInfra",
      "accuracy": 0.5527496666666667,
      "accuracy_std_dev": 0.01040809115811132,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.004829589225589226,
      "avg_duration_per_task_ms": 647930.4865319865,
      "avg_output_tokens_per_task": 17015.548821548822,
      "last_run_timestamp": "2026-10-08T23:30:03.872Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-2.5-flash-lite",
      "model_slug": "google/gemini-2.5-flash-lite",
      "model_name": "Google: Gemini 2.5 Flash Lite",
      "provider_name": "auto-routing",
      "accuracy": 0.5521885000000001,
      "accuracy_std_dev": 0.005050500000000013,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.01778226515151515,
      "avg_duration_per_task_ms": 116667.00757575757,
      "avg_output_tokens_per_task": 44273.378787878784,
      "last_run_timestamp": "2026-10-01T19:15:36.386Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-32b-04-28",
      "model_slug": "qwen/qwen3-32b",
      "model_name": "Qwen: Qwen3 32B",
      "provider_name": "Tenstorrent",
      "accuracy": 0.5469078490566038,
      "accuracy_std_dev": 0.022963942121508662,
      "run_count": 10,
      "total_questions": 1908,
      "avg_cost_per_task_usd": 0.005774233752620544,
      "avg_duration_per_task_ms": 784958.8862683438,
      "avg_output_tokens_per_task": 14661.248427672956,
      "last_run_timestamp": "2026-11-08T04:29:32.011Z"
    },
    {
      "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": 0.5462965,
      "accuracy_std_dev": 0.08670050000000001,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.006357126262626262,
      "avg_duration_per_task_ms": 124531.47474747474,
      "avg_output_tokens_per_task": 5650.492424242424,
      "last_run_timestamp": "2026-08-29T20:57:58.443Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-vl-8b-instruct",
      "model_slug": "qwen/qwen3-vl-8b-instruct",
      "model_name": "Qwen: Qwen3 VL 8B Instruct",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.538721,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.005473,
      "avg_duration_per_task_ms": 117831.5808080808,
      "avg_output_tokens_per_task": 11819.661616161617,
      "last_run_timestamp": "2026-10-04T03:06:12.419Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-chat-v3",
      "model_slug": "deepseek/deepseek-chat",
      "model_name": "DeepSeek: DeepSeek V3",
      "provider_name": "DeepInfra",
      "accuracy": 0.5373176666666667,
      "accuracy_std_dev": 0.013403444242266851,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0019087592592592591,
      "avg_duration_per_task_ms": 97924.95622895623,
      "avg_output_tokens_per_task": 1886.641414141414,
      "last_run_timestamp": "2026-10-08T22:10:22.349Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-chat-v3-0324",
      "model_slug": "deepseek/deepseek-chat-v3-0324",
      "model_name": "DeepSeek: DeepSeek V3 0324",
      "provider_name": "DeepInfra",
      "accuracy": 0.5332481099744245,
      "accuracy_std_dev": 0.020375000000000032,
      "run_count": 2,
      "total_questions": 391,
      "avg_cost_per_task_usd": 0.004206966751918158,
      "avg_duration_per_task_ms": 160590.6675191816,
      "avg_output_tokens_per_task": 4524.092071611253,
      "last_run_timestamp": "2026-08-29T20:34:10.117Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4o-2024-08-06",
      "model_slug": "openai/gpt-4o-2024-08-06",
      "model_name": "OpenAI: GPT-4o (2024-08-06)",
      "provider_name": "OpenAI",
      "accuracy": 0.5280586666666666,
      "accuracy_std_dev": 0.004827144451490511,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.01799902861952862,
      "avg_duration_per_task_ms": 14939.786195286195,
      "avg_output_tokens_per_task": 1607.080808080808,
      "last_run_timestamp": "2026-10-02T23:16:52.382Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4o",
      "model_slug": "openai/gpt-4o",
      "model_name": "OpenAI: GPT-4o",
      "provider_name": "OpenAI",
      "accuracy": 0.5266555,
      "accuracy_std_dev": 0.023895888382383546,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.01547567760942761,
      "avg_duration_per_task_ms": 13716.89898989899,
      "avg_output_tokens_per_task": 1354.1801346801346,
      "last_run_timestamp": "2026-10-02T23:17:37.980Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": "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": "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": 0.5165396259541984,
      "accuracy_std_dev": 0.0053874999999999895,
      "run_count": 2,
      "total_questions": 393,
      "avg_cost_per_task_usd": 0.015734966921119593,
      "avg_duration_per_task_ms": 637615.4732824428,
      "avg_output_tokens_per_task": 40479.653944020356,
      "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": "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": "openai/gpt-4o",
      "model_slug": "openai/gpt-4o",
      "model_name": "OpenAI: GPT-4o",
      "provider_name": "auto-routing",
      "accuracy": 0.510943,
      "accuracy_std_dev": 0.0008420000000000094,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.013169762626262625,
      "avg_duration_per_task_ms": 12925.411616161617,
      "avg_output_tokens_per_task": 1121.608585858586,
      "last_run_timestamp": "2026-10-01T16:03:45.472Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "DigitalOcean",
      "accuracy": 0.508418,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.030087782828282828,
      "avg_duration_per_task_ms": 3926274.5656565656,
      "avg_output_tokens_per_task": 21326.782828282827,
      "last_run_timestamp": "2026-08-22T17:37:26.215Z"
    },
    {
      "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": 0.5077319793814432,
      "accuracy_std_dev": 0.009572000000000025,
      "run_count": 2,
      "total_questions": 388,
      "avg_cost_per_task_usd": 0.011120894329896908,
      "avg_duration_per_task_ms": 426473.57216494845,
      "avg_output_tokens_per_task": 27736.909793814433,
      "last_run_timestamp": "2026-08-30T00:23:23.003Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4.1-nano-2025-04-14",
      "model_slug": "openai/gpt-4.1-nano",
      "model_name": "OpenAI: GPT-4.1 Nano",
      "provider_name": "OpenAI",
      "accuracy": 0.5050510000000001,
      "accuracy_std_dev": 0.002749143987983661,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0008273518518518518,
      "avg_duration_per_task_ms": 18715.93265993266,
      "avg_output_tokens_per_task": 1873.8602693602693,
      "last_run_timestamp": "2026-10-02T23:16:50.042Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "inclusionai/ling-3.0-flash-fin-20260827",
      "model_slug": "inclusionai/ling-3.0-flash-fin",
      "model_name": "inclusionAI: Ling 3.0 Flash Fin",
      "provider_name": "auto-routing",
      "accuracy": 0.505051,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.009755954545454545,
      "avg_duration_per_task_ms": 700759.0555555555,
      "avg_output_tokens_per_task": 64762.606060606064,
      "last_run_timestamp": "2026-10-01T08:24:17.043Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4.1-nano-2025-04-14",
      "model_slug": "openai/gpt-4.1-nano",
      "model_name": "OpenAI: GPT-4.1 Nano",
      "provider_name": "Azure",
      "accuracy": 0.4995792500000001,
      "accuracy_std_dev": 0.02025051336849268,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.0008581388888888889,
      "avg_duration_per_task_ms": 12992.723484848484,
      "avg_output_tokens_per_task": 1853.0088383838383,
      "last_run_timestamp": "2026-10-01T19:12:29.219Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "Groq",
      "accuracy": 0.49951899999999994,
      "accuracy_std_dev": 0.012685149528934545,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.0023819927849927846,
      "avg_duration_per_task_ms": 9173.090187590187,
      "avg_output_tokens_per_task": 2368.319624819625,
      "last_run_timestamp": "2026-10-31T16:04:29.124Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-14b-04-28",
      "model_slug": "qwen/qwen3-14b",
      "model_name": "Qwen: Qwen3 14B",
      "provider_name": "Alibaba",
      "accuracy": 0.49242400000000003,
      "accuracy_std_dev": 0.0008419999999999817,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.0017634166666666666,
      "avg_duration_per_task_ms": 30891.38888888889,
      "avg_output_tokens_per_task": 1731.7373737373737,
      "last_run_timestamp": "2026-10-01T16:05:56.758Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4.1-nano-2025-04-14",
      "model_slug": "openai/gpt-4.1-nano",
      "model_name": "OpenAI: GPT-4.1 Nano",
      "provider_name": "auto-routing",
      "accuracy": 0.49074049999999997,
      "accuracy_std_dev": 0.009259500000000004,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.0008383863636363636,
      "avg_duration_per_task_ms": 22708.954545454544,
      "avg_output_tokens_per_task": 1902.2954545454545,
      "last_run_timestamp": "2026-10-01T19:13:37.127Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "DigitalOcean",
      "accuracy": 0.4902728888888888,
      "accuracy_std_dev": 0.1180581110088911,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.003499423120089787,
      "avg_duration_per_task_ms": 205174.32940516275,
      "avg_output_tokens_per_task": 8349.423120089787,
      "last_run_timestamp": "2026-11-13T10:41:52.262Z"
    },
    {
      "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": 0.48436714285714283,
      "accuracy_std_dev": 0.012361779684964115,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.0026505512265512263,
      "avg_duration_per_task_ms": 49856.58441558442,
      "avg_output_tokens_per_task": 2925.426406926407,
      "last_run_timestamp": "2026-10-31T16:04:29.124Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "AkashML",
      "accuracy": 0.4836306971153846,
      "accuracy_std_dev": 0.010925716722180446,
      "run_count": 15,
      "total_questions": 2912,
      "avg_cost_per_task_usd": 0.0013556023351648351,
      "avg_duration_per_task_ms": 59721.06524725275,
      "avg_output_tokens_per_task": 2363.5096153846152,
      "last_run_timestamp": "2026-10-31T16:01:53.654Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "SambaNova",
      "accuracy": 0.48172171428571425,
      "accuracy_std_dev": 0.01107352294844762,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.0023519292929292932,
      "avg_duration_per_task_ms": 16333.764790764792,
      "avg_output_tokens_per_task": 2210.2922077922076,
      "last_run_timestamp": "2026-10-31T16:05:14.701Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.7-flash-20260119",
      "model_slug": "z-ai/glm-4.7-flash",
      "model_name": "Z.ai: GLM 4.7 Flash",
      "provider_name": "DeepInfra",
      "accuracy": 0.4787955114503817,
      "accuracy_std_dev": 0.007795999999999997,
      "run_count": 2,
      "total_questions": 393,
      "avg_cost_per_task_usd": 0.011332788804071247,
      "avg_duration_per_task_ms": 362429.45801526716,
      "avg_output_tokens_per_task": 28246.409669211196,
      "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": "Parasail",
      "accuracy": 0.47715228571428564,
      "accuracy_std_dev": 0.014837295958055795,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.0013270916305916304,
      "avg_duration_per_task_ms": 44547.831168831166,
      "avg_output_tokens_per_task": 2383.0901875901877,
      "last_run_timestamp": "2026-10-31T16:36:11.519Z"
    },
    {
      "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": "auto-routing",
      "accuracy": 0.4754687142857143,
      "accuracy_std_dev": 0.010214854935188212,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.0013394170274170275,
      "avg_duration_per_task_ms": 85516.39105339105,
      "avg_output_tokens_per_task": 2360.8838383838383,
      "last_run_timestamp": "2026-10-31T16:23:41.139Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "Together",
      "accuracy": 0.475084,
      "accuracy_std_dev": 0.006683163113376778,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.0031415444444444444,
      "avg_duration_per_task_ms": 73694.87777777777,
      "avg_output_tokens_per_task": 2214.6858585858586,
      "last_run_timestamp": "2026-10-17T16:05:55.476Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "Novita",
      "accuracy": 0.47234228571428577,
      "accuracy_std_dev": 0.009696421014466937,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.0010906248196248197,
      "avg_duration_per_task_ms": 63821.896103896106,
      "avg_output_tokens_per_task": 2454.1998556998556,
      "last_run_timestamp": "2026-10-31T16:31:54.106Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "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": "Cloudflare",
      "accuracy": 0.46955114615384613,
      "accuracy_std_dev": 0.06298705219874558,
      "run_count": 8,
      "total_questions": 1560,
      "avg_cost_per_task_usd": 0.0048902935897435895,
      "avg_duration_per_task_ms": 46682.213461538464,
      "avg_output_tokens_per_task": 2063.5615384615385,
      "last_run_timestamp": "2026-10-31T16:05:36.136Z"
    },
    {
      "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": 0.468855,
      "accuracy_std_dev": 0.012625999999999998,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.001083489898989899,
      "avg_duration_per_task_ms": 264261.16161616164,
      "avg_output_tokens_per_task": 2445.7929292929293,
      "last_run_timestamp": "2026-08-29T20:09:29.346Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "Crusoe",
      "accuracy": 0.468855,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.002208848484848485,
      "avg_duration_per_task_ms": 36376.982323232325,
      "avg_output_tokens_per_task": 2683.0429292929293,
      "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": 0.46412025,
      "accuracy_std_dev": 0.011095032793439591,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.0007977228535353536,
      "avg_duration_per_task_ms": 144411.35542929292,
      "avg_output_tokens_per_task": 2242.354797979798,
      "last_run_timestamp": "2026-10-31T16:02:14.361Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen-2.5-72b-instruct",
      "model_slug": "qwen/qwen-2.5-72b-instruct",
      "model_name": "Qwen2.5 72B Instruct",
      "provider_name": "Novita",
      "accuracy": 0.461279,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0011232171717171719,
      "avg_duration_per_task_ms": 136788.7626262626,
      "avg_output_tokens_per_task": 2036.2272727272727,
      "last_run_timestamp": "2026-10-01T16:27:00.375Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen-2.5-72b-instruct",
      "model_slug": "qwen/qwen-2.5-72b-instruct",
      "model_name": "Qwen2.5 72B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.45959600000000006,
      "accuracy_std_dev": 0.01346799999999998,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.0010278964646464647,
      "avg_duration_per_task_ms": 96379.95959595959,
      "avg_output_tokens_per_task": 1848.388888888889,
      "last_run_timestamp": "2026-10-01T16:47:11.191Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen2.5-vl-72b-instruct",
      "model_slug": "qwen/qwen2.5-vl-72b-instruct",
      "model_name": "Qwen: Qwen2.5 VL 72B Instruct",
      "provider_name": "Parasail",
      "accuracy": 0.457071,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0017076717171717172,
      "avg_duration_per_task_ms": 27081.62626262626,
      "avg_output_tokens_per_task": 1337.1363636363637,
      "last_run_timestamp": "2026-09-01T13:03:42.840Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen2.5-vl-72b-instruct",
      "model_slug": "qwen/qwen2.5-vl-72b-instruct",
      "model_name": "Qwen: Qwen2.5 VL 72B Instruct",
      "provider_name": "Nebius",
      "accuracy": 0.442761,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0013494898989898992,
      "avg_duration_per_task_ms": 69228.69696969698,
      "avg_output_tokens_per_task": 1528.5050505050506,
      "last_run_timestamp": "2026-09-01T12:41:25.955Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4o-mini",
      "model_slug": "openai/gpt-4o-mini",
      "model_name": "OpenAI: GPT-4o-mini",
      "provider_name": "auto-routing",
      "accuracy": 0.43602700000000005,
      "accuracy_std_dev": 0.006734000000000018,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.0010042146464646464,
      "avg_duration_per_task_ms": 21558.090909090908,
      "avg_output_tokens_per_task": 1481.030303030303,
      "last_run_timestamp": "2026-10-01T19:16:07.087Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4o-mini",
      "model_slug": "openai/gpt-4o-mini",
      "model_name": "OpenAI: GPT-4o-mini",
      "provider_name": "Azure",
      "accuracy": 0.43139700000000003,
      "accuracy_std_dev": 0.012951914704011906,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.0010746982323232325,
      "avg_duration_per_task_ms": 24227.715909090908,
      "avg_output_tokens_per_task": 1509.2487373737374,
      "last_run_timestamp": "2026-10-01T19:05:18.003Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen2.5-vl-72b-instruct",
      "model_slug": "qwen/qwen2.5-vl-72b-instruct",
      "model_name": "Qwen: Qwen2.5 VL 72B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.429293,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0015700353535353536,
      "avg_duration_per_task_ms": 59122.85858585859,
      "avg_output_tokens_per_task": 1664.8737373737374,
      "last_run_timestamp": "2026-09-01T13:21:35.512Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4o-mini",
      "model_slug": "openai/gpt-4o-mini",
      "model_name": "OpenAI: GPT-4o-mini",
      "provider_name": "OpenAI",
      "accuracy": 0.4276093333333334,
      "accuracy_std_dev": 0.00629925838668507,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0010292508417508418,
      "avg_duration_per_task_ms": 16725.542087542086,
      "avg_output_tokens_per_task": 1523.1902356902358,
      "last_run_timestamp": "2026-10-02T23:19:57.879Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-3-12b-it",
      "model_slug": "google/gemma-3-12b-it",
      "model_name": "Google: Gemma 3 12B",
      "provider_name": "DeepInfra",
      "accuracy": 0.405724,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0003259848484848485,
      "avg_duration_per_task_ms": 29056.348484848484,
      "avg_output_tokens_per_task": 1917.1515151515152,
      "last_run_timestamp": "2026-10-08T22:01:16.852Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "tencent/hy-mt2-30b-a3b-20260521",
      "model_slug": "tencent/hy-mt2-30b-a3b",
      "model_name": "Tencent: Hy-MT2-30B-A3B",
      "provider_name": "Reka",
      "accuracy": 0.388889,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.00023147979797979797,
      "avg_duration_per_task_ms": 8604.353535353535,
      "avg_output_tokens_per_task": 710.540404040404,
      "last_run_timestamp": "2026-09-24T06:00:03.613Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "amazon/nova-micro-v1",
      "model_slug": "amazon/nova-micro-v1",
      "model_name": "Amazon: Nova Micro 1.0",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.3855215,
      "accuracy_std_dev": 0.011293261320362695,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.0001567070707070707,
      "avg_duration_per_task_ms": 7111.70202020202,
      "avg_output_tokens_per_task": 932.0808080808081,
      "last_run_timestamp": "2026-10-01T19:39:32.831Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-3-12b-it",
      "model_slug": "google/gemma-3-12b-it",
      "model_name": "Google: Gemma 3 12B",
      "provider_name": "NextBit",
      "accuracy": 0.38047125000000004,
      "accuracy_std_dev": 0.012081350470353048,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.0003843409090909091,
      "avg_duration_per_task_ms": 52179.61616161616,
      "avg_output_tokens_per_task": 2261.531565656566,
      "last_run_timestamp": "2026-11-01T11:50:48.836Z"
    },
    {
      "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": 0.3787201875,
      "accuracy_std_dev": 0.3723545,
      "run_count": 2,
      "total_questions": 224,
      "avg_cost_per_task_usd": 0.0059331517857142855,
      "avg_duration_per_task_ms": 145883.42857142858,
      "avg_output_tokens_per_task": 5127.165178571428,
      "last_run_timestamp": "2026-11-11T05:27:07.237Z"
    },
    {
      "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": 0.3633958571428571,
      "accuracy_std_dev": 0.08617342972074823,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.008518821067821068,
      "avg_duration_per_task_ms": 483581.14285714284,
      "avg_output_tokens_per_task": 21062.341991341993,
      "last_run_timestamp": "2026-10-31T16:45:48.945Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Reka",
      "accuracy": 0.346801,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.008616373737373737,
      "avg_duration_per_task_ms": 1179181.5303030303,
      "avg_output_tokens_per_task": 32995.232323232325,
      "last_run_timestamp": "2026-10-03T00:19:24.096Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.1-8b-instruct",
      "model_slug": "meta-llama/llama-3.1-8b-instruct",
      "model_name": "Meta: Llama 3.1 8B Instruct",
      "provider_name": "Groq",
      "accuracy": 0.2912128823529411,
      "accuracy_std_dev": 0.01689256477789889,
      "run_count": 7,
      "total_questions": 1377,
      "avg_cost_per_task_usd": 0.0029356129266521424,
      "avg_duration_per_task_ms": 41493.66521423384,
      "avg_output_tokens_per_task": 36404.61365286855,
      "last_run_timestamp": "2026-10-31T16:04:29.122Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "sao10k/l3-lunaris-8b",
      "model_slug": "sao10k/l3-lunaris-8b",
      "model_name": "Sao10K: Llama 3 8B Lunaris",
      "provider_name": "DeepInfra",
      "accuracy": 0.2873176666666667,
      "accuracy_std_dev": 0.0007938452130121062,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.00006128619528619528,
      "avg_duration_per_task_ms": 10231.205387205388,
      "avg_output_tokens_per_task": 580.9023569023569,
      "last_run_timestamp": "2026-10-08T22:00:27.722Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.2-3b-instruct",
      "model_slug": "meta-llama/llama-3.2-3b-instruct",
      "model_name": "Meta: Llama 3.2 3B Instruct",
      "provider_name": "Cloudflare",
      "accuracy": 0.286195,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0005941969696969697,
      "avg_duration_per_task_ms": 14998.156565656565,
      "avg_output_tokens_per_task": 1642.1363636363637,
      "last_run_timestamp": "2026-10-13T18:56:05.001Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.1-8b-instruct",
      "model_slug": "meta-llama/llama-3.1-8b-instruct",
      "model_name": "Meta: Llama 3.1 8B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.2770562857142857,
      "accuracy_std_dev": 0.021947671114424378,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.0010397142857142857,
      "avg_duration_per_task_ms": 152294.55627705628,
      "avg_output_tokens_per_task": 15530.19769119769,
      "last_run_timestamp": "2026-10-31T16:08:31.064Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "sao10k/l3-lunaris-8b",
      "model_slug": "sao10k/l3-lunaris-8b",
      "model_name": "Sao10K: Llama 3 8B Lunaris",
      "provider_name": "auto-routing",
      "accuracy": 0.27665533333333336,
      "accuracy_std_dev": 0.0007938452130121063,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.00006201683501683503,
      "avg_duration_per_task_ms": 9496.395622895623,
      "avg_output_tokens_per_task": 580.2053872053872,
      "last_run_timestamp": "2026-10-01T20:43:57.173Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "sao10k/l3-lunaris-8b",
      "model_slug": "sao10k/l3-lunaris-8b",
      "model_name": "Sao10K: Llama 3 8B Lunaris",
      "provider_name": "Novita",
      "accuracy": 0.2745902896174863,
      "accuracy_std_dev": 0.009941000000000005,
      "run_count": 2,
      "total_questions": 366,
      "avg_cost_per_task_usd": 0.0000641912568306011,
      "avg_duration_per_task_ms": 8433.43169398907,
      "avg_output_tokens_per_task": 532.0136612021857,
      "last_run_timestamp": "2026-10-01T16:01:04.877Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "sao10k/l3-lunaris-8b",
      "model_slug": "sao10k/l3-lunaris-8b",
      "model_name": "Sao10K: Llama 3 8B Lunaris",
      "provider_name": "Parasail",
      "accuracy": 0.27161569453924916,
      "accuracy_std_dev": 0.020450238993441835,
      "run_count": 3,
      "total_questions": 586,
      "avg_cost_per_task_usd": 0.000055156996587030725,
      "avg_duration_per_task_ms": 10158.571672354949,
      "avg_output_tokens_per_task": 550.3515358361775,
      "last_run_timestamp": "2026-10-01T20:47:49.128Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-3-4b-it",
      "model_slug": "google/gemma-3-4b-it",
      "model_name": "Google: Gemma 3 4B",
      "provider_name": "Parasail",
      "accuracy": 0.271044,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.00011306565656565657,
      "avg_duration_per_task_ms": 10846.11616161616,
      "avg_output_tokens_per_task": 1804.4646464646464,
      "last_run_timestamp": "2026-10-15T03:00:03.049Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.1-8b-instruct",
      "model_slug": "meta-llama/llama-3.1-8b-instruct",
      "model_name": "Meta: Llama 3.1 8B Instruct",
      "provider_name": "CoreWeave",
      "accuracy": 0.26953032780708364,
      "accuracy_std_dev": 0.014445882398999181,
      "run_count": 7,
      "total_questions": 1327,
      "avg_cost_per_task_usd": 0.011133856819894499,
      "avg_duration_per_task_ms": 459953.84174830443,
      "avg_output_tokens_per_task": 49838.27731725697,
      "last_run_timestamp": "2026-10-31T16:04:29.124Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.1-8b-instruct",
      "model_slug": "meta-llama/llama-3.1-8b-instruct",
      "model_name": "Meta: Llama 3.1 8B Instruct",
      "provider_name": "DeepInfra",
      "accuracy": 0.26473062500000005,
      "accuracy_std_dev": 0.010705448612009439,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.0003948459595959595,
      "avg_duration_per_task_ms": 362944.5946969697,
      "avg_output_tokens_per_task": 9468.12436868687,
      "last_run_timestamp": "2026-10-31T16:08:31.068Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.1-8b-instruct",
      "model_slug": "meta-llama/llama-3.1-8b-instruct",
      "model_name": "Meta: Llama 3.1 8B Instruct",
      "provider_name": "Novita",
      "accuracy": 0.25901885714285716,
      "accuracy_std_dev": 0.013953091766226396,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.00042360028860028863,
      "avg_duration_per_task_ms": 73749.8658008658,
      "avg_output_tokens_per_task": 8148.3650793650795,
      "last_run_timestamp": "2026-10-31T16:55:07.715Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-nano-banana-2.1-20261006",
      "model_slug": "google/gemini-nano-banana-2.1",
      "model_name": "Google: Nano Banana 2.1",
      "provider_name": "Google",
      "accuracy": 0.213805,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.17523430303030305,
      "avg_duration_per_task_ms": 78031.78787878787,
      "avg_output_tokens_per_task": 14613.282828282829,
      "last_run_timestamp": "2026-11-12T21:01:32.027Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "inclusionai/ling-3.0-flash-fin-20260827",
      "model_slug": "inclusionai/ling-3.0-flash-fin",
      "model_name": "inclusionAI: Ling 3.0 Flash Fin",
      "provider_name": "DeepInfra",
      "accuracy": 0.17174282233502539,
      "accuracy_std_dev": 0.01788397279378631,
      "run_count": 3,
      "total_questions": 591,
      "avg_cost_per_task_usd": 0.019897788494077834,
      "avg_duration_per_task_ms": 1784246.4196277496,
      "avg_output_tokens_per_task": 110294.06598984772,
      "last_run_timestamp": "2026-10-08T23:13:58.126Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.1-flash-lite-image-20260630",
      "model_slug": "google/gemini-3.1-flash-lite-image",
      "model_name": "Google: Nano Banana 2 Lite (Gemini 3.1 Flash Lite Image)",
      "provider_name": "Google AI Studio",
      "accuracy": 0.1161615,
      "accuracy_std_dev": 0.0016835000000000044,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.017850737373737373,
      "avg_duration_per_task_ms": 46853.982323232325,
      "avg_output_tokens_per_task": 11776.444444444445,
      "last_run_timestamp": "2026-10-01T16:00:58.389Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-chat-v3-0324",
      "model_slug": "deepseek/deepseek-chat-v3-0324",
      "model_name": "DeepSeek: DeepSeek V3 0324",
      "provider_name": "Novita",
      "accuracy": 0.11531999999999999,
      "accuracy_std_dev": 0.027778000000000004,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.0013887626262626262,
      "avg_duration_per_task_ms": 56176.38383838384,
      "avg_output_tokens_per_task": 1211.2525252525252,
      "last_run_timestamp": "2026-08-29T18:43:09.114Z"
    },
    {
      "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-3.1-8b-instruct",
      "model_slug": "meta-llama/llama-3.1-8b-instruct",
      "model_name": "Meta: Llama 3.1 8B Instruct",
      "provider_name": "Cloudflare",
      "accuracy": 0.09483766666666665,
      "accuracy_std_dev": 0.1317391787847825,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0006702474747474747,
      "avg_duration_per_task_ms": 243969.58922558924,
      "avg_output_tokens_per_task": 2175.9141414141413,
      "last_run_timestamp": "2026-10-24T16:03:19.773Z"
    },
    {
      "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": 0.0033670000000000006,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0034259191919191923,
      "avg_duration_per_task_ms": 169432.20707070708,
      "avg_output_tokens_per_task": 3225.4444444444443,
      "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": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.003385919191919192,
      "avg_duration_per_task_ms": 136907.34343434343,
      "avg_output_tokens_per_task": 3185.4444444444443,
      "last_run_timestamp": "2026-10-31T16:04:43.943Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "OpenInference",
      "accuracy": 0.816667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.05960144,
      "avg_duration_per_task_ms": 162346.76,
      "avg_output_tokens_per_task": 42454.8,
      "last_run_timestamp": "2026-10-17T19:19:25.280Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-5-fable-20260609",
      "model_slug": "anthropic/claude-fable-5",
      "model_name": "Anthropic: Claude Fable 5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.8000000909090909,
      "accuracy_std_dev": 0.020695063581156234,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.8742483381818181,
      "avg_duration_per_task_ms": 136193.65272727274,
      "avg_output_tokens_per_task": 5706.489090909091,
      "last_run_timestamp": "2026-10-31T12:24:55.848Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.1-20251113",
      "model_slug": "openai/gpt-5.1",
      "model_name": "OpenAI: GPT-5.1",
      "provider_name": "auto-routing",
      "accuracy": 0.8,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.39313749000000003,
      "avg_duration_per_task_ms": 477750.16,
      "avg_output_tokens_per_task": 34082.08,
      "last_run_timestamp": "2026-10-01T16:00:45.071Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7995237142857143,
      "accuracy_std_dev": 0.010754053185583914,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.14178823428571427,
      "avg_duration_per_task_ms": 2712952.931428571,
      "avg_output_tokens_per_task": 28444.37714285714,
      "last_run_timestamp": "2026-10-31T20:31:41.691Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "BaseTen",
      "accuracy": 0.7981221830985916,
      "accuracy_std_dev": 0.05639978873679898,
      "run_count": 3,
      "total_questions": 142,
      "avg_cost_per_task_usd": 0.007624964788732395,
      "avg_duration_per_task_ms": 30939.802816901407,
      "avg_output_tokens_per_task": 8761.732394366198,
      "last_run_timestamp": "2026-10-31T11:56:20.980Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-5-fable-20260609",
      "model_slug": "anthropic/claude-fable-5",
      "model_name": "Anthropic: Claude Fable 5",
      "provider_name": "Azure",
      "accuracy": 0.798,
      "accuracy_std_dev": 0.022910589499181372,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.8759728400000001,
      "avg_duration_per_task_ms": 127096.316,
      "avg_output_tokens_per_task": 5702.834,
      "last_run_timestamp": "2026-10-31T11:20:38.480Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "Phala",
      "accuracy": 0.7978929137931035,
      "accuracy_std_dev": 0.03017695806811711,
      "run_count": 7,
      "total_questions": 348,
      "avg_cost_per_task_usd": 0.20533228735632184,
      "avg_duration_per_task_ms": 772776.3793103448,
      "avg_output_tokens_per_task": 35305.502873563215,
      "last_run_timestamp": "2026-10-31T19:01:09.612Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "Alibaba",
      "accuracy": 0.7936961919770773,
      "accuracy_std_dev": 0.02409553422592355,
      "run_count": 7,
      "total_questions": 349,
      "avg_cost_per_task_usd": 0.18211524641833812,
      "avg_duration_per_task_ms": 451041.38395415473,
      "avg_output_tokens_per_task": 30422.72206303725,
      "last_run_timestamp": "2026-10-31T19:12:41.306Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "inclusionai/ling-3.1-flash-20261002",
      "model_slug": "inclusionai/ling-3.1-flash",
      "model_name": "inclusionAI: Ling 3.1 Flash",
      "provider_name": "Novita",
      "accuracy": 0.793333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.00828808,
      "avg_duration_per_task_ms": 196026.56,
      "avg_output_tokens_per_task": 16303,
      "last_run_timestamp": "2026-10-14T19:21:35.384Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "inclusionai/ling-3.1-flash-20261002",
      "model_slug": "inclusionai/ling-3.1-flash",
      "model_name": "inclusionAI: Ling 3.1 Flash",
      "provider_name": "DeepInfra",
      "accuracy": 0.793333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.03146312,
      "avg_duration_per_task_ms": 636210.72,
      "avg_output_tokens_per_task": 16745.44,
      "last_run_timestamp": "2026-10-14T19:17:59.742Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-5-fable-20260609",
      "model_slug": "anthropic/claude-fable-5",
      "model_name": "Anthropic: Claude Fable 5",
      "provider_name": "Anthropic",
      "accuracy": 0.7926667,
      "accuracy_std_dev": 0.017751370527652206,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.81432549,
      "avg_duration_per_task_ms": 121567.298,
      "avg_output_tokens_per_task": 5714.138,
      "last_run_timestamp": "2026-10-31T11:18:52.340Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "Alibaba",
      "accuracy": 0.7924998750000001,
      "accuracy_std_dev": 0.02591377395535768,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.082230975,
      "avg_duration_per_task_ms": 213864.49,
      "avg_output_tokens_per_task": 26801.935,
      "last_run_timestamp": "2026-10-31T19:22:16.313Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "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": 0.7914285714285715,
      "accuracy_std_dev": 0.01807021076529105,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.20560172000000002,
      "avg_duration_per_task_ms": 767128.9885714286,
      "avg_output_tokens_per_task": 32379.805714285714,
      "last_run_timestamp": "2026-10-31T19:47:27.263Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-fable-5.1-20260831",
      "model_slug": "anthropic/claude-fable-5.1",
      "model_name": "Anthropic: Claude Fable 5.1",
      "provider_name": "Anthropic",
      "accuracy": 0.7906666,
      "accuracy_std_dev": 0.009977824303925176,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.814870324,
      "avg_duration_per_task_ms": 182126.652,
      "avg_output_tokens_per_task": 7046.908,
      "last_run_timestamp": "2026-10-01T08:05:22.427Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-5-fable-20260609",
      "model_slug": "anthropic/claude-fable-5",
      "model_name": "Anthropic: Claude Fable 5",
      "provider_name": "auto-routing",
      "accuracy": 0.79,
      "accuracy_std_dev": 0.026204223579415577,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.857535498,
      "avg_duration_per_task_ms": 115142.934,
      "avg_output_tokens_per_task": 5717.478,
      "last_run_timestamp": "2026-10-31T11:38:36.784Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Chutes",
      "accuracy": 0.789795887755102,
      "accuracy_std_dev": 0.02205636874011677,
      "run_count": 10,
      "total_questions": 490,
      "avg_cost_per_task_usd": 0.07972918775510204,
      "avg_duration_per_task_ms": 1143421.8387755102,
      "avg_output_tokens_per_task": 25766.961224489794,
      "last_run_timestamp": "2026-10-31T20:22:41.827Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-5-fable-20260609",
      "model_slug": "anthropic/claude-fable-5",
      "model_name": "Anthropic: Claude Fable 5",
      "provider_name": "Google",
      "accuracy": 0.7892064761904761,
      "accuracy_std_dev": 0.022488265074455106,
      "run_count": 21,
      "total_questions": 1050,
      "avg_cost_per_task_usd": 0.8947322371428572,
      "avg_duration_per_task_ms": 91208.88952380953,
      "avg_output_tokens_per_task": 5717.874285714286,
      "last_run_timestamp": "2026-11-11T21:16:12.438Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "Xiaomi",
      "accuracy": 0.788889,
      "accuracy_std_dev": 0.015234869040023505,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.057788776666666666,
      "avg_duration_per_task_ms": 893038.1233333333,
      "avg_output_tokens_per_task": 40447.416666666664,
      "last_run_timestamp": "2026-10-31T13:13:33.031Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Wafer",
      "accuracy": 0.7882883714036617,
      "accuracy_std_dev": 0.015477690995371508,
      "run_count": 23,
      "total_questions": 1147,
      "avg_cost_per_task_usd": 0.0821072258064516,
      "avg_duration_per_task_ms": 149668.3635571055,
      "avg_output_tokens_per_task": 12209.048823016565,
      "last_run_timestamp": "2026-10-31T12:32:50.229Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-sonnet-5.5-20260928",
      "model_slug": "anthropic/claude-sonnet-5.5",
      "model_name": "Anthropic: Claude Sonnet 5.5",
      "provider_name": "auto-routing",
      "accuracy": 0.788,
      "accuracy_std_dev": 0.02246951515275753,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.15925035199999998,
      "avg_duration_per_task_ms": 71120.884,
      "avg_output_tokens_per_task": 6245.168,
      "last_run_timestamp": "2026-10-31T11:10:36.361Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-4-argon-20261013",
      "model_slug": "google/gemini-4-argon",
      "model_name": "Google: Gemini 4 Argon",
      "provider_name": "Google",
      "accuracy": 0.786667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.49828548,
      "avg_duration_per_task_ms": 348417.68,
      "avg_output_tokens_per_task": 21971.04,
      "last_run_timestamp": "2026-10-14T15:16:05.578Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3.1-pro-20261023",
      "model_slug": "moonshotai/kimi-k3.1-pro",
      "model_name": "MoonshotAI: Kimi K3.1 Pro",
      "provider_name": "DigitalOcean",
      "accuracy": 0.786667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.37777382000000004,
      "avg_duration_per_task_ms": 196216.66,
      "avg_output_tokens_per_task": 12248.68,
      "last_run_timestamp": "2026-11-07T03:08:27.754Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3.1-pro-20261023",
      "model_slug": "moonshotai/kimi-k3.1-pro",
      "model_name": "MoonshotAI: Kimi K3.1 Pro",
      "provider_name": "Fireworks",
      "accuracy": 0.786667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.20422622,
      "avg_duration_per_task_ms": 176471.64,
      "avg_output_tokens_per_task": 10146.04,
      "last_run_timestamp": "2026-11-05T15:41:01.946Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "DeepInfra",
      "accuracy": 0.786127239884393,
      "accuracy_std_dev": 0.023660230134130145,
      "run_count": 7,
      "total_questions": 346,
      "avg_cost_per_task_usd": 0.09665253179190753,
      "avg_duration_per_task_ms": 380894.4450867052,
      "avg_output_tokens_per_task": 16737.754335260117,
      "last_run_timestamp": "2026-10-31T12:56:48.075Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "BaseTen",
      "accuracy": 0.7859650631578948,
      "accuracy_std_dev": 0.03135199999999999,
      "run_count": 2,
      "total_questions": 95,
      "avg_cost_per_task_usd": 0.09807105263157893,
      "avg_duration_per_task_ms": 69953.93684210526,
      "avg_output_tokens_per_task": 8018.557894736842,
      "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/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Reka",
      "accuracy": 0.7858794166666667,
      "accuracy_std_dev": 0.021093449409710124,
      "run_count": 3,
      "total_questions": 144,
      "avg_cost_per_task_usd": 0.07943493055555556,
      "avg_duration_per_task_ms": 334937.25694444444,
      "avg_output_tokens_per_task": 18674.54861111111,
      "last_run_timestamp": "2026-10-31T19:43:30.408Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "Parasail",
      "accuracy": 0.7845118821548821,
      "accuracy_std_dev": 0.03356639684143456,
      "run_count": 6,
      "total_questions": 297,
      "avg_cost_per_task_usd": 0.19079432323232323,
      "avg_duration_per_task_ms": 638899.1245791246,
      "avg_output_tokens_per_task": 30741.198653198655,
      "last_run_timestamp": "2026-10-31T19:27:20.607Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Together",
      "accuracy": 0.7844443333333334,
      "accuracy_std_dev": 0.01632984089872824,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.0785912688888889,
      "avg_duration_per_task_ms": 47362.52222222222,
      "avg_output_tokens_per_task": 4720.888888888889,
      "last_run_timestamp": "2026-10-31T11:58:26.584Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-sonnet-5.5-20260928",
      "model_slug": "anthropic/claude-sonnet-5.5",
      "model_name": "Anthropic: Claude Sonnet 5.5",
      "provider_name": "Azure",
      "accuracy": 0.7833333333333333,
      "accuracy_std_dev": 0.011706202852429232,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.1729425,
      "avg_duration_per_task_ms": 66541.35,
      "avg_output_tokens_per_task": 6203.445,
      "last_run_timestamp": "2026-10-31T12:38:55.501Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.5-flash-lite-20260721",
      "model_slug": "google/gemini-3.5-flash-lite",
      "model_name": "Google: Gemini 3.5 Flash Lite",
      "provider_name": "auto-routing",
      "accuracy": 0.7833330000000001,
      "accuracy_std_dev": 0.009999999999999953,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.07443635,
      "avg_duration_per_task_ms": 98397.15,
      "avg_output_tokens_per_task": 21518.67,
      "last_run_timestamp": "2026-10-01T19:04:42.463Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Novita",
      "accuracy": 0.7825,
      "accuracy_std_dev": 0.023790007240436065,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.08953825,
      "avg_duration_per_task_ms": 274829.5175,
      "avg_output_tokens_per_task": 19992.6725,
      "last_run_timestamp": "2026-10-31T12:57:09.932Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Novita",
      "accuracy": 0.7822223333333334,
      "accuracy_std_dev": 0.01571327975171178,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.011317603333333334,
      "avg_duration_per_task_ms": 599501.3033333333,
      "avg_output_tokens_per_task": 21464.33,
      "last_run_timestamp": "2026-10-31T14:08:11.994Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-opus-5-20260723",
      "model_slug": "anthropic/claude-opus-5",
      "model_name": "Anthropic: Claude Opus 5",
      "provider_name": "auto-routing",
      "accuracy": 0.782,
      "accuracy_std_dev": 0.01521706644527781,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.470768748,
      "avg_duration_per_task_ms": 115538.554,
      "avg_output_tokens_per_task": 6965.978,
      "last_run_timestamp": "2026-10-31T12:17:12.454Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-122b-a10b-20260224",
      "model_slug": "qwen/qwen3.5-122b-a10b",
      "model_name": "Qwen: Qwen3.5-122B-A10B",
      "provider_name": "Novita",
      "accuracy": 0.7818181878787879,
      "accuracy_std_dev": 0.019584458716288278,
      "run_count": 10,
      "total_questions": 495,
      "avg_cost_per_task_usd": 0.19799780808080808,
      "avg_duration_per_task_ms": 511549.30303030304,
      "avg_output_tokens_per_task": 27725.535353535353,
      "last_run_timestamp": "2026-10-31T18:53:29.091Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "Novita",
      "accuracy": 0.781333,
      "accuracy_std_dev": 0.027129319932501044,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.0587014,
      "avg_duration_per_task_ms": 871009.496,
      "avg_output_tokens_per_task": 40629.384,
      "last_run_timestamp": "2026-10-31T13:41:45.309Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Fireworks",
      "accuracy": 0.7811976407035177,
      "accuracy_std_dev": 0.02110393286871771,
      "run_count": 16,
      "total_questions": 796,
      "avg_cost_per_task_usd": 0.09273568467336683,
      "avg_duration_per_task_ms": 66515.5364321608,
      "avg_output_tokens_per_task": 4679.810301507538,
      "last_run_timestamp": "2026-10-31T13:12:01.731Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Crusoe",
      "accuracy": 0.7808333749999999,
      "accuracy_std_dev": 0.010769114145758466,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0823930725,
      "avg_duration_per_task_ms": 41435.8875,
      "avg_output_tokens_per_task": 4744.8375,
      "last_run_timestamp": "2026-10-31T12:40:46.044Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-sonnet-5.5-20260928",
      "model_slug": "anthropic/claude-sonnet-5.5",
      "model_name": "Anthropic: Claude Sonnet 5.5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7800001,
      "accuracy_std_dev": 0.02108170349592273,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.174296886,
      "avg_duration_per_task_ms": 72081.318,
      "avg_output_tokens_per_task": 6231.684,
      "last_run_timestamp": "2026-10-31T12:10:32.453Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-sonnet-5.5-20260928",
      "model_slug": "anthropic/claude-sonnet-5.5",
      "model_name": "Anthropic: Claude Sonnet 5.5",
      "provider_name": "Google",
      "accuracy": 0.78,
      "accuracy_std_dev": 0.021081938910409114,
      "run_count": 18,
      "total_questions": 900,
      "avg_cost_per_task_usd": 0.18010515333333332,
      "avg_duration_per_task_ms": 64530.58666666667,
      "avg_output_tokens_per_task": 6325.823333333334,
      "last_run_timestamp": "2026-10-31T12:07:29.560Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Together",
      "accuracy": 0.78,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.08351372,
      "avg_duration_per_task_ms": 190050,
      "avg_output_tokens_per_task": 8493.48,
      "last_run_timestamp": "2026-08-22T08:39:22.990Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.78,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.0042799399999999994,
      "avg_duration_per_task_ms": 627315.22,
      "avg_output_tokens_per_task": 12675.1,
      "last_run_timestamp": "2026-10-31T21:10:18.835Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.7786668,
      "accuracy_std_dev": 0.01484714413481597,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.014840148000000001,
      "avg_duration_per_task_ms": 379677.016,
      "avg_output_tokens_per_task": 21736.348,
      "last_run_timestamp": "2026-10-31T12:28:58.923Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Near AI",
      "accuracy": 0.7786666,
      "accuracy_std_dev": 0.010666575000439441,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.031286299999999996,
      "avg_duration_per_task_ms": 221994.964,
      "avg_output_tokens_per_task": 12740.492,
      "last_run_timestamp": "2026-10-31T12:48:16.034Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "BaseTen",
      "accuracy": 0.7784256831875607,
      "accuracy_std_dev": 0.029691712716069543,
      "run_count": 21,
      "total_questions": 1029,
      "avg_cost_per_task_usd": 0.07680297376093294,
      "avg_duration_per_task_ms": 74568.79203109816,
      "avg_output_tokens_per_task": 4283.289601554908,
      "last_run_timestamp": "2026-10-31T13:24:01.335Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "ModelRun",
      "accuracy": 0.77833325,
      "accuracy_std_dev": 0.005527984369324845,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.12165100000000001,
      "avg_duration_per_task_ms": 123453.655,
      "avg_output_tokens_per_task": 4597.39,
      "last_run_timestamp": "2026-10-28T05:13:02.136Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Z.AI",
      "accuracy": 0.7779999000000001,
      "accuracy_std_dev": 0.020011142453393298,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.081736628,
      "avg_duration_per_task_ms": 135707.96,
      "avg_output_tokens_per_task": 4671.7,
      "last_run_timestamp": "2026-10-31T12:38:49.692Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "AtlasCloud",
      "accuracy": 0.777778,
      "accuracy_std_dev": 0.01257094435593444,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.016507473333333335,
      "avg_duration_per_task_ms": 92510.76666666666,
      "avg_output_tokens_per_task": 10861.266666666666,
      "last_run_timestamp": "2026-10-31T11:43:38.949Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Nebius",
      "accuracy": 0.777459306590258,
      "accuracy_std_dev": 0.02712463143584833,
      "run_count": 7,
      "total_questions": 349,
      "avg_cost_per_task_usd": 0.17213766475644698,
      "avg_duration_per_task_ms": 55665.0888252149,
      "avg_output_tokens_per_task": 4643.535816618911,
      "last_run_timestamp": "2026-10-31T12:26:01.294Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-sonnet-5.5-20260928",
      "model_slug": "anthropic/claude-sonnet-5.5",
      "model_name": "Anthropic: Claude Sonnet 5.5",
      "provider_name": "Anthropic",
      "accuracy": 0.7766666666666666,
      "accuracy_std_dev": 0.014782246611692315,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.17363205,
      "avg_duration_per_task_ms": 73619.57333333333,
      "avg_output_tokens_per_task": 6340.086666666667,
      "last_run_timestamp": "2026-10-31T12:24:45.929Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7766665833333334,
      "accuracy_std_dev": 0.0379571517689318,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.10135336166666668,
      "avg_duration_per_task_ms": 172908.82,
      "avg_output_tokens_per_task": 6639.301666666666,
      "last_run_timestamp": "2026-10-31T11:53:06.149Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "CoreWeave",
      "accuracy": 0.7766665,
      "accuracy_std_dev": 0.03666649999999999,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.15718574,
      "avg_duration_per_task_ms": 517790.44,
      "avg_output_tokens_per_task": 27035.53,
      "last_run_timestamp": "2026-08-29T08:26:04.467Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "InferenceNet",
      "accuracy": 0.7766665,
      "accuracy_std_dev": 0.016666499999999973,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.017346550000000002,
      "avg_duration_per_task_ms": 335231.13,
      "avg_output_tokens_per_task": 20113.88,
      "last_run_timestamp": "2026-10-03T08:17:48.643Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Alibaba",
      "accuracy": 0.7764815,
      "accuracy_std_dev": 0.02583109973193382,
      "run_count": 18,
      "total_questions": 900,
      "avg_cost_per_task_usd": 0.13240519666666667,
      "avg_duration_per_task_ms": 319498.1977777778,
      "avg_output_tokens_per_task": 19319.296666666665,
      "last_run_timestamp": "2026-10-31T13:25:38.758Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "CoreWeave",
      "accuracy": 0.7760940000000001,
      "accuracy_std_dev": 0.020936325423770052,
      "run_count": 4,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.06970348989898989,
      "avg_duration_per_task_ms": 95095.22222222222,
      "avg_output_tokens_per_task": 4292.19696969697,
      "last_run_timestamp": "2026-10-31T12:44:06.053Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "DeepInfra",
      "accuracy": 0.7760001999999999,
      "accuracy_std_dev": 0.028472283599318148,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.06820797999999999,
      "avg_duration_per_task_ms": 658315.768,
      "avg_output_tokens_per_task": 13133.756,
      "last_run_timestamp": "2026-10-31T13:30:16.905Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.7-opus-20260416",
      "model_slug": "anthropic/claude-opus-4.7",
      "model_name": "Anthropic: Claude Opus 4.7",
      "provider_name": "Anthropic",
      "accuracy": 0.7759999000000001,
      "accuracy_std_dev": 0.01890341768278953,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.43835202,
      "avg_duration_per_task_ms": 86044.028,
      "avg_output_tokens_per_task": 5677.174,
      "last_run_timestamp": "2026-10-31T12:26:17.134Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7758335,
      "accuracy_std_dev": 0.016308646311082982,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.15398643500000003,
      "avg_duration_per_task_ms": 330861.565,
      "avg_output_tokens_per_task": 19948.0325,
      "last_run_timestamp": "2026-10-31T13:41:05.689Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Cloudflare",
      "accuracy": 0.7757695461847389,
      "accuracy_std_dev": 0.016169521055368324,
      "run_count": 10,
      "total_questions": 498,
      "avg_cost_per_task_usd": 0.08735203012048193,
      "avg_duration_per_task_ms": 139136.04417670684,
      "avg_output_tokens_per_task": 4443.712851405622,
      "last_run_timestamp": "2026-10-31T12:29:23.419Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "StreamLake",
      "accuracy": 0.7757574545454545,
      "accuracy_std_dev": 0.02074823590031709,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.08103076363636365,
      "avg_duration_per_task_ms": 336159.44,
      "avg_output_tokens_per_task": 20208.643636363635,
      "last_run_timestamp": "2026-10-31T12:24:53.587Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta/muse-glimmer-30b-20260810",
      "model_slug": "meta/muse-glimmer-30b",
      "model_name": "Meta: Muse Glimmer 30B",
      "provider_name": "DeepInfra",
      "accuracy": 0.7757142857142857,
      "accuracy_std_dev": 0.033557166846673904,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.03545617428571429,
      "avg_duration_per_task_ms": 502388.40285714285,
      "avg_output_tokens_per_task": 13844.954285714286,
      "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": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "DeepInfra",
      "accuracy": 0.7753625384615385,
      "accuracy_std_dev": 0.017323958686038136,
      "run_count": 6,
      "total_questions": 299,
      "avg_cost_per_task_usd": 0.06710363879598663,
      "avg_duration_per_task_ms": 1125092.9698996656,
      "avg_output_tokens_per_task": 36573.67558528428,
      "last_run_timestamp": "2026-10-24T13:26:35.887Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-122b-a10b-20260224",
      "model_slug": "qwen/qwen3.5-122b-a10b",
      "model_name": "Qwen: Qwen3.5-122B-A10B",
      "provider_name": "auto-routing",
      "accuracy": 0.7753334,
      "accuracy_std_dev": 0.024046292392799363,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.16927985399999998,
      "avg_duration_per_task_ms": 561845.348,
      "avg_output_tokens_per_task": 35586.574,
      "last_run_timestamp": "2026-10-31T19:08:11.310Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "auto-routing",
      "accuracy": 0.7753334,
      "accuracy_std_dev": 0.026339740994170767,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.065886822,
      "avg_duration_per_task_ms": 257998.344,
      "avg_output_tokens_per_task": 14164.592,
      "last_run_timestamp": "2026-10-31T12:39:11.628Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "AtlasCloud",
      "accuracy": 0.775,
      "accuracy_std_dev": 0.01787294665828779,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.09322250500000001,
      "avg_duration_per_task_ms": 167383.47,
      "avg_output_tokens_per_task": 5123.47,
      "last_run_timestamp": "2026-10-31T12:32:51.992Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-opus-5-20260723",
      "model_slug": "anthropic/claude-opus-5",
      "model_name": "Anthropic: Claude Opus 5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.774712655172414,
      "accuracy_std_dev": 0.021326058387893743,
      "run_count": 29,
      "total_questions": 1450,
      "avg_cost_per_task_usd": 0.5006132268965519,
      "avg_duration_per_task_ms": 112500.0448275862,
      "avg_output_tokens_per_task": 7031.657931034482,
      "last_run_timestamp": "2026-10-31T11:36:20.802Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.8-opus-20260528",
      "model_slug": "anthropic/claude-opus-4.8",
      "model_name": "Anthropic: Claude Opus 4.8",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.7745454545454545,
      "accuracy_std_dev": 0.02425759601131023,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.48932630545454553,
      "avg_duration_per_task_ms": 124123.69272727273,
      "avg_output_tokens_per_task": 7977.827272727272,
      "last_run_timestamp": "2026-10-31T11:58:26.584Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "DeepInfra",
      "accuracy": 0.7742857142857142,
      "accuracy_std_dev": 0.026945723560596418,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.21623159142857146,
      "avg_duration_per_task_ms": 562306.5571428571,
      "avg_output_tokens_per_task": 12602.945714285714,
      "last_run_timestamp": "2026-10-10T11:06:43.832Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Friendli",
      "accuracy": 0.7739394545454545,
      "accuracy_std_dev": 0.019165412105066386,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.06798592909090909,
      "avg_duration_per_task_ms": 45561.432727272724,
      "avg_output_tokens_per_task": 4731.818181818182,
      "last_run_timestamp": "2026-10-31T12:41:41.331Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Reka",
      "accuracy": 0.7733335,
      "accuracy_std_dev": 0.00666650000000002,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.11407636,
      "avg_duration_per_task_ms": 360681.75,
      "avg_output_tokens_per_task": 19997.86,
      "last_run_timestamp": "2026-10-01T22:40:52.402Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Io Net",
      "accuracy": 0.7733335,
      "accuracy_std_dev": 0.013333499999999998,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.10867763,
      "avg_duration_per_task_ms": 521628.03,
      "avg_output_tokens_per_task": 12092.47,
      "last_run_timestamp": "2026-09-19T08:11:59.798Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Tenstorrent",
      "accuracy": 0.7733335,
      "accuracy_std_dev": 0.03711848895429339,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0407917975,
      "avg_duration_per_task_ms": 116084.2975,
      "avg_output_tokens_per_task": 4762.3475,
      "last_run_timestamp": "2026-11-08T04:06:44.769Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "DekaLLM",
      "accuracy": 0.7733334,
      "accuracy_std_dev": 0.02108173511929225,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.05645720600000001,
      "avg_duration_per_task_ms": 156313.234,
      "avg_output_tokens_per_task": 16426.692,
      "last_run_timestamp": "2026-10-31T11:34:45.444Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "DeepInfra",
      "accuracy": 0.773333375,
      "accuracy_std_dev": 0.02236071704428047,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.04385152750000001,
      "avg_duration_per_task_ms": 107910.09,
      "avg_output_tokens_per_task": 4940.6725,
      "last_run_timestamp": "2026-10-31T12:00:41.768Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-sonnet-5.5-20260928",
      "model_slug": "anthropic/claude-sonnet-5.5",
      "model_name": "Anthropic: Claude Sonnet 5.5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.7733333333333333,
      "accuracy_std_dev": 0.02309381831909905,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.16708652333333332,
      "avg_duration_per_task_ms": 79748.64333333333,
      "avg_output_tokens_per_task": 6314.02,
      "last_run_timestamp": "2026-10-31T11:48:36.920Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "auto-routing",
      "accuracy": 0.7733333,
      "accuracy_std_dev": 0.026997983465622016,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.10150856999999999,
      "avg_duration_per_task_ms": 290647.242,
      "avg_output_tokens_per_task": 21805.618,
      "last_run_timestamp": "2026-10-31T11:29:11.154Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "ModelRun",
      "accuracy": 0.77333325,
      "accuracy_std_dev": 0.029814193116492343,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.06547900499999999,
      "avg_duration_per_task_ms": 79239.345,
      "avg_output_tokens_per_task": 12411.3875,
      "last_run_timestamp": "2026-10-31T11:09:23.381Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Io Net",
      "accuracy": 0.7733331666666667,
      "accuracy_std_dev": 0.021081851068448003,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.05726335,
      "avg_duration_per_task_ms": 158508.54333333333,
      "avg_output_tokens_per_task": 4588.7266666666665,
      "last_run_timestamp": "2026-10-31T13:11:06.716Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Reka",
      "accuracy": 0.773333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.012345539999999999,
      "avg_duration_per_task_ms": 657478.84,
      "avg_output_tokens_per_task": 14196.68,
      "last_run_timestamp": "2026-10-03T00:00:06.435Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Sail Research",
      "accuracy": 0.773333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.00804126,
      "avg_duration_per_task_ms": 386984.54,
      "avg_output_tokens_per_task": 13398.72,
      "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": "moonshotai/kimi-k3.1-pro-20261023",
      "model_slug": "moonshotai/kimi-k3.1-pro",
      "model_name": "MoonshotAI: Kimi K3.1 Pro",
      "provider_name": "River",
      "accuracy": 0.773333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.19214888,
      "avg_duration_per_task_ms": 135761,
      "avg_output_tokens_per_task": 9989.38,
      "last_run_timestamp": "2026-11-13T06:41:18.869Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3.1-pro-20261023",
      "model_slug": "moonshotai/kimi-k3.1-pro",
      "model_name": "MoonshotAI: Kimi K3.1 Pro",
      "provider_name": "Relace",
      "accuracy": 0.773333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.14084568,
      "avg_duration_per_task_ms": 130912.8,
      "avg_output_tokens_per_task": 10689.44,
      "last_run_timestamp": "2026-11-05T22:50:26.968Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3.1-pro-20261023",
      "model_slug": "moonshotai/kimi-k3.1-pro",
      "model_name": "MoonshotAI: Kimi K3.1 Pro",
      "provider_name": "Nebius",
      "accuracy": 0.773333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.2468412,
      "avg_duration_per_task_ms": 828355.82,
      "avg_output_tokens_per_task": 10570.46,
      "last_run_timestamp": "2026-11-04T22:42:55.359Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "GMICloud",
      "accuracy": 0.773333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.26629643999999997,
      "avg_duration_per_task_ms": 438741.58,
      "avg_output_tokens_per_task": 27876.64,
      "last_run_timestamp": "2026-08-29T16:58:30.455Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "Makora",
      "accuracy": 0.773333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.11485508,
      "avg_duration_per_task_ms": 381375.48,
      "avg_output_tokens_per_task": 14199.1,
      "last_run_timestamp": "2026-10-13T03:50:19.636Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "StreamLake",
      "accuracy": 0.773333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.057549260000000005,
      "avg_duration_per_task_ms": 140710.34,
      "avg_output_tokens_per_task": 4743.94,
      "last_run_timestamp": "2026-09-11T04:09:34.110Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-opus-5-20260723",
      "model_slug": "anthropic/claude-opus-5",
      "model_name": "Anthropic: Claude Opus 5",
      "provider_name": "Google",
      "accuracy": 0.7726666333333335,
      "accuracy_std_dev": 0.022400415672457697,
      "run_count": 30,
      "total_questions": 1500,
      "avg_cost_per_task_usd": 0.49225354466666665,
      "avg_duration_per_task_ms": 105640.478,
      "avg_output_tokens_per_task": 7051.117333333334,
      "last_run_timestamp": "2026-10-31T12:26:08.030Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "auto-routing",
      "accuracy": 0.7725926666666666,
      "accuracy_std_dev": 0.033176625054396364,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.05704026,
      "avg_duration_per_task_ms": 115075.10666666667,
      "avg_output_tokens_per_task": 6558.98,
      "last_run_timestamp": "2026-10-31T12:50:03.839Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Alibaba",
      "accuracy": 0.7723335,
      "accuracy_std_dev": 0.025212516733757476,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.18855182,
      "avg_duration_per_task_ms": 469904.514,
      "avg_output_tokens_per_task": 27186.904,
      "last_run_timestamp": "2026-10-31T19:55:50.954Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.7-flash-20260813",
      "model_slug": "google/gemini-3.7-flash",
      "model_name": "Google: Gemini 3.7 Flash",
      "provider_name": "Google",
      "accuracy": 0.7721088979591837,
      "accuracy_std_dev": 0.008055499999999993,
      "run_count": 2,
      "total_questions": 98,
      "avg_cost_per_task_usd": 0.18560600000000002,
      "avg_duration_per_task_ms": 213412.75510204083,
      "avg_output_tokens_per_task": 17105.88775510204,
      "last_run_timestamp": "2026-10-15T21:26:28.195Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Relace",
      "accuracy": 0.7720001999999999,
      "accuracy_std_dev": 0.024367767418456705,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.14564374000000002,
      "avg_duration_per_task_ms": 314489.068,
      "avg_output_tokens_per_task": 20798.24,
      "last_run_timestamp": "2026-10-31T13:16:57.928Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-fable-5.1-20260831",
      "model_slug": "anthropic/claude-fable-5.1",
      "model_name": "Anthropic: Claude Fable 5.1",
      "provider_name": "Azure",
      "accuracy": 0.772,
      "accuracy_std_dev": 0.02679938639596063,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.808827636,
      "avg_duration_per_task_ms": 183544.5,
      "avg_output_tokens_per_task": 6940.7,
      "last_run_timestamp": "2026-10-01T08:03:00.737Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.7-opus-20260416",
      "model_slug": "anthropic/claude-opus-4.7",
      "model_name": "Anthropic: Claude Opus 4.7",
      "provider_name": "Azure",
      "accuracy": 0.772,
      "accuracy_std_dev": 0.015719782301291577,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.438245934,
      "avg_duration_per_task_ms": 92813.308,
      "avg_output_tokens_per_task": 5707.604,
      "last_run_timestamp": "2026-10-31T12:52:03.935Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Parasail",
      "accuracy": 0.7718182727272727,
      "accuracy_std_dev": 0.031151820244762325,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.06332707454545455,
      "avg_duration_per_task_ms": 243359.18363636362,
      "avg_output_tokens_per_task": 11661.423636363636,
      "last_run_timestamp": "2026-10-31T12:57:09.934Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Baidu",
      "accuracy": 0.771666625,
      "accuracy_std_dev": 0.021537947777919213,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0856126475,
      "avg_duration_per_task_ms": 364320.3175,
      "avg_output_tokens_per_task": 20467.8275,
      "last_run_timestamp": "2026-11-10T19:01:03.247Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "StreamLake",
      "accuracy": 0.7716665,
      "accuracy_std_dev": 0.0016665000000000152,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.050993250000000004,
      "avg_duration_per_task_ms": 677393.46,
      "avg_output_tokens_per_task": 30421.97,
      "last_run_timestamp": "2026-10-31T13:28:05.897Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7715774062499999,
      "accuracy_std_dev": 0.022644299042712767,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.14963372544642858,
      "avg_duration_per_task_ms": 341867.6473214286,
      "avg_output_tokens_per_task": 20209.140625,
      "last_run_timestamp": "2026-10-31T14:04:49.550Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "DeepInfra",
      "accuracy": 0.7714602976588629,
      "accuracy_std_dev": 0.028416062527791094,
      "run_count": 6,
      "total_questions": 299,
      "avg_cost_per_task_usd": 0.08344332775919731,
      "avg_duration_per_task_ms": 249520.2474916388,
      "avg_output_tokens_per_task": 12314.311036789297,
      "last_run_timestamp": "2026-10-31T19:22:52.905Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "Venice",
      "accuracy": 0.771428857142857,
      "accuracy_std_dev": 0.019095107469331168,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.3362465285714286,
      "avg_duration_per_task_ms": 611059.8085714285,
      "avg_output_tokens_per_task": 28432.711428571427,
      "last_run_timestamp": "2026-10-31T19:49:31.051Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Novita",
      "accuracy": 0.7713334,
      "accuracy_std_dev": 0.020011225740568715,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.051776854,
      "avg_duration_per_task_ms": 108112.036,
      "avg_output_tokens_per_task": 4665.94,
      "last_run_timestamp": "2026-10-31T13:28:16.093Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-fable-5.1-20260831",
      "model_slug": "anthropic/claude-fable-5.1",
      "model_name": "Anthropic: Claude Fable 5.1",
      "provider_name": "Google",
      "accuracy": 0.7711111666666667,
      "accuracy_std_dev": 0.014229407910107222,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.7819622033333333,
      "avg_duration_per_task_ms": 110376.29,
      "avg_output_tokens_per_task": 6891.35,
      "last_run_timestamp": "2026-11-11T21:09:52.293Z"
    },
    {
      "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": 0.771,
      "accuracy_std_dev": 0.018651732123317666,
      "run_count": 20,
      "total_questions": 1000,
      "avg_cost_per_task_usd": 0.47884973,
      "avg_duration_per_task_ms": 100219.903,
      "avg_output_tokens_per_task": 7083.76,
      "last_run_timestamp": "2026-10-31T12:54:18.844Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "DeepInfra",
      "accuracy": 0.7709447032085562,
      "accuracy_std_dev": 0.032091449469013114,
      "run_count": 15,
      "total_questions": 748,
      "avg_cost_per_task_usd": 0.03325723529411764,
      "avg_duration_per_task_ms": 455906.7553475936,
      "avg_output_tokens_per_task": 12250.7807486631,
      "last_run_timestamp": "2026-10-31T19:33:25.394Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Venice",
      "accuracy": 0.770833375,
      "accuracy_std_dev": 0.02283704805758343,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.2110248925,
      "avg_duration_per_task_ms": 381381.2675,
      "avg_output_tokens_per_task": 17629.1275,
      "last_run_timestamp": "2026-10-31T12:14:30.525Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta/muse-glimmer-30b-20260810",
      "model_slug": "meta/muse-glimmer-30b",
      "model_name": "Meta: Muse Glimmer 30B",
      "provider_name": "Parasail",
      "accuracy": 0.7706935637583894,
      "accuracy_std_dev": 0.013644816256244217,
      "run_count": 3,
      "total_questions": 149,
      "avg_cost_per_task_usd": 0.09587820805369127,
      "avg_duration_per_task_ms": 288021.6644295302,
      "avg_output_tokens_per_task": 13114.979865771813,
      "last_run_timestamp": "2026-08-29T08:48:12.957Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Phala",
      "accuracy": 0.7706668,
      "accuracy_std_dev": 0.027194791985966728,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.13800210599999999,
      "avg_duration_per_task_ms": 324537.226,
      "avg_output_tokens_per_task": 20086.83,
      "last_run_timestamp": "2026-10-31T13:20:51.002Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.8-opus-20260528",
      "model_slug": "anthropic/claude-opus-4.8",
      "model_name": "Anthropic: Claude Opus 4.8",
      "provider_name": "Anthropic",
      "accuracy": 0.7706667,
      "accuracy_std_dev": 0.025854399153915776,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.5051475320000001,
      "avg_duration_per_task_ms": 119868.748,
      "avg_output_tokens_per_task": 7988.984,
      "last_run_timestamp": "2026-10-31T11:50:39.610Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "NextBit",
      "accuracy": 0.7706666,
      "accuracy_std_dev": 0.0204828947622156,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.10916058,
      "avg_duration_per_task_ms": 258128.42,
      "avg_output_tokens_per_task": 11359.708,
      "last_run_timestamp": "2026-10-03T16:07:05.245Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "CoreWeave",
      "accuracy": 0.7704079387755102,
      "accuracy_std_dev": 0.002986000000000044,
      "run_count": 2,
      "total_questions": 98,
      "avg_cost_per_task_usd": 0.1301587857142857,
      "avg_duration_per_task_ms": 225452.5,
      "avg_output_tokens_per_task": 42669.47959183674,
      "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": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "auto-routing",
      "accuracy": 0.7700001,
      "accuracy_std_dev": 0.023333328572880462,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.145186808,
      "avg_duration_per_task_ms": 851216.382,
      "avg_output_tokens_per_task": 32358.94,
      "last_run_timestamp": "2026-10-31T20:06:11.663Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Wafer",
      "accuracy": 0.7700000714285714,
      "accuracy_std_dev": 0.014638549891415803,
      "run_count": 14,
      "total_questions": 700,
      "avg_cost_per_task_usd": 0.14041841571428573,
      "avg_duration_per_task_ms": 235377.64428571428,
      "avg_output_tokens_per_task": 25220.13142857143,
      "last_run_timestamp": "2026-10-31T13:20:49.855Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-opus-5-20260723",
      "model_slug": "anthropic/claude-opus-5",
      "model_name": "Anthropic: Claude Opus 5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.77,
      "accuracy_std_dev": 0.01639789959720451,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.464249572,
      "avg_duration_per_task_ms": 116894.73,
      "avg_output_tokens_per_task": 7042.312,
      "last_run_timestamp": "2026-10-31T11:40:34.382Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.1-pro-preview-20260219",
      "model_slug": "google/gemini-3.1-pro-preview",
      "model_name": "Google: Gemini 3.1 Pro Preview",
      "provider_name": "auto-routing",
      "accuracy": 0.77,
      "accuracy_std_dev": 0.0033330000000000304,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.35129818,
      "avg_duration_per_task_ms": 177012.35,
      "avg_output_tokens_per_task": 12679.64,
      "last_run_timestamp": "2026-10-01T19:06:01.702Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3.1-pro-20261023",
      "model_slug": "moonshotai/kimi-k3.1-pro",
      "model_name": "MoonshotAI: Kimi K3.1 Pro",
      "provider_name": "Moonshot AI",
      "accuracy": 0.77,
      "accuracy_std_dev": 0.016666999999999987,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.28240516,
      "avg_duration_per_task_ms": 307072.21,
      "avg_output_tokens_per_task": 10658.56,
      "last_run_timestamp": "2026-11-04T18:56:58.927Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "DeepInfra",
      "accuracy": 0.769999875,
      "accuracy_std_dev": 0.03382961802340332,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0406579,
      "avg_duration_per_task_ms": 682767.975,
      "avg_output_tokens_per_task": 12586.08,
      "last_run_timestamp": "2026-10-31T12:34:48.818Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "DeepInfra",
      "accuracy": 0.7699214781144781,
      "accuracy_std_dev": 0.02849514533750542,
      "run_count": 6,
      "total_questions": 297,
      "avg_cost_per_task_usd": 0.010507353535353536,
      "avg_duration_per_task_ms": 977066.3771043771,
      "avg_output_tokens_per_task": 18208.95286195286,
      "last_run_timestamp": "2026-10-31T12:24:58.436Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Parasail",
      "accuracy": 0.7696666,
      "accuracy_std_dev": 0.019913617899317042,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.07924518799999998,
      "avg_duration_per_task_ms": 90727.314,
      "avg_output_tokens_per_task": 4394.952,
      "last_run_timestamp": "2026-11-05T00:52:28.447Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Reka",
      "accuracy": 0.7695237142857143,
      "accuracy_std_dev": 0.02386662296246901,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.11461242285714283,
      "avg_duration_per_task_ms": 472929.0742857143,
      "avg_output_tokens_per_task": 11699.722857142857,
      "last_run_timestamp": "2026-10-23T07:26:58.509Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Morph",
      "accuracy": 0.7693333,
      "accuracy_std_dev": 0.039350196385914006,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.087123052,
      "avg_duration_per_task_ms": 282737.078,
      "avg_output_tokens_per_task": 16723.342,
      "last_run_timestamp": "2026-10-31T19:58:49.610Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Modal",
      "accuracy": 0.7693333,
      "accuracy_std_dev": 0.027680754646685506,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.080204858,
      "avg_duration_per_task_ms": 55518.588,
      "avg_output_tokens_per_task": 4833.54,
      "last_run_timestamp": "2026-10-31T12:18:44.291Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Together",
      "accuracy": 0.769166625,
      "accuracy_std_dev": 0.028272094359710526,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.014342402500000002,
      "avg_duration_per_task_ms": 227243.235,
      "avg_output_tokens_per_task": 15618.9,
      "last_run_timestamp": "2026-10-17T11:00:48.267Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "AkashML",
      "accuracy": 0.7690024251781471,
      "accuracy_std_dev": 0.02344176596675554,
      "run_count": 17,
      "total_questions": 842,
      "avg_cost_per_task_usd": 0.05268543230403801,
      "avg_duration_per_task_ms": 269011.41448931117,
      "avg_output_tokens_per_task": 11083.264845605701,
      "last_run_timestamp": "2026-10-31T12:29:57.959Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "PrimaLabs",
      "accuracy": 0.7689656482758621,
      "accuracy_std_dev": 0.012187977929090621,
      "run_count": 3,
      "total_questions": 145,
      "avg_cost_per_task_usd": 0.5075254827586206,
      "avg_duration_per_task_ms": 227003.33793103448,
      "avg_output_tokens_per_task": 39823.972413793104,
      "last_run_timestamp": "2026-11-06T16:54:34.176Z"
    },
    {
      "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": 0.7688888333333334,
      "accuracy_std_dev": 0.013698652724710655,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.15528754666666666,
      "avg_duration_per_task_ms": 428069.5,
      "avg_output_tokens_per_task": 20104.073333333334,
      "last_run_timestamp": "2026-10-31T20:34:04.127Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "auto-routing",
      "accuracy": 0.7685714285714286,
      "accuracy_std_dev": 0.020995671729853154,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.062093528571428566,
      "avg_duration_per_task_ms": 230920.66857142857,
      "avg_output_tokens_per_task": 13057.314285714287,
      "last_run_timestamp": "2026-10-31T20:15:14.788Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Chutes",
      "accuracy": 0.7684848181818182,
      "accuracy_std_dev": 0.034709847551636586,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.05697716181818182,
      "avg_duration_per_task_ms": 301468.96545454545,
      "avg_output_tokens_per_task": 12590.405454545455,
      "last_run_timestamp": "2026-10-31T11:18:52.425Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3.1-pro-20261023",
      "model_slug": "moonshotai/kimi-k3.1-pro",
      "model_name": "MoonshotAI: Kimi K3.1 Pro",
      "provider_name": "Phala",
      "accuracy": 0.7683335,
      "accuracy_std_dev": 0.018333500000000003,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.18849928,
      "avg_duration_per_task_ms": 346868.4,
      "avg_output_tokens_per_task": 10339.12,
      "last_run_timestamp": "2026-11-05T18:26:07.849Z"
    },
    {
      "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": 0.7683333999999999,
      "accuracy_std_dev": 0.023154082915114577,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.127404986,
      "avg_duration_per_task_ms": 482373.346,
      "avg_output_tokens_per_task": 21637.406,
      "last_run_timestamp": "2026-11-05T02:16:26.851Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Venice",
      "accuracy": 0.768333375,
      "accuracy_std_dev": 0.027638527357375157,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.12531909,
      "avg_duration_per_task_ms": 166385.7025,
      "avg_output_tokens_per_task": 7956.0025,
      "last_run_timestamp": "2026-10-31T13:24:47.637Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Phala",
      "accuracy": 0.7681482222222222,
      "accuracy_std_dev": 0.017435059205697385,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.052077991111111104,
      "avg_duration_per_task_ms": 111760.7,
      "avg_output_tokens_per_task": 4431.0888888888885,
      "last_run_timestamp": "2026-10-31T13:22:10.288Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.7680001999999999,
      "accuracy_std_dev": 0.021457811839980336,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.49101839,
      "avg_duration_per_task_ms": 123065.604,
      "avg_output_tokens_per_task": 8037.726,
      "last_run_timestamp": "2026-10-31T12:30:31.663Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "auto-routing",
      "accuracy": 0.7677776666666667,
      "accuracy_std_dev": 0.043104839177934014,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.06842919666666666,
      "avg_duration_per_task_ms": 974829.46,
      "avg_output_tokens_per_task": 44158.48333333333,
      "last_run_timestamp": "2026-10-31T13:18:29.756Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Decart",
      "accuracy": 0.7675,
      "accuracy_std_dev": 0.019561171859068173,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0578507225,
      "avg_duration_per_task_ms": 50974.42,
      "avg_output_tokens_per_task": 4677.1725,
      "last_run_timestamp": "2026-10-24T12:36:03.010Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Alibaba",
      "accuracy": 0.7674509411764706,
      "accuracy_std_dev": 0.021678839985437112,
      "run_count": 17,
      "total_questions": 850,
      "avg_cost_per_task_usd": 0.12603030588235295,
      "avg_duration_per_task_ms": 92367.94588235294,
      "avg_output_tokens_per_task": 4470.871764705882,
      "last_run_timestamp": "2026-10-31T12:50:48.724Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.8-opus-20260528",
      "model_slug": "anthropic/claude-opus-4.8",
      "model_name": "Anthropic: Claude Opus 4.8",
      "provider_name": "Google",
      "accuracy": 0.7673332999999999,
      "accuracy_std_dev": 0.02871832890118551,
      "run_count": 30,
      "total_questions": 1500,
      "avg_cost_per_task_usd": 0.5323030653333332,
      "avg_duration_per_task_ms": 111144.59533333333,
      "avg_output_tokens_per_task": 8030.952666666667,
      "last_run_timestamp": "2026-10-31T12:31:31.431Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.7-opus-20260416",
      "model_slug": "anthropic/claude-opus-4.7",
      "model_name": "Anthropic: Claude Opus 4.7",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.7672726363636364,
      "accuracy_std_dev": 0.030278820027065213,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.43524433636363635,
      "avg_duration_per_task_ms": 90898.93818181819,
      "avg_output_tokens_per_task": 5582.549090909091,
      "last_run_timestamp": "2026-10-31T12:33:27.768Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.7-opus-20260416",
      "model_slug": "anthropic/claude-opus-4.7",
      "model_name": "Anthropic: Claude Opus 4.7",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7669999000000001,
      "accuracy_std_dev": 0.030657633815903015,
      "run_count": 20,
      "total_questions": 1000,
      "avg_cost_per_task_usd": 0.449202301,
      "avg_duration_per_task_ms": 78155.357,
      "avg_output_tokens_per_task": 5628.041,
      "last_run_timestamp": "2026-10-31T12:41:58.029Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "amazon/nova-micro-v1",
      "model_slug": "amazon/nova-micro-v1",
      "model_name": "Amazon: Nova Micro 1.0",
      "provider_name": "auto-routing",
      "accuracy": 0.7666669999999999,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 1.2750355,
      "avg_duration_per_task_ms": 98712.99,
      "avg_output_tokens_per_task": 5708.24,
      "last_run_timestamp": "2026-10-01T19:04:24.509Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "Novita",
      "accuracy": 0.7666669999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.0337152,
      "avg_duration_per_task_ms": 192245.9,
      "avg_output_tokens_per_task": 17702.16,
      "last_run_timestamp": "2026-09-01T12:41:31.535Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "ModelRun",
      "accuracy": 0.7666669999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.01300872,
      "avg_duration_per_task_ms": 90405.02,
      "avg_output_tokens_per_task": 13808.24,
      "last_run_timestamp": "2026-10-29T03:35:28.909Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Nebius",
      "accuracy": 0.7666669999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.0734938,
      "avg_duration_per_task_ms": 215054.84,
      "avg_output_tokens_per_task": 13096.66,
      "last_run_timestamp": "2026-11-04T18:29:36.971Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Featherless",
      "accuracy": 0.7666669999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.0650557,
      "avg_duration_per_task_ms": 127920.32,
      "avg_output_tokens_per_task": 4266.18,
      "last_run_timestamp": "2026-11-14T00:27:40.926Z"
    },
    {
      "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": "anthropic/claude-4.8-opus-20260528",
      "model_slug": "anthropic/claude-opus-4.8",
      "model_name": "Anthropic: Claude Opus 4.8",
      "provider_name": "Azure",
      "accuracy": 0.7666667,
      "accuracy_std_dev": 0.02658317763943958,
      "run_count": 20,
      "total_questions": 1000,
      "avg_cost_per_task_usd": 0.5167896780000001,
      "avg_duration_per_task_ms": 124380.271,
      "avg_output_tokens_per_task": 7924.367,
      "last_run_timestamp": "2026-10-31T13:09:20.558Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Relace",
      "accuracy": 0.7666666666666667,
      "accuracy_std_dev": 0.010886757164351354,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.031067933333333332,
      "avg_duration_per_task_ms": 60278.37,
      "avg_output_tokens_per_task": 4752.3966666666665,
      "last_run_timestamp": "2026-10-31T12:25:28.648Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Inceptron",
      "accuracy": 0.7666666666666667,
      "accuracy_std_dev": 0.023517736937610882,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.04758358666666667,
      "avg_duration_per_task_ms": 110427.16444444444,
      "avg_output_tokens_per_task": 4684.946666666667,
      "last_run_timestamp": "2026-10-31T12:08:49.485Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.8-flash-20260902",
      "model_slug": "google/gemini-3.8-flash",
      "model_name": "Google: Gemini 3.8 Flash",
      "provider_name": "Google",
      "accuracy": 0.7666665,
      "accuracy_std_dev": 0.013333499999999998,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.27949224,
      "avg_duration_per_task_ms": 363191.33,
      "avg_output_tokens_per_task": 35460.86,
      "last_run_timestamp": "2026-10-15T21:31:19.497Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Makora",
      "accuracy": 0.766068730941704,
      "accuracy_std_dev": 0.02506483163840778,
      "run_count": 9,
      "total_questions": 446,
      "avg_cost_per_task_usd": 0.06749501121076233,
      "avg_duration_per_task_ms": 62014.139013452914,
      "avg_output_tokens_per_task": 4517.719730941704,
      "last_run_timestamp": "2026-10-31T13:21:43.928Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "DekaLLM",
      "accuracy": 0.7660001999999999,
      "accuracy_std_dev": 0.0260256591340162,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.12242998000000002,
      "avg_duration_per_task_ms": 384915.524,
      "avg_output_tokens_per_task": 20927.716,
      "last_run_timestamp": "2026-10-31T13:14:15.095Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Mistral",
      "accuracy": 0.7659459729729727,
      "accuracy_std_dev": 0.021296341804228094,
      "run_count": 37,
      "total_questions": 1850,
      "avg_cost_per_task_usd": 0.06441325405405407,
      "avg_duration_per_task_ms": 66619.0454054054,
      "avg_output_tokens_per_task": 4623.922162162162,
      "last_run_timestamp": "2026-11-11T03:15:32.401Z"
    },
    {
      "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": 0.76583325,
      "accuracy_std_dev": 0.03273434952672042,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0269725475,
      "avg_duration_per_task_ms": 160221.945,
      "avg_output_tokens_per_task": 9025.64,
      "last_run_timestamp": "2026-10-24T19:16:17.608Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Venice",
      "accuracy": 0.7657144285714286,
      "accuracy_std_dev": 0.02990523505942035,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.14187513999999998,
      "avg_duration_per_task_ms": 291309.5857142857,
      "avg_output_tokens_per_task": 12316.085714285715,
      "last_run_timestamp": "2026-10-31T19:58:31.913Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7655555,
      "accuracy_std_dev": 0.04258615812972569,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.10541816000000001,
      "avg_duration_per_task_ms": 274154.0233333333,
      "avg_output_tokens_per_task": 11441.496666666666,
      "last_run_timestamp": "2026-10-31T21:36:17.452Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "Tencent",
      "accuracy": 0.7653333999999999,
      "accuracy_std_dev": 0.023999911113168737,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.088749776,
      "avg_duration_per_task_ms": 386636.328,
      "avg_output_tokens_per_task": 14600.352,
      "last_run_timestamp": "2026-10-31T12:27:43.549Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.7652380952380953,
      "accuracy_std_dev": 0.027344026724583557,
      "run_count": 21,
      "total_questions": 1050,
      "avg_cost_per_task_usd": 0.044898270476190466,
      "avg_duration_per_task_ms": 87109.40476190476,
      "avg_output_tokens_per_task": 4336.9914285714285,
      "last_run_timestamp": "2026-11-06T23:32:54.061Z"
    },
    {
      "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": 0.7651852222222222,
      "accuracy_std_dev": 0.027893614278292685,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.09080310222222222,
      "avg_duration_per_task_ms": 365591.88666666666,
      "avg_output_tokens_per_task": 18878.693333333333,
      "last_run_timestamp": "2026-11-10T08:00:49.752Z"
    },
    {
      "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": 0.7651852222222222,
      "accuracy_std_dev": 0.023471135705030368,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.05947763333333333,
      "avg_duration_per_task_ms": 1101862.8644444444,
      "avg_output_tokens_per_task": 56201.85555555556,
      "last_run_timestamp": "2026-11-13T02:43:05.270Z"
    },
    {
      "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": 0.7648888,
      "accuracy_std_dev": 0.023153787094699365,
      "run_count": 15,
      "total_questions": 750,
      "avg_cost_per_task_usd": 0.03474511733333333,
      "avg_duration_per_task_ms": 66218.656,
      "avg_output_tokens_per_task": 4837.296,
      "last_run_timestamp": "2026-11-04T21:48:00.430Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Baidu",
      "accuracy": 0.7647619999999999,
      "accuracy_std_dev": 0.026116980863579387,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.04505181428571428,
      "avg_duration_per_task_ms": 210436.56857142856,
      "avg_output_tokens_per_task": 13672.442857142858,
      "last_run_timestamp": "2026-10-31T20:06:41.224Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Cloudflare",
      "accuracy": 0.7646666,
      "accuracy_std_dev": 0.016343743623784605,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.035043700000000004,
      "avg_duration_per_task_ms": 223770.176,
      "avg_output_tokens_per_task": 15530.04,
      "last_run_timestamp": "2026-10-31T12:46:55.135Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.7644444333333335,
      "accuracy_std_dev": 0.01919359203950342,
      "run_count": 30,
      "total_questions": 1500,
      "avg_cost_per_task_usd": 0.5300209846666668,
      "avg_duration_per_task_ms": 117525.25933333334,
      "avg_output_tokens_per_task": 8055.547333333333,
      "last_run_timestamp": "2026-10-31T12:26:06.842Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-flash-20260826",
      "model_slug": "qwen/qwen3.8-flash",
      "model_name": "Qwen: Qwen3.8 Flash",
      "provider_name": "Alibaba",
      "accuracy": 0.7644443333333334,
      "accuracy_std_dev": 0.04565023765059233,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.01549936,
      "avg_duration_per_task_ms": 297786.8,
      "avg_output_tokens_per_task": 16565.086666666666,
      "last_run_timestamp": "2026-10-28T04:10:41.154Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "SambaNova",
      "accuracy": 0.76439777486911,
      "accuracy_std_dev": 0.015263579494993957,
      "run_count": 4,
      "total_questions": 191,
      "avg_cost_per_task_usd": 0.06284435078534031,
      "avg_duration_per_task_ms": 156780.0209424084,
      "avg_output_tokens_per_task": 9570.141361256545,
      "last_run_timestamp": "2026-10-31T19:43:06.929Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Friendli",
      "accuracy": 0.764166625,
      "accuracy_std_dev": 0.019418799956083154,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0292764325,
      "avg_duration_per_task_ms": 122942.345,
      "avg_output_tokens_per_task": 12042.5375,
      "last_run_timestamp": "2026-10-31T19:08:17.847Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7641665000000001,
      "accuracy_std_dev": 0.021586383312171588,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0185151775,
      "avg_duration_per_task_ms": 270096.6575,
      "avg_output_tokens_per_task": 14323.5575,
      "last_run_timestamp": "2026-10-31T12:30:14.640Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.7640001,
      "accuracy_std_dev": 0.02235083095748345,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.431202254,
      "avg_duration_per_task_ms": 89253.752,
      "avg_output_tokens_per_task": 5552.148,
      "last_run_timestamp": "2026-10-31T11:48:47.453Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.7-opus-20260416",
      "model_slug": "anthropic/claude-opus-4.7",
      "model_name": "Anthropic: Claude Opus 4.7",
      "provider_name": "Google",
      "accuracy": 0.7640000666666666,
      "accuracy_std_dev": 0.021609382051836254,
      "run_count": 30,
      "total_questions": 1500,
      "avg_cost_per_task_usd": 0.5330969666666666,
      "avg_duration_per_task_ms": 69978.30333333333,
      "avg_output_tokens_per_task": 5705.826,
      "last_run_timestamp": "2026-10-31T12:33:51.242Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-opus-5-20260723",
      "model_slug": "anthropic/claude-opus-5",
      "model_name": "Anthropic: Claude Opus 5",
      "provider_name": "Anthropic",
      "accuracy": 0.764,
      "accuracy_std_dev": 0.02091261182635972,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.476702984,
      "avg_duration_per_task_ms": 107995.534,
      "avg_output_tokens_per_task": 7135.662,
      "last_run_timestamp": "2026-10-31T11:27:21.008Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "Novita",
      "accuracy": 0.764,
      "accuracy_std_dev": 0.020483142122242875,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.080228768,
      "avg_duration_per_task_ms": 494377.212,
      "avg_output_tokens_per_task": 14669.972,
      "last_run_timestamp": "2026-10-31T12:27:34.738Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7637037777777778,
      "accuracy_std_dev": 0.021798347556015337,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.06269429111111112,
      "avg_duration_per_task_ms": 113855.65555555555,
      "avg_output_tokens_per_task": 4583.311111111111,
      "last_run_timestamp": "2026-10-31T12:40:08.732Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.6-flash-20260721",
      "model_slug": "google/gemini-3.6-flash",
      "model_name": "Google: Gemini 3.6 Flash",
      "provider_name": "Google",
      "accuracy": 0.7633335,
      "accuracy_std_dev": 0.0033334999999999892,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.178977705,
      "avg_duration_per_task_ms": 157076.125,
      "avg_output_tokens_per_task": 23034.1,
      "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": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Xiaomi",
      "accuracy": 0.7633332500000001,
      "accuracy_std_dev": 0.013743584364622665,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.01189864,
      "avg_duration_per_task_ms": 642869.1,
      "avg_output_tokens_per_task": 23140.98,
      "last_run_timestamp": "2026-10-24T13:28:08.470Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "CoreWeave",
      "accuracy": 0.763333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.025058120000000003,
      "avg_duration_per_task_ms": 106870.58,
      "avg_output_tokens_per_task": 12582.44,
      "last_run_timestamp": "2026-11-04T18:35:21.465Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "BaseTen",
      "accuracy": 0.763333,
      "accuracy_std_dev": 0.010000000000000009,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.020903190000000002,
      "avg_duration_per_task_ms": 62428.04,
      "avg_output_tokens_per_task": 8093.56,
      "last_run_timestamp": "2026-10-31T12:43:49.743Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Cloudflare",
      "accuracy": 0.7630302727272728,
      "accuracy_std_dev": 0.03170123269496259,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.15556449636363637,
      "avg_duration_per_task_ms": 593095.6363636364,
      "avg_output_tokens_per_task": 12420.629090909091,
      "last_run_timestamp": "2026-10-31T13:02:37.599Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "NextBit",
      "accuracy": 0.7628571428571429,
      "accuracy_std_dev": 0.023328310442945685,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.11986269428571429,
      "avg_duration_per_task_ms": 415507.14,
      "avg_output_tokens_per_task": 18962.734285714287,
      "last_run_timestamp": "2026-10-03T08:33:31.578Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "GMICloud",
      "accuracy": 0.7627778333333333,
      "accuracy_std_dev": 0.03313374040368652,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.07174823666666666,
      "avg_duration_per_task_ms": 893696.3066666666,
      "avg_output_tokens_per_task": 51834.78666666667,
      "last_run_timestamp": "2026-10-27T03:58:39.326Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Z.AI",
      "accuracy": 0.7626666,
      "accuracy_std_dev": 0.02954472165785287,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.009234846,
      "avg_duration_per_task_ms": 195074.734,
      "avg_output_tokens_per_task": 4475.264,
      "last_run_timestamp": "2026-10-31T12:25:34.861Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7626665,
      "accuracy_std_dev": 0.026025525932245826,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.022020616000000003,
      "avg_duration_per_task_ms": 287031.818,
      "avg_output_tokens_per_task": 15706.274,
      "last_run_timestamp": "2026-10-31T13:19:25.796Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.7625418160535117,
      "accuracy_std_dev": 0.02147753020972668,
      "run_count": 12,
      "total_questions": 598,
      "avg_cost_per_task_usd": 0.09006360702341139,
      "avg_duration_per_task_ms": 48936.09866220736,
      "avg_output_tokens_per_task": 4348.224080267559,
      "last_run_timestamp": "2026-10-31T12:39:14.326Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Crusoe",
      "accuracy": 0.7625417625418061,
      "accuracy_std_dev": 0.01910679234955175,
      "run_count": 6,
      "total_questions": 299,
      "avg_cost_per_task_usd": 0.03341546153846154,
      "avg_duration_per_task_ms": 144903.0602006689,
      "avg_output_tokens_per_task": 21777.678929765887,
      "last_run_timestamp": "2026-10-31T12:18:15.192Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "BaseTen",
      "accuracy": 0.7623932051282051,
      "accuracy_std_dev": 0.018251804130003154,
      "run_count": 4,
      "total_questions": 195,
      "avg_cost_per_task_usd": 0.24276578461538462,
      "avg_duration_per_task_ms": 125754.82051282052,
      "avg_output_tokens_per_task": 7748.235897435898,
      "last_run_timestamp": "2026-10-31T11:39:17.172Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Io Net",
      "accuracy": 0.7623223984962407,
      "accuracy_std_dev": 0.0190669296226313,
      "run_count": 8,
      "total_questions": 399,
      "avg_cost_per_task_usd": 0.011861127819548874,
      "avg_duration_per_task_ms": 357840.4411027569,
      "avg_output_tokens_per_task": 17112.51127819549,
      "last_run_timestamp": "2026-10-31T11:47:32.862Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "DeepInfra",
      "accuracy": 0.7622223333333333,
      "accuracy_std_dev": 0.030469494967188978,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.029655940000000006,
      "avg_duration_per_task_ms": 256385.88,
      "avg_output_tokens_per_task": 18856.086666666666,
      "last_run_timestamp": "2026-10-08T22:27:05.549Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Cohere",
      "accuracy": 0.762177653295129,
      "accuracy_std_dev": 0.016008036097981727,
      "run_count": 7,
      "total_questions": 349,
      "avg_cost_per_task_usd": 0.027902553008595985,
      "avg_duration_per_task_ms": 104077.65616045846,
      "avg_output_tokens_per_task": 17491.627507163324,
      "last_run_timestamp": "2026-10-31T11:55:25.455Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "DekaLLM",
      "accuracy": 0.7621775042979944,
      "accuracy_std_dev": 0.026194934880387766,
      "run_count": 7,
      "total_questions": 349,
      "avg_cost_per_task_usd": 0.06229761318051576,
      "avg_duration_per_task_ms": 372098.9226361032,
      "avg_output_tokens_per_task": 49911.94555873926,
      "last_run_timestamp": "2026-10-31T19:59:18.193Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Morph",
      "accuracy": 0.7619782998454405,
      "accuracy_std_dev": 0.039147489254698645,
      "run_count": 13,
      "total_questions": 647,
      "avg_cost_per_task_usd": 0.06417079598145287,
      "avg_duration_per_task_ms": 203709.1777434312,
      "avg_output_tokens_per_task": 11216.542503863988,
      "last_run_timestamp": "2026-10-31T14:18:08.529Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "DeepInfra",
      "accuracy": 0.7619046734693877,
      "accuracy_std_dev": 0.00894270029465374,
      "run_count": 5,
      "total_questions": 245,
      "avg_cost_per_task_usd": 0.03367711020408164,
      "avg_duration_per_task_ms": 2596340.9346938776,
      "avg_output_tokens_per_task": 51764.93877551021,
      "last_run_timestamp": "2026-10-09T04:05:56.641Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "DekaLLM",
      "accuracy": 0.761784404040404,
      "accuracy_std_dev": 0.046884570628299496,
      "run_count": 4,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.03727159090909091,
      "avg_duration_per_task_ms": 138498.65151515152,
      "avg_output_tokens_per_task": 4381.035353535353,
      "last_run_timestamp": "2026-10-31T12:40:46.071Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "GMICloud",
      "accuracy": 0.76166675,
      "accuracy_std_dev": 0.025404142635027726,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.023970375,
      "avg_duration_per_task_ms": 166710.515,
      "avg_output_tokens_per_task": 16443.75,
      "last_run_timestamp": "2026-11-10T06:15:38.651Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.5-opus-20251124",
      "model_slug": "anthropic/claude-opus-4.5",
      "model_name": "Anthropic: Claude Opus 4.5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.7612120909090909,
      "accuracy_std_dev": 0.02288640210595304,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.54388782,
      "avg_duration_per_task_ms": 180374.2618181818,
      "avg_output_tokens_per_task": 10717.156363636364,
      "last_run_timestamp": "2026-10-31T12:58:32.935Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta/muse-glimmer-30b-20260810",
      "model_slug": "meta/muse-glimmer-30b",
      "model_name": "Meta: Muse Glimmer 30B",
      "provider_name": "Phala",
      "accuracy": 0.7611111666666667,
      "accuracy_std_dev": 0.013005103465135887,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.033468996666666674,
      "avg_duration_per_task_ms": 272357.00333333336,
      "avg_output_tokens_per_task": 13650.26,
      "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": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "GMICloud",
      "accuracy": 0.7611111666666667,
      "accuracy_std_dev": 0.014098336301571996,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.01045996,
      "avg_duration_per_task_ms": 648808.1933333334,
      "avg_output_tokens_per_task": 22246.43666666667,
      "last_run_timestamp": "2026-11-10T14:26:17.627Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Reka",
      "accuracy": 0.7609786762860727,
      "accuracy_std_dev": 0.02511202768812936,
      "run_count": 16,
      "total_questions": 797,
      "avg_cost_per_task_usd": 0.054708953575909665,
      "avg_duration_per_task_ms": 274645.58720200753,
      "avg_output_tokens_per_task": 12041.819322459221,
      "last_run_timestamp": "2026-10-31T12:07:50.786Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7609525714285714,
      "accuracy_std_dev": 0.07435901694742715,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.2507432571428571,
      "avg_duration_per_task_ms": 493938.8057142857,
      "avg_output_tokens_per_task": 29106.445714285714,
      "last_run_timestamp": "2026-10-31T19:26:18.381Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Novita",
      "accuracy": 0.7609524285714285,
      "accuracy_std_dev": 0.016106319095818472,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.12609460857142857,
      "avg_duration_per_task_ms": 262407.6285714286,
      "avg_output_tokens_per_task": 20587.46,
      "last_run_timestamp": "2026-10-31T19:46:00.618Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.7609522857142857,
      "accuracy_std_dev": 0.01649574672206044,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.08085598000000001,
      "avg_duration_per_task_ms": 105369.85714285714,
      "avg_output_tokens_per_task": 22435.354285714286,
      "last_run_timestamp": "2026-10-01T19:05:47.886Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Novita",
      "accuracy": 0.7607886785714285,
      "accuracy_std_dev": 0.017367969160654122,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.018826165178571424,
      "avg_duration_per_task_ms": 120325.66071428571,
      "avg_output_tokens_per_task": 14433.707589285714,
      "last_run_timestamp": "2026-10-31T12:07:52.801Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "GMICloud",
      "accuracy": 0.7607406666666667,
      "accuracy_std_dev": 0.038289998766721775,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.008870675555555556,
      "avg_duration_per_task_ms": 170336.9,
      "avg_output_tokens_per_task": 15026.084444444445,
      "last_run_timestamp": "2026-11-10T06:27:33.470Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Relace",
      "accuracy": 0.7607406666666667,
      "accuracy_std_dev": 0.034633054297881376,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.006436806666666667,
      "avg_duration_per_task_ms": 125547.04444444444,
      "avg_output_tokens_per_task": 4262.8533333333335,
      "last_run_timestamp": "2026-10-24T14:25:23.734Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.7606667,
      "accuracy_std_dev": 0.027398352749937356,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.535814928,
      "avg_duration_per_task_ms": 173096.566,
      "avg_output_tokens_per_task": 10678.3,
      "last_run_timestamp": "2026-10-31T11:55:38.336Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Sail Research",
      "accuracy": 0.76066665,
      "accuracy_std_dev": 0.019194930466857663,
      "run_count": 20,
      "total_questions": 1000,
      "avg_cost_per_task_usd": 0.011314683,
      "avg_duration_per_task_ms": 308174.577,
      "avg_output_tokens_per_task": 15148.268,
      "last_run_timestamp": "2026-11-06T22:45:20.558Z"
    },
    {
      "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": 0.7606666,
      "accuracy_std_dev": 0.030470293002201318,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.047939388,
      "avg_duration_per_task_ms": 181256.216,
      "avg_output_tokens_per_task": 19517.584,
      "last_run_timestamp": "2026-10-31T12:26:48.848Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-122b-a10b-20260224",
      "model_slug": "qwen/qwen3.5-122b-a10b",
      "model_name": "Qwen: Qwen3.5-122B-A10B",
      "provider_name": "DeepInfra",
      "accuracy": 0.7606664000000001,
      "accuracy_std_dev": 0.028315632117966207,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.224639588,
      "avg_duration_per_task_ms": 748495.616,
      "avg_output_tokens_per_task": 60239.316,
      "last_run_timestamp": "2026-09-26T08:05:12.736Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "AkashML",
      "accuracy": 0.7604166354166666,
      "accuracy_std_dev": 0.010476531320050537,
      "run_count": 4,
      "total_questions": 192,
      "avg_cost_per_task_usd": 0.29691326041666666,
      "avg_duration_per_task_ms": 310718.984375,
      "avg_output_tokens_per_task": 7030.182291666667,
      "last_run_timestamp": "2026-10-31T12:13:14.359Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Mancer 2",
      "accuracy": 0.7603703333333334,
      "accuracy_std_dev": 0.033126907327897526,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.08524436222222224,
      "avg_duration_per_task_ms": 566210.18,
      "avg_output_tokens_per_task": 12520.904444444444,
      "last_run_timestamp": "2026-10-31T12:26:16.049Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Alibaba",
      "accuracy": 0.7600002499999999,
      "accuracy_std_dev": 0.027080189561513434,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.02989183,
      "avg_duration_per_task_ms": 125270.795,
      "avg_output_tokens_per_task": 12269.195,
      "last_run_timestamp": "2026-10-31T12:01:23.978Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "auto-routing",
      "accuracy": 0.7600001999999999,
      "accuracy_std_dev": 0.03126929626582598,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.114411384,
      "avg_duration_per_task_ms": 874872.748,
      "avg_output_tokens_per_task": 24203.156,
      "last_run_timestamp": "2026-10-31T12:44:40.958Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Cloudflare",
      "accuracy": 0.76,
      "accuracy_std_dev": 0.022253774427851893,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.03227686,
      "avg_duration_per_task_ms": 210388.82571428572,
      "avg_output_tokens_per_task": 14164.22,
      "last_run_timestamp": "2026-10-31T19:55:08.023Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Pearl Research",
      "accuracy": 0.76,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.11851576,
      "avg_duration_per_task_ms": 319323.58,
      "avg_output_tokens_per_task": 19182.22,
      "last_run_timestamp": "2026-11-04T15:08:37.819Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "DekaLLM",
      "accuracy": 0.76,
      "accuracy_std_dev": 0.014401594564491825,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.019761673333333334,
      "avg_duration_per_task_ms": 291193.25333333336,
      "avg_output_tokens_per_task": 17811.393333333333,
      "last_run_timestamp": "2026-10-10T11:23:23.352Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "GLM 5.3",
      "provider_name": "digitalocean",
      "accuracy": 0.76,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.01597332,
      "avg_duration_per_task_ms": 42256.58,
      "avg_output_tokens_per_task": 1764.66,
      "last_run_timestamp": "2026-10-01T23:50:45.025Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.1-pro-preview-20260219",
      "model_slug": "google/gemini-3.1-pro-preview",
      "model_name": "Google: Gemini 3.1 Pro Preview",
      "provider_name": "Google",
      "accuracy": 0.76,
      "accuracy_std_dev": 0.01333300000000004,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.3443486,
      "avg_duration_per_task_ms": 184596.7,
      "avg_output_tokens_per_task": 12502.76,
      "last_run_timestamp": "2026-10-01T19:04:40.361Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3.1-pro-20261023",
      "model_slug": "moonshotai/kimi-k3.1-pro",
      "model_name": "MoonshotAI: Kimi K3.1 Pro",
      "provider_name": "Decart",
      "accuracy": 0.76,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.14539668,
      "avg_duration_per_task_ms": 189204.7,
      "avg_output_tokens_per_task": 10161.9,
      "last_run_timestamp": "2026-11-11T15:19:13.582Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3.1-pro-20261023",
      "model_slug": "moonshotai/kimi-k3.1-pro",
      "model_name": "MoonshotAI: Kimi K3.1 Pro",
      "provider_name": "Morph",
      "accuracy": 0.76,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.27690942,
      "avg_duration_per_task_ms": 230447.6,
      "avg_output_tokens_per_task": 10636.8,
      "last_run_timestamp": "2026-11-05T06:25:26.737Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.1-20251113",
      "model_slug": "openai/gpt-5.1",
      "model_name": "OpenAI: GPT-5.1",
      "provider_name": "OpenAI",
      "accuracy": 0.76,
      "accuracy_std_dev": 0.005443582705045132,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.40005456666666667,
      "avg_duration_per_task_ms": 394539.76666666666,
      "avg_output_tokens_per_task": 34690.113333333335,
      "last_run_timestamp": "2026-10-02T23:12:57.355Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Novita",
      "accuracy": 0.76,
      "accuracy_std_dev": 0.025991443038046205,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.08891490399999999,
      "avg_duration_per_task_ms": 301406.882,
      "avg_output_tokens_per_task": 12044.934,
      "last_run_timestamp": "2026-10-31T12:46:14.833Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "GMICloud",
      "accuracy": 0.76,
      "accuracy_std_dev": 0.010886348913509402,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.06985152,
      "avg_duration_per_task_ms": 702954.7,
      "avg_output_tokens_per_task": 12942.693333333333,
      "last_run_timestamp": "2026-11-10T07:27:08.340Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "Cohere",
      "accuracy": 0.76,
      "accuracy_std_dev": 0.0066669999999999785,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.10151575999999998,
      "avg_duration_per_task_ms": 434936.03,
      "avg_output_tokens_per_task": 13962.46,
      "last_run_timestamp": "2026-10-24T12:32:07.608Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Makora",
      "accuracy": 0.76,
      "accuracy_std_dev": 0.0066669999999999785,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.07038507,
      "avg_duration_per_task_ms": 256302.49,
      "avg_output_tokens_per_task": 10045.27,
      "last_run_timestamp": "2026-08-26T00:00:05.638Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "SambaNova",
      "accuracy": 0.76,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.14464874,
      "avg_duration_per_task_ms": 109820.9,
      "avg_output_tokens_per_task": 4147.16,
      "last_run_timestamp": "2026-11-14T00:56:57.828Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Novita",
      "accuracy": 0.7599999000000001,
      "accuracy_std_dev": 0.030840204640858015,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.008153706000000002,
      "avg_duration_per_task_ms": 200971.108,
      "avg_output_tokens_per_task": 4219.412,
      "last_run_timestamp": "2026-10-31T12:49:39.925Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "DeepInfra",
      "accuracy": 0.7599998,
      "accuracy_std_dev": 0.007302967435227963,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.007451167999999999,
      "avg_duration_per_task_ms": 324203.24,
      "avg_output_tokens_per_task": 12454.056,
      "last_run_timestamp": "2026-10-31T20:10:56.893Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "DeepInfra",
      "accuracy": 0.759846288184438,
      "accuracy_std_dev": 0.015452349289717374,
      "run_count": 7,
      "total_questions": 347,
      "avg_cost_per_task_usd": 0.15869831700288184,
      "avg_duration_per_task_ms": 791662.6916426513,
      "avg_output_tokens_per_task": 30669.08069164265,
      "last_run_timestamp": "2026-10-31T19:07:30.175Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Ionstream",
      "accuracy": 0.7595079765100671,
      "accuracy_std_dev": 0.030194432004533367,
      "run_count": 12,
      "total_questions": 596,
      "avg_cost_per_task_usd": 0.04601521476510067,
      "avg_duration_per_task_ms": 392651.7818791946,
      "avg_output_tokens_per_task": 12917.01845637584,
      "last_run_timestamp": "2026-11-10T17:13:11.399Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "CoreWeave",
      "accuracy": 0.7594444166666666,
      "accuracy_std_dev": 0.033715523315080306,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.09267692833333334,
      "avg_duration_per_task_ms": 244983.77333333335,
      "avg_output_tokens_per_task": 12483.865,
      "last_run_timestamp": "2026-10-31T12:53:55.102Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Wafer",
      "accuracy": 0.7592155882352941,
      "accuracy_std_dev": 0.027618637596970693,
      "run_count": 17,
      "total_questions": 850,
      "avg_cost_per_task_usd": 0.011689743529411765,
      "avg_duration_per_task_ms": 166650.9011764706,
      "avg_output_tokens_per_task": 5474.670588235294,
      "last_run_timestamp": "2026-10-31T12:46:12.537Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "NextBit",
      "accuracy": 0.759157608058608,
      "accuracy_std_dev": 0.022493444068575088,
      "run_count": 11,
      "total_questions": 546,
      "avg_cost_per_task_usd": 0.01790187728937729,
      "avg_duration_per_task_ms": 191122.336996337,
      "avg_output_tokens_per_task": 6046.5641025641025,
      "last_run_timestamp": "2026-11-01T11:24:36.664Z"
    },
    {
      "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": 0.7590475714285715,
      "accuracy_std_dev": 0.02882413325103582,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.021705314285714285,
      "avg_duration_per_task_ms": 232157.9742857143,
      "avg_output_tokens_per_task": 12154.794285714286,
      "last_run_timestamp": "2026-10-31T19:23:14.281Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3.1-pro-20261023",
      "model_slug": "moonshotai/kimi-k3.1-pro",
      "model_name": "MoonshotAI: Kimi K3.1 Pro",
      "provider_name": "InferenceNet",
      "accuracy": 0.7590360803212852,
      "accuracy_std_dev": 0.01520064937560236,
      "run_count": 5,
      "total_questions": 249,
      "avg_cost_per_task_usd": 0.21346128514056226,
      "avg_duration_per_task_ms": 312634.5903614458,
      "avg_output_tokens_per_task": 10222.072289156626,
      "last_run_timestamp": "2026-11-05T00:24:27.672Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "Alibaba",
      "accuracy": 0.7587878181818182,
      "accuracy_std_dev": 0.01653114932933463,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.17482494545454544,
      "avg_duration_per_task_ms": 315756.4018181818,
      "avg_output_tokens_per_task": 10740.534545454546,
      "last_run_timestamp": "2026-10-31T11:13:27.901Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.5-opus-20251124",
      "model_slug": "anthropic/claude-opus-4.5",
      "model_name": "Anthropic: Claude Opus 4.5",
      "provider_name": "Anthropic",
      "accuracy": 0.7586668,
      "accuracy_std_dev": 0.024184373284416548,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.5305409400000001,
      "avg_duration_per_task_ms": 181423.388,
      "avg_output_tokens_per_task": 10458.596,
      "last_run_timestamp": "2026-10-31T12:38:33.924Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "OpenInference",
      "accuracy": 0.7583335,
      "accuracy_std_dev": 0.011666500000000024,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01736758,
      "avg_duration_per_task_ms": 642594.16,
      "avg_output_tokens_per_task": 12204.57,
      "last_run_timestamp": "2026-08-29T16:47:05.637Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Relace",
      "accuracy": 0.7583333333333333,
      "accuracy_std_dev": 0.03125451671181126,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.06620967333333333,
      "avg_duration_per_task_ms": 161431.49,
      "avg_output_tokens_per_task": 10316.483333333334,
      "last_run_timestamp": "2026-10-31T19:43:05.932Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.1-20251113",
      "model_slug": "openai/gpt-5.1",
      "model_name": "OpenAI: GPT-5.1",
      "provider_name": "Azure",
      "accuracy": 0.7583332500000001,
      "accuracy_std_dev": 0.014433785597946933,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.42515429,
      "avg_duration_per_task_ms": 460450.54,
      "avg_output_tokens_per_task": 34342.89,
      "last_run_timestamp": "2026-10-01T16:01:13.647Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7580951428571429,
      "accuracy_std_dev": 0.032014759919354034,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.10769543714285713,
      "avg_duration_per_task_ms": 240797.41714285716,
      "avg_output_tokens_per_task": 11172.16,
      "last_run_timestamp": "2026-10-31T20:03:31.357Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Alibaba",
      "accuracy": 0.7578571428571429,
      "accuracy_std_dev": 0.029048460957159017,
      "run_count": 14,
      "total_questions": 700,
      "avg_cost_per_task_usd": 0.09093985428571429,
      "avg_duration_per_task_ms": 189688.76714285713,
      "avg_output_tokens_per_task": 11033.007142857143,
      "last_run_timestamp": "2026-10-31T19:54:53.711Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "ParetoAlpha",
      "accuracy": 0.7577776666666667,
      "accuracy_std_dev": 0.013698571612479267,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.28612407999999995,
      "avg_duration_per_task_ms": 2065607.3133333332,
      "avg_output_tokens_per_task": 10347.066666666668,
      "last_run_timestamp": "2026-10-31T11:54:15.773Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7577317113402062,
      "accuracy_std_dev": 0.03673749999999998,
      "run_count": 2,
      "total_questions": 97,
      "avg_cost_per_task_usd": 0.07915294845360824,
      "avg_duration_per_task_ms": 496514.40206185565,
      "avg_output_tokens_per_task": 7669.20618556701,
      "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-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "CoreWeave",
      "accuracy": 0.7575757272727273,
      "accuracy_std_dev": 0.0216999667829277,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.02146501818181818,
      "avg_duration_per_task_ms": 118448.94,
      "avg_output_tokens_per_task": 17795.93090909091,
      "last_run_timestamp": "2026-10-31T12:41:20.328Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Cerebras",
      "accuracy": 0.7575,
      "accuracy_std_dev": 0.05171382256708548,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.26214604,
      "avg_duration_per_task_ms": 26048.875,
      "avg_output_tokens_per_task": 12109.95,
      "last_run_timestamp": "2026-10-31T11:18:07.571Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7575,
      "accuracy_std_dev": 0.023078766095699303,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.012481125,
      "avg_duration_per_task_ms": 215467.87,
      "avg_output_tokens_per_task": 4890.4975,
      "last_run_timestamp": "2026-10-31T11:56:22.568Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "GMICloud",
      "accuracy": 0.7573335999999999,
      "accuracy_std_dev": 0.009043086986201126,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.015731108,
      "avg_duration_per_task_ms": 124933.624,
      "avg_output_tokens_per_task": 11968.716,
      "last_run_timestamp": "2026-11-10T06:08:17.524Z"
    },
    {
      "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": 0.7573335,
      "accuracy_std_dev": 0.016653408061114705,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.071170548,
      "avg_duration_per_task_ms": 532347.09,
      "avg_output_tokens_per_task": 31376.552,
      "last_run_timestamp": "2026-10-31T20:08:33.748Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "NextBit",
      "accuracy": 0.7573333999999999,
      "accuracy_std_dev": 0.03991088294989224,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.023026752,
      "avg_duration_per_task_ms": 161149.652,
      "avg_output_tokens_per_task": 18473.992,
      "last_run_timestamp": "2026-10-03T08:55:31.322Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "InferenceNet",
      "accuracy": 0.7568888666666667,
      "accuracy_std_dev": 0.02909656905746029,
      "run_count": 15,
      "total_questions": 750,
      "avg_cost_per_task_usd": 0.008429334666666666,
      "avg_duration_per_task_ms": 120051.87733333334,
      "avg_output_tokens_per_task": 4480.442666666667,
      "last_run_timestamp": "2026-11-04T21:55:13.332Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta/muse-spark-open-20261015",
      "model_slug": "meta/muse-spark-open",
      "model_name": "Meta: Muse Spark Open",
      "provider_name": "auto-routing",
      "accuracy": 0.7566669999999999,
      "accuracy_std_dev": 0.010000000000000009,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.08839812,
      "avg_duration_per_task_ms": 73380.11,
      "avg_output_tokens_per_task": 8719.83,
      "last_run_timestamp": "2026-10-31T12:44:24.100Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3.1-pro-20261023",
      "model_slug": "moonshotai/kimi-k3.1-pro",
      "model_name": "MoonshotAI: Kimi K3.1 Pro",
      "provider_name": "Wafer",
      "accuracy": 0.7566669999999999,
      "accuracy_std_dev": 0.010000000000000009,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.17797051000000003,
      "avg_duration_per_task_ms": 171587.38,
      "avg_output_tokens_per_task": 10362.94,
      "last_run_timestamp": "2026-11-05T21:56:52.017Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7566669999999999,
      "accuracy_std_dev": 0.010000000000000009,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.06446149,
      "avg_duration_per_task_ms": 318929.97,
      "avg_output_tokens_per_task": 13284.66,
      "last_run_timestamp": "2026-10-31T12:25:36.318Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "CoreWeave",
      "accuracy": 0.7566668,
      "accuracy_std_dev": 0.01819650760338369,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.151380274,
      "avg_duration_per_task_ms": 165330.14,
      "avg_output_tokens_per_task": 19976.098,
      "last_run_timestamp": "2026-10-31T12:20:13.275Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "Darkbloom",
      "accuracy": 0.7566666666666667,
      "accuracy_std_dev": 0.016216087122223912,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.02569377333333333,
      "avg_duration_per_task_ms": 474320.2833333333,
      "avg_output_tokens_per_task": 22069.101666666666,
      "last_run_timestamp": "2026-11-13T02:23:53.892Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7566666666666667,
      "accuracy_std_dev": 0.022026835486641215,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.06927284666666667,
      "avg_duration_per_task_ms": 313911.26,
      "avg_output_tokens_per_task": 14319.526666666667,
      "last_run_timestamp": "2026-10-31T12:43:48.271Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "StreamLake",
      "accuracy": 0.756666625,
      "accuracy_std_dev": 0.03366510310001107,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.01679233,
      "avg_duration_per_task_ms": 291946.0025,
      "avg_output_tokens_per_task": 10730.4225,
      "last_run_timestamp": "2026-10-31T20:10:47.431Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Baidu",
      "accuracy": 0.756666625,
      "accuracy_std_dev": 0.015985992133251373,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.052604632500000005,
      "avg_duration_per_task_ms": 71429.835,
      "avg_output_tokens_per_task": 4345.545,
      "last_run_timestamp": "2026-11-10T18:36:01.112Z"
    },
    {
      "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": 0.7566665,
      "accuracy_std_dev": 0.0251662251622421,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.03758577333333333,
      "avg_duration_per_task_ms": 199140.7,
      "avg_output_tokens_per_task": 13605.943333333333,
      "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": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7566665,
      "accuracy_std_dev": 0.03666649999999999,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.07760444,
      "avg_duration_per_task_ms": 178261.76,
      "avg_output_tokens_per_task": 9204.98,
      "last_run_timestamp": "2026-08-29T16:11:58.076Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Crusoe",
      "accuracy": 0.7566665,
      "accuracy_std_dev": 0.0033334999999999892,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.09622449,
      "avg_duration_per_task_ms": 171236.24,
      "avg_output_tokens_per_task": 9651.1,
      "last_run_timestamp": "2026-08-29T16:31:41.554Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Venice",
      "accuracy": 0.7564471518624641,
      "accuracy_std_dev": 0.017879759130688233,
      "run_count": 7,
      "total_questions": 349,
      "avg_cost_per_task_usd": 0.010146441260744987,
      "avg_duration_per_task_ms": 321562.229226361,
      "avg_output_tokens_per_task": 12668.309455587392,
      "last_run_timestamp": "2026-10-31T19:48:48.460Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Reka",
      "accuracy": 0.7564444666666666,
      "accuracy_std_dev": 0.023958871344776582,
      "run_count": 15,
      "total_questions": 750,
      "avg_cost_per_task_usd": 0.04415017200000001,
      "avg_duration_per_task_ms": 138405.18933333334,
      "avg_output_tokens_per_task": 4510.322666666667,
      "last_run_timestamp": "2026-10-31T12:43:47.323Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "StreamLake",
      "accuracy": 0.7563635454545454,
      "accuracy_std_dev": 0.01641972318648099,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.005844881818181818,
      "avg_duration_per_task_ms": 265682.2036363636,
      "avg_output_tokens_per_task": 15989.905454545455,
      "last_run_timestamp": "2026-10-31T13:22:04.008Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Mancer 2",
      "accuracy": 0.7562962222222223,
      "accuracy_std_dev": 0.035011314652957475,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.04747446444444444,
      "avg_duration_per_task_ms": 235698.6,
      "avg_output_tokens_per_task": 14674.364444444445,
      "last_run_timestamp": "2026-10-31T12:45:59.472Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Alibaba",
      "accuracy": 0.7562961111111112,
      "accuracy_std_dev": 0.02966657881426859,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.03134207111111111,
      "avg_duration_per_task_ms": 175492.92222222223,
      "avg_output_tokens_per_task": 14182.893333333333,
      "last_run_timestamp": "2026-10-31T11:54:37.503Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Darkbloom",
      "accuracy": 0.7562620173410404,
      "accuracy_std_dev": 0.01469578908589609,
      "run_count": 7,
      "total_questions": 346,
      "avg_cost_per_task_usd": 0.03183617341040463,
      "avg_duration_per_task_ms": 505021.5144508671,
      "avg_output_tokens_per_task": 12785.520231213874,
      "last_run_timestamp": "2026-11-13T02:24:13.374Z"
    },
    {
      "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": 0.7561904285714286,
      "accuracy_std_dev": 0.020735037227271297,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.1716927257142857,
      "avg_duration_per_task_ms": 187054.57428571428,
      "avg_output_tokens_per_task": 11177.605714285713,
      "last_run_timestamp": "2026-10-31T19:18:33.522Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "auto-routing",
      "accuracy": 0.7561904285714286,
      "accuracy_std_dev": 0.03184434649468365,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.03082173142857143,
      "avg_duration_per_task_ms": 402850.4971428571,
      "avg_output_tokens_per_task": 11747.214285714286,
      "last_run_timestamp": "2026-10-31T19:35:25.708Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Cloudflare",
      "accuracy": 0.7560607272727272,
      "accuracy_std_dev": 0.029432815279705333,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.18950213272727276,
      "avg_duration_per_task_ms": 498615.7418181818,
      "avg_output_tokens_per_task": 20980.29090909091,
      "last_run_timestamp": "2026-10-31T14:12:32.839Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "StreamLake",
      "accuracy": 0.755833375,
      "accuracy_std_dev": 0.03072051092892785,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0448964025,
      "avg_duration_per_task_ms": 187316.3775,
      "avg_output_tokens_per_task": 9579.4275,
      "last_run_timestamp": "2026-10-31T20:17:26.781Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Parasail",
      "accuracy": 0.7558333333333334,
      "accuracy_std_dev": 0.021131098726495254,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.01632131833333333,
      "avg_duration_per_task_ms": 115260.22166666666,
      "avg_output_tokens_per_task": 4430.61,
      "last_run_timestamp": "2026-10-31T12:09:39.076Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.7557471896551724,
      "accuracy_std_dev": 0.051344765491603474,
      "run_count": 7,
      "total_questions": 348,
      "avg_cost_per_task_usd": 0.07958818965517242,
      "avg_duration_per_task_ms": 709509.3793103448,
      "avg_output_tokens_per_task": 47576.2183908046,
      "last_run_timestamp": "2026-10-31T18:44:20.050Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.5-opus-20251124",
      "model_slug": "anthropic/claude-opus-4.5",
      "model_name": "Anthropic: Claude Opus 4.5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7556667499999999,
      "accuracy_std_dev": 0.03178233349342839,
      "run_count": 20,
      "total_questions": 1000,
      "avg_cost_per_task_usd": 0.5530219200000001,
      "avg_duration_per_task_ms": 181245.853,
      "avg_output_tokens_per_task": 10575.78,
      "last_run_timestamp": "2026-10-31T12:37:18.963Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.7555554444444444,
      "accuracy_std_dev": 0.018856167739030906,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.0086613,
      "avg_duration_per_task_ms": 117282.64,
      "avg_output_tokens_per_task": 5380.108888888889,
      "last_run_timestamp": "2026-10-31T13:20:27.978Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Together",
      "accuracy": 0.7554545454545455,
      "accuracy_std_dev": 0.03830433230509108,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.16801238363636364,
      "avg_duration_per_task_ms": 309650.71636363637,
      "avg_output_tokens_per_task": 24145.685454545455,
      "last_run_timestamp": "2026-10-31T13:11:05.620Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Novita",
      "accuracy": 0.7553333999999999,
      "accuracy_std_dev": 0.017650954020675496,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.040253192,
      "avg_duration_per_task_ms": 172625.032,
      "avg_output_tokens_per_task": 16547.004,
      "last_run_timestamp": "2026-10-31T12:45:56.690Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Venice",
      "accuracy": 0.7552381428571429,
      "accuracy_std_dev": 0.023015459657919207,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.021888451428571428,
      "avg_duration_per_task_ms": 479502.3685714286,
      "avg_output_tokens_per_task": 13683.494285714285,
      "last_run_timestamp": "2026-10-31T19:04:16.045Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Makora",
      "accuracy": 0.7552378571428572,
      "accuracy_std_dev": 0.014126057041900291,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.009880265714285713,
      "avg_duration_per_task_ms": 245506.26571428572,
      "avg_output_tokens_per_task": 16074.99142857143,
      "last_run_timestamp": "2026-10-31T12:27:44.073Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta/muse-spark-open-20261015",
      "model_slug": "meta/muse-spark-open",
      "model_name": "Meta: Muse Spark Open",
      "provider_name": "Modal",
      "accuracy": 0.7551024285714285,
      "accuracy_std_dev": 0.011805499999999969,
      "run_count": 2,
      "total_questions": 98,
      "avg_cost_per_task_usd": 0.27501580612244897,
      "avg_duration_per_task_ms": 65997.72448979592,
      "avg_output_tokens_per_task": 8842.591836734693,
      "last_run_timestamp": "2026-10-31T12:36:21.331Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "DeepInfra",
      "accuracy": 0.7550335167785235,
      "accuracy_std_dev": 0.023364430687122125,
      "run_count": 6,
      "total_questions": 298,
      "avg_cost_per_task_usd": 0.012232758389261746,
      "avg_duration_per_task_ms": 267603.8926174497,
      "avg_output_tokens_per_task": 4166.208053691275,
      "last_run_timestamp": "2026-10-31T12:14:04.084Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "DigitalOcean",
      "accuracy": 0.754999875,
      "accuracy_std_dev": 0.022791430530121965,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.011049522499999999,
      "avg_duration_per_task_ms": 1077704.9225,
      "avg_output_tokens_per_task": 13220.1325,
      "last_run_timestamp": "2026-10-31T20:29:45.923Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Modal",
      "accuracy": 0.7549998,
      "accuracy_std_dev": 0.029636688974310207,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.014858368,
      "avg_duration_per_task_ms": 71120.474,
      "avg_output_tokens_per_task": 4319.272,
      "last_run_timestamp": "2026-10-31T11:35:00.891Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Baidu",
      "accuracy": 0.7547045887096774,
      "accuracy_std_dev": 0.03743255869747622,
      "run_count": 5,
      "total_questions": 248,
      "avg_cost_per_task_usd": 0.02217318548387097,
      "avg_duration_per_task_ms": 116333.60483870968,
      "avg_output_tokens_per_task": 15344.233870967742,
      "last_run_timestamp": "2026-11-10T18:13:04.202Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Venice",
      "accuracy": 0.7547041935483871,
      "accuracy_std_dev": 0.023888446777469652,
      "run_count": 10,
      "total_questions": 496,
      "avg_cost_per_task_usd": 0.01950188306451613,
      "avg_duration_per_task_ms": 216554.71370967742,
      "avg_output_tokens_per_task": 4241.487903225807,
      "last_run_timestamp": "2026-10-31T11:51:07.775Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.6-opus-20260205",
      "model_slug": "anthropic/claude-opus-4.6",
      "model_name": "Anthropic: Claude Opus 4.6",
      "provider_name": "Anthropic",
      "accuracy": 0.7546666,
      "accuracy_std_dev": 0.025785348123304443,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.5201454280000001,
      "avg_duration_per_task_ms": 216038.234,
      "avg_output_tokens_per_task": 9883.556,
      "last_run_timestamp": "2026-10-31T12:37:18.962Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Alibaba",
      "accuracy": 0.75416675,
      "accuracy_std_dev": 0.019096569464219906,
      "run_count": 13,
      "total_questions": 640,
      "avg_cost_per_task_usd": 0.0324113640625,
      "avg_duration_per_task_ms": 163294.796875,
      "avg_output_tokens_per_task": 15840.8265625,
      "last_run_timestamp": "2026-10-31T11:41:35.064Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.5-opus-20251124",
      "model_slug": "anthropic/claude-opus-4.5",
      "model_name": "Anthropic: Claude Opus 4.5",
      "provider_name": "Google",
      "accuracy": 0.754,
      "accuracy_std_dev": 0.031895600533615925,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.535401688,
      "avg_duration_per_task_ms": 188314.086,
      "avg_output_tokens_per_task": 10622.22,
      "last_run_timestamp": "2026-10-31T12:48:37.862Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "GMICloud",
      "accuracy": 0.7538888333333333,
      "accuracy_std_dev": 0.021162299055290663,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.06070119499999999,
      "avg_duration_per_task_ms": 164537.05,
      "avg_output_tokens_per_task": 4635.8516666666665,
      "last_run_timestamp": "2026-11-10T06:54:00.557Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.6-flash-20260721",
      "model_slug": "google/gemini-3.6-flash",
      "model_name": "Google: Gemini 3.6 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.7533335,
      "accuracy_std_dev": 0.026666500000000037,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.16966541999999996,
      "avg_duration_per_task_ms": 147439,
      "avg_output_tokens_per_task": 22903.86,
      "last_run_timestamp": "2026-10-01T19:04:26.676Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Azure",
      "accuracy": 0.7533334285714286,
      "accuracy_std_dev": 0.028507921484874254,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.02505947714285714,
      "avg_duration_per_task_ms": 206845.73428571428,
      "avg_output_tokens_per_task": 13108.18,
      "last_run_timestamp": "2026-10-31T19:02:15.155Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-opus-5.5-20260921",
      "model_slug": "anthropic/claude-opus-5.5",
      "model_name": "Anthropic: Claude Opus 5.5",
      "provider_name": "auto-routing",
      "accuracy": 0.7533333333333333,
      "accuracy_std_dev": 0.01721340993398913,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.33945626,
      "avg_duration_per_task_ms": 116389.55333333333,
      "avg_output_tokens_per_task": 7044.336666666667,
      "last_run_timestamp": "2026-10-31T12:36:01.147Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Wafer",
      "accuracy": 0.7533333333333333,
      "accuracy_std_dev": 0.03925236487426234,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.03720142666666666,
      "avg_duration_per_task_ms": 244561.18666666668,
      "avg_output_tokens_per_task": 13729.346666666666,
      "last_run_timestamp": "2026-10-31T19:28:27.377Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Relace",
      "accuracy": 0.7533333333333333,
      "accuracy_std_dev": 0.02277092037860767,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.020851383333333334,
      "avg_duration_per_task_ms": 133350.03333333333,
      "avg_output_tokens_per_task": 13555.86,
      "last_run_timestamp": "2026-10-31T19:21:19.082Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Parasail",
      "accuracy": 0.7533333333333333,
      "accuracy_std_dev": 0.005443174461857933,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.018967786666666667,
      "avg_duration_per_task_ms": 85971.64666666667,
      "avg_output_tokens_per_task": 11730.173333333334,
      "last_run_timestamp": "2026-10-31T12:40:46.055Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Alibaba",
      "accuracy": 0.7533333333333333,
      "accuracy_std_dev": 0.029186609056590034,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.09943221,
      "avg_duration_per_task_ms": 284246.225,
      "avg_output_tokens_per_task": 12177.723333333333,
      "last_run_timestamp": "2026-10-31T12:41:58.031Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-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": 0.7533333333333333,
      "accuracy_std_dev": 0.02000009259423011,
      "run_count": 18,
      "total_questions": 900,
      "avg_cost_per_task_usd": 0.009942297777777777,
      "avg_duration_per_task_ms": 208204.5977777778,
      "avg_output_tokens_per_task": 4241.026666666667,
      "last_run_timestamp": "2026-11-06T22:25:00.575Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-opus-5.5-20260921",
      "model_slug": "anthropic/claude-opus-5.5",
      "model_name": "Anthropic: Claude Opus 5.5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.7533332857142857,
      "accuracy_std_dev": 0.015118704911601446,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.3460215285714286,
      "avg_duration_per_task_ms": 119668.60857142857,
      "avg_output_tokens_per_task": 7072.9971428571425,
      "last_run_timestamp": "2026-10-31T12:24:55.233Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "DeepInfra",
      "accuracy": 0.7533332727272728,
      "accuracy_std_dev": 0.023952899912084696,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.007874056363636363,
      "avg_duration_per_task_ms": 231905.71636363637,
      "avg_output_tokens_per_task": 15094.398181818182,
      "last_run_timestamp": "2026-10-31T13:30:46.598Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Decart",
      "accuracy": 0.753333,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01197713,
      "avg_duration_per_task_ms": 212560.45,
      "avg_output_tokens_per_task": 13957.42,
      "last_run_timestamp": "2026-10-24T11:49:49.163Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.3-codex-20260224",
      "model_slug": "openai/gpt-5.3-codex",
      "model_name": "OpenAI: GPT-5.3-Codex",
      "provider_name": "OpenAI",
      "accuracy": 0.753333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.29971074000000003,
      "avg_duration_per_task_ms": 180697.5,
      "avg_output_tokens_per_task": 13569.24,
      "last_run_timestamp": "2026-10-02T23:12:05.500Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "x-ai/grok-4.7-20260916",
      "model_slug": "x-ai/grok-4.7",
      "model_name": "SpaceXAI: Grok 4.7",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.753333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.29315726000000003,
      "avg_duration_per_task_ms": 206111.18,
      "avg_output_tokens_per_task": 11166.84,
      "last_run_timestamp": "2026-11-03T04:31:24.038Z"
    },
    {
      "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": "River",
      "accuracy": 0.753333,
      "accuracy_std_dev": 0.020000000000000018,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.06050348,
      "avg_duration_per_task_ms": 93599.96,
      "avg_output_tokens_per_task": 4522.22,
      "last_run_timestamp": "2026-11-12T21:06:33.901Z"
    },
    {
      "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": 0.7532544781065089,
      "accuracy_std_dev": 0.023852454489848872,
      "run_count": 17,
      "total_questions": 845,
      "avg_cost_per_task_usd": 0.08912921893491124,
      "avg_duration_per_task_ms": 600544.8769230769,
      "avg_output_tokens_per_task": 21467.282840236687,
      "last_run_timestamp": "2026-11-07T02:09:29.421Z"
    },
    {
      "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": 0.7527289420654912,
      "accuracy_std_dev": 0.024253123484087077,
      "run_count": 8,
      "total_questions": 397,
      "avg_cost_per_task_usd": 0.06776491435768262,
      "avg_duration_per_task_ms": 1043193.3853904282,
      "avg_output_tokens_per_task": 77465.15365239295,
      "last_run_timestamp": "2026-10-24T19:38:23.433Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Fireworks",
      "accuracy": 0.752715030075188,
      "accuracy_std_dev": 0.009940333684036977,
      "run_count": 8,
      "total_questions": 399,
      "avg_cost_per_task_usd": 0.02512772681704261,
      "avg_duration_per_task_ms": 75665.01253132832,
      "avg_output_tokens_per_task": 12251.558897243107,
      "last_run_timestamp": "2026-10-31T12:26:39.122Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "StreamLake",
      "accuracy": 0.7526667,
      "accuracy_std_dev": 0.02374861961062159,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.007248376000000002,
      "avg_duration_per_task_ms": 194889.812,
      "avg_output_tokens_per_task": 4384.802,
      "last_run_timestamp": "2026-10-31T13:46:40.932Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "InferenceNet",
      "accuracy": 0.7525287586206897,
      "accuracy_std_dev": 0.020184097544858358,
      "run_count": 29,
      "total_questions": 1450,
      "avg_cost_per_task_usd": 0.2696313268965517,
      "avg_duration_per_task_ms": 233088.7579310345,
      "avg_output_tokens_per_task": 11456.733793103449,
      "last_run_timestamp": "2026-11-04T21:30:17.879Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-opus-5.5-20260921",
      "model_slug": "anthropic/claude-opus-5.5",
      "model_name": "Anthropic: Claude Opus 5.5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7523809047619048,
      "accuracy_std_dev": 0.017612644913472624,
      "run_count": 21,
      "total_questions": 1050,
      "avg_cost_per_task_usd": 0.36889707904761904,
      "avg_duration_per_task_ms": 141180.9876190476,
      "avg_output_tokens_per_task": 7127.063809523809,
      "last_run_timestamp": "2026-10-31T12:26:17.125Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Wafer",
      "accuracy": 0.752222375,
      "accuracy_std_dev": 0.02258048225941397,
      "run_count": 24,
      "total_questions": 1200,
      "avg_cost_per_task_usd": 0.24382183083333336,
      "avg_duration_per_task_ms": 219317.55666666667,
      "avg_output_tokens_per_task": 10147.890833333333,
      "last_run_timestamp": "2026-10-31T12:43:33.299Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Venice",
      "accuracy": 0.752,
      "accuracy_std_dev": 0.0324961125890467,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.018297658,
      "avg_duration_per_task_ms": 209683.558,
      "avg_output_tokens_per_task": 15226.814,
      "last_run_timestamp": "2026-10-31T12:48:59.031Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "Google",
      "accuracy": 0.752,
      "accuracy_std_dev": 0.018570921118781412,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.12117575600000001,
      "avg_duration_per_task_ms": 124707.216,
      "avg_output_tokens_per_task": 10483.008,
      "last_run_timestamp": "2026-10-17T19:14:08.120Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-122b-a10b-20260224",
      "model_slug": "qwen/qwen3.5-122b-a10b",
      "model_name": "Qwen: Qwen3.5-122B-A10B",
      "provider_name": "Alibaba",
      "accuracy": 0.7517208112094396,
      "accuracy_std_dev": 0.045868971980311604,
      "run_count": 7,
      "total_questions": 339,
      "avg_cost_per_task_usd": 0.09701299410029499,
      "avg_duration_per_task_ms": 405555.37758112093,
      "avg_output_tokens_per_task": 19035.16814159292,
      "last_run_timestamp": "2026-10-31T19:29:13.230Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta/muse-spark-open-20261015",
      "model_slug": "meta/muse-spark-open",
      "model_name": "Meta: Muse Spark Open",
      "provider_name": "Nebius",
      "accuracy": 0.7516778724832214,
      "accuracy_std_dev": 0.030780262062995275,
      "run_count": 3,
      "total_questions": 149,
      "avg_cost_per_task_usd": 0.08526638255033557,
      "avg_duration_per_task_ms": 172059.79865771811,
      "avg_output_tokens_per_task": 8282.208053691274,
      "last_run_timestamp": "2026-10-31T11:50:00.858Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Alibaba",
      "accuracy": 0.7514694483627204,
      "accuracy_std_dev": 0.034873185045067215,
      "run_count": 8,
      "total_questions": 397,
      "avg_cost_per_task_usd": 0.10740512846347608,
      "avg_duration_per_task_ms": 153833.89168765742,
      "avg_output_tokens_per_task": 9153.158690176322,
      "last_run_timestamp": "2026-10-31T19:46:27.571Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.7514287142857142,
      "accuracy_std_dev": 0.02844421520758676,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.032982819999999996,
      "avg_duration_per_task_ms": 221535.9714285714,
      "avg_output_tokens_per_task": 35206.688571428575,
      "last_run_timestamp": "2026-10-31T11:46:18.454Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Parasail",
      "accuracy": 0.751256015075377,
      "accuracy_std_dev": 0.006852721115549665,
      "run_count": 4,
      "total_questions": 199,
      "avg_cost_per_task_usd": 0.07795362814070352,
      "avg_duration_per_task_ms": 228603.50753768845,
      "avg_output_tokens_per_task": 10368.422110552763,
      "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-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "GMICloud",
      "accuracy": 0.7511110833333333,
      "accuracy_std_dev": 0.03446230540619305,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.09647276666666667,
      "avg_duration_per_task_ms": 304912.8616666667,
      "avg_output_tokens_per_task": 18589.79,
      "last_run_timestamp": "2026-11-10T08:38:21.830Z"
    },
    {
      "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": 0.7508046206896551,
      "accuracy_std_dev": 0.028089208249744292,
      "run_count": 29,
      "total_questions": 1450,
      "avg_cost_per_task_usd": 0.20928471310344823,
      "avg_duration_per_task_ms": 114151.32896551724,
      "avg_output_tokens_per_task": 7693.8193103448275,
      "last_run_timestamp": "2026-10-31T12:48:39.187Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Decart",
      "accuracy": 0.7506668,
      "accuracy_std_dev": 0.025854254755455656,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.264866352,
      "avg_duration_per_task_ms": 264537.464,
      "avg_output_tokens_per_task": 10440.212,
      "last_run_timestamp": "2026-10-24T11:49:28.605Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "Parasail",
      "accuracy": 0.7506666,
      "accuracy_std_dev": 0.016384222765819578,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.036424934,
      "avg_duration_per_task_ms": 269977.466,
      "avg_output_tokens_per_task": 21216.128,
      "last_run_timestamp": "2026-10-31T20:23:01.980Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Near AI",
      "accuracy": 0.7506666,
      "accuracy_std_dev": 0.03785357285435553,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.007070539999999999,
      "avg_duration_per_task_ms": 292989.264,
      "avg_output_tokens_per_task": 3954.392,
      "last_run_timestamp": "2026-10-31T13:04:37.434Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7504761428571428,
      "accuracy_std_dev": 0.026869944603210353,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.011859305714285716,
      "avg_duration_per_task_ms": 203490.8142857143,
      "avg_output_tokens_per_task": 13513.24,
      "last_run_timestamp": "2026-10-31T20:16:35.156Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7504761428571428,
      "accuracy_std_dev": 0.04779025223509368,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.020752462857142853,
      "avg_duration_per_task_ms": 333009.8742857143,
      "avg_output_tokens_per_task": 11171.528571428571,
      "last_run_timestamp": "2026-10-31T13:00:18.168Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "Darkbloom",
      "accuracy": 0.7500001506849315,
      "accuracy_std_dev": 0.02552374915424634,
      "run_count": 3,
      "total_questions": 146,
      "avg_cost_per_task_usd": 0.027398205479452053,
      "avg_duration_per_task_ms": 1306114.2808219178,
      "avg_output_tokens_per_task": 60026.84246575343,
      "last_run_timestamp": "2026-11-13T02:06:40.995Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "CoreWeave",
      "accuracy": 0.75,
      "accuracy_std_dev": 0.023333000000000048,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01955629,
      "avg_duration_per_task_ms": 235357.66,
      "avg_output_tokens_per_task": 12724.88,
      "last_run_timestamp": "2026-08-29T16:33:29.042Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Ionstream",
      "accuracy": 0.75,
      "accuracy_std_dev": 0.030000000000000027,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.08026519,
      "avg_duration_per_task_ms": 649894.08,
      "avg_output_tokens_per_task": 11165.01,
      "last_run_timestamp": "2026-08-29T17:19:42.622Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-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": "meta/muse-glimmer-30b-20260810",
      "model_slug": "meta/muse-glimmer-30b",
      "model_name": "Meta: Muse Glimmer 30B",
      "provider_name": "Together",
      "accuracy": 0.75,
      "accuracy_std_dev": 0.01666660000320003,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.041500640000000005,
      "avg_duration_per_task_ms": 199607.65333333332,
      "avg_output_tokens_per_task": 12674.073333333334,
      "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": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Io Net",
      "accuracy": 0.75,
      "accuracy_std_dev": 0.0033330000000000304,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.37141135,
      "avg_duration_per_task_ms": 606870.4,
      "avg_output_tokens_per_task": 13812.73,
      "last_run_timestamp": "2026-11-11T05:59:06.349Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.4-20260305",
      "model_slug": "openai/gpt-5.4",
      "model_name": "OpenAI: GPT-5.4",
      "provider_name": "auto-routing",
      "accuracy": 0.75,
      "accuracy_std_dev": 0.0033330000000000304,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.29130333999999997,
      "avg_duration_per_task_ms": 176962.37,
      "avg_output_tokens_per_task": 10386.26,
      "last_run_timestamp": "2026-10-01T19:04:36.169Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "DeepInfra",
      "accuracy": 0.75,
      "accuracy_std_dev": 0.028631403807707387,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.04503071555555555,
      "avg_duration_per_task_ms": 450907.20666666667,
      "avg_output_tokens_per_task": 22758.244444444445,
      "last_run_timestamp": "2026-10-24T19:38:23.863Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7499999999999999,
      "accuracy_std_dev": 0.03349968490896593,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.028200379999999997,
      "avg_duration_per_task_ms": 163955.21,
      "avg_output_tokens_per_task": 15323.823333333334,
      "last_run_timestamp": "2026-10-31T12:28:07.595Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-opus-5.5-20260921",
      "model_slug": "anthropic/claude-opus-5.5",
      "model_name": "Anthropic: Claude Opus 5.5",
      "provider_name": "Azure",
      "accuracy": 0.749524,
      "accuracy_std_dev": 0.021633867377662402,
      "run_count": 14,
      "total_questions": 700,
      "avg_cost_per_task_usd": 0.35241304,
      "avg_duration_per_task_ms": 87048.52428571429,
      "avg_output_tokens_per_task": 6993.714285714285,
      "last_run_timestamp": "2026-10-31T11:29:30.760Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "CoreWeave",
      "accuracy": 0.7495238571428572,
      "accuracy_std_dev": 0.021338220105372136,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.016605834285714285,
      "avg_duration_per_task_ms": 72972.84285714285,
      "avg_output_tokens_per_task": 4253.202857142857,
      "last_run_timestamp": "2026-10-31T12:03:51.953Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "DeepInfra",
      "accuracy": 0.7493333,
      "accuracy_std_dev": 0.015549210771289982,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.012228522,
      "avg_duration_per_task_ms": 315949.068,
      "avg_output_tokens_per_task": 16300.002,
      "last_run_timestamp": "2026-10-31T11:27:38.689Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Fireworks",
      "accuracy": 0.7492774046242774,
      "accuracy_std_dev": 0.025230016427736986,
      "run_count": 14,
      "total_questions": 692,
      "avg_cost_per_task_usd": 0.013838482658959535,
      "avg_duration_per_task_ms": 56300.32225433526,
      "avg_output_tokens_per_task": 4085.725433526012,
      "last_run_timestamp": "2026-10-31T12:14:05.375Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-sonnet-5-20260630",
      "model_slug": "anthropic/claude-sonnet-5",
      "model_name": "Anthropic: Claude Sonnet 5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.749091,
      "accuracy_std_dev": 0.0224325111713902,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.1957560618181818,
      "avg_duration_per_task_ms": 114755.80545454545,
      "avg_output_tokens_per_task": 7618.518181818182,
      "last_run_timestamp": "2026-10-31T11:08:14.469Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Ionstream",
      "accuracy": 0.7490679239373601,
      "accuracy_std_dev": 0.028614120182218948,
      "run_count": 9,
      "total_questions": 447,
      "avg_cost_per_task_usd": 0.09606371140939597,
      "avg_duration_per_task_ms": 393542.0715883669,
      "avg_output_tokens_per_task": 24831.977628635348,
      "last_run_timestamp": "2026-10-31T12:39:48.592Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "DeepInfra",
      "accuracy": 0.748889,
      "accuracy_std_dev": 0.015713326891527487,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.03620722,
      "avg_duration_per_task_ms": 114735.25333333333,
      "avg_output_tokens_per_task": 14087.753333333334,
      "last_run_timestamp": "2026-10-31T11:35:01.591Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "ModelRun",
      "accuracy": 0.7488888333333333,
      "accuracy_std_dev": 0.024993786884054944,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.08968307666666667,
      "avg_duration_per_task_ms": 142393.51,
      "avg_output_tokens_per_task": 12273.018333333333,
      "last_run_timestamp": "2026-10-31T18:56:42.115Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7487373080808083,
      "accuracy_std_dev": 0.03476268720171639,
      "run_count": 8,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.08973175505050504,
      "avg_duration_per_task_ms": 588220.7929292929,
      "avg_output_tokens_per_task": 9930.64898989899,
      "last_run_timestamp": "2026-10-31T20:25:10.002Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-sonnet-5-20260630",
      "model_slug": "anthropic/claude-sonnet-5",
      "model_name": "Anthropic: Claude Sonnet 5",
      "provider_name": "auto-routing",
      "accuracy": 0.7486668,
      "accuracy_std_dev": 0.03141830181534325,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.200451768,
      "avg_duration_per_task_ms": 115869.634,
      "avg_output_tokens_per_task": 7737.526,
      "last_run_timestamp": "2026-10-31T11:54:55.856Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Parasail",
      "accuracy": 0.7486667,
      "accuracy_std_dev": 0.02733343658616678,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.018688805999999995,
      "avg_duration_per_task_ms": 207184.068,
      "avg_output_tokens_per_task": 15779.308,
      "last_run_timestamp": "2026-10-31T12:32:12.444Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7485715714285714,
      "accuracy_std_dev": 0.02069105868213987,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.013271285714285713,
      "avg_duration_per_task_ms": 244472.46,
      "avg_output_tokens_per_task": 14692.394285714287,
      "last_run_timestamp": "2026-10-31T19:19:01.715Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-opus-5.5-20260921",
      "model_slug": "anthropic/claude-opus-5.5",
      "model_name": "Anthropic: Claude Opus 5.5",
      "provider_name": "Anthropic",
      "accuracy": 0.7485714285714286,
      "accuracy_std_dev": 0.023560522731632144,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.34892749714285715,
      "avg_duration_per_task_ms": 98456.37428571428,
      "avg_output_tokens_per_task": 7086.754285714285,
      "last_run_timestamp": "2026-10-31T11:27:20.548Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7485712857142858,
      "accuracy_std_dev": 0.027994178209632344,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.18495269714285717,
      "avg_duration_per_task_ms": 176865.04,
      "avg_output_tokens_per_task": 9737.897142857142,
      "last_run_timestamp": "2026-10-31T20:09:19.251Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Wafer",
      "accuracy": 0.748333375,
      "accuracy_std_dev": 0.024664456546503833,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0092696725,
      "avg_duration_per_task_ms": 91481.9675,
      "avg_output_tokens_per_task": 13410.6275,
      "last_run_timestamp": "2026-10-31T13:03:50.748Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Modal",
      "accuracy": 0.748322067114094,
      "accuracy_std_dev": 0.02148614665313445,
      "run_count": 3,
      "total_questions": 149,
      "avg_cost_per_task_usd": 0.029814261744966444,
      "avg_duration_per_task_ms": 257681.3489932886,
      "avg_output_tokens_per_task": 10900.818791946309,
      "last_run_timestamp": "2026-10-31T12:26:05.030Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Baidu",
      "accuracy": 0.7481482222222222,
      "accuracy_std_dev": 0.02927572719953733,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.012146982222222224,
      "avg_duration_per_task_ms": 103906.89555555556,
      "avg_output_tokens_per_task": 16269.204444444444,
      "last_run_timestamp": "2026-11-10T18:35:45.854Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "auto-routing",
      "accuracy": 0.7481482222222222,
      "accuracy_std_dev": 0.020068464436510978,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.19089935333333333,
      "avg_duration_per_task_ms": 286228.22222222225,
      "avg_output_tokens_per_task": 11572.273333333333,
      "last_run_timestamp": "2026-10-31T12:37:16.632Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Io Net",
      "accuracy": 0.748148,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 45,
      "avg_cost_per_task_usd": 0.021196088888888888,
      "avg_duration_per_task_ms": 92100.66666666667,
      "avg_output_tokens_per_task": 8167.488888888889,
      "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-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "auto-routing",
      "accuracy": 0.7480002,
      "accuracy_std_dev": 0.020396130352593862,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.030171100000000003,
      "avg_duration_per_task_ms": 197857.312,
      "avg_output_tokens_per_task": 18663.832,
      "last_run_timestamp": "2026-10-03T08:18:33.163Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-sonnet-5-20260630",
      "model_slug": "anthropic/claude-sonnet-5",
      "model_name": "Anthropic: Claude Sonnet 5",
      "provider_name": "Anthropic",
      "accuracy": 0.7480001,
      "accuracy_std_dev": 0.037333425000795194,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.19706498,
      "avg_duration_per_task_ms": 95603.86,
      "avg_output_tokens_per_task": 7573.922,
      "last_run_timestamp": "2026-10-31T11:28:40.153Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "BaseTen",
      "accuracy": 0.74791675,
      "accuracy_std_dev": 0.031222055284486006,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.01487202,
      "avg_duration_per_task_ms": 51526.7675,
      "avg_output_tokens_per_task": 4120.2525,
      "last_run_timestamp": "2026-10-31T13:13:13.022Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Inceptron",
      "accuracy": 0.7478510515759312,
      "accuracy_std_dev": 0.031000808296901307,
      "run_count": 7,
      "total_questions": 349,
      "avg_cost_per_task_usd": 0.018540833810888252,
      "avg_duration_per_task_ms": 126193.2005730659,
      "avg_output_tokens_per_task": 4035.12893982808,
      "last_run_timestamp": "2026-10-31T11:27:58.748Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Crusoe",
      "accuracy": 0.7477776666666667,
      "accuracy_std_dev": 0.028393017748891882,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.020487799999999997,
      "avg_duration_per_task_ms": 112394.16,
      "avg_output_tokens_per_task": 6879.653333333334,
      "last_run_timestamp": "2026-10-31T12:42:03.915Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "x-ai/grok-4.6-20260810",
      "model_slug": "x-ai/grok-4.6",
      "model_name": "SpaceXAI: Grok 4.6",
      "provider_name": "xAI",
      "accuracy": 0.747619,
      "accuracy_std_dev": 0.025305279849978234,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.33171734571428574,
      "avg_duration_per_task_ms": 235739.5942857143,
      "avg_output_tokens_per_task": 12164.34,
      "last_run_timestamp": "2026-10-01T16:01:04.865Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "Modal",
      "accuracy": 0.7472221666666666,
      "accuracy_std_dev": 0.011613686586906374,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.17439429333333334,
      "avg_duration_per_task_ms": 161420.59666666668,
      "avg_output_tokens_per_task": 12499.9,
      "last_run_timestamp": "2026-10-10T11:19:48.441Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Wafer",
      "accuracy": 0.7470370555555556,
      "accuracy_std_dev": 0.025285697851359517,
      "run_count": 18,
      "total_questions": 900,
      "avg_cost_per_task_usd": 0.02032471444444444,
      "avg_duration_per_task_ms": 102872.56777777777,
      "avg_output_tokens_per_task": 14307.606666666667,
      "last_run_timestamp": "2026-10-31T13:23:12.329Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Relace",
      "accuracy": 0.746667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.00930136,
      "avg_duration_per_task_ms": 89811.58,
      "avg_output_tokens_per_task": 12020.36,
      "last_run_timestamp": "2026-10-24T15:18:33.444Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3.1-pro-20261023",
      "model_slug": "moonshotai/kimi-k3.1-pro",
      "model_name": "MoonshotAI: Kimi K3.1 Pro",
      "provider_name": "DeepInfra",
      "accuracy": 0.746667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.22191687999999998,
      "avg_duration_per_task_ms": 2285676.56,
      "avg_output_tokens_per_task": 10495.84,
      "last_run_timestamp": "2026-11-05T06:48:08.646Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3.1-pro-20261023",
      "model_slug": "moonshotai/kimi-k3.1-pro",
      "model_name": "MoonshotAI: Kimi K3.1 Pro",
      "provider_name": "AkashML",
      "accuracy": 0.746667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.21601631999999998,
      "avg_duration_per_task_ms": 312892.28,
      "avg_output_tokens_per_task": 10774.7,
      "last_run_timestamp": "2026-11-05T06:21:39.264Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3.1-pro-20261023",
      "model_slug": "moonshotai/kimi-k3.1-pro",
      "model_name": "MoonshotAI: Kimi K3.1 Pro",
      "provider_name": "Alibaba",
      "accuracy": 0.746667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.33231256000000003,
      "avg_duration_per_task_ms": 225567.04,
      "avg_output_tokens_per_task": 11123.54,
      "last_run_timestamp": "2026-11-04T21:00:38.446Z"
    },
    {
      "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": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "Venice",
      "accuracy": 0.7466667499999999,
      "accuracy_std_dev": 0.012472414008021857,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.070158785,
      "avg_duration_per_task_ms": 1726878.96,
      "avg_output_tokens_per_task": 17573.88,
      "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": "anthropic/claude-4.5-opus-20251124",
      "model_slug": "anthropic/claude-opus-4.5",
      "model_name": "Anthropic: Claude Opus 4.5",
      "provider_name": "Azure",
      "accuracy": 0.7466667,
      "accuracy_std_dev": 0.02529832669189804,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.5260258940000001,
      "avg_duration_per_task_ms": 187136.702,
      "avg_output_tokens_per_task": 10321.958,
      "last_run_timestamp": "2026-10-31T11:53:29.846Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Phala",
      "accuracy": 0.7466666000000001,
      "accuracy_std_dev": 0.035527816958546725,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.032979988,
      "avg_duration_per_task_ms": 209798.362,
      "avg_output_tokens_per_task": 16884.604,
      "last_run_timestamp": "2026-10-31T12:25:12.423Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Reka",
      "accuracy": 0.7466665625000001,
      "accuracy_std_dev": 0.022360772946078904,
      "run_count": 16,
      "total_questions": 800,
      "avg_cost_per_task_usd": 0.013514558749999999,
      "avg_duration_per_task_ms": 139486.7775,
      "avg_output_tokens_per_task": 15486.175,
      "last_run_timestamp": "2026-10-31T13:03:01.037Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "DeepInfra",
      "accuracy": 0.7466665,
      "accuracy_std_dev": 0.026666500000000037,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.00520861,
      "avg_duration_per_task_ms": 148091.91,
      "avg_output_tokens_per_task": 12193.49,
      "last_run_timestamp": "2026-10-31T11:47:33.345Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta/muse-glimmer-30b-20260810",
      "model_slug": "meta/muse-glimmer-30b",
      "model_name": "Meta: Muse Glimmer 30B",
      "provider_name": "Fireworks",
      "accuracy": 0.7466665,
      "accuracy_std_dev": 0.01054097823970812,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.047252215,
      "avg_duration_per_task_ms": 265241.545,
      "avg_output_tokens_per_task": 13256.62,
      "last_run_timestamp": "2026-09-19T08:01:36.427Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.5-20260423",
      "model_slug": "openai/gpt-5.5",
      "model_name": "OpenAI: GPT-5.5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7466665,
      "accuracy_std_dev": 0.00666650000000002,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.58250779,
      "avg_duration_per_task_ms": 102747.13,
      "avg_output_tokens_per_task": 7546.85,
      "last_run_timestamp": "2026-10-01T19:04:42.459Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "River",
      "accuracy": 0.7466665,
      "accuracy_std_dev": 0.00666650000000002,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.007670949999999999,
      "avg_duration_per_task_ms": 53013.89,
      "avg_output_tokens_per_task": 4043,
      "last_run_timestamp": "2026-11-12T20:51:11.231Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Morph",
      "accuracy": 0.7466663333333334,
      "accuracy_std_dev": 0.033993463423951875,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.01425808,
      "avg_duration_per_task_ms": 139346.08,
      "avg_output_tokens_per_task": 12850.746666666666,
      "last_run_timestamp": "2026-10-31T12:18:46.817Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-sonnet-5-20260630",
      "model_slug": "anthropic/claude-sonnet-5",
      "model_name": "Anthropic: Claude Sonnet 5",
      "provider_name": "Azure",
      "accuracy": 0.74633335,
      "accuracy_std_dev": 0.028478019009887265,
      "run_count": 20,
      "total_questions": 1000,
      "avg_cost_per_task_usd": 0.201189951,
      "avg_duration_per_task_ms": 95684.807,
      "avg_output_tokens_per_task": 7554.449,
      "last_run_timestamp": "2026-10-31T12:38:13.068Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "DeepInfra",
      "accuracy": 0.7463126578171091,
      "accuracy_std_dev": 0.020825554044618038,
      "run_count": 7,
      "total_questions": 339,
      "avg_cost_per_task_usd": 0.15311067846607668,
      "avg_duration_per_task_ms": 683576.4837758112,
      "avg_output_tokens_per_task": 30552.457227138642,
      "last_run_timestamp": "2026-10-31T19:55:58.516Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.5-sonnet-20250929",
      "model_slug": "anthropic/claude-sonnet-4.5",
      "model_name": "Anthropic: Claude Sonnet 4.5",
      "provider_name": "Anthropic",
      "accuracy": 0.7460002,
      "accuracy_std_dev": 0.031191292232929373,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.34538107000000007,
      "avg_duration_per_task_ms": 224701.758,
      "avg_output_tokens_per_task": 10858.334,
      "last_run_timestamp": "2026-10-31T12:24:26.627Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Decart",
      "accuracy": 0.74583325,
      "accuracy_std_dev": 0.03402396010442498,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.02499106,
      "avg_duration_per_task_ms": 929627.78,
      "avg_output_tokens_per_task": 52531.49,
      "last_run_timestamp": "2026-10-24T12:31:49.408Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.7457779,
      "accuracy_std_dev": 0.024569333893222824,
      "run_count": 30,
      "total_questions": 1500,
      "avg_cost_per_task_usd": 0.3685412306666667,
      "avg_duration_per_task_ms": 148474.26666666666,
      "avg_output_tokens_per_task": 11703.939333333334,
      "last_run_timestamp": "2026-10-31T12:39:28.528Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "CoreWeave",
      "accuracy": 0.7457142857142857,
      "accuracy_std_dev": 0.02770085585430936,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.025092839999999998,
      "avg_duration_per_task_ms": 285709.0542857143,
      "avg_output_tokens_per_task": 10724.948571428571,
      "last_run_timestamp": "2026-10-31T19:15:14.763Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Venice",
      "accuracy": 0.7457142857142857,
      "accuracy_std_dev": 0.035353609893330494,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.0900297257142857,
      "avg_duration_per_task_ms": 239520.15714285715,
      "avg_output_tokens_per_task": 6359.951428571429,
      "last_run_timestamp": "2026-10-31T19:33:49.408Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-sonnet-5-20260630",
      "model_slug": "anthropic/claude-sonnet-5",
      "model_name": "Anthropic: Claude Sonnet 5",
      "provider_name": "Google",
      "accuracy": 0.7453333,
      "accuracy_std_dev": 0.032313302845886877,
      "run_count": 30,
      "total_questions": 1500,
      "avg_cost_per_task_usd": 0.20589792800000006,
      "avg_duration_per_task_ms": 90844.94666666667,
      "avg_output_tokens_per_task": 7426.582,
      "last_run_timestamp": "2026-10-31T13:20:20.506Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "Novita",
      "accuracy": 0.7453333,
      "accuracy_std_dev": 0.02454934079379731,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.17119798400000003,
      "avg_duration_per_task_ms": 320387.726,
      "avg_output_tokens_per_task": 11023.856,
      "last_run_timestamp": "2026-10-31T12:25:36.323Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "StreamLake",
      "accuracy": 0.7451853333333333,
      "accuracy_std_dev": 0.031234315132203188,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.01557802666666667,
      "avg_duration_per_task_ms": 147555.71333333335,
      "avg_output_tokens_per_task": 15042.41111111111,
      "last_run_timestamp": "2026-10-31T12:25:02.563Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "Together",
      "accuracy": 0.7451852222222222,
      "accuracy_std_dev": 0.017716056438394953,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.21677278888888887,
      "avg_duration_per_task_ms": 284791.18666666665,
      "avg_output_tokens_per_task": 11557.691111111111,
      "last_run_timestamp": "2026-10-31T11:21:59.486Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "SambaNova",
      "accuracy": 0.745,
      "accuracy_std_dev": 0.026822730826670128,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.15284815,
      "avg_duration_per_task_ms": 494527.49,
      "avg_output_tokens_per_task": 28246.935,
      "last_run_timestamp": "2026-10-17T19:47:21.934Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Phala",
      "accuracy": 0.744999875,
      "accuracy_std_dev": 0.030867001888900313,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0114373625,
      "avg_duration_per_task_ms": 404697.66,
      "avg_output_tokens_per_task": 5754.185,
      "last_run_timestamp": "2026-10-31T12:24:48.023Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Cohere",
      "accuracy": 0.744898,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 49,
      "avg_cost_per_task_usd": 0.046561530612244904,
      "avg_duration_per_task_ms": 36257.63265306123,
      "avg_output_tokens_per_task": 3739.4285714285716,
      "last_run_timestamp": "2026-11-13T00:52:20.873Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3.1-pro-20261023",
      "model_slug": "moonshotai/kimi-k3.1-pro",
      "model_name": "MoonshotAI: Kimi K3.1 Pro",
      "provider_name": "PrimaLabs",
      "accuracy": 0.7447914583333333,
      "accuracy_std_dev": 0.005435000000000023,
      "run_count": 2,
      "total_questions": 96,
      "avg_cost_per_task_usd": 0.13105378125,
      "avg_duration_per_task_ms": 273825.1666666667,
      "avg_output_tokens_per_task": 8026.104166666667,
      "last_run_timestamp": "2026-11-06T15:59:08.297Z"
    },
    {
      "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": 0.7446667,
      "accuracy_std_dev": 0.01956190515798501,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.3469682859999999,
      "avg_duration_per_task_ms": 215223.822,
      "avg_output_tokens_per_task": 11882.416,
      "last_run_timestamp": "2026-10-31T13:31:12.841Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Wafer",
      "accuracy": 0.7446667,
      "accuracy_std_dev": 0.01956180291818729,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.055664519999999995,
      "avg_duration_per_task_ms": 332682.316,
      "avg_output_tokens_per_task": 14458.648,
      "last_run_timestamp": "2026-10-31T11:42:33.248Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Together",
      "accuracy": 0.7444443333333334,
      "accuracy_std_dev": 0.02199873172606905,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.02140772,
      "avg_duration_per_task_ms": 53184.22,
      "avg_output_tokens_per_task": 14364.18,
      "last_run_timestamp": "2026-10-31T11:41:11.192Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "Reka",
      "accuracy": 0.7444443333333334,
      "accuracy_std_dev": 0.006285236475706844,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.04041968666666667,
      "avg_duration_per_task_ms": 291588.68666666665,
      "avg_output_tokens_per_task": 24330.48,
      "last_run_timestamp": "2026-09-30T11:25:05.946Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "DekaLLM",
      "accuracy": 0.7444443333333334,
      "accuracy_std_dev": 0.03823241466202323,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.012960726666666665,
      "avg_duration_per_task_ms": 94310.08666666667,
      "avg_output_tokens_per_task": 4403.8,
      "last_run_timestamp": "2026-11-11T15:58:45.893Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-122b-a10b-20260224",
      "model_slug": "qwen/qwen3.5-122b-a10b",
      "model_name": "Qwen: Qwen3.5-122B-A10B",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7440000999999999,
      "accuracy_std_dev": 0.10836965996112564,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.150848968,
      "avg_duration_per_task_ms": 558886.734,
      "avg_output_tokens_per_task": 27996,
      "last_run_timestamp": "2026-10-31T19:37:30.834Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Phala",
      "accuracy": 0.744,
      "accuracy_std_dev": 0.04013309523821955,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.12688716400000002,
      "avg_duration_per_task_ms": 700323.218,
      "avg_output_tokens_per_task": 36239.088,
      "last_run_timestamp": "2026-10-31T12:47:12.433Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Io Net",
      "accuracy": 0.743738,
      "accuracy_std_dev": 0.01688456971344439,
      "run_count": 7,
      "total_questions": 346,
      "avg_cost_per_task_usd": 0.00938970520231214,
      "avg_duration_per_task_ms": 347914.32658959535,
      "avg_output_tokens_per_task": 3877.916184971098,
      "last_run_timestamp": "2026-10-03T08:46:08.334Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.743502793220339,
      "accuracy_std_dev": 0.018213736559866647,
      "run_count": 6,
      "total_questions": 295,
      "avg_cost_per_task_usd": 0.010024135593220337,
      "avg_duration_per_task_ms": 123315.84745762713,
      "avg_output_tokens_per_task": 13505.108474576271,
      "last_run_timestamp": "2026-11-02T17:51:25.369Z"
    },
    {
      "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": 0.7433335,
      "accuracy_std_dev": 0.01666650000000003,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.06358472,
      "avg_duration_per_task_ms": 288690.34,
      "avg_output_tokens_per_task": 8551,
      "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": "x-ai/grok-4.6-20260810",
      "model_slug": "x-ai/grok-4.6",
      "model_name": "SpaceXAI: Grok 4.6",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7433335,
      "accuracy_std_dev": 0.043333500000000025,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.64823893,
      "avg_duration_per_task_ms": 264133.83,
      "avg_output_tokens_per_task": 10511,
      "last_run_timestamp": "2026-10-01T16:00:37.003Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Cloudflare",
      "accuracy": 0.7433333333333333,
      "accuracy_std_dev": 0.017532868606008445,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.012350623333333333,
      "avg_duration_per_task_ms": 125234.99333333333,
      "avg_output_tokens_per_task": 4198.8,
      "last_run_timestamp": "2026-10-03T08:23:09.816Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-opus-5.5-20260921",
      "model_slug": "anthropic/claude-opus-5.5",
      "model_name": "Anthropic: Claude Opus 5.5",
      "provider_name": "Google",
      "accuracy": 0.7431746190476191,
      "accuracy_std_dev": 0.018008818983714862,
      "run_count": 21,
      "total_questions": 1050,
      "avg_cost_per_task_usd": 0.36043625523809525,
      "avg_duration_per_task_ms": 86437.01714285715,
      "avg_output_tokens_per_task": 7106.278095238095,
      "last_run_timestamp": "2026-10-31T12:25:51.224Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.6-opus-20260205",
      "model_slug": "anthropic/claude-opus-4.6",
      "model_name": "Anthropic: Claude Opus 4.6",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.7430302727272727,
      "accuracy_std_dev": 0.027760033635986,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.5088923236363636,
      "avg_duration_per_task_ms": 192892.78363636363,
      "avg_output_tokens_per_task": 9984.365454545454,
      "last_run_timestamp": "2026-10-31T12:42:53.834Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "AkashML",
      "accuracy": 0.7430122052565706,
      "accuracy_std_dev": 0.013125368013739409,
      "run_count": 16,
      "total_questions": 799,
      "avg_cost_per_task_usd": 0.04480421151439299,
      "avg_duration_per_task_ms": 334888.2340425532,
      "avg_output_tokens_per_task": 25744.07634543179,
      "last_run_timestamp": "2026-10-31T19:20:57.312Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Morph",
      "accuracy": 0.7427351282051282,
      "accuracy_std_dev": 0.028431218171008805,
      "run_count": 4,
      "total_questions": 195,
      "avg_cost_per_task_usd": 0.013693692307692308,
      "avg_duration_per_task_ms": 1000472.5128205129,
      "avg_output_tokens_per_task": 18094.641025641027,
      "last_run_timestamp": "2026-11-05T09:28:04.757Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7426666000000001,
      "accuracy_std_dev": 0.04434197309367279,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.16675142799999998,
      "avg_duration_per_task_ms": 302047.356,
      "avg_output_tokens_per_task": 9682.606,
      "last_run_timestamp": "2026-10-31T11:21:43.861Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Sail Research",
      "accuracy": 0.7425926666666667,
      "accuracy_std_dev": 0.03165311687654157,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.013586744444444443,
      "avg_duration_per_task_ms": 236886.98,
      "avg_output_tokens_per_task": 22693.746666666666,
      "last_run_timestamp": "2026-10-24T12:29:34.092Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Morph",
      "accuracy": 0.7425926111111111,
      "accuracy_std_dev": 0.0225683017579448,
      "run_count": 18,
      "total_questions": 900,
      "avg_cost_per_task_usd": 0.4177886388888889,
      "avg_duration_per_task_ms": 618406.2588888889,
      "avg_output_tokens_per_task": 10668.22,
      "last_run_timestamp": "2026-10-31T11:24:57.351Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Mara",
      "accuracy": 0.7425405959079282,
      "accuracy_std_dev": 0.026283120985422872,
      "run_count": 8,
      "total_questions": 391,
      "avg_cost_per_task_usd": 0.12785004603580563,
      "avg_duration_per_task_ms": 203562.63938618926,
      "avg_output_tokens_per_task": 10880.268542199488,
      "last_run_timestamp": "2026-10-31T19:27:07.786Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Baidu",
      "accuracy": 0.7425,
      "accuracy_std_dev": 0.02470671298959049,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.007974519999999999,
      "avg_duration_per_task_ms": 152350.215,
      "avg_output_tokens_per_task": 12723.1225,
      "last_run_timestamp": "2026-11-10T18:17:30.522Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Alibaba",
      "accuracy": 0.7424241818181818,
      "accuracy_std_dev": 0.03530779679235989,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.36784247818181814,
      "avg_duration_per_task_ms": 381853.54363636364,
      "avg_output_tokens_per_task": 9773.532727272728,
      "last_run_timestamp": "2026-10-31T12:43:33.294Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-sol-20260709",
      "model_slug": "openai/gpt-5.6-sol",
      "model_name": "OpenAI: GPT-5.6 Sol",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7422223333333333,
      "accuracy_std_dev": 0.016629618082071397,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.27693500666666665,
      "avg_duration_per_task_ms": 100083.4,
      "avg_output_tokens_per_task": 5197.013333333333,
      "last_run_timestamp": "2026-10-30T00:15:31.633Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Cohere",
      "accuracy": 0.7422223333333333,
      "accuracy_std_dev": 0.011330936570096763,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.01793052666666667,
      "avg_duration_per_task_ms": 86484.18,
      "avg_output_tokens_per_task": 20778.286666666667,
      "last_run_timestamp": "2026-10-31T11:37:00.194Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Friendli",
      "accuracy": 0.7420636446886446,
      "accuracy_std_dev": 0.016469967104193704,
      "run_count": 11,
      "total_questions": 546,
      "avg_cost_per_task_usd": 0.01720449084249084,
      "avg_duration_per_task_ms": 82040.02197802198,
      "avg_output_tokens_per_task": 4114.483516483517,
      "last_run_timestamp": "2026-10-31T12:00:42.348Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "DeepInfra",
      "accuracy": 0.7420434974874371,
      "accuracy_std_dev": 0.018804943831343932,
      "run_count": 8,
      "total_questions": 398,
      "avg_cost_per_task_usd": 0.046914160804020104,
      "avg_duration_per_task_ms": 685430.3140703518,
      "avg_output_tokens_per_task": 14198.173366834171,
      "last_run_timestamp": "2026-10-31T19:48:13.441Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Venice",
      "accuracy": 0.742,
      "accuracy_std_dev": 0.02599994871918021,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.143269796,
      "avg_duration_per_task_ms": 585731.59,
      "avg_output_tokens_per_task": 25891.308,
      "last_run_timestamp": "2026-10-31T19:26:38.406Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "ModelRun",
      "accuracy": 0.7419049285714285,
      "accuracy_std_dev": 0.027305278452197487,
      "run_count": 14,
      "total_questions": 700,
      "avg_cost_per_task_usd": 0.05683837428571428,
      "avg_duration_per_task_ms": 141278.7842857143,
      "avg_output_tokens_per_task": 10137.55,
      "last_run_timestamp": "2026-10-31T19:13:33.926Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "GMICloud",
      "accuracy": 0.7418181818181818,
      "accuracy_std_dev": 0.027279408610685823,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.019760812727272727,
      "avg_duration_per_task_ms": 152641.17272727273,
      "avg_output_tokens_per_task": 15736.389090909091,
      "last_run_timestamp": "2026-11-10T06:11:05.037Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Phala",
      "accuracy": 0.741666625,
      "accuracy_std_dev": 0.025331164930464133,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.021842495,
      "avg_duration_per_task_ms": 141327.99,
      "avg_output_tokens_per_task": 16250.775,
      "last_run_timestamp": "2026-10-31T12:24:54.205Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "CoreWeave",
      "accuracy": 0.741666625,
      "accuracy_std_dev": 0.030866893899198455,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.021115824999999998,
      "avg_duration_per_task_ms": 150544.0025,
      "avg_output_tokens_per_task": 8219.325,
      "last_run_timestamp": "2026-10-31T18:57:59.813Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Relace",
      "accuracy": 0.7416481202672606,
      "accuracy_std_dev": 0.022563192218990436,
      "run_count": 9,
      "total_questions": 449,
      "avg_cost_per_task_usd": 0.21138012694877506,
      "avg_duration_per_task_ms": 402109.59020044544,
      "avg_output_tokens_per_task": 9801.538975501113,
      "last_run_timestamp": "2026-10-31T11:26:03.627Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7414841313868613,
      "accuracy_std_dev": 0.027678383181115592,
      "run_count": 11,
      "total_questions": 548,
      "avg_cost_per_task_usd": 0.04167549270072993,
      "avg_duration_per_task_ms": 154993.6697080292,
      "avg_output_tokens_per_task": 8579.950729927008,
      "last_run_timestamp": "2026-10-31T19:46:31.278Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta/muse-spark-open-20261015",
      "model_slug": "meta/muse-spark-open",
      "model_name": "Meta: Muse Spark Open",
      "provider_name": "Fireworks",
      "accuracy": 0.7413334,
      "accuracy_std_dev": 0.02436751571744648,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.097602524,
      "avg_duration_per_task_ms": 217804.16,
      "avg_output_tokens_per_task": 9920.02,
      "last_run_timestamp": "2026-10-31T14:40:25.043Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "auto-routing",
      "accuracy": 0.7413334,
      "accuracy_std_dev": 0.025086522278705756,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.025878452000000003,
      "avg_duration_per_task_ms": 132781.264,
      "avg_output_tokens_per_task": 9514.518,
      "last_run_timestamp": "2026-10-31T19:14:12.566Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.7413333,
      "accuracy_std_dev": 0.025438450338218323,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.51193901,
      "avg_duration_per_task_ms": 192863.266,
      "avg_output_tokens_per_task": 10096.234,
      "last_run_timestamp": "2026-10-31T12:10:56.773Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "DeepInfra",
      "accuracy": 0.7413333,
      "accuracy_std_dev": 0.03095523075685272,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.40647999399999996,
      "avg_duration_per_task_ms": 588538.78,
      "avg_output_tokens_per_task": 9896.416,
      "last_run_timestamp": "2026-10-31T12:27:47.241Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Modal",
      "accuracy": 0.74125,
      "accuracy_std_dev": 0.021980774429032325,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.35810955499999997,
      "avg_duration_per_task_ms": 189257.0575,
      "avg_output_tokens_per_task": 10848.9475,
      "last_run_timestamp": "2026-10-31T11:51:20.518Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.6-sonnet-20260217",
      "model_slug": "anthropic/claude-sonnet-4.6",
      "model_name": "Anthropic: Claude Sonnet 4.6",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.7412121818181818,
      "accuracy_std_dev": 0.0275340250329006,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.3449351727272727,
      "avg_duration_per_task_ms": 200135.41454545455,
      "avg_output_tokens_per_task": 11800.205454545454,
      "last_run_timestamp": "2026-10-31T13:12:27.229Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Mistral",
      "accuracy": 0.7411111666666667,
      "accuracy_std_dev": 0.009749878126018938,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.029943926666666662,
      "avg_duration_per_task_ms": 81311.67,
      "avg_output_tokens_per_task": 13159.393333333333,
      "last_run_timestamp": "2026-11-11T03:07:37.121Z"
    },
    {
      "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": 0.7411111666666667,
      "accuracy_std_dev": 0.04950373847975749,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.1105783,
      "avg_duration_per_task_ms": 224383.69333333333,
      "avg_output_tokens_per_task": 11802.466666666667,
      "last_run_timestamp": "2026-10-24T19:31:21.574Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Nebius",
      "accuracy": 0.7410254102564102,
      "accuracy_std_dev": 0.015687179899124656,
      "run_count": 4,
      "total_questions": 195,
      "avg_cost_per_task_usd": 0.010222671794871796,
      "avg_duration_per_task_ms": 114819.93846153846,
      "avg_output_tokens_per_task": 3734.005128205128,
      "last_run_timestamp": "2026-10-31T12:06:01.335Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Baidu",
      "accuracy": 0.7409909864864865,
      "accuracy_std_dev": 0.009883309915655236,
      "run_count": 3,
      "total_questions": 148,
      "avg_cost_per_task_usd": 0.014433466216216218,
      "avg_duration_per_task_ms": 89927.77702702703,
      "avg_output_tokens_per_task": 10790.993243243243,
      "last_run_timestamp": "2026-11-10T18:11:41.495Z"
    },
    {
      "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": 0.7409637389558233,
      "accuracy_std_dev": 0.023850320811259554,
      "run_count": 5,
      "total_questions": 249,
      "avg_cost_per_task_usd": 0.12855561847389557,
      "avg_duration_per_task_ms": 62087.51004016064,
      "avg_output_tokens_per_task": 9516.100401606425,
      "last_run_timestamp": "2026-10-31T19:12:18.660Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "GMICloud",
      "accuracy": 0.7409525714285714,
      "accuracy_std_dev": 0.01094194948492327,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.03426931142857143,
      "avg_duration_per_task_ms": 134514.65142857144,
      "avg_output_tokens_per_task": 15886.077142857142,
      "last_run_timestamp": "2026-11-10T06:15:52.514Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "auto-routing",
      "accuracy": 0.7409522857142857,
      "accuracy_std_dev": 0.03499266784307123,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.0443475,
      "avg_duration_per_task_ms": 181001.05142857143,
      "avg_output_tokens_per_task": 9386.614285714286,
      "last_run_timestamp": "2026-10-31T20:15:42.614Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "Baidu",
      "accuracy": 0.740833375,
      "accuracy_std_dev": 0.01898462645759391,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.06345218250000001,
      "avg_duration_per_task_ms": 202212.49,
      "avg_output_tokens_per_task": 11837.4125,
      "last_run_timestamp": "2026-11-10T19:24:51.597Z"
    },
    {
      "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": 0.7406666000000001,
      "accuracy_std_dev": 0.024303563525540885,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.5231801380000001,
      "avg_duration_per_task_ms": 223642.216,
      "avg_output_tokens_per_task": 10003.868,
      "last_run_timestamp": "2026-10-31T13:21:45.027Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.6-opus-20260205",
      "model_slug": "anthropic/claude-opus-4.6",
      "model_name": "Anthropic: Claude Opus 4.6",
      "provider_name": "Google",
      "accuracy": 0.7406666000000001,
      "accuracy_std_dev": 0.026741542381470827,
      "run_count": 20,
      "total_questions": 1000,
      "avg_cost_per_task_usd": 0.5445005100000001,
      "avg_duration_per_task_ms": 200696.784,
      "avg_output_tokens_per_task": 9760.559,
      "last_run_timestamp": "2026-10-31T13:01:06.466Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Together",
      "accuracy": 0.7406666000000001,
      "accuracy_std_dev": 0.018961987708043703,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.014243233999999999,
      "avg_duration_per_task_ms": 72756.944,
      "avg_output_tokens_per_task": 4429.298,
      "last_run_timestamp": "2026-10-31T12:42:18.186Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "AkashML",
      "accuracy": 0.74014702004008,
      "accuracy_std_dev": 0.022155950173486136,
      "run_count": 10,
      "total_questions": 499,
      "avg_cost_per_task_usd": 0.06005350100200401,
      "avg_duration_per_task_ms": 140448.37274549098,
      "avg_output_tokens_per_task": 4119.851703406814,
      "last_run_timestamp": "2026-10-31T12:27:45.919Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Cloudflare",
      "accuracy": 0.740000125,
      "accuracy_std_dev": 0.026034122908394196,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.080340995,
      "avg_duration_per_task_ms": 330284.965,
      "avg_output_tokens_per_task": 11392.905,
      "last_run_timestamp": "2026-10-31T21:26:40.342Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "amazon/nova-micro-v1",
      "model_slug": "amazon/nova-micro-v1",
      "model_name": "Amazon: Nova Micro 1.0",
      "provider_name": "Anthropic",
      "accuracy": 0.74,
      "accuracy_std_dev": 0.0066669999999999785,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 1.23335095,
      "avg_duration_per_task_ms": 96091.26,
      "avg_output_tokens_per_task": 5555.9,
      "last_run_timestamp": "2026-10-01T19:04:26.700Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.74,
      "accuracy_std_dev": 0.027325104453597256,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.328687276,
      "avg_duration_per_task_ms": 196120.964,
      "avg_output_tokens_per_task": 11587.26,
      "last_run_timestamp": "2026-10-31T12:07:21.658Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Crusoe",
      "accuracy": 0.74,
      "accuracy_std_dev": 0.026666999999999996,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.14729807,
      "avg_duration_per_task_ms": 386775.27,
      "avg_output_tokens_per_task": 24280.02,
      "last_run_timestamp": "2026-10-31T19:58:44.255Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "AtlasCloud",
      "accuracy": 0.74,
      "accuracy_std_dev": 0.03511903567511501,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.023083885,
      "avg_duration_per_task_ms": 150345.7,
      "avg_output_tokens_per_task": 16221.6925,
      "last_run_timestamp": "2026-10-31T11:30:43.640Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Near AI",
      "accuracy": 0.74,
      "accuracy_std_dev": 0.0066669999999999785,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01862194,
      "avg_duration_per_task_ms": 339225.77,
      "avg_output_tokens_per_task": 12354.38,
      "last_run_timestamp": "2026-10-31T13:22:50.108Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.1-pro-preview-20260219",
      "model_slug": "google/gemini-3.1-pro-preview",
      "model_name": "Google: Gemini 3.1 Pro Preview",
      "provider_name": "Google AI Studio",
      "accuracy": 0.74,
      "accuracy_std_dev": 0.0066669999999999785,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.37915447,
      "avg_duration_per_task_ms": 148084.91,
      "avg_output_tokens_per_task": 12525.4,
      "last_run_timestamp": "2026-10-01T19:04:55.722Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "KT",
      "accuracy": 0.74,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.01977924,
      "avg_duration_per_task_ms": 63877.22,
      "avg_output_tokens_per_task": 11452.5,
      "last_run_timestamp": "2026-11-13T08:16:50.682Z"
    },
    {
      "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": 0.74,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.00321776,
      "avg_duration_per_task_ms": 67631.78,
      "avg_output_tokens_per_task": 11380.95,
      "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": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "Nebius",
      "accuracy": 0.74,
      "accuracy_std_dev": 0.026666999999999996,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.06914866,
      "avg_duration_per_task_ms": 152605.49,
      "avg_output_tokens_per_task": 14774.68,
      "last_run_timestamp": "2026-10-31T12:25:52.830Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "DeepInfra",
      "accuracy": 0.74,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.0873214,
      "avg_duration_per_task_ms": 347037.32,
      "avg_output_tokens_per_task": 8891.3,
      "last_run_timestamp": "2026-08-22T16:12:20.461Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Sail Research",
      "accuracy": 0.74,
      "accuracy_std_dev": 0.020000000000000018,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.049803139999999996,
      "avg_duration_per_task_ms": 363448.84,
      "avg_output_tokens_per_task": 8150.99,
      "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": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Reka",
      "accuracy": 0.74,
      "accuracy_std_dev": 0.019860653454128848,
      "run_count": 16,
      "total_questions": 800,
      "avg_cost_per_task_usd": 0.01624762875,
      "avg_duration_per_task_ms": 161764.35125,
      "avg_output_tokens_per_task": 4237.58,
      "last_run_timestamp": "2026-10-31T11:33:42.887Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Makora",
      "accuracy": 0.7399329261744967,
      "accuracy_std_dev": 0.017150616176167584,
      "run_count": 6,
      "total_questions": 298,
      "avg_cost_per_task_usd": 0.24555212751677855,
      "avg_duration_per_task_ms": 357614.23825503356,
      "avg_output_tokens_per_task": 12408.355704697986,
      "last_run_timestamp": "2026-10-31T12:24:39.056Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "Venice",
      "accuracy": 0.739687695652174,
      "accuracy_std_dev": 0.02325959803799046,
      "run_count": 6,
      "total_questions": 299,
      "avg_cost_per_task_usd": 0.050316361204013386,
      "avg_duration_per_task_ms": 654799.304347826,
      "avg_output_tokens_per_task": 13915.548494983277,
      "last_run_timestamp": "2026-10-24T19:40:41.765Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "StreamLake",
      "accuracy": 0.73958325,
      "accuracy_std_dev": 0.022756993736816403,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.02112607,
      "avg_duration_per_task_ms": 318544.7,
      "avg_output_tokens_per_task": 11340.385,
      "last_run_timestamp": "2026-10-17T11:09:59.070Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7395555333333333,
      "accuracy_std_dev": 0.016947170784004716,
      "run_count": 15,
      "total_questions": 750,
      "avg_cost_per_task_usd": 0.013297554666666666,
      "avg_duration_per_task_ms": 100947.12933333333,
      "avg_output_tokens_per_task": 10746.964,
      "last_run_timestamp": "2026-10-31T12:45:41.031Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Alibaba",
      "accuracy": 0.7394557102040816,
      "accuracy_std_dev": 0.01129971226890311,
      "run_count": 5,
      "total_questions": 245,
      "avg_cost_per_task_usd": 0.010599469387755103,
      "avg_duration_per_task_ms": 125314.78775510204,
      "avg_output_tokens_per_task": 12060.526530612246,
      "last_run_timestamp": "2026-10-24T19:21:18.651Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.5-sonnet-20250929",
      "model_slug": "anthropic/claude-sonnet-4.5",
      "model_name": "Anthropic: Claude Sonnet 4.5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.7393938181818183,
      "accuracy_std_dev": 0.027627475312607935,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.34960052363636374,
      "avg_duration_per_task_ms": 219975.07454545455,
      "avg_output_tokens_per_task": 10861.59090909091,
      "last_run_timestamp": "2026-10-31T12:46:18.490Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "InferenceNet",
      "accuracy": 0.7392308461538462,
      "accuracy_std_dev": 0.0224239080065011,
      "run_count": 13,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.017090104615384617,
      "avg_duration_per_task_ms": 299674.4630769231,
      "avg_output_tokens_per_task": 22554.26307692308,
      "last_run_timestamp": "2026-11-04T21:17:38.304Z"
    },
    {
      "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": 0.739166625,
      "accuracy_std_dev": 0.02757564218172218,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.020439862500000003,
      "avg_duration_per_task_ms": 430063.965,
      "avg_output_tokens_per_task": 11905.0125,
      "last_run_timestamp": "2026-10-17T11:29:07.053Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Relace",
      "accuracy": 0.738889,
      "accuracy_std_dev": 0.024695368688615833,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.01095846,
      "avg_duration_per_task_ms": 321436.35333333333,
      "avg_output_tokens_per_task": 22570.04,
      "last_run_timestamp": "2026-10-31T12:52:06.048Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.5-20260423",
      "model_slug": "openai/gpt-5.5",
      "model_name": "OpenAI: GPT-5.5",
      "provider_name": "Azure",
      "accuracy": 0.7388888333333333,
      "accuracy_std_dev": 0.026504076091654706,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.4952929133333333,
      "avg_duration_per_task_ms": 170578.00666666665,
      "avg_output_tokens_per_task": 7379.15,
      "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": "Decart",
      "accuracy": 0.7386668,
      "accuracy_std_dev": 0.03166503013357166,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.009092168000000001,
      "avg_duration_per_task_ms": 101935.708,
      "avg_output_tokens_per_task": 5027.116,
      "last_run_timestamp": "2026-10-17T11:17:02.460Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.7386666000000001,
      "accuracy_std_dev": 0.027616400924812788,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.077249096,
      "avg_duration_per_task_ms": 122620.824,
      "avg_output_tokens_per_task": 8779.518,
      "last_run_timestamp": "2026-10-31T20:18:12.373Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "auto-routing",
      "accuracy": 0.7385184444444445,
      "accuracy_std_dev": 0.028765060349555692,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.24696524444444445,
      "avg_duration_per_task_ms": 265213.1511111111,
      "avg_output_tokens_per_task": 10562.944444444445,
      "last_run_timestamp": "2026-10-31T11:38:53.393Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "inclusionai/ling-3.0-flash-20260723",
      "model_slug": "inclusionai/ling-3.0-flash",
      "model_name": "inclusionAI: Ling 3.0 Flash",
      "provider_name": "DeepInfra",
      "accuracy": 0.7383335,
      "accuracy_std_dev": 0.0016665000000000152,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.00751252,
      "avg_duration_per_task_ms": 320365.75,
      "avg_output_tokens_per_task": 15536.72,
      "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": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Morph",
      "accuracy": 0.7380000999999999,
      "accuracy_std_dev": 0.02844855708274146,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.016056571999999998,
      "avg_duration_per_task_ms": 534586.188,
      "avg_output_tokens_per_task": 17634.696,
      "last_run_timestamp": "2026-10-31T12:03:50.388Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.6-sonnet-20260217",
      "model_slug": "anthropic/claude-sonnet-4.6",
      "model_name": "Anthropic: Claude Sonnet 4.6",
      "provider_name": "Anthropic",
      "accuracy": 0.738,
      "accuracy_std_dev": 0.020232086728758366,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.345152742,
      "avg_duration_per_task_ms": 207631.326,
      "avg_output_tokens_per_task": 11819.48,
      "last_run_timestamp": "2026-10-31T13:36:32.496Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "DeepInfra",
      "accuracy": 0.7379031612903226,
      "accuracy_std_dev": 0.033855790905545226,
      "run_count": 5,
      "total_questions": 248,
      "avg_cost_per_task_usd": 0.06981228629032259,
      "avg_duration_per_task_ms": 509430.0443548387,
      "avg_output_tokens_per_task": 10842.524193548386,
      "last_run_timestamp": "2026-10-24T19:42:30.512Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Parasail",
      "accuracy": 0.7378277303370787,
      "accuracy_std_dev": 0.033381118228698506,
      "run_count": 11,
      "total_questions": 534,
      "avg_cost_per_task_usd": 0.025130365168539327,
      "avg_duration_per_task_ms": 93504.49625468165,
      "avg_output_tokens_per_task": 8311.610486891386,
      "last_run_timestamp": "2026-11-06T06:42:03.203Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7377779999999999,
      "accuracy_std_dev": 0.027933050471917086,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.02205162,
      "avg_duration_per_task_ms": 230910.26,
      "avg_output_tokens_per_task": 11297.92,
      "last_run_timestamp": "2026-10-31T12:27:40.071Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Cloudflare",
      "accuracy": 0.7377779999999999,
      "accuracy_std_dev": 0.020608136952831695,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.009694766666666667,
      "avg_duration_per_task_ms": 179217.14666666667,
      "avg_output_tokens_per_task": 12873.993333333334,
      "last_run_timestamp": "2026-10-31T11:06:39.132Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Parasail",
      "accuracy": 0.7377776666666667,
      "accuracy_std_dev": 0.020366893981273733,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.3511924911111111,
      "avg_duration_per_task_ms": 195777.59333333332,
      "avg_output_tokens_per_task": 9122.282222222222,
      "last_run_timestamp": "2026-10-31T11:28:52.482Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Makora",
      "accuracy": 0.7376126486486487,
      "accuracy_std_dev": 0.027927098584230564,
      "run_count": 6,
      "total_questions": 296,
      "avg_cost_per_task_usd": 0.011551577702702704,
      "avg_duration_per_task_ms": 61368.75337837838,
      "avg_output_tokens_per_task": 4262.564189189189,
      "last_run_timestamp": "2026-10-31T12:25:51.222Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "Novita",
      "accuracy": 0.73750025,
      "accuracy_std_dev": 0.047015028625828785,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0101418575,
      "avg_duration_per_task_ms": 462885.0975,
      "avg_output_tokens_per_task": 15094.72,
      "last_run_timestamp": "2026-10-17T11:26:35.790Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Morph",
      "accuracy": 0.737499875,
      "accuracy_std_dev": 0.022592425906913488,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.01732587,
      "avg_duration_per_task_ms": 193821.065,
      "avg_output_tokens_per_task": 4493.1475,
      "last_run_timestamp": "2026-10-31T12:42:00.537Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.7373335,
      "accuracy_std_dev": 0.029393808872788148,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.5392687700000001,
      "avg_duration_per_task_ms": 178582.538,
      "avg_output_tokens_per_task": 10153.252,
      "last_run_timestamp": "2026-10-31T12:28:38.525Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Ionstream",
      "accuracy": 0.7373334,
      "accuracy_std_dev": 0.020482933819157834,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.039714615999999994,
      "avg_duration_per_task_ms": 174773.632,
      "avg_output_tokens_per_task": 27822.08,
      "last_run_timestamp": "2026-10-21T12:36:51.217Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "auto-routing",
      "accuracy": 0.737143,
      "accuracy_std_dev": 0.024907972114508722,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.04305575428571429,
      "avg_duration_per_task_ms": 613136.2114285714,
      "avg_output_tokens_per_task": 16198.417142857143,
      "last_run_timestamp": "2026-10-31T19:47:52.982Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7371428571428571,
      "accuracy_std_dev": 0.03452293332359381,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.05083306285714286,
      "avg_duration_per_task_ms": 397117.17714285717,
      "avg_output_tokens_per_task": 11318.654285714285,
      "last_run_timestamp": "2026-10-31T21:59:10.427Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "GMICloud",
      "accuracy": 0.7371427142857143,
      "accuracy_std_dev": 0.04561487206325298,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.006840197142857143,
      "avg_duration_per_task_ms": 433214.3028571429,
      "avg_output_tokens_per_task": 14593.82,
      "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-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "StreamLake",
      "accuracy": 0.7370185075376885,
      "accuracy_std_dev": 0.03341638611800355,
      "run_count": 8,
      "total_questions": 398,
      "avg_cost_per_task_usd": 0.004073236180904522,
      "avg_duration_per_task_ms": 239891.2864321608,
      "avg_output_tokens_per_task": 12177.208542713568,
      "last_run_timestamp": "2026-10-31T19:49:31.140Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Inceptron",
      "accuracy": 0.7367941894353369,
      "accuracy_std_dev": 0.02953072467422912,
      "run_count": 11,
      "total_questions": 549,
      "avg_cost_per_task_usd": 0.01213822586520947,
      "avg_duration_per_task_ms": 405302.62841530057,
      "avg_output_tokens_per_task": 14294.963570127504,
      "last_run_timestamp": "2026-10-31T11:58:45.245Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "StreamLake",
      "accuracy": 0.736667,
      "accuracy_std_dev": 0.010000000000000009,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.02084315,
      "avg_duration_per_task_ms": 478625.81,
      "avg_output_tokens_per_task": 14704.35,
      "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": "openai/gpt-5.5-20260423",
      "model_slug": "openai/gpt-5.5",
      "model_name": "OpenAI: GPT-5.5",
      "provider_name": "auto-routing",
      "accuracy": 0.736667,
      "accuracy_std_dev": 0.010000000000000009,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.47698523000000004,
      "avg_duration_per_task_ms": 128774.34,
      "avg_output_tokens_per_task": 7206.5,
      "last_run_timestamp": "2026-10-01T19:04:40.362Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Together",
      "accuracy": 0.7366666000000001,
      "accuracy_std_dev": 0.027039148574613085,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.29887413199999996,
      "avg_duration_per_task_ms": 294262.764,
      "avg_output_tokens_per_task": 9417.472,
      "last_run_timestamp": "2026-10-31T13:04:39.070Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "DeepInfra",
      "accuracy": 0.7366665,
      "accuracy_std_dev": 0.036666500000000046,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.03538476,
      "avg_duration_per_task_ms": 350978.71,
      "avg_output_tokens_per_task": 8276.33,
      "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": "reflection/beam-1.0-501b-a23b-20261016",
      "model_slug": "reflection/beam-1.0-501b-a23b",
      "model_name": "Reflection: Beam 1.0 501B A23B",
      "provider_name": "Io Net",
      "accuracy": 0.7366665,
      "accuracy_std_dev": 0.0033334999999999892,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.02539905,
      "avg_duration_per_task_ms": 216631.23,
      "avg_output_tokens_per_task": 14336.27,
      "last_run_timestamp": "2026-10-24T12:02:36.303Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7365900689655173,
      "accuracy_std_dev": 0.040120795309227075,
      "run_count": 7,
      "total_questions": 348,
      "avg_cost_per_task_usd": 0.07452031609195402,
      "avg_duration_per_task_ms": 904419.2068965518,
      "avg_output_tokens_per_task": 14631.514367816091,
      "last_run_timestamp": "2026-10-24T20:35:30.856Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Phala",
      "accuracy": 0.7361905714285714,
      "accuracy_std_dev": 0.03340122609750018,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.14018482571428573,
      "avg_duration_per_task_ms": 334656.22,
      "avg_output_tokens_per_task": 11876.488571428572,
      "last_run_timestamp": "2026-10-31T21:07:15.251Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7361904285714286,
      "accuracy_std_dev": 0.03649010551547657,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.07199701142857143,
      "avg_duration_per_task_ms": 222190.2542857143,
      "avg_output_tokens_per_task": 9529.64,
      "last_run_timestamp": "2026-10-31T19:53:34.188Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "AtlasCloud",
      "accuracy": 0.736111045045045,
      "accuracy_std_dev": 0.04484669172680976,
      "run_count": 9,
      "total_questions": 444,
      "avg_cost_per_task_usd": 0.02811942117117117,
      "avg_duration_per_task_ms": 111185.40990990991,
      "avg_output_tokens_per_task": 10611.61936936937,
      "last_run_timestamp": "2026-10-31T19:48:53.852Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.736,
      "accuracy_std_dev": 0.011623577573191504,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.02217724,
      "avg_duration_per_task_ms": 74896.184,
      "avg_output_tokens_per_task": 4355.024,
      "last_run_timestamp": "2026-10-31T12:05:37.260Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.6-sonnet-20260217",
      "model_slug": "anthropic/claude-sonnet-4.6",
      "model_name": "Anthropic: Claude Sonnet 4.6",
      "provider_name": "Google",
      "accuracy": 0.7359999666666666,
      "accuracy_std_dev": 0.028627806219924643,
      "run_count": 30,
      "total_questions": 1500,
      "avg_cost_per_task_usd": 0.3641733786666666,
      "avg_duration_per_task_ms": 196840.98666666666,
      "avg_output_tokens_per_task": 11882.770666666667,
      "last_run_timestamp": "2026-10-31T13:17:20.726Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "Google",
      "accuracy": 0.7359999333333334,
      "accuracy_std_dev": 0.0231964569463145,
      "run_count": 15,
      "total_questions": 750,
      "avg_cost_per_task_usd": 0.013030309333333335,
      "avg_duration_per_task_ms": 73246.75466666666,
      "avg_output_tokens_per_task": 10779.536,
      "last_run_timestamp": "2026-10-31T11:42:11.484Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Moonshot AI",
      "accuracy": 0.7359360302267003,
      "accuracy_std_dev": 0.023828191210399825,
      "run_count": 8,
      "total_questions": 397,
      "avg_cost_per_task_usd": 0.28617903778337533,
      "avg_duration_per_task_ms": 407022.72040302266,
      "avg_output_tokens_per_task": 9678.146095717884,
      "last_run_timestamp": "2026-10-24T12:29:09.762Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Makora",
      "accuracy": 0.7359261111111111,
      "accuracy_std_dev": 0.02318882707227026,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.04132428666666666,
      "avg_duration_per_task_ms": 135748.28,
      "avg_output_tokens_per_task": 25779.64,
      "last_run_timestamp": "2026-10-31T12:16:10.183Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.735833375,
      "accuracy_std_dev": 0.03463102033559473,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.030614205000000002,
      "avg_duration_per_task_ms": 479244.2275,
      "avg_output_tokens_per_task": 15408.6125,
      "last_run_timestamp": "2026-10-17T13:55:17.194Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "Xiaomi",
      "accuracy": 0.73583325,
      "accuracy_std_dev": 0.025590284073012965,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0082457875,
      "avg_duration_per_task_ms": 453019.37,
      "avg_output_tokens_per_task": 14752.6375,
      "last_run_timestamp": "2026-10-17T12:09:02.862Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "NextBit",
      "accuracy": 0.7355556666666666,
      "accuracy_std_dev": 0.01369867976201439,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.01667712,
      "avg_duration_per_task_ms": 656503.2166666667,
      "avg_output_tokens_per_task": 11802.333333333334,
      "last_run_timestamp": "2026-11-03T00:15:40.640Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.5-20260423",
      "model_slug": "openai/gpt-5.5",
      "model_name": "OpenAI: GPT-5.5",
      "provider_name": "OpenAI",
      "accuracy": 0.7355556666666666,
      "accuracy_std_dev": 0.04157400066013479,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.48569376,
      "avg_duration_per_task_ms": 125885.67333333334,
      "avg_output_tokens_per_task": 7296.606666666667,
      "last_run_timestamp": "2026-10-02T23:08:08.058Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Relace",
      "accuracy": 0.7352940588235294,
      "accuracy_std_dev": 0.03872239054432369,
      "run_count": 17,
      "total_questions": 850,
      "avg_cost_per_task_usd": 0.009413557647058823,
      "avg_duration_per_task_ms": 215238.48941176472,
      "avg_output_tokens_per_task": 15440.829411764706,
      "last_run_timestamp": "2026-11-02T18:06:38.075Z"
    },
    {
      "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": 0.7352381428571428,
      "accuracy_std_dev": 0.029106006893170788,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.051658899999999994,
      "avg_duration_per_task_ms": 348562.78857142857,
      "avg_output_tokens_per_task": 9821.065714285714,
      "last_run_timestamp": "2026-10-31T21:01:26.628Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Fireworks",
      "accuracy": 0.7349398099062917,
      "accuracy_std_dev": 0.02455646627618342,
      "run_count": 15,
      "total_questions": 747,
      "avg_cost_per_task_usd": 0.05549164792503348,
      "avg_duration_per_task_ms": 157974.1485943775,
      "avg_output_tokens_per_task": 30263.555555555555,
      "last_run_timestamp": "2026-10-31T12:41:59.422Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Phala",
      "accuracy": 0.7349397911646587,
      "accuracy_std_dev": 0.047428285455095254,
      "run_count": 10,
      "total_questions": 498,
      "avg_cost_per_task_usd": 0.2969604819277108,
      "avg_duration_per_task_ms": 386565.9397590361,
      "avg_output_tokens_per_task": 11351.530120481928,
      "last_run_timestamp": "2026-10-31T12:42:56.001Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.7346668,
      "accuracy_std_dev": 0.03066679130525396,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.029810941999999993,
      "avg_duration_per_task_ms": 101760.856,
      "avg_output_tokens_per_task": 10584.172,
      "last_run_timestamp": "2026-10-31T19:09:49.337Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7344446666666666,
      "accuracy_std_dev": 0.03183677894965017,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.07159142333333333,
      "avg_duration_per_task_ms": 344988.6066666667,
      "avg_output_tokens_per_task": 11458.973333333333,
      "last_run_timestamp": "2026-10-24T22:27:21.158Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7344443333333334,
      "accuracy_std_dev": 0.019404823126108514,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.056632510000000004,
      "avg_duration_per_task_ms": 908364.2133333334,
      "avg_output_tokens_per_task": 15944.676666666666,
      "last_run_timestamp": "2026-10-31T20:02:53.411Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "CoreWeave",
      "accuracy": 0.7342859999999999,
      "accuracy_std_dev": 0.027701017933849065,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.08147224571428571,
      "avg_duration_per_task_ms": 77662.75714285714,
      "avg_output_tokens_per_task": 11731.842857142858,
      "last_run_timestamp": "2026-10-31T19:56:16.658Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Novita",
      "accuracy": 0.7342857142857143,
      "accuracy_std_dev": 0.03331958219903675,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.09951040285714285,
      "avg_duration_per_task_ms": 257635.84285714285,
      "avg_output_tokens_per_task": 12551.72,
      "last_run_timestamp": "2026-10-31T20:07:04.555Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.05052255713253118,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.009341564444444445,
      "avg_duration_per_task_ms": 247292.54444444444,
      "avg_output_tokens_per_task": 15978.137777777778,
      "last_run_timestamp": "2026-10-31T11:40:30.784Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Inceptron",
      "accuracy": 0.7338094285714286,
      "accuracy_std_dev": 0.034752133656416974,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.06415493714285714,
      "avg_duration_per_task_ms": 159387.71142857143,
      "avg_output_tokens_per_task": 11195.86,
      "last_run_timestamp": "2026-10-31T19:24:16.029Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "Novita",
      "accuracy": 0.733796,
      "accuracy_std_dev": 0.0059017228557995535,
      "run_count": 3,
      "total_questions": 144,
      "avg_cost_per_task_usd": 0.20467558333333333,
      "avg_duration_per_task_ms": 370954.7361111111,
      "avg_output_tokens_per_task": 23212.9375,
      "last_run_timestamp": "2026-10-24T19:18:46.938Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Mistral",
      "accuracy": 0.73375,
      "accuracy_std_dev": 0.027810549558575794,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.018745945,
      "avg_duration_per_task_ms": 93462.3375,
      "avg_output_tokens_per_task": 4950.955,
      "last_run_timestamp": "2026-11-11T03:16:45.253Z"
    },
    {
      "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": 0.7336684371859297,
      "accuracy_std_dev": 0.0449684893006689,
      "run_count": 8,
      "total_questions": 398,
      "avg_cost_per_task_usd": 0.033184525125628146,
      "avg_duration_per_task_ms": 1207877.3844221106,
      "avg_output_tokens_per_task": 52852.572864321606,
      "last_run_timestamp": "2026-10-31T20:03:31.183Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.1-terminus",
      "model_slug": "deepseek/deepseek-v3.1-terminus",
      "model_name": "DeepSeek: DeepSeek V3.1 Terminus",
      "provider_name": "Novita",
      "accuracy": 0.7333335,
      "accuracy_std_dev": 0.026666499999999982,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.06130771,
      "avg_duration_per_task_ms": 487797.15,
      "avg_output_tokens_per_task": 14346.49,
      "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": "google/gemini-3-flash-preview-20251217",
      "model_slug": "google/gemini-3-flash-preview",
      "model_name": "Google: Gemini 3 Flash Preview",
      "provider_name": "Google",
      "accuracy": 0.7333335,
      "accuracy_std_dev": 0.00666650000000002,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.12301775,
      "avg_duration_per_task_ms": 155307.55,
      "avg_output_tokens_per_task": 18812.75,
      "last_run_timestamp": "2026-10-01T19:05:47.885Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta/muse-spark-open-20261015",
      "model_slug": "meta/muse-spark-open",
      "model_name": "Meta: Muse Spark Open",
      "provider_name": "Together",
      "accuracy": 0.7333335,
      "accuracy_std_dev": 0.00666650000000002,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.08901925000000001,
      "avg_duration_per_task_ms": 93612.32,
      "avg_output_tokens_per_task": 8634.62,
      "last_run_timestamp": "2026-10-31T11:29:29.503Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Phala",
      "accuracy": 0.7333335,
      "accuracy_std_dev": 0.013333499999999998,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.08394800999999999,
      "avg_duration_per_task_ms": 134148.76,
      "avg_output_tokens_per_task": 8723.94,
      "last_run_timestamp": "2026-10-17T19:55:39.408Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-sol-20260709",
      "model_slug": "openai/gpt-5.6-sol",
      "model_name": "OpenAI: GPT-5.6 Sol",
      "provider_name": "auto-routing",
      "accuracy": 0.7333335,
      "accuracy_std_dev": 0.026666499999999982,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.12687428,
      "avg_duration_per_task_ms": 110480.49,
      "avg_output_tokens_per_task": 5152.54,
      "last_run_timestamp": "2026-10-01T19:04:55.696Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7333334285714286,
      "accuracy_std_dev": 0.03342828381500203,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.10974002285714286,
      "avg_duration_per_task_ms": 171262.26571428572,
      "avg_output_tokens_per_task": 9719.968571428572,
      "last_run_timestamp": "2026-10-31T20:27:27.677Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7333334,
      "accuracy_std_dev": 0.03451234883690184,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.30572972400000004,
      "avg_duration_per_task_ms": 282521.796,
      "avg_output_tokens_per_task": 10813.856,
      "last_run_timestamp": "2026-10-31T12:51:11.835Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "StreamLake",
      "accuracy": 0.733333375,
      "accuracy_std_dev": 0.03109126351138492,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.06630269500000001,
      "avg_duration_per_task_ms": 258633.95,
      "avg_output_tokens_per_task": 9560.47,
      "last_run_timestamp": "2026-10-31T20:31:43.469Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Mancer 2",
      "accuracy": 0.7333333333333333,
      "accuracy_std_dev": 0.014401903168987377,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.06451945333333334,
      "avg_duration_per_task_ms": 876085.0533333333,
      "avg_output_tokens_per_task": 15144.64,
      "last_run_timestamp": "2026-10-31T12:31:12.958Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Crusoe",
      "accuracy": 0.7333333333333333,
      "accuracy_std_dev": 0.03569390445004426,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.021479246666666667,
      "avg_duration_per_task_ms": 120781.28,
      "avg_output_tokens_per_task": 11581.666666666666,
      "last_run_timestamp": "2026-10-31T11:22:01.481Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7333333333333333,
      "accuracy_std_dev": 0.06834903523337645,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.15942455777777778,
      "avg_duration_per_task_ms": 1424663.4244444445,
      "avg_output_tokens_per_task": 28406.333333333332,
      "last_run_timestamp": "2026-10-24T20:45:53.033Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "GMICloud",
      "accuracy": 0.7333333333333333,
      "accuracy_std_dev": 0.035136551965154465,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.04706574222222222,
      "avg_duration_per_task_ms": 216014.41555555555,
      "avg_output_tokens_per_task": 10077.462222222222,
      "last_run_timestamp": "2026-11-10T07:12:51.901Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "InferenceNet",
      "accuracy": 0.7333333333333333,
      "accuracy_std_dev": 0.05760637824832321,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.04898614,
      "avg_duration_per_task_ms": 106829.63333333333,
      "avg_output_tokens_per_task": 9897.326666666666,
      "last_run_timestamp": "2026-10-10T11:51:48.927Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7333333333333333,
      "accuracy_std_dev": 0.024944382578492966,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.09260214,
      "avg_duration_per_task_ms": 326558.9733333333,
      "avg_output_tokens_per_task": 4729.153333333334,
      "last_run_timestamp": "2026-10-31T12:28:59.426Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Cohere",
      "accuracy": 0.7333333333333333,
      "accuracy_std_dev": 0.009428090415820642,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.024898453333333334,
      "avg_duration_per_task_ms": 37126.9,
      "avg_output_tokens_per_task": 4153.473333333333,
      "last_run_timestamp": "2026-10-31T12:07:50.460Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Relace",
      "accuracy": 0.73333325,
      "accuracy_std_dev": 0.0368176738970226,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.013768695000000003,
      "avg_duration_per_task_ms": 81021.62,
      "avg_output_tokens_per_task": 13622.935,
      "last_run_timestamp": "2026-11-01T12:08:57.072Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "Anthropic",
      "accuracy": 0.7333332,
      "accuracy_std_dev": 0.032659904062320846,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.012294704000000002,
      "avg_duration_per_task_ms": 85675.876,
      "avg_output_tokens_per_task": 10698.828,
      "last_run_timestamp": "2026-10-31T11:35:21.238Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "NextBit",
      "accuracy": 0.7333330000000001,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01186947,
      "avg_duration_per_task_ms": 176514.74,
      "avg_output_tokens_per_task": 14201.61,
      "last_run_timestamp": "2026-08-29T16:50:07.898Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Decart",
      "accuracy": 0.7333330000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.00602072,
      "avg_duration_per_task_ms": 517268.78,
      "avg_output_tokens_per_task": 18678.62,
      "last_run_timestamp": "2026-08-22T08:38:00.859Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7333330000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.058005,
      "avg_duration_per_task_ms": 191243.62,
      "avg_output_tokens_per_task": 30812.32,
      "last_run_timestamp": "2026-09-12T08:09:12.993Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "DekaLLM",
      "accuracy": 0.7333330000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.01115704,
      "avg_duration_per_task_ms": 102700.12,
      "avg_output_tokens_per_task": 12077.2,
      "last_run_timestamp": "2026-11-04T09:13:50.416Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3-flash-preview-20251217",
      "model_slug": "google/gemini-3-flash-preview",
      "model_name": "Google: Gemini 3 Flash Preview",
      "provider_name": "auto-routing",
      "accuracy": 0.7333330000000001,
      "accuracy_std_dev": 0.020000000000000018,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.12365041999999998,
      "avg_duration_per_task_ms": 156560.04,
      "avg_output_tokens_per_task": 19704.13,
      "last_run_timestamp": "2026-10-01T19:04:26.702Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.5-flash-lite-20260721",
      "model_slug": "google/gemini-3.5-flash-lite",
      "model_name": "Google: Gemini 3.5 Flash Lite",
      "provider_name": "Google AI Studio",
      "accuracy": 0.7333330000000001,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.10928178,
      "avg_duration_per_task_ms": 75479.44,
      "avg_output_tokens_per_task": 18824.05,
      "last_run_timestamp": "2026-10-01T19:05:46.177Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3.1-pro-20261023",
      "model_slug": "moonshotai/kimi-k3.1-pro",
      "model_name": "MoonshotAI: Kimi K3.1 Pro",
      "provider_name": "Parasail",
      "accuracy": 0.7333330000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.2484202,
      "avg_duration_per_task_ms": 207733.62,
      "avg_output_tokens_per_task": 10608.24,
      "last_run_timestamp": "2026-11-04T21:15:40.182Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3.1-pro-20261023",
      "model_slug": "moonshotai/kimi-k3.1-pro",
      "model_name": "MoonshotAI: Kimi K3.1 Pro",
      "provider_name": "Together",
      "accuracy": 0.7333330000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.2865894,
      "avg_duration_per_task_ms": 249378.06,
      "avg_output_tokens_per_task": 16191.68,
      "last_run_timestamp": "2026-11-04T18:44:37.678Z"
    },
    {
      "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": 0.7333330000000001,
      "accuracy_std_dev": 0.020000000000000018,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.055889249999999994,
      "avg_duration_per_task_ms": 201945.25,
      "avg_output_tokens_per_task": 9141.8,
      "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-flashx-20260918",
      "model_slug": "z-ai/glm-5.3-flashx",
      "model_name": "Z.ai: GLM 5.3 FlashX",
      "provider_name": "Novita",
      "accuracy": 0.7333330000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.027065839999999997,
      "avg_duration_per_task_ms": 104408.46,
      "avg_output_tokens_per_task": 4213.84,
      "last_run_timestamp": "2026-11-12T21:20:43.637Z"
    },
    {
      "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": 0.7330556666666667,
      "accuracy_std_dev": 0.016468863254301695,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.1880071,
      "avg_duration_per_task_ms": 216665.43166666667,
      "avg_output_tokens_per_task": 9600.853333333333,
      "last_run_timestamp": "2026-11-06T22:35:15.892Z"
    },
    {
      "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": 0.733019156654888,
      "accuracy_std_dev": 0.02874817787961395,
      "run_count": 17,
      "total_questions": 849,
      "avg_cost_per_task_usd": 0.12840728386336864,
      "avg_duration_per_task_ms": 141161.32508833922,
      "avg_output_tokens_per_task": 9290.04593639576,
      "last_run_timestamp": "2026-10-24T19:48:18.775Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Tenstorrent",
      "accuracy": 0.7329929795918368,
      "accuracy_std_dev": 0.05409699999999995,
      "run_count": 2,
      "total_questions": 98,
      "avg_cost_per_task_usd": 0.003721785714285715,
      "avg_duration_per_task_ms": 82527.76530612246,
      "avg_output_tokens_per_task": 3412.2551020408164,
      "last_run_timestamp": "2026-11-08T04:25:13.787Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Modal",
      "accuracy": 0.732916625,
      "accuracy_std_dev": 0.018888392049996622,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0482453475,
      "avg_duration_per_task_ms": 162563.185,
      "avg_output_tokens_per_task": 29218.5525,
      "last_run_timestamp": "2026-10-31T12:25:59.335Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Inceptron",
      "accuracy": 0.7325939704251386,
      "accuracy_std_dev": 0.03159346942038068,
      "run_count": 11,
      "total_questions": 541,
      "avg_cost_per_task_usd": 0.06591540850277265,
      "avg_duration_per_task_ms": 143626.1885397412,
      "avg_output_tokens_per_task": 8437.54898336414,
      "last_run_timestamp": "2026-10-31T19:10:53.479Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.732381,
      "accuracy_std_dev": 0.016106167044245587,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.11714905142857145,
      "avg_duration_per_task_ms": 334799.0685714286,
      "avg_output_tokens_per_task": 9713.528571428571,
      "last_run_timestamp": "2026-10-31T20:34:53.249Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "Parasail",
      "accuracy": 0.7321772312138729,
      "accuracy_std_dev": 0.08403648462496778,
      "run_count": 7,
      "total_questions": 346,
      "avg_cost_per_task_usd": 0.1502556156069364,
      "avg_duration_per_task_ms": 2018284.2369942197,
      "avg_output_tokens_per_task": 112906.43641618497,
      "last_run_timestamp": "2026-10-31T19:11:32.112Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "CoreWeave",
      "accuracy": 0.732,
      "accuracy_std_dev": 0.02124973202654565,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.036015575999999994,
      "avg_duration_per_task_ms": 148623.836,
      "avg_output_tokens_per_task": 32577.608,
      "last_run_timestamp": "2026-10-31T12:40:24.644Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "auto-routing",
      "accuracy": 0.73166675,
      "accuracy_std_dev": 0.04146614750537478,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.010049615000000001,
      "avg_duration_per_task_ms": 411186.61,
      "avg_output_tokens_per_task": 14357.37,
      "last_run_timestamp": "2026-10-17T13:13:33.362Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Crusoe",
      "accuracy": 0.7316665,
      "accuracy_std_dev": 0.018333500000000003,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01274185,
      "avg_duration_per_task_ms": 154997.2,
      "avg_output_tokens_per_task": 13696.96,
      "last_run_timestamp": "2026-10-31T19:54:14.421Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "auto-routing",
      "accuracy": 0.7314284285714286,
      "accuracy_std_dev": 0.019424805892445475,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.06280418,
      "avg_duration_per_task_ms": 232822.65142857144,
      "avg_output_tokens_per_task": 12506.968571428572,
      "last_run_timestamp": "2026-10-31T20:18:31.003Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "Azure",
      "accuracy": 0.7313332,
      "accuracy_std_dev": 0.020231941744676914,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.012560144,
      "avg_duration_per_task_ms": 74716.332,
      "avg_output_tokens_per_task": 10525.228,
      "last_run_timestamp": "2026-10-31T11:19:23.393Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7311827782258065,
      "accuracy_std_dev": 0.07837008853695393,
      "run_count": 10,
      "total_questions": 496,
      "avg_cost_per_task_usd": 0.009883709677419354,
      "avg_duration_per_task_ms": 170348.1431451613,
      "avg_output_tokens_per_task": 4275.5161290322585,
      "last_run_timestamp": "2026-10-31T12:31:51.998Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Wafer",
      "accuracy": 0.7311637573964497,
      "accuracy_std_dev": 0.02813377440238986,
      "run_count": 17,
      "total_questions": 845,
      "avg_cost_per_task_usd": 0.024811487573964494,
      "avg_duration_per_task_ms": 209572.125443787,
      "avg_output_tokens_per_task": 21468.289940828403,
      "last_run_timestamp": "2026-10-31T11:27:16.882Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "auto-routing",
      "accuracy": 0.7311110000000001,
      "accuracy_std_dev": 0.029731211518088756,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.012172700000000002,
      "avg_duration_per_task_ms": 217797.45666666667,
      "avg_output_tokens_per_task": 14731.27,
      "last_run_timestamp": "2026-10-31T19:46:05.353Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "OpenInference",
      "accuracy": 0.7311110000000001,
      "accuracy_std_dev": 0.020608136952831695,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.028709900000000003,
      "avg_duration_per_task_ms": 638312.6933333334,
      "avg_output_tokens_per_task": 61738.18,
      "last_run_timestamp": "2026-10-31T12:28:49.803Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.4-20260305",
      "model_slug": "openai/gpt-5.4",
      "model_name": "OpenAI: GPT-5.4",
      "provider_name": "Azure",
      "accuracy": 0.7311110000000001,
      "accuracy_std_dev": 0.014740588127118051,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.29708465,
      "avg_duration_per_task_ms": 186510.67,
      "avg_output_tokens_per_task": 10592.956666666667,
      "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": "openai/gpt-5.6-sol-20260709",
      "model_slug": "openai/gpt-5.6-sol",
      "model_name": "OpenAI: GPT-5.6 Sol",
      "provider_name": "OpenAI",
      "accuracy": 0.7311110000000001,
      "accuracy_std_dev": 0.013698950130089036,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.12928292,
      "avg_duration_per_task_ms": 109953.54,
      "avg_output_tokens_per_task": 5131.753333333333,
      "last_run_timestamp": "2026-10-02T23:08:37.533Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "Xiaomi",
      "accuracy": 0.730833375,
      "accuracy_std_dev": 0.028074913427905263,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.025226807499999997,
      "avg_duration_per_task_ms": 435604.5225,
      "avg_output_tokens_per_task": 16317.23,
      "last_run_timestamp": "2026-10-17T11:15:55.168Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Crusoe",
      "accuracy": 0.7304761428571429,
      "accuracy_std_dev": 0.032043090320150264,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.09725585142857143,
      "avg_duration_per_task_ms": 317413.97142857144,
      "avg_output_tokens_per_task": 11001.088571428572,
      "last_run_timestamp": "2026-10-31T20:47:51.733Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "GMICloud",
      "accuracy": 0.7303704444444444,
      "accuracy_std_dev": 0.04158720192054578,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.028855546666666666,
      "avg_duration_per_task_ms": 158061.0911111111,
      "avg_output_tokens_per_task": 13689.44,
      "last_run_timestamp": "2026-11-10T06:43:24.233Z"
    },
    {
      "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": 0.7303333,
      "accuracy_std_dev": 0.025427430945929256,
      "run_count": 20,
      "total_questions": 1000,
      "avg_cost_per_task_usd": 0.3636253240000001,
      "avg_duration_per_task_ms": 225306.613,
      "avg_output_tokens_per_task": 10854.177,
      "last_run_timestamp": "2026-10-31T12:43:31.737Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Reka",
      "accuracy": 0.7300002499999999,
      "accuracy_std_dev": 0.028867484592314224,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.03407485,
      "avg_duration_per_task_ms": 264292.92,
      "avg_output_tokens_per_task": 34254.08,
      "last_run_timestamp": "2026-10-31T13:21:05.335Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "Novita",
      "accuracy": 0.73,
      "accuracy_std_dev": 0.016666999999999987,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.06303523999999999,
      "avg_duration_per_task_ms": 517277.01,
      "avg_output_tokens_per_task": 14652.21,
      "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": "deepseek/deepseek-v3.1-terminus",
      "model_slug": "deepseek/deepseek-v3.1-terminus",
      "model_name": "DeepSeek: DeepSeek V3.1 Terminus",
      "provider_name": "auto-routing",
      "accuracy": 0.73,
      "accuracy_std_dev": 0.0033330000000000304,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.06290177,
      "avg_duration_per_task_ms": 485845.77,
      "avg_output_tokens_per_task": 14483.09,
      "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": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Nebius",
      "accuracy": 0.73,
      "accuracy_std_dev": 0.0268328157299975,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.074570136,
      "avg_duration_per_task_ms": 246781.984,
      "avg_output_tokens_per_task": 30037.412,
      "last_run_timestamp": "2026-10-31T11:52:51.154Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3-flash-preview-20251217",
      "model_slug": "google/gemini-3-flash-preview",
      "model_name": "Google: Gemini 3 Flash Preview",
      "provider_name": "Google AI Studio",
      "accuracy": 0.73,
      "accuracy_std_dev": 0.030000000000000027,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.12856867000000002,
      "avg_duration_per_task_ms": 121477.86,
      "avg_output_tokens_per_task": 19799.68,
      "last_run_timestamp": "2026-10-01T19:05:00.104Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.6-flash-20260721",
      "model_slug": "google/gemini-3.6-flash",
      "model_name": "Google: Gemini 3.6 Flash",
      "provider_name": "Google AI Studio",
      "accuracy": 0.73,
      "accuracy_std_dev": 0.04333300000000001,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.24468226999999998,
      "avg_duration_per_task_ms": 158524.06,
      "avg_output_tokens_per_task": 22552.35,
      "last_run_timestamp": "2026-10-01T19:04:55.721Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3.1-pro-20261023",
      "model_slug": "moonshotai/kimi-k3.1-pro",
      "model_name": "MoonshotAI: Kimi K3.1 Pro",
      "provider_name": "Sail Research",
      "accuracy": 0.73,
      "accuracy_std_dev": 0.0033330000000000304,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.22321662,
      "avg_duration_per_task_ms": 277142.38,
      "avg_output_tokens_per_task": 10411.9,
      "last_run_timestamp": "2026-11-06T22:14:35.733Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3.1-pro-20261023",
      "model_slug": "moonshotai/kimi-k3.1-pro",
      "model_name": "MoonshotAI: Kimi K3.1 Pro",
      "provider_name": "Modal",
      "accuracy": 0.73,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.14279748,
      "avg_duration_per_task_ms": 129435.32,
      "avg_output_tokens_per_task": 7981.12,
      "last_run_timestamp": "2026-11-04T21:11:46.281Z"
    },
    {
      "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": 0.73,
      "accuracy_std_dev": 0.036667000000000005,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.06843740999999999,
      "avg_duration_per_task_ms": 1343672.46,
      "avg_output_tokens_per_task": 8825.5,
      "last_run_timestamp": "2026-08-29T16:20:58.762Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "KT",
      "accuracy": 0.73,
      "accuracy_std_dev": 0.0033330000000000304,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01420625,
      "avg_duration_per_task_ms": 63360.51,
      "avg_output_tokens_per_task": 4011.33,
      "last_run_timestamp": "2026-11-11T02:06:30.445Z"
    },
    {
      "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": 0.7299997500000001,
      "accuracy_std_dev": 0.017320556188745798,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.32059346000000005,
      "avg_duration_per_task_ms": 156347.09,
      "avg_output_tokens_per_task": 21297.645,
      "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": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Nebius",
      "accuracy": 0.729654525083612,
      "accuracy_std_dev": 0.01804542249436126,
      "run_count": 6,
      "total_questions": 299,
      "avg_cost_per_task_usd": 0.18312612040133777,
      "avg_duration_per_task_ms": 543481.8829431438,
      "avg_output_tokens_per_task": 8776.65551839465,
      "last_run_timestamp": "2026-10-31T21:08:40.848Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.7293334,
      "accuracy_std_dev": 0.026533159130416407,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.012403696,
      "avg_duration_per_task_ms": 92314.648,
      "avg_output_tokens_per_task": 10764.472,
      "last_run_timestamp": "2026-10-31T12:41:01.831Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Together",
      "accuracy": 0.7291665,
      "accuracy_std_dev": 0.023318429765959808,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0316769325,
      "avg_duration_per_task_ms": 89151.6375,
      "avg_output_tokens_per_task": 22436.51,
      "last_run_timestamp": "2026-10-31T11:37:01.556Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "ModelRun",
      "accuracy": 0.7288890000000001,
      "accuracy_std_dev": 0.03142665378305492,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.02386675333333333,
      "avg_duration_per_task_ms": 151228.52,
      "avg_output_tokens_per_task": 22554.02666666667,
      "last_run_timestamp": "2026-10-28T06:40:49.113Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Novita",
      "accuracy": 0.728889,
      "accuracy_std_dev": 0.01911599421078241,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.017266966666666664,
      "avg_duration_per_task_ms": 122400.14,
      "avg_output_tokens_per_task": 11879.513333333334,
      "last_run_timestamp": "2026-10-31T11:33:22.952Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "Together",
      "accuracy": 0.728889,
      "accuracy_std_dev": 0.01257094435593444,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.04855964666666667,
      "avg_duration_per_task_ms": 115027.81333333334,
      "avg_output_tokens_per_task": 15070.966666666667,
      "last_run_timestamp": "2026-10-31T12:30:34.970Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "auto-routing",
      "accuracy": 0.7288886666666667,
      "accuracy_std_dev": 0.017497738831695454,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.033088113333333335,
      "avg_duration_per_task_ms": 134354.10666666666,
      "avg_output_tokens_per_task": 14586.926666666666,
      "last_run_timestamp": "2026-10-31T12:32:08.127Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.7283335,
      "accuracy_std_dev": 0.024626886091898288,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.04604961333333333,
      "avg_duration_per_task_ms": 118630.39,
      "avg_output_tokens_per_task": 26239.556666666667,
      "last_run_timestamp": "2026-10-31T12:19:28.999Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Parasail",
      "accuracy": 0.7283333333333334,
      "accuracy_std_dev": 0.03857423338908441,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.07535822,
      "avg_duration_per_task_ms": 168347.59666666668,
      "avg_output_tokens_per_task": 11190.87,
      "last_run_timestamp": "2026-10-31T20:16:51.242Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "auto-routing",
      "accuracy": 0.72833325,
      "accuracy_std_dev": 0.041197373838480274,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.011627464999999998,
      "avg_duration_per_task_ms": 86463.365,
      "avg_output_tokens_per_task": 10669.235,
      "last_run_timestamp": "2026-10-31T12:31:32.731Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Phala",
      "accuracy": 0.72833325,
      "accuracy_std_dev": 0.03411588502717615,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.17819147999999999,
      "avg_duration_per_task_ms": 442725.015,
      "avg_output_tokens_per_task": 9652.53,
      "last_run_timestamp": "2026-10-31T22:21:14.715Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "Friendli",
      "accuracy": 0.728,
      "accuracy_std_dev": 0.018086336157442175,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.10671685199999999,
      "avg_duration_per_task_ms": 361548.212,
      "avg_output_tokens_per_task": 14424.78,
      "last_run_timestamp": "2026-10-17T19:16:53.251Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.728,
      "accuracy_std_dev": 0.02362657637492154,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.027497304,
      "avg_duration_per_task_ms": 93734.552,
      "avg_output_tokens_per_task": 12939.792,
      "last_run_timestamp": "2026-10-31T12:24:43.560Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Venice",
      "accuracy": 0.728,
      "accuracy_std_dev": 0.03194425700497665,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.080962124,
      "avg_duration_per_task_ms": 531036.2,
      "avg_output_tokens_per_task": 12180.928,
      "last_run_timestamp": "2026-10-24T23:08:37.435Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Relace",
      "accuracy": 0.7279999333333333,
      "accuracy_std_dev": 0.024092455251846428,
      "run_count": 15,
      "total_questions": 750,
      "avg_cost_per_task_usd": 0.024600684000000005,
      "avg_duration_per_task_ms": 183990.44,
      "avg_output_tokens_per_task": 23628.08,
      "last_run_timestamp": "2026-11-02T17:49:35.292Z"
    },
    {
      "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": "Alibaba",
      "accuracy": 0.7277167582417582,
      "accuracy_std_dev": 0.03220902548783545,
      "run_count": 11,
      "total_questions": 546,
      "avg_cost_per_task_usd": 0.0777849358974359,
      "avg_duration_per_task_ms": 190736.38461538462,
      "avg_output_tokens_per_task": 8353.329670329671,
      "last_run_timestamp": "2026-10-31T19:01:31.914Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Baidu",
      "accuracy": 0.7275001249999999,
      "accuracy_std_dev": 0.027373361190386816,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.065442425,
      "avg_duration_per_task_ms": 198229.1,
      "avg_output_tokens_per_task": 13024.92,
      "last_run_timestamp": "2026-11-10T19:01:19.915Z"
    },
    {
      "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": 0.7271345697356426,
      "accuracy_std_dev": 0.02666397891606509,
      "run_count": 22,
      "total_questions": 1097,
      "avg_cost_per_task_usd": 0.4311616608933454,
      "avg_duration_per_task_ms": 251082.97538742024,
      "avg_output_tokens_per_task": 10876.747493163171,
      "last_run_timestamp": "2026-10-31T12:24:25.742Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "BaseTen",
      "accuracy": 0.7270409540816326,
      "accuracy_std_dev": 0.02754153279934688,
      "run_count": 8,
      "total_questions": 392,
      "avg_cost_per_task_usd": 0.04606436989795917,
      "avg_duration_per_task_ms": 81576.88520408163,
      "avg_output_tokens_per_task": 19891.04081632653,
      "last_run_timestamp": "2026-10-31T12:26:53.018Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "River",
      "accuracy": 0.7268042474226803,
      "accuracy_std_dev": 0.04829799999999995,
      "run_count": 2,
      "total_questions": 97,
      "avg_cost_per_task_usd": 0.005463453608247422,
      "avg_duration_per_task_ms": 80743.0412371134,
      "avg_output_tokens_per_task": 7162.484536082474,
      "last_run_timestamp": "2026-11-12T20:24:40.836Z"
    },
    {
      "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": "Nebius",
      "accuracy": 0.726667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.07819403999999999,
      "avg_duration_per_task_ms": 528014.2,
      "avg_output_tokens_per_task": 10299.58,
      "last_run_timestamp": "2026-11-03T22:19:06.647Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Nebius",
      "accuracy": 0.726667,
      "accuracy_std_dev": 0.019999999999999962,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.24019618,
      "avg_duration_per_task_ms": 932058.35,
      "avg_output_tokens_per_task": 24816.04,
      "last_run_timestamp": "2026-10-31T11:58:06.550Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Inferact vLLM",
      "accuracy": 0.726667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.00647162,
      "avg_duration_per_task_ms": 74169.8,
      "avg_output_tokens_per_task": 10062.22,
      "last_run_timestamp": "2026-11-13T21:15:53.722Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Fireworks",
      "accuracy": 0.726667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.21341358,
      "avg_duration_per_task_ms": 101630.84,
      "avg_output_tokens_per_task": 8827.48,
      "last_run_timestamp": "2026-08-26T21:00:03.913Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-astra-pro-20260903",
      "model_slug": "openai/gpt-6-astra-pro",
      "model_name": "OpenAI: GPT-6 Astra Pro",
      "provider_name": "auto-routing",
      "accuracy": 0.726667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 2.91364046,
      "avg_duration_per_task_ms": 331183.04,
      "avg_output_tokens_per_task": 12422.4,
      "last_run_timestamp": "2026-10-01T08:02:52.271Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.7-plus-20260602",
      "model_slug": "qwen/qwen3.7-plus",
      "model_name": "Qwen: Qwen3.7 Plus",
      "provider_name": "Alibaba",
      "accuracy": 0.726667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.05448868,
      "avg_duration_per_task_ms": 350679.6,
      "avg_output_tokens_per_task": 17629,
      "last_run_timestamp": "2026-10-01T08:04:38.753Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Sail Research",
      "accuracy": 0.72666675,
      "accuracy_std_dev": 0.014907082584714534,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0051741124999999995,
      "avg_duration_per_task_ms": 172373.7025,
      "avg_output_tokens_per_task": 11668.895,
      "last_run_timestamp": "2026-11-06T22:18:15.008Z"
    },
    {
      "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": 0.7266667,
      "accuracy_std_dev": 0.04442226168724414,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.028172072,
      "avg_duration_per_task_ms": 117004.198,
      "avg_output_tokens_per_task": 10046.564,
      "last_run_timestamp": "2026-10-31T19:22:00.212Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Novita",
      "accuracy": 0.7266666666666667,
      "accuracy_std_dev": 0.024944382578492966,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.01469154,
      "avg_duration_per_task_ms": 161228.33333333334,
      "avg_output_tokens_per_task": 15812.58,
      "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": "openai/gpt-5.4-20260305",
      "model_slug": "openai/gpt-5.4",
      "model_name": "OpenAI: GPT-5.4",
      "provider_name": "OpenAI",
      "accuracy": 0.7266666666666667,
      "accuracy_std_dev": 0.014401903168987377,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.30249026,
      "avg_duration_per_task_ms": 183822.13333333333,
      "avg_output_tokens_per_task": 10747.233333333334,
      "last_run_timestamp": "2026-10-02T23:13:07.333Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-luna-20260709",
      "model_slug": "openai/gpt-5.6-luna",
      "model_name": "OpenAI: GPT-5.6 Luna",
      "provider_name": "OpenAI",
      "accuracy": 0.7266666666666667,
      "accuracy_std_dev": 0.014401440259764123,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.013071513333333333,
      "avg_duration_per_task_ms": 97597.57333333333,
      "avg_output_tokens_per_task": 4935.306666666666,
      "last_run_timestamp": "2026-10-02T23:07:05.152Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Venice",
      "accuracy": 0.726666625,
      "accuracy_std_dev": 0.030550477356898618,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.02624409,
      "avg_duration_per_task_ms": 262347.155,
      "avg_output_tokens_per_task": 8877.63,
      "last_run_timestamp": "2026-10-24T19:14:23.110Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.7266665714285715,
      "accuracy_std_dev": 0.02225403120372413,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.09929396571428571,
      "avg_duration_per_task_ms": 163639.28,
      "avg_output_tokens_per_task": 52902.33142857143,
      "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": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Modal",
      "accuracy": 0.7266665,
      "accuracy_std_dev": 0.026666500000000037,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.018476000000000003,
      "avg_duration_per_task_ms": 104934.72,
      "avg_output_tokens_per_task": 10917.83,
      "last_run_timestamp": "2026-10-24T11:48:43.775Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.5-flash-20260519",
      "model_slug": "google/gemini-3.5-flash",
      "model_name": "Google: Gemini 3.5 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.7266665,
      "accuracy_std_dev": 0.026666500000000037,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.43167194999999997,
      "avg_duration_per_task_ms": 195075.98,
      "avg_output_tokens_per_task": 31329.14,
      "last_run_timestamp": "2026-10-01T19:04:40.362Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "Io Net",
      "accuracy": 0.7266665,
      "accuracy_std_dev": 0.013333499999999998,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.06725601,
      "avg_duration_per_task_ms": 640810.38,
      "avg_output_tokens_per_task": 15599.02,
      "last_run_timestamp": "2026-10-27T14:13:59.381Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "CoreWeave",
      "accuracy": 0.7266663333333333,
      "accuracy_std_dev": 0.018856180831641284,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.024500659999999997,
      "avg_duration_per_task_ms": 72543.00666666667,
      "avg_output_tokens_per_task": 15175.193333333333,
      "last_run_timestamp": "2026-10-31T11:34:42.984Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Nebius",
      "accuracy": 0.7264369310344827,
      "accuracy_std_dev": 0.006057869592521776,
      "run_count": 3,
      "total_questions": 145,
      "avg_cost_per_task_usd": 0.012712737931034484,
      "avg_duration_per_task_ms": 160320.86896551723,
      "avg_output_tokens_per_task": 8003.303448275862,
      "last_run_timestamp": "2026-10-31T12:11:53.396Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "Darkbloom",
      "accuracy": 0.7261904081632652,
      "accuracy_std_dev": 0.026678340975171223,
      "run_count": 4,
      "total_questions": 196,
      "avg_cost_per_task_usd": 0.016112545918367346,
      "avg_duration_per_task_ms": 812871.9897959183,
      "avg_output_tokens_per_task": 11587.295918367347,
      "last_run_timestamp": "2026-11-13T04:05:40.636Z"
    },
    {
      "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": 0.7260605454545455,
      "accuracy_std_dev": 0.026431521795571147,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.07336239636363638,
      "avg_duration_per_task_ms": 233666.03636363638,
      "avg_output_tokens_per_task": 8533.292727272727,
      "last_run_timestamp": "2026-10-31T20:10:58.659Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "StepFun",
      "accuracy": 0.7255556666666666,
      "accuracy_std_dev": 0.016960557004087128,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.012908883333333333,
      "avg_duration_per_task_ms": 53544.27,
      "avg_output_tokens_per_task": 10836.613333333333,
      "last_run_timestamp": "2026-11-02T17:45:53.936Z"
    },
    {
      "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": 0.7253333,
      "accuracy_std_dev": 0.02454917785609123,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.09142225,
      "avg_duration_per_task_ms": 693347.86,
      "avg_output_tokens_per_task": 9534.488,
      "last_run_timestamp": "2026-10-31T19:40:42.793Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-sol-20260709",
      "model_slug": "openai/gpt-5.6-sol",
      "model_name": "OpenAI: GPT-5.6 Sol",
      "provider_name": "Azure",
      "accuracy": 0.725,
      "accuracy_std_dev": 0.019220966564318938,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.30439079,
      "avg_duration_per_task_ms": 120258.43333333333,
      "avg_output_tokens_per_task": 5186.016666666666,
      "last_run_timestamp": "2026-10-01T19:05:16.562Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Particle AI",
      "accuracy": 0.725,
      "accuracy_std_dev": 0.017240198505237704,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.01070377,
      "avg_duration_per_task_ms": 64833.36,
      "avg_output_tokens_per_task": 4228.4,
      "last_run_timestamp": "2026-11-03T02:02:55.434Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "StreamLake",
      "accuracy": 0.7244444444444444,
      "accuracy_std_dev": 0.022662439694648512,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.06561171777777777,
      "avg_duration_per_task_ms": 200024.29111111112,
      "avg_output_tokens_per_task": 9471.135555555555,
      "last_run_timestamp": "2026-10-31T20:17:42.375Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Phala",
      "accuracy": 0.7244443333333334,
      "accuracy_std_dev": 0.006285236475706843,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.018751046666666667,
      "avg_duration_per_task_ms": 183683.96666666667,
      "avg_output_tokens_per_task": 11588.8,
      "last_run_timestamp": "2026-10-31T11:59:31.865Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5-mini-2025-08-07",
      "model_slug": "openai/gpt-5-mini",
      "model_name": "OpenAI: GPT-5 Mini",
      "provider_name": "OpenAI",
      "accuracy": 0.7244443333333334,
      "accuracy_std_dev": 0.02199873172606905,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.12152117999999999,
      "avg_duration_per_task_ms": 471173.63333333336,
      "avg_output_tokens_per_task": 54771.27333333333,
      "last_run_timestamp": "2026-10-02T23:27:09.304Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "Venice",
      "accuracy": 0.724,
      "accuracy_std_dev": 0.02351363670723862,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.26574487199999997,
      "avg_duration_per_task_ms": 177434.862,
      "avg_output_tokens_per_task": 11402.85,
      "last_run_timestamp": "2026-10-31T12:56:09.161Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "Z.AI",
      "accuracy": 0.724,
      "accuracy_std_dev": 0.028158109155268212,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.05089722,
      "avg_duration_per_task_ms": 400509.008,
      "avg_output_tokens_per_task": 9631.752,
      "last_run_timestamp": "2026-10-24T20:04:11.026Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7238094285714286,
      "accuracy_std_dev": 0.03062471250410662,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.1589893657142857,
      "avg_duration_per_task_ms": 227082.73428571428,
      "avg_output_tokens_per_task": 9576.722857142857,
      "last_run_timestamp": "2026-10-31T19:57:59.015Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta/muse-spark-open-20261015",
      "model_slug": "meta/muse-spark-open",
      "model_name": "Meta: Muse Spark Open",
      "provider_name": "Makora",
      "accuracy": 0.723360643442623,
      "accuracy_std_dev": 0.019132270021092652,
      "run_count": 5,
      "total_questions": 244,
      "avg_cost_per_task_usd": 0.0756424631147541,
      "avg_duration_per_task_ms": 76631.30327868853,
      "avg_output_tokens_per_task": 8337.151639344262,
      "last_run_timestamp": "2026-10-31T11:46:18.075Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-terra-20260709",
      "model_slug": "openai/gpt-5.6-terra",
      "model_name": "OpenAI: GPT-5.6 Terra",
      "provider_name": "auto-routing",
      "accuracy": 0.7233335,
      "accuracy_std_dev": 0.0033334999999999892,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.11131322,
      "avg_duration_per_task_ms": 83286.18,
      "avg_output_tokens_per_task": 3736.91,
      "last_run_timestamp": "2026-10-01T19:04:40.360Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "StreamLake",
      "accuracy": 0.7233335,
      "accuracy_std_dev": 0.04333349999999997,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.11253119,
      "avg_duration_per_task_ms": 335490.34,
      "avg_output_tokens_per_task": 10275.49,
      "last_run_timestamp": "2026-08-30T00:01:09.490Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "InferenceNet",
      "accuracy": 0.723333375,
      "accuracy_std_dev": 0.026457591854406843,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.007071109999999999,
      "avg_duration_per_task_ms": 91771.1525,
      "avg_output_tokens_per_task": 9121.6525,
      "last_run_timestamp": "2026-11-04T21:12:04.761Z"
    },
    {
      "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": 0.7233333,
      "accuracy_std_dev": 0.031162630755602133,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.3478321060000001,
      "avg_duration_per_task_ms": 232644.894,
      "avg_output_tokens_per_task": 10841.002,
      "last_run_timestamp": "2026-10-31T12:28:19.030Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "auto-routing",
      "accuracy": 0.7233330000000001,
      "accuracy_std_dev": 0.010000000000000009,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.02918011,
      "avg_duration_per_task_ms": 177614.61,
      "avg_output_tokens_per_task": 35023.7,
      "last_run_timestamp": "2026-10-31T11:38:56.927Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.4-20260305",
      "model_slug": "openai/gpt-5.4",
      "model_name": "OpenAI: GPT-5.4",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7233330000000001,
      "accuracy_std_dev": 0.010000000000000009,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.33860155,
      "avg_duration_per_task_ms": 113561.71,
      "avg_output_tokens_per_task": 10711.4,
      "last_run_timestamp": "2026-10-01T19:04:58.091Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.7228916987951808,
      "accuracy_std_dev": 0.08587293804779245,
      "run_count": 10,
      "total_questions": 498,
      "avg_cost_per_task_usd": 0.040060716867469884,
      "avg_duration_per_task_ms": 271599.5261044177,
      "avg_output_tokens_per_task": 22537.02610441767,
      "last_run_timestamp": "2026-10-31T19:30:45.204Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta/muse-spark-open-20261015",
      "model_slug": "meta/muse-spark-open",
      "model_name": "Meta: Muse Spark Open",
      "provider_name": "BaseTen",
      "accuracy": 0.7228067157894738,
      "accuracy_std_dev": 0.002917999999999976,
      "run_count": 2,
      "total_questions": 95,
      "avg_cost_per_task_usd": 0.06760965263157895,
      "avg_duration_per_task_ms": 49831.26315789474,
      "avg_output_tokens_per_task": 5886.4315789473685,
      "last_run_timestamp": "2026-10-31T12:12:21.006Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "OpenInference",
      "accuracy": 0.722650186440678,
      "accuracy_std_dev": 0.03806732610044826,
      "run_count": 13,
      "total_questions": 649,
      "avg_cost_per_task_usd": 0.024282243451463792,
      "avg_duration_per_task_ms": 277851.95069337444,
      "avg_output_tokens_per_task": 9180.11093990755,
      "last_run_timestamp": "2026-10-31T13:24:52.766Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Venice",
      "accuracy": 0.7225,
      "accuracy_std_dev": 0.029615564602924586,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.05079202000000001,
      "avg_duration_per_task_ms": 151914.885,
      "avg_output_tokens_per_task": 28118.625,
      "last_run_timestamp": "2026-10-31T12:23:57.101Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "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": 0.722,
      "accuracy_std_dev": 0.030265381765310677,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.332162946,
      "avg_duration_per_task_ms": 207983.252,
      "avg_output_tokens_per_task": 10350.054,
      "last_run_timestamp": "2026-10-31T12:49:56.822Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.722,
      "accuracy_std_dev": 0.03148898925815181,
      "run_count": 20,
      "total_questions": 1000,
      "avg_cost_per_task_usd": 0.350364916,
      "avg_duration_per_task_ms": 202841.697,
      "avg_output_tokens_per_task": 10188.288,
      "last_run_timestamp": "2026-10-31T12:42:19.688Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "DeepInfra",
      "accuracy": 0.722,
      "accuracy_std_dev": 0.040309938119525816,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.054631136,
      "avg_duration_per_task_ms": 317758.82,
      "avg_output_tokens_per_task": 11839.54,
      "last_run_timestamp": "2026-09-26T08:12:00.292Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Nebius",
      "accuracy": 0.72166675,
      "accuracy_std_dev": 0.012801898733683979,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.024317345,
      "avg_duration_per_task_ms": 116922.04,
      "avg_output_tokens_per_task": 11395.83,
      "last_run_timestamp": "2026-10-31T12:46:14.030Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Particle AI",
      "accuracy": 0.7216665,
      "accuracy_std_dev": 0.031754396022440755,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.018527804999999998,
      "avg_duration_per_task_ms": 90190.985,
      "avg_output_tokens_per_task": 8919.29,
      "last_run_timestamp": "2026-11-03T01:55:17.248Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Ionstream",
      "accuracy": 0.7213335999999999,
      "accuracy_std_dev": 0.027455065511486212,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.060558396,
      "avg_duration_per_task_ms": 335487.684,
      "avg_output_tokens_per_task": 77986.164,
      "last_run_timestamp": "2026-11-10T17:50:20.037Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Reka",
      "accuracy": 0.7213334,
      "accuracy_std_dev": 0.01146975927559073,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.012899843999999999,
      "avg_duration_per_task_ms": 152142.784,
      "avg_output_tokens_per_task": 13063.108,
      "last_run_timestamp": "2026-10-31T13:10:04.012Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "OpenInference",
      "accuracy": 0.7211110000000001,
      "accuracy_std_dev": 0.04794518990124728,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.011416513333333333,
      "avg_duration_per_task_ms": 885570.78,
      "avg_output_tokens_per_task": 25900.693333333333,
      "last_run_timestamp": "2026-10-17T10:32:07.729Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Makora",
      "accuracy": 0.7210882448979592,
      "accuracy_std_dev": 0.0076780598851532815,
      "run_count": 5,
      "total_questions": 245,
      "avg_cost_per_task_usd": 0.009489710204081632,
      "avg_duration_per_task_ms": 145499.9755102041,
      "avg_output_tokens_per_task": 16260.81224489796,
      "last_run_timestamp": "2026-09-26T08:06:43.933Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "auto-routing",
      "accuracy": 0.7209522857142857,
      "accuracy_std_dev": 0.02348344522493778,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.06015369714285714,
      "avg_duration_per_task_ms": 338155.90571428573,
      "avg_output_tokens_per_task": 10030,
      "last_run_timestamp": "2026-10-31T20:35:24.659Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "DeepInfra",
      "accuracy": 0.7206894068965518,
      "accuracy_std_dev": 0.04434366958954421,
      "run_count": 3,
      "total_questions": 145,
      "avg_cost_per_task_usd": 0.015247455172413793,
      "avg_duration_per_task_ms": 270487.4206896552,
      "avg_output_tokens_per_task": 5712.675862068965,
      "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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7206666,
      "accuracy_std_dev": 0.037172284778313,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.07504917600000001,
      "avg_duration_per_task_ms": 152339.734,
      "avg_output_tokens_per_task": 8284.966,
      "last_run_timestamp": "2026-10-31T19:53:09.865Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7206058181818182,
      "accuracy_std_dev": 0.06497575192305928,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.2521624945454546,
      "avg_duration_per_task_ms": 431862.3145454546,
      "avg_output_tokens_per_task": 10421.192727272728,
      "last_run_timestamp": "2026-10-31T12:33:51.242Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Fireworks",
      "accuracy": 0.72,
      "accuracy_std_dev": 0.030000000000000027,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.026063740000000002,
      "avg_duration_per_task_ms": 127597.92,
      "avg_output_tokens_per_task": 12968.38,
      "last_run_timestamp": "2026-09-19T08:13:24.062Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Io Net",
      "accuracy": 0.72,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.12671924,
      "avg_duration_per_task_ms": 566714.78,
      "avg_output_tokens_per_task": 16714.02,
      "last_run_timestamp": "2026-09-28T18:00:05.892Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "inclusionai/ling-3.0-flash-vl-20260910",
      "model_slug": "inclusionai/ling-3.0-flash-vl",
      "model_name": "inclusionAI: Ling 3.0 Flash VL",
      "provider_name": "Novita",
      "accuracy": 0.72,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.00317077,
      "avg_duration_per_task_ms": 128719.23,
      "avg_output_tokens_per_task": 15118.03,
      "last_run_timestamp": "2026-10-01T08:00:27.045Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "Venice",
      "accuracy": 0.72,
      "accuracy_std_dev": 0.01247214674785381,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.08792842000000001,
      "avg_duration_per_task_ms": 120135.21,
      "avg_output_tokens_per_task": 16462.21,
      "last_run_timestamp": "2026-10-31T11:24:22.174Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.72,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.00807376,
      "avg_duration_per_task_ms": 275288.98,
      "avg_output_tokens_per_task": 19939.9,
      "last_run_timestamp": "2026-11-01T18:47:31.518Z"
    },
    {
      "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-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Tenstorrent",
      "accuracy": 0.72,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.01515436,
      "avg_duration_per_task_ms": 412586.86,
      "avg_output_tokens_per_task": 21062.22,
      "last_run_timestamp": "2026-11-08T06:16:39.274Z"
    },
    {
      "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": 0.7196295555555555,
      "accuracy_std_dev": 0.023540924295603994,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.04364436,
      "avg_duration_per_task_ms": 154693.13555555555,
      "avg_output_tokens_per_task": 29598.053333333333,
      "last_run_timestamp": "2026-11-05T00:27:56.809Z"
    },
    {
      "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": 0.7190477142857142,
      "accuracy_std_dev": 0.026047291719454555,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.21749334,
      "avg_duration_per_task_ms": 1099812.5457142857,
      "avg_output_tokens_per_task": 14108.717142857144,
      "last_run_timestamp": "2026-10-31T19:47:23.162Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "auto-routing",
      "accuracy": 0.7190477142857142,
      "accuracy_std_dev": 0.014443512619010297,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.052984574285714285,
      "avg_duration_per_task_ms": 384227.2342857143,
      "avg_output_tokens_per_task": 9361.582857142857,
      "last_run_timestamp": "2026-10-31T20:46:45.843Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "Novita",
      "accuracy": 0.7190475714285715,
      "accuracy_std_dev": 0.021507714169951352,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.09843814571428573,
      "avg_duration_per_task_ms": 197643.15714285715,
      "avg_output_tokens_per_task": 13285.197142857143,
      "last_run_timestamp": "2026-10-31T20:36:53.013Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "DeepInfra",
      "accuracy": 0.71833325,
      "accuracy_std_dev": 0.03378866920563018,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.03334888,
      "avg_duration_per_task_ms": 414661.68,
      "avg_output_tokens_per_task": 8985.465,
      "last_run_timestamp": "2026-10-31T21:01:10.349Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Novita",
      "accuracy": 0.718,
      "accuracy_std_dev": 0.022120970444354387,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.081670762,
      "avg_duration_per_task_ms": 224518.662,
      "avg_output_tokens_per_task": 9223.09,
      "last_run_timestamp": "2026-10-31T20:11:34.826Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Novita",
      "accuracy": 0.718,
      "accuracy_std_dev": 0.027170482351257574,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.049269784,
      "avg_duration_per_task_ms": 173368.438,
      "avg_output_tokens_per_task": 9812.098,
      "last_run_timestamp": "2026-10-31T20:01:58.593Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "StreamLake",
      "accuracy": 0.717778,
      "accuracy_std_dev": 0.046930323977857506,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.045013793333333336,
      "avg_duration_per_task_ms": 196458.52,
      "avg_output_tokens_per_task": 18662.633333333335,
      "last_run_timestamp": "2026-10-31T19:03:35.054Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7173333000000001,
      "accuracy_std_dev": 0.023702007417305403,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.08080122399999999,
      "avg_duration_per_task_ms": 234769.354,
      "avg_output_tokens_per_task": 8765.008,
      "last_run_timestamp": "2026-10-31T20:37:11.918Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "Novita",
      "accuracy": 0.717143,
      "accuracy_std_dev": 0.02802638374817556,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.026408662857142856,
      "avg_duration_per_task_ms": 224207.81142857144,
      "avg_output_tokens_per_task": 10096.908571428572,
      "last_run_timestamp": "2026-10-31T20:20:16.475Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.7171427142857143,
      "accuracy_std_dev": 0.03802358976708421,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.10077471142857143,
      "avg_duration_per_task_ms": 217139.6,
      "avg_output_tokens_per_task": 10145.26857142857,
      "last_run_timestamp": "2026-10-31T20:09:58.111Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "DekaLLM",
      "accuracy": 0.716667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.01519146,
      "avg_duration_per_task_ms": 819823.24,
      "avg_output_tokens_per_task": 11470.58,
      "last_run_timestamp": "2026-10-31T20:19:18.995Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5-mini-2025-08-07",
      "model_slug": "openai/gpt-5-mini",
      "model_name": "OpenAI: GPT-5 Mini",
      "provider_name": "auto-routing",
      "accuracy": 0.7166669999999999,
      "accuracy_std_dev": 0.009999999999999953,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.1280142,
      "avg_duration_per_task_ms": 531647.04,
      "avg_output_tokens_per_task": 56970.52,
      "last_run_timestamp": "2026-10-01T19:04:42.411Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "StreamLake",
      "accuracy": 0.716666625,
      "accuracy_std_dev": 0.029627375668195347,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0693026925,
      "avg_duration_per_task_ms": 283497.055,
      "avg_output_tokens_per_task": 12604.555,
      "last_run_timestamp": "2026-10-31T19:48:30.443Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Ambient",
      "accuracy": 0.7166665000000001,
      "accuracy_std_dev": 0.01666650000000003,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.00699117,
      "avg_duration_per_task_ms": 151328.11,
      "avg_output_tokens_per_task": 13794.25,
      "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": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Venice",
      "accuracy": 0.7166665000000001,
      "accuracy_std_dev": 0.0033334999999999892,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.022047789999999998,
      "avg_duration_per_task_ms": 63248.07,
      "avg_output_tokens_per_task": 8992.32,
      "last_run_timestamp": "2026-10-31T12:43:48.267Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "StreamLake",
      "accuracy": 0.7166665000000001,
      "accuracy_std_dev": 0.0033334999999999892,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.015074639999999999,
      "avg_duration_per_task_ms": 124785.23,
      "avg_output_tokens_per_task": 11839.88,
      "last_run_timestamp": "2026-10-31T12:28:39.529Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "Novita",
      "accuracy": 0.7166665000000001,
      "accuracy_std_dev": 0.043333500000000025,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.048402060000000004,
      "avg_duration_per_task_ms": 204313.4,
      "avg_output_tokens_per_task": 8310.75,
      "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": "openai/gpt-5.6-sol-pro-20260709",
      "model_slug": "openai/gpt-5.6-sol-pro",
      "model_name": "OpenAI: GPT-5.6 Sol Pro",
      "provider_name": "OpenAI",
      "accuracy": 0.7166665000000001,
      "accuracy_std_dev": 0.03666649999999999,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.71669186,
      "avg_duration_per_task_ms": 277432.68,
      "avg_output_tokens_per_task": 22919.22,
      "last_run_timestamp": "2026-10-02T23:08:29.026Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Relace",
      "accuracy": 0.7166665000000001,
      "accuracy_std_dev": 0.02603405888927041,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.13610847,
      "avg_duration_per_task_ms": 230555.42,
      "avg_output_tokens_per_task": 11694.86,
      "last_run_timestamp": "2026-10-31T19:46:31.005Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "Novita",
      "accuracy": 0.7161904285714286,
      "accuracy_std_dev": 0.02133854528350812,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.053568791428571426,
      "avg_duration_per_task_ms": 335614.3914285714,
      "avg_output_tokens_per_task": 9870.974285714286,
      "last_run_timestamp": "2026-10-31T20:51:04.120Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Venice",
      "accuracy": 0.7159998000000001,
      "accuracy_std_dev": 0.03295793548085197,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.24450749400000002,
      "avg_duration_per_task_ms": 314057.314,
      "avg_output_tokens_per_task": 21575.564,
      "last_run_timestamp": "2026-10-31T12:46:35.433Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Together",
      "accuracy": 0.7155556666666666,
      "accuracy_std_dev": 0.03189478845969807,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.01325666,
      "avg_duration_per_task_ms": 86099.67333333334,
      "avg_output_tokens_per_task": 8966.793333333333,
      "last_run_timestamp": "2026-10-31T13:08:15.957Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "Venice",
      "accuracy": 0.7153333000000001,
      "accuracy_std_dev": 0.03183982342303425,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.09070920600000001,
      "avg_duration_per_task_ms": 284476.36,
      "avg_output_tokens_per_task": 9245.778,
      "last_run_timestamp": "2026-10-31T19:57:59.098Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "Venice",
      "accuracy": 0.7152381428571428,
      "accuracy_std_dev": 0.022172375614898876,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.04021302285714286,
      "avg_duration_per_task_ms": 623020.8714285714,
      "avg_output_tokens_per_task": 9178.234285714285,
      "last_run_timestamp": "2026-10-31T19:41:08.087Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "Z.AI",
      "accuracy": 0.7152381428571428,
      "accuracy_std_dev": 0.03866263877783441,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.10695628285714286,
      "avg_duration_per_task_ms": 188318.44857142857,
      "avg_output_tokens_per_task": 12311.014285714286,
      "last_run_timestamp": "2026-10-31T20:14:27.576Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Chutes",
      "accuracy": 0.715238,
      "accuracy_std_dev": 0.0354049972784311,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.05114133714285715,
      "avg_duration_per_task_ms": 426873.76857142855,
      "avg_output_tokens_per_task": 11755.502857142857,
      "last_run_timestamp": "2026-10-31T20:58:51.733Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Baidu",
      "accuracy": 0.7147368947368421,
      "accuracy_std_dev": 0.0235532410014658,
      "run_count": 19,
      "total_questions": 950,
      "avg_cost_per_task_usd": 0.07154260947368421,
      "avg_duration_per_task_ms": 115435.01157894736,
      "avg_output_tokens_per_task": 8767.831578947369,
      "last_run_timestamp": "2026-11-10T19:27:07.187Z"
    },
    {
      "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": 0.7143545606936417,
      "accuracy_std_dev": 0.030307639134686294,
      "run_count": 7,
      "total_questions": 346,
      "avg_cost_per_task_usd": 0.08907920231213873,
      "avg_duration_per_task_ms": 87722.67919075144,
      "avg_output_tokens_per_task": 7278.56647398844,
      "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": "z-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": 0.7139393636363637,
      "accuracy_std_dev": 0.07661811901688757,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.010148356363636365,
      "avg_duration_per_task_ms": 242007.0509090909,
      "avg_output_tokens_per_task": 8765.507272727273,
      "last_run_timestamp": "2026-11-10T08:18:49.964Z"
    },
    {
      "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": 0.71375,
      "accuracy_std_dev": 0.03794860559691226,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.09171088000000001,
      "avg_duration_per_task_ms": 170840.095,
      "avg_output_tokens_per_task": 11418.88,
      "last_run_timestamp": "2026-11-10T20:09:17.079Z"
    },
    {
      "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": 0.7135715,
      "accuracy_std_dev": 0.03209528384302411,
      "run_count": 28,
      "total_questions": 1400,
      "avg_cost_per_task_usd": 0.1159875742857143,
      "avg_duration_per_task_ms": 125459.01928571428,
      "avg_output_tokens_per_task": 8558.635714285714,
      "last_run_timestamp": "2026-10-31T19:56:11.088Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "Minimax",
      "accuracy": 0.7133333571428572,
      "accuracy_std_dev": 0.024168666585380864,
      "run_count": 14,
      "total_questions": 700,
      "avg_cost_per_task_usd": 0.038873661428571425,
      "avg_duration_per_task_ms": 230582.26428571428,
      "avg_output_tokens_per_task": 9732.062857142857,
      "last_run_timestamp": "2026-10-31T20:38:36.972Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Novita",
      "accuracy": 0.7133332857142858,
      "accuracy_std_dev": 0.026427450856552726,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.13545089142857142,
      "avg_duration_per_task_ms": 267525.3171428571,
      "avg_output_tokens_per_task": 10590.037142857143,
      "last_run_timestamp": "2026-10-31T20:31:35.900Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-fable-5.2-20261109",
      "model_slug": "anthropic/claude-fable-5.2",
      "model_name": "Anthropic: Claude Fable 5.2",
      "provider_name": "Anthropic",
      "accuracy": 0.713333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.78360724,
      "avg_duration_per_task_ms": 157424.24,
      "avg_output_tokens_per_task": 9303.06,
      "last_run_timestamp": "2026-11-10T18:06:53.999Z"
    },
    {
      "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": "River",
      "accuracy": 0.713333,
      "accuracy_std_dev": 0.010000000000000009,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.011907570000000001,
      "avg_duration_per_task_ms": 93169.09,
      "avg_output_tokens_per_task": 17553.86,
      "last_run_timestamp": "2026-11-12T21:15:48.475Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Cloudflare",
      "accuracy": 0.713333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.05452906,
      "avg_duration_per_task_ms": 272872.18,
      "avg_output_tokens_per_task": 39719.44,
      "last_run_timestamp": "2026-10-13T20:01:29.821Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.1-flash-lite-20260507",
      "model_slug": "google/gemini-3.1-flash-lite",
      "model_name": "Google: Gemini 3.1 Flash Lite",
      "provider_name": "auto-routing",
      "accuracy": 0.713333,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.08007529,
      "avg_duration_per_task_ms": 140906.22,
      "avg_output_tokens_per_task": 42659.23,
      "last_run_timestamp": "2026-10-01T19:04:52.649Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "inclusionai/ling-3.0-flash-vl-20260910",
      "model_slug": "inclusionai/ling-3.0-flash-vl",
      "model_name": "inclusionAI: Ling 3.0 Flash VL",
      "provider_name": "DeepInfra",
      "accuracy": 0.713333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.006318460000000001,
      "avg_duration_per_task_ms": 234030.92,
      "avg_output_tokens_per_task": 14773.36,
      "last_run_timestamp": "2026-10-01T08:05:57.971Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Fireworks",
      "accuracy": 0.713333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.01617958,
      "avg_duration_per_task_ms": 141600.68,
      "avg_output_tokens_per_task": 14742.64,
      "last_run_timestamp": "2026-10-31T12:42:39.534Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "GMICloud",
      "accuracy": 0.713333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.09868048,
      "avg_duration_per_task_ms": 318461,
      "avg_output_tokens_per_task": 11794.24,
      "last_run_timestamp": "2026-08-29T17:18:09.782Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-flash-20260826",
      "model_slug": "qwen/qwen3.8-flash",
      "model_name": "Qwen: Qwen3.8 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.713333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.01665106,
      "avg_duration_per_task_ms": 327560.6,
      "avg_output_tokens_per_task": 17227.24,
      "last_run_timestamp": "2026-10-01T08:00:41.459Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.8-flash-20260902",
      "model_slug": "google/gemini-3.8-flash",
      "model_name": "Google: Gemini 3.8 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.712766,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 47,
      "avg_cost_per_task_usd": 0.21022357446808512,
      "avg_duration_per_task_ms": 261220.25531914894,
      "avg_output_tokens_per_task": 27609.17021276596,
      "last_run_timestamp": "2026-10-01T08:02:55.088Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7126666,
      "accuracy_std_dev": 0.028975068211136257,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.07464725,
      "avg_duration_per_task_ms": 125244.052,
      "avg_output_tokens_per_task": 9446.146,
      "last_run_timestamp": "2026-10-31T19:18:12.479Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "ModelRun",
      "accuracy": 0.7125,
      "accuracy_std_dev": 0.006401128377091036,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.066283255,
      "avg_duration_per_task_ms": 74239.48,
      "avg_output_tokens_per_task": 8403.98,
      "last_run_timestamp": "2026-09-26T08:08:19.642Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "Alibaba",
      "accuracy": 0.712222,
      "accuracy_std_dev": 0.01640534390577251,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.04505285333333333,
      "avg_duration_per_task_ms": 337842.44,
      "avg_output_tokens_per_task": 13141.253333333334,
      "last_run_timestamp": "2026-10-04T00:00:34.898Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "OpenInference",
      "accuracy": 0.7117822371364653,
      "accuracy_std_dev": 0.056118924341169225,
      "run_count": 9,
      "total_questions": 447,
      "avg_cost_per_task_usd": 0.0462677941834452,
      "avg_duration_per_task_ms": 1033622.2237136465,
      "avg_output_tokens_per_task": 60076.49888143177,
      "last_run_timestamp": "2026-10-31T13:23:14.602Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Cohere",
      "accuracy": 0.7111113333333333,
      "accuracy_std_dev": 0.006285236475706792,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.06754876666666666,
      "avg_duration_per_task_ms": 139695.35333333333,
      "avg_output_tokens_per_task": 33581.94,
      "last_run_timestamp": "2026-10-31T12:01:01.015Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Crusoe",
      "accuracy": 0.7111113333333333,
      "accuracy_std_dev": 0.0332591470759883,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.10563259333333332,
      "avg_duration_per_task_ms": 87886.20666666667,
      "avg_output_tokens_per_task": 8897.613333333333,
      "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": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "GMICloud",
      "accuracy": 0.7103703333333333,
      "accuracy_std_dev": 0.018354812708266867,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.022485688888888886,
      "avg_duration_per_task_ms": 293573.69333333336,
      "avg_output_tokens_per_task": 10080.99111111111,
      "last_run_timestamp": "2026-11-10T07:15:16.990Z"
    },
    {
      "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": 0.71,
      "accuracy_std_dev": 0.036667000000000005,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.011451959999999999,
      "avg_duration_per_task_ms": 477432.72,
      "avg_output_tokens_per_task": 17288.47,
      "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-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Cohere",
      "accuracy": 0.71,
      "accuracy_std_dev": 0.016666999999999987,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.015407820000000001,
      "avg_duration_per_task_ms": 62181.78,
      "avg_output_tokens_per_task": 11850.44,
      "last_run_timestamp": "2026-10-31T11:31:06.991Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "ModelRun",
      "accuracy": 0.71,
      "accuracy_std_dev": 0.01795033271279393,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.0134254,
      "avg_duration_per_task_ms": 120168.52,
      "avg_output_tokens_per_task": 11670.58,
      "last_run_timestamp": "2026-10-31T12:25:39.138Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-terra-20260709",
      "model_slug": "openai/gpt-5.6-terra",
      "model_name": "OpenAI: GPT-5.6 Terra",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.71,
      "accuracy_std_dev": 0.010000000000000009,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.12724396999999998,
      "avg_duration_per_task_ms": 73218.95,
      "avg_output_tokens_per_task": 3826.45,
      "last_run_timestamp": "2026-10-01T19:04:38.304Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7095238571428572,
      "accuracy_std_dev": 0.057993713007602476,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.030501814285714284,
      "avg_duration_per_task_ms": 305801.77142857143,
      "avg_output_tokens_per_task": 10004.028571428571,
      "last_run_timestamp": "2026-10-31T20:10:49.937Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "Venice",
      "accuracy": 0.7095237142857143,
      "accuracy_std_dev": 0.03614029729387676,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.06154300571428571,
      "avg_duration_per_task_ms": 247462.57428571428,
      "avg_output_tokens_per_task": 25975.411428571428,
      "last_run_timestamp": "2026-10-31T19:22:37.818Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7093334,
      "accuracy_std_dev": 0.07236922366199598,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.10543298799999999,
      "avg_duration_per_task_ms": 199005.244,
      "avg_output_tokens_per_task": 11270.708,
      "last_run_timestamp": "2026-10-17T19:18:15.219Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "auto-routing",
      "accuracy": 0.7093333,
      "accuracy_std_dev": 0.02924233156931914,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.047131844,
      "avg_duration_per_task_ms": 100777.312,
      "avg_output_tokens_per_task": 8394.926,
      "last_run_timestamp": "2026-10-31T19:54:42.760Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "CoreWeave",
      "accuracy": 0.709091,
      "accuracy_std_dev": 0.0405291288424421,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.07860464545454546,
      "avg_duration_per_task_ms": 111991.2290909091,
      "avg_output_tokens_per_task": 8595.541818181819,
      "last_run_timestamp": "2026-10-31T19:45:55.386Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Novita",
      "accuracy": 0.708889,
      "accuracy_std_dev": 0.0031423825355929936,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.017471633333333333,
      "avg_duration_per_task_ms": 219758.70666666667,
      "avg_output_tokens_per_task": 16040.473333333333,
      "last_run_timestamp": "2026-10-31T12:30:16.930Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "auto-routing",
      "accuracy": 0.7085714285714285,
      "accuracy_std_dev": 0.012706521619761615,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.027396339999999998,
      "avg_duration_per_task_ms": 231738.34285714285,
      "avg_output_tokens_per_task": 10497.88,
      "last_run_timestamp": "2026-10-31T19:47:12.389Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Wafer",
      "accuracy": 0.7083333,
      "accuracy_std_dev": 0.024843901555713817,
      "run_count": 20,
      "total_questions": 1000,
      "avg_cost_per_task_usd": 0.09730978499999998,
      "avg_duration_per_task_ms": 109439.124,
      "avg_output_tokens_per_task": 9725.372,
      "last_run_timestamp": "2026-10-31T19:37:51.988Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Nebius",
      "accuracy": 0.7083331666666667,
      "accuracy_std_dev": 0.017400462957602263,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.25373078000000004,
      "avg_duration_per_task_ms": 176420.9,
      "avg_output_tokens_per_task": 10605.743333333334,
      "last_run_timestamp": "2026-10-31T19:48:27.256Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Phala",
      "accuracy": 0.7080952857142857,
      "accuracy_std_dev": 0.122226283771085,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.23844771999999997,
      "avg_duration_per_task_ms": 1324378.657142857,
      "avg_output_tokens_per_task": 65425.98,
      "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": "openai/gpt-5.6-terra-20260709",
      "model_slug": "openai/gpt-5.6-terra",
      "model_name": "OpenAI: GPT-5.6 Terra",
      "provider_name": "Azure",
      "accuracy": 0.7077778333333333,
      "accuracy_std_dev": 0.012422535280914084,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.11194547666666667,
      "avg_duration_per_task_ms": 92493.15666666666,
      "avg_output_tokens_per_task": 3647.7966666666666,
      "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": "inclusionai/ling-3.0-flash-fin-20260827",
      "model_slug": "inclusionai/ling-3.0-flash-fin",
      "model_name": "inclusionAI: Ling 3.0 Flash Fin",
      "provider_name": "auto-routing",
      "accuracy": 0.707483,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 49,
      "avg_cost_per_task_usd": 0.004842469387755102,
      "avg_duration_per_task_ms": 161558.55102040817,
      "avg_output_tokens_per_task": 17024.020408163266,
      "last_run_timestamp": "2026-10-01T08:01:49.153Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.5-flash-20260519",
      "model_slug": "google/gemini-3.5-flash",
      "model_name": "Google: Gemini 3.5 Flash",
      "provider_name": "Google AI Studio",
      "accuracy": 0.706667,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.44632813,
      "avg_duration_per_task_ms": 132811.5,
      "avg_output_tokens_per_task": 20668.24,
      "last_run_timestamp": "2026-10-01T19:04:52.658Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-sol-pro-20260709",
      "model_slug": "openai/gpt-5.6-sol-pro",
      "model_name": "OpenAI: GPT-5.6 Sol Pro",
      "provider_name": "auto-routing",
      "accuracy": 0.706667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.71176192,
      "avg_duration_per_task_ms": 285827.28,
      "avg_output_tokens_per_task": 22895.76,
      "last_run_timestamp": "2026-10-01T16:00:34.660Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "thinkingmachines/inkling-2-small-20261107",
      "model_slug": "thinkingmachines/inkling-2-small",
      "model_name": "Thinking Machines: Inkling 2 Small",
      "provider_name": "DeepInfra",
      "accuracy": 0.706667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.023244639999999997,
      "avg_duration_per_task_ms": 174558.52,
      "avg_output_tokens_per_task": 8529.56,
      "last_run_timestamp": "2026-11-13T12:51:19.323Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-astra-pro-20260903",
      "model_slug": "openai/gpt-6-astra-pro",
      "model_name": "OpenAI: GPT-6 Astra Pro",
      "provider_name": "OpenAI",
      "accuracy": 0.7066669999999999,
      "accuracy_std_dev": 0.019999999999999962,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 3.04321209,
      "avg_duration_per_task_ms": 317897.29,
      "avg_output_tokens_per_task": 13272.53,
      "last_run_timestamp": "2026-10-02T23:13:16.494Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6.1-sol-20260929",
      "model_slug": "openai/gpt-6.1-sol",
      "model_name": "OpenAI: GPT-6.1 Sol",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7066669999999999,
      "accuracy_std_dev": 0.019999999999999962,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.1042244,
      "avg_duration_per_task_ms": 226024.67,
      "avg_output_tokens_per_task": 3543.23,
      "last_run_timestamp": "2026-10-30T00:17:12.390Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7066666666666667,
      "accuracy_std_dev": 0.014401440259764123,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.015417799999999999,
      "avg_duration_per_task_ms": 75266.33333333333,
      "avg_output_tokens_per_task": 12018.02,
      "last_run_timestamp": "2026-10-31T11:20:39.476Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5-2025-08-07",
      "model_slug": "openai/gpt-5",
      "model_name": "OpenAI: GPT-5",
      "provider_name": "OpenAI",
      "accuracy": 0.7066666666666667,
      "accuracy_std_dev": 0.005443174461857933,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.5994524266666668,
      "avg_duration_per_task_ms": 636541.5466666666,
      "avg_output_tokens_per_task": 54354.86,
      "last_run_timestamp": "2026-10-02T23:22:41.519Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-luna-pro-20260709",
      "model_slug": "openai/gpt-5.6-luna-pro",
      "model_name": "OpenAI: GPT-5.6 Luna Pro",
      "provider_name": "OpenAI",
      "accuracy": 0.7066666666666667,
      "accuracy_std_dev": 0.01962587106064057,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.08575882,
      "avg_duration_per_task_ms": 316120.8933333333,
      "avg_output_tokens_per_task": 32977.22666666667,
      "last_run_timestamp": "2026-10-02T23:12:53.530Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Phala",
      "accuracy": 0.706666625,
      "accuracy_std_dev": 0.020000041668315968,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.10048072499999999,
      "avg_duration_per_task_ms": 168728.9075,
      "avg_output_tokens_per_task": 11337.64,
      "last_run_timestamp": "2026-10-17T19:53:39.136Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Novita",
      "accuracy": 0.7063038567335244,
      "accuracy_std_dev": 0.03586910330405172,
      "run_count": 7,
      "total_questions": 349,
      "avg_cost_per_task_usd": 0.02153851862464183,
      "avg_duration_per_task_ms": 628235.6618911175,
      "avg_output_tokens_per_task": 9341.048710601719,
      "last_run_timestamp": "2026-10-31T19:57:15.103Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "Makora",
      "accuracy": 0.7061111666666666,
      "accuracy_std_dev": 0.023602995992152234,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.023716863333333334,
      "avg_duration_per_task_ms": 111467.81333333334,
      "avg_output_tokens_per_task": 29595.083333333332,
      "last_run_timestamp": "2026-10-31T19:52:48.234Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "SambaNova",
      "accuracy": 0.7060303366834171,
      "accuracy_std_dev": 0.0301180502813512,
      "run_count": 4,
      "total_questions": 199,
      "avg_cost_per_task_usd": 0.030429407035175878,
      "avg_duration_per_task_ms": 140666.3216080402,
      "avg_output_tokens_per_task": 10689.924623115578,
      "last_run_timestamp": "2026-11-11T19:39:53.430Z"
    },
    {
      "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": 0.7055558333333333,
      "accuracy_std_dev": 0.032754340019894884,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.012937629999999999,
      "avg_duration_per_task_ms": 95536.39333333333,
      "avg_output_tokens_per_task": 4842.643333333333,
      "last_run_timestamp": "2026-10-01T19:05:16.550Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Chutes",
      "accuracy": 0.7055555,
      "accuracy_std_dev": 0.03408387439709088,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.09377007666666666,
      "avg_duration_per_task_ms": 923124.9966666667,
      "avg_output_tokens_per_task": 9445.45,
      "last_run_timestamp": "2026-10-31T21:09:07.436Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5-mini-2025-08-07",
      "model_slug": "openai/gpt-5-mini",
      "model_name": "OpenAI: GPT-5 Mini",
      "provider_name": "Azure",
      "accuracy": 0.70499975,
      "accuracy_std_dev": 0.011902485735656231,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.13260128,
      "avg_duration_per_task_ms": 477326.39,
      "avg_output_tokens_per_task": 55745.89,
      "last_run_timestamp": "2026-10-01T19:05:16.547Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Tenstorrent",
      "accuracy": 0.7048148888888889,
      "accuracy_std_dev": 0.14567945297081722,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.0241112,
      "avg_duration_per_task_ms": 290526.02444444445,
      "avg_output_tokens_per_task": 11444.402222222223,
      "last_run_timestamp": "2026-11-08T03:52:11.162Z"
    },
    {
      "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": 0.704762,
      "accuracy_std_dev": 0.03486274480219332,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.07204742,
      "avg_duration_per_task_ms": 281672.5857142857,
      "avg_output_tokens_per_task": 11399.114285714286,
      "last_run_timestamp": "2026-10-31T21:34:09.712Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Together",
      "accuracy": 0.7046667,
      "accuracy_std_dev": 0.036642272492436935,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.09915010799999999,
      "avg_duration_per_task_ms": 70205.334,
      "avg_output_tokens_per_task": 8678.166,
      "last_run_timestamp": "2026-10-31T19:33:25.981Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Makora",
      "accuracy": 0.7039998000000001,
      "accuracy_std_dev": 0.025508095447524113,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.021704752,
      "avg_duration_per_task_ms": 119683.016,
      "avg_output_tokens_per_task": 11904.78,
      "last_run_timestamp": "2026-10-31T12:25:42.323Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7033335,
      "accuracy_std_dev": 0.03666649999999999,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.053050639999999996,
      "avg_duration_per_task_ms": 750494.53,
      "avg_output_tokens_per_task": 9139.19,
      "last_run_timestamp": "2026-08-29T16:47:05.512Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-luna-pro-20260709",
      "model_slug": "openai/gpt-5.6-luna-pro",
      "model_name": "OpenAI: GPT-5.6 Luna Pro",
      "provider_name": "auto-routing",
      "accuracy": 0.7033335,
      "accuracy_std_dev": 0.0033335000000000448,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.08534018,
      "avg_duration_per_task_ms": 347658.09,
      "avg_output_tokens_per_task": 32908.85,
      "last_run_timestamp": "2026-10-01T16:00:47.532Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "Moonshot AI",
      "accuracy": 0.703333,
      "accuracy_std_dev": 0.010000000000000009,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.045981190000000005,
      "avg_duration_per_task_ms": 151977.07,
      "avg_output_tokens_per_task": 8039.38,
      "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": "qwen/qwen3.7-plus-20260602",
      "model_slug": "qwen/qwen3.7-plus",
      "model_name": "Qwen: Qwen3.7 Plus",
      "provider_name": "auto-routing",
      "accuracy": 0.703333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.05425128,
      "avg_duration_per_task_ms": 396980.46,
      "avg_output_tokens_per_task": 20673.9,
      "last_run_timestamp": "2026-10-01T08:05:20.277Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Nebius",
      "accuracy": 0.7025,
      "accuracy_std_dev": 0.030173028796426787,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.2793706775,
      "avg_duration_per_task_ms": 496473.2875,
      "avg_output_tokens_per_task": 9455.475,
      "last_run_timestamp": "2026-10-31T22:14:51.015Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "Io Net",
      "accuracy": 0.7023811326530611,
      "accuracy_std_dev": 0.047340802625087566,
      "run_count": 4,
      "total_questions": 196,
      "avg_cost_per_task_usd": 0.014528989795918367,
      "avg_duration_per_task_ms": 320290.7346938776,
      "avg_output_tokens_per_task": 12552.510204081633,
      "last_run_timestamp": "2026-11-11T06:41:01.794Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-terra-20260709",
      "model_slug": "openai/gpt-5.6-terra",
      "model_name": "OpenAI: GPT-5.6 Terra",
      "provider_name": "OpenAI",
      "accuracy": 0.7022223333333333,
      "accuracy_std_dev": 0.021999169459070246,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.10818771333333332,
      "avg_duration_per_task_ms": 79804.25333333333,
      "avg_output_tokens_per_task": 3542.24,
      "last_run_timestamp": "2026-10-02T23:10:36.500Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-sol-20260922",
      "model_slug": "openai/gpt-6-sol",
      "model_name": "OpenAI: GPT-6 Sol",
      "provider_name": "OpenAI",
      "accuracy": 0.7022223333333333,
      "accuracy_std_dev": 0.011330936570096727,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.13357544,
      "avg_duration_per_task_ms": 105586.74666666667,
      "avg_output_tokens_per_task": 5260.826666666667,
      "last_run_timestamp": "2026-10-02T23:07:17.250Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Nebius",
      "accuracy": 0.7018973089430894,
      "accuracy_std_dev": 0.04349030049378826,
      "run_count": 5,
      "total_questions": 246,
      "avg_cost_per_task_usd": 0.029020682926829265,
      "avg_duration_per_task_ms": 391991.35365853657,
      "avg_output_tokens_per_task": 21482.62601626016,
      "last_run_timestamp": "2026-11-12T15:21:11.557Z"
    },
    {
      "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": 0.7018181818181818,
      "accuracy_std_dev": 0.033525796984135756,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.04067215272727272,
      "avg_duration_per_task_ms": 142417.50727272726,
      "avg_output_tokens_per_task": 8505.803636363637,
      "last_run_timestamp": "2026-10-31T20:18:13.909Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.7009522857142857,
      "accuracy_std_dev": 0.04471111302801668,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.06413979428571429,
      "avg_duration_per_task_ms": 345333.78285714285,
      "avg_output_tokens_per_task": 10011.842857142858,
      "last_run_timestamp": "2026-10-31T20:44:35.166Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7006667,
      "accuracy_std_dev": 0.16001239243699222,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.061942158000000004,
      "avg_duration_per_task_ms": 147418.05,
      "avg_output_tokens_per_task": 16768.652,
      "last_run_timestamp": "2026-10-31T19:40:48.790Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "StreamLake",
      "accuracy": 0.7006060000000001,
      "accuracy_std_dev": 0.025968777609765438,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.03481650181818182,
      "avg_duration_per_task_ms": 169300.62181818183,
      "avg_output_tokens_per_task": 8963.998181818182,
      "last_run_timestamp": "2026-10-31T20:27:27.569Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.7000001,
      "accuracy_std_dev": 0.02170498150863068,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.124576664,
      "avg_duration_per_task_ms": 163368.62,
      "avg_output_tokens_per_task": 11905.278,
      "last_run_timestamp": "2026-10-31T19:55:04.254Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4-sonnet-20250522",
      "model_slug": "anthropic/claude-sonnet-4",
      "model_name": "Anthropic: Claude Sonnet 4",
      "provider_name": "auto-routing",
      "accuracy": 0.7,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.29769207999999997,
      "avg_duration_per_task_ms": 128375.6,
      "avg_output_tokens_per_task": 8500.86,
      "last_run_timestamp": "2026-09-01T12:41:24.616Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Phala",
      "accuracy": 0.7,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.02395168,
      "avg_duration_per_task_ms": 160119.2,
      "avg_output_tokens_per_task": 14308.04,
      "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": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Google",
      "accuracy": 0.7,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.0355865,
      "avg_duration_per_task_ms": 154491.8,
      "avg_output_tokens_per_task": 12574.4,
      "last_run_timestamp": "2026-10-31T11:15:31.400Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "GLM 5.3 Flash",
      "provider_name": "digitalocean",
      "accuracy": 0.7,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.00357748,
      "avg_duration_per_task_ms": 48793.08,
      "avg_output_tokens_per_task": 1567.76,
      "last_run_timestamp": "2026-10-01T23:50:45.022Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "inclusionai/ling-3.0-flash-vl-20260910",
      "model_slug": "inclusionai/ling-3.0-flash-vl",
      "model_name": "inclusionAI: Ling 3.0 Flash VL",
      "provider_name": "auto-routing",
      "accuracy": 0.7,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.0030046400000000003,
      "avg_duration_per_task_ms": 98848.12,
      "avg_output_tokens_per_task": 17113.74,
      "last_run_timestamp": "2026-10-01T08:00:19.901Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.7,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.12726106,
      "avg_duration_per_task_ms": 181226.44,
      "avg_output_tokens_per_task": 9838.56,
      "last_run_timestamp": "2026-10-04T03:00:16.405Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "StreamLake",
      "accuracy": 0.7,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.0161093,
      "avg_duration_per_task_ms": 215051.72,
      "avg_output_tokens_per_task": 15584.08,
      "last_run_timestamp": "2026-10-31T12:24:55.227Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "auto-routing",
      "accuracy": 0.7,
      "accuracy_std_dev": 0.03266000904120248,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.055787651428571436,
      "avg_duration_per_task_ms": 343671.10285714286,
      "avg_output_tokens_per_task": 9284.377142857144,
      "last_run_timestamp": "2026-10-31T22:10:13.617Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "Phala",
      "accuracy": 0.7,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.07376152,
      "avg_duration_per_task_ms": 750952.84,
      "avg_output_tokens_per_task": 11282.28,
      "last_run_timestamp": "2026-08-29T16:44:16.336Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5-2025-08-07",
      "model_slug": "openai/gpt-5",
      "model_name": "OpenAI: GPT-5",
      "provider_name": "Azure",
      "accuracy": 0.7,
      "accuracy_std_dev": 0.004714280910170736,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.634261665,
      "avg_duration_per_task_ms": 673734.145,
      "avg_output_tokens_per_task": 52923.67,
      "last_run_timestamp": "2026-10-01T16:01:02.669Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6.1-sol-20260929",
      "model_slug": "openai/gpt-6.1-sol",
      "model_name": "OpenAI: GPT-6.1 Sol",
      "provider_name": "auto-routing",
      "accuracy": 0.7,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.1012818,
      "avg_duration_per_task_ms": 144103.72,
      "avg_output_tokens_per_task": 3638.14,
      "last_run_timestamp": "2026-10-01T08:03:00.822Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6.1-terra-20261026",
      "model_slug": "openai/gpt-6.1-terra",
      "model_name": "OpenAI: GPT-6.1 Terra",
      "provider_name": "OpenAI",
      "accuracy": 0.7,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.05778106,
      "avg_duration_per_task_ms": 103456.94,
      "avg_output_tokens_per_task": 4693.4,
      "last_run_timestamp": "2026-10-28T00:06:02.481Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Parasail",
      "accuracy": 0.6977776666666667,
      "accuracy_std_dev": 0.011330936570096727,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.013071946666666667,
      "avg_duration_per_task_ms": 88036.78,
      "avg_output_tokens_per_task": 11978.486666666666,
      "last_run_timestamp": "2026-10-31T11:55:17.583Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-sol-20260922",
      "model_slug": "openai/gpt-6-sol",
      "model_name": "OpenAI: GPT-6 Sol",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.6977776666666667,
      "accuracy_std_dev": 0.01911642045177101,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.14221055333333335,
      "avg_duration_per_task_ms": 130884.03333333334,
      "avg_output_tokens_per_task": 4996.426666666666,
      "last_run_timestamp": "2026-10-30T00:14:00.742Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Parasail",
      "accuracy": 0.6976666,
      "accuracy_std_dev": 0.024722754252712212,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.11522000200000002,
      "avg_duration_per_task_ms": 90418.644,
      "avg_output_tokens_per_task": 9963.432,
      "last_run_timestamp": "2026-10-31T19:57:32.812Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "BaseTen",
      "accuracy": 0.6972222117283948,
      "accuracy_std_dev": 0.0265134534342492,
      "run_count": 33,
      "total_questions": 1620,
      "avg_cost_per_task_usd": 0.10315858827160493,
      "avg_duration_per_task_ms": 72358.56851851851,
      "avg_output_tokens_per_task": 8108.612962962963,
      "last_run_timestamp": "2026-10-31T20:07:10.133Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-luna-20260709",
      "model_slug": "openai/gpt-5.6-luna",
      "model_name": "OpenAI: GPT-5.6 Luna",
      "provider_name": "auto-routing",
      "accuracy": 0.696667,
      "accuracy_std_dev": 0.010000000000000009,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01321477,
      "avg_duration_per_task_ms": 116452.62,
      "avg_output_tokens_per_task": 5139.02,
      "last_run_timestamp": "2026-10-01T19:04:36.191Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Decart",
      "accuracy": 0.6966665384615385,
      "accuracy_std_dev": 0.021162820164747506,
      "run_count": 13,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.1543893523076923,
      "avg_duration_per_task_ms": 76695.08,
      "avg_output_tokens_per_task": 10682.713846153847,
      "last_run_timestamp": "2026-10-24T19:06:47.526Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-astra-20260903",
      "model_slug": "openai/gpt-6-astra",
      "model_name": "OpenAI: GPT-6 Astra",
      "provider_name": "Azure",
      "accuracy": 0.6966665000000001,
      "accuracy_std_dev": 0.0033334999999999892,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.62488186,
      "avg_duration_per_task_ms": 152081.62,
      "avg_output_tokens_per_task": 3927.45,
      "last_run_timestamp": "2026-10-01T08:05:41.499Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "Venice",
      "accuracy": 0.6961905714285714,
      "accuracy_std_dev": 0.027338489642977408,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.045278854285714286,
      "avg_duration_per_task_ms": 549308.6714285715,
      "avg_output_tokens_per_task": 9042.874285714286,
      "last_run_timestamp": "2026-10-31T20:03:54.586Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Venice",
      "accuracy": 0.696,
      "accuracy_std_dev": 0.036049398208569304,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.11530058400000001,
      "avg_duration_per_task_ms": 126878.5,
      "avg_output_tokens_per_task": 9663.674,
      "last_run_timestamp": "2026-10-31T21:47:15.954Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "InferenceNet",
      "accuracy": 0.6957762602739725,
      "accuracy_std_dev": 0.0313433490574777,
      "run_count": 6,
      "total_questions": 292,
      "avg_cost_per_task_usd": 0.011891609589041095,
      "avg_duration_per_task_ms": 271916.02397260274,
      "avg_output_tokens_per_task": 12921.842465753425,
      "last_run_timestamp": "2026-11-04T21:26:44.978Z"
    },
    {
      "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": 0.6957576363636363,
      "accuracy_std_dev": 0.027600759894916997,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.07744747818181819,
      "avg_duration_per_task_ms": 159605.20727272728,
      "avg_output_tokens_per_task": 9554.076363636364,
      "last_run_timestamp": "2026-10-31T19:41:05.183Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.69499975,
      "accuracy_std_dev": 0.014813722993815583,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.02174445,
      "avg_duration_per_task_ms": 448918.0325,
      "avg_output_tokens_per_task": 18617.01,
      "last_run_timestamp": "2026-10-31T19:50:01.310Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Friendli",
      "accuracy": 0.6945454545454546,
      "accuracy_std_dev": 0.05278641023693957,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.11215429818181818,
      "avg_duration_per_task_ms": 85066.3509090909,
      "avg_output_tokens_per_task": 9316.025454545455,
      "last_run_timestamp": "2026-10-31T19:47:27.755Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Ambient",
      "accuracy": 0.6942857142857143,
      "accuracy_std_dev": 0.025305290602753317,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.049070077142857146,
      "avg_duration_per_task_ms": 249231.63428571427,
      "avg_output_tokens_per_task": 8147.505714285714,
      "last_run_timestamp": "2026-10-31T20:09:59.492Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Friendli",
      "accuracy": 0.6941667499999999,
      "accuracy_std_dev": 0.05957600071494812,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0977737325,
      "avg_duration_per_task_ms": 138301.6425,
      "avg_output_tokens_per_task": 9130.415,
      "last_run_timestamp": "2026-10-31T19:19:57.503Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "StreamLake",
      "accuracy": 0.6933334999999999,
      "accuracy_std_dev": 0.026666499999999982,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.04279356,
      "avg_duration_per_task_ms": 161286,
      "avg_output_tokens_per_task": 8223.72,
      "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": 0.6933334999999999,
      "accuracy_std_dev": 0.018257236010689022,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.08853248000000001,
      "avg_duration_per_task_ms": 338024.125,
      "avg_output_tokens_per_task": 32557.16,
      "last_run_timestamp": "2026-10-01T16:01:04.829Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-luna-pro-20260922",
      "model_slug": "openai/gpt-6-luna-pro",
      "model_name": "OpenAI: GPT-6 Luna Pro",
      "provider_name": "OpenAI",
      "accuracy": 0.6933333333333334,
      "accuracy_std_dev": 0.014401903168987411,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.03953364,
      "avg_duration_per_task_ms": 309829.13333333336,
      "avg_output_tokens_per_task": 32261.04,
      "last_run_timestamp": "2026-10-02T23:10:33.891Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "DeepInfra",
      "accuracy": 0.693333,
      "accuracy_std_dev": 0.020000000000000018,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.02093496,
      "avg_duration_per_task_ms": 185229.42,
      "avg_output_tokens_per_task": 9197.38,
      "last_run_timestamp": "2026-08-29T16:45:30.973Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Parasail",
      "accuracy": 0.693333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.054513559999999996,
      "avg_duration_per_task_ms": 221323.18,
      "avg_output_tokens_per_task": 6973.32,
      "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": "openai/gpt-6-sol-20260922",
      "model_slug": "openai/gpt-6-sol",
      "model_name": "OpenAI: GPT-6 Sol",
      "provider_name": "auto-routing",
      "accuracy": 0.693333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.12999802000000002,
      "avg_duration_per_task_ms": 96250.54,
      "avg_output_tokens_per_task": 4982.06,
      "last_run_timestamp": "2026-10-01T08:05:35.867Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "reflection/beam-1.0-501b-a23b-20261016",
      "model_slug": "reflection/beam-1.0-501b-a23b",
      "model_name": "Reflection: Beam 1.0 501B A23B",
      "provider_name": "auto-routing",
      "accuracy": 0.693333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.02754302,
      "avg_duration_per_task_ms": 209408.8,
      "avg_output_tokens_per_task": 14914.54,
      "last_run_timestamp": "2026-10-24T12:03:24.771Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.7-flash-20260119",
      "model_slug": "z-ai/glm-4.7-flash",
      "model_name": "Z.ai: GLM 4.7 Flash",
      "provider_name": "Venice",
      "accuracy": 0.693333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.00621348,
      "avg_duration_per_task_ms": 180391.96,
      "avg_output_tokens_per_task": 9984.56,
      "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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Cloudflare",
      "accuracy": 0.6930555833333333,
      "accuracy_std_dev": 0.0348665084368422,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.12472750499999999,
      "avg_duration_per_task_ms": 130389.57833333334,
      "avg_output_tokens_per_task": 9036.746666666666,
      "last_run_timestamp": "2026-10-31T20:01:39.270Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "DeepInfra",
      "accuracy": 0.6926297185929648,
      "accuracy_std_dev": 0.014566408256670558,
      "run_count": 4,
      "total_questions": 199,
      "avg_cost_per_task_usd": 0.04924894974874372,
      "avg_duration_per_task_ms": 416469.09547738696,
      "avg_output_tokens_per_task": 8270.989949748744,
      "last_run_timestamp": "2026-10-31T21:01:44.534Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "DigitalOcean",
      "accuracy": 0.6918933673469387,
      "accuracy_std_dev": 0.11325810487735402,
      "run_count": 12,
      "total_questions": 588,
      "avg_cost_per_task_usd": 0.3106834897959183,
      "avg_duration_per_task_ms": 1590211.7602040817,
      "avg_output_tokens_per_task": 51719.284013605444,
      "last_run_timestamp": "2026-10-31T12:54:39.225Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "BaseTen",
      "accuracy": 0.6909721458333333,
      "accuracy_std_dev": 0.0024635000000000074,
      "run_count": 2,
      "total_questions": 96,
      "avg_cost_per_task_usd": 0.0246803125,
      "avg_duration_per_task_ms": 48638.177083333336,
      "avg_output_tokens_per_task": 11672.458333333334,
      "last_run_timestamp": "2026-10-31T11:12:51.723Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.1-flash-lite-20260507",
      "model_slug": "google/gemini-3.1-flash-lite",
      "model_name": "Google: Gemini 3.1 Flash Lite",
      "provider_name": "Google AI Studio",
      "accuracy": 0.69,
      "accuracy_std_dev": 0.003332999999999975,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.10882898999999999,
      "avg_duration_per_task_ms": 135482.83,
      "avg_output_tokens_per_task": 51335.48,
      "last_run_timestamp": "2026-10-01T19:05:16.548Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Minimax",
      "accuracy": 0.69,
      "accuracy_std_dev": 0.009999999999999953,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01750064,
      "avg_duration_per_task_ms": 190406.46,
      "avg_output_tokens_per_task": 15714.76,
      "last_run_timestamp": "2026-10-31T12:31:30.661Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.69,
      "accuracy_std_dev": 0.016667000000000043,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01877389,
      "avg_duration_per_task_ms": 125629.58,
      "avg_output_tokens_per_task": 15732.46,
      "last_run_timestamp": "2026-10-31T11:20:36.173Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-astra-pro-20260903",
      "model_slug": "openai/gpt-6-astra-pro",
      "model_name": "OpenAI: GPT-6 Astra Pro",
      "provider_name": "Azure",
      "accuracy": 0.69,
      "accuracy_std_dev": 0.009999999999999953,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 2.99530482,
      "avg_duration_per_task_ms": 320001.13,
      "avg_output_tokens_per_task": 12475.35,
      "last_run_timestamp": "2026-10-01T08:05:20.373Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "thinkingmachines/inkling-20260715",
      "model_slug": "thinkingmachines/inkling",
      "model_name": "Thinking Machines: Inkling",
      "provider_name": "BaseTen",
      "accuracy": 0.69,
      "accuracy_std_dev": 0.04333300000000001,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.07366,
      "avg_duration_per_task_ms": 63569.01,
      "avg_output_tokens_per_task": 3649.5,
      "last_run_timestamp": "2026-09-12T08:11:26.919Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Mistral",
      "accuracy": 0.6899999117647057,
      "accuracy_std_dev": 0.026333763959163144,
      "run_count": 34,
      "total_questions": 1700,
      "avg_cost_per_task_usd": 0.10328957000000001,
      "avg_duration_per_task_ms": 78259.03647058824,
      "avg_output_tokens_per_task": 9525.96294117647,
      "last_run_timestamp": "2026-11-11T03:23:11.016Z"
    },
    {
      "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": 0.6895238571428572,
      "accuracy_std_dev": 0.0390126672850483,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.06955698857142856,
      "avg_duration_per_task_ms": 75132.57428571429,
      "avg_output_tokens_per_task": 10137.50857142857,
      "last_run_timestamp": "2026-10-31T19:23:52.545Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "GMICloud",
      "accuracy": 0.687499875,
      "accuracy_std_dev": 0.10511576786029475,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.01592138,
      "avg_duration_per_task_ms": 247628.705,
      "avg_output_tokens_per_task": 11536.2,
      "last_run_timestamp": "2026-11-10T07:17:44.193Z"
    },
    {
      "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": 0.6872223333333333,
      "accuracy_std_dev": 0.08398462234374948,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.044577226666666664,
      "avg_duration_per_task_ms": 505556.9866666667,
      "avg_output_tokens_per_task": 15808.26,
      "last_run_timestamp": "2026-10-31T19:37:26.520Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Wafer",
      "accuracy": 0.6866665000000001,
      "accuracy_std_dev": 0.026666499999999982,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01561806,
      "avg_duration_per_task_ms": 117250.14,
      "avg_output_tokens_per_task": 14519.81,
      "last_run_timestamp": "2026-10-24T12:06:50.408Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "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": 0.6847617142857143,
      "accuracy_std_dev": 0.047771144717629,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.01437896,
      "avg_duration_per_task_ms": 347803.89714285714,
      "avg_output_tokens_per_task": 20220.391428571427,
      "last_run_timestamp": "2026-10-31T20:18:22.353Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "GMICloud",
      "accuracy": 0.6844444166666667,
      "accuracy_std_dev": 0.04779057451432995,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.10127122666666667,
      "avg_duration_per_task_ms": 175036.245,
      "avg_output_tokens_per_task": 7013.575,
      "last_run_timestamp": "2026-11-10T06:53:22.469Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6.1-terra-20261026",
      "model_slug": "openai/gpt-6.1-terra",
      "model_name": "OpenAI: GPT-6.1 Terra",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.6844443333333333,
      "accuracy_std_dev": 0.016178098288186456,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.06359924333333333,
      "avg_duration_per_task_ms": 60968.18666666667,
      "avg_output_tokens_per_task": 4743.64,
      "last_run_timestamp": "2026-11-02T17:36:19.479Z"
    },
    {
      "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": 0.6841667499999999,
      "accuracy_std_dev": 0.019982613972838976,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.00873134,
      "avg_duration_per_task_ms": 302344.745,
      "avg_output_tokens_per_task": 8518.3375,
      "last_run_timestamp": "2026-11-03T03:04:12.747Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.6833334999999999,
      "accuracy_std_dev": 0.023333500000000007,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.018626550000000002,
      "avg_duration_per_task_ms": 239747.89,
      "avg_output_tokens_per_task": 15949.04,
      "last_run_timestamp": "2026-10-31T19:36:36.917Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Venice",
      "accuracy": 0.6833334999999999,
      "accuracy_std_dev": 0.03666649999999999,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.0168126,
      "avg_duration_per_task_ms": 195045.15,
      "avg_output_tokens_per_task": 9021.87,
      "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": "openai/gpt-6.1-sol-20260929",
      "model_slug": "openai/gpt-6.1-sol",
      "model_name": "OpenAI: GPT-6.1 Sol",
      "provider_name": "Azure",
      "accuracy": 0.6833333333333333,
      "accuracy_std_dev": 0.012619766977070334,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.10710260999999997,
      "avg_duration_per_task_ms": 135398.91,
      "avg_output_tokens_per_task": 4027.3166666666666,
      "last_run_timestamp": "2026-10-01T08:06:09.665Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-luna-20260709",
      "model_slug": "openai/gpt-5.6-luna",
      "model_name": "OpenAI: GPT-5.6 Luna",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.6822223333333333,
      "accuracy_std_dev": 0.021998563367234276,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.014046139999999999,
      "avg_duration_per_task_ms": 71689.16,
      "avg_output_tokens_per_task": 5022.28,
      "last_run_timestamp": "2026-10-30T00:15:40.872Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6.1-sol-20260929",
      "model_slug": "openai/gpt-6.1-sol",
      "model_name": "OpenAI: GPT-6.1 Sol",
      "provider_name": "OpenAI",
      "accuracy": 0.6822223333333333,
      "accuracy_std_dev": 0.011330936570096727,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.09746202,
      "avg_duration_per_task_ms": 132162.47333333333,
      "avg_output_tokens_per_task": 3507.6066666666666,
      "last_run_timestamp": "2026-10-02T23:10:34.290Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.6817421155778893,
      "accuracy_std_dev": 0.023817302869644456,
      "run_count": 8,
      "total_questions": 398,
      "avg_cost_per_task_usd": 0.02058146231155779,
      "avg_duration_per_task_ms": 307187.45477386937,
      "avg_output_tokens_per_task": 16771.46984924623,
      "last_run_timestamp": "2026-10-31T20:14:20.109Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "thinkingmachines/inkling-20260715",
      "model_slug": "thinkingmachines/inkling",
      "model_name": "Thinking Machines: Inkling",
      "provider_name": "Together",
      "accuracy": 0.6813334,
      "accuracy_std_dev": 0.0026667999999999913,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.08704178,
      "avg_duration_per_task_ms": 82752.556,
      "avg_output_tokens_per_task": 3576.016,
      "last_run_timestamp": "2026-10-01T16:01:04.859Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "CoreWeave",
      "accuracy": 0.68125,
      "accuracy_std_dev": 0.0202715728731147,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0174918175,
      "avg_duration_per_task_ms": 123091.5525,
      "avg_output_tokens_per_task": 21340.095,
      "last_run_timestamp": "2026-10-31T11:46:38.147Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-luna-20260922",
      "model_slug": "openai/gpt-6-luna",
      "model_name": "OpenAI: GPT-6 Luna",
      "provider_name": "Azure",
      "accuracy": 0.6811111666666666,
      "accuracy_std_dev": 0.008088865998326882,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.00903582,
      "avg_duration_per_task_ms": 105469.92666666667,
      "avg_output_tokens_per_task": 8459.866666666667,
      "last_run_timestamp": "2026-10-01T08:05:22.516Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "AtlasCloud",
      "accuracy": 0.6809524285714286,
      "accuracy_std_dev": 0.09813702854661528,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.09356050285714285,
      "avg_duration_per_task_ms": 244745.57714285713,
      "avg_output_tokens_per_task": 21782.17142857143,
      "last_run_timestamp": "2026-10-31T19:46:05.991Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.6808944146341463,
      "accuracy_std_dev": 0.01806743167802219,
      "run_count": 5,
      "total_questions": 246,
      "avg_cost_per_task_usd": 0.033741605691056914,
      "avg_duration_per_task_ms": 453897.35772357724,
      "avg_output_tokens_per_task": 18491.044715447155,
      "last_run_timestamp": "2026-10-17T19:21:01.980Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-fable-5.2-20261109",
      "model_slug": "anthropic/claude-fable-5.2",
      "model_name": "Anthropic: Claude Fable 5.2",
      "provider_name": "Google",
      "accuracy": 0.68,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.79080038,
      "avg_duration_per_task_ms": 125754.98,
      "avg_output_tokens_per_task": 9356.94,
      "last_run_timestamp": "2026-11-10T18:07:28.806Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-fable-5.2-20261109",
      "model_slug": "anthropic/claude-fable-5.2",
      "model_name": "Anthropic: Claude Fable 5.2",
      "provider_name": "Azure",
      "accuracy": 0.68,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.77161558,
      "avg_duration_per_task_ms": 159103.76,
      "avg_output_tokens_per_task": 9152.76,
      "last_run_timestamp": "2026-11-10T18:07:20.112Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-astra-20260903",
      "model_slug": "openai/gpt-6-astra",
      "model_name": "OpenAI: GPT-6 Astra",
      "provider_name": "auto-routing",
      "accuracy": 0.68,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.58024224,
      "avg_duration_per_task_ms": 141487.98,
      "avg_output_tokens_per_task": 3731.58,
      "last_run_timestamp": "2026-10-01T08:05:19.702Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-luna-20260922",
      "model_slug": "openai/gpt-6-luna",
      "model_name": "OpenAI: GPT-6 Luna",
      "provider_name": "OpenAI",
      "accuracy": 0.68,
      "accuracy_std_dev": 0.03810292105162893,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.00883898,
      "avg_duration_per_task_ms": 114396.12,
      "avg_output_tokens_per_task": 8841.28,
      "last_run_timestamp": "2026-10-02T23:08:44.329Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "DekaLLM",
      "accuracy": 0.6799998571428572,
      "accuracy_std_dev": 0.0198405229080613,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.015568199999999999,
      "avg_duration_per_task_ms": 259045.84285714285,
      "avg_output_tokens_per_task": 18504.7,
      "last_run_timestamp": "2026-10-31T20:09:24.664Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "NextBit",
      "accuracy": 0.6793334,
      "accuracy_std_dev": 0.03508080956933576,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.016317446,
      "avg_duration_per_task_ms": 294258.966,
      "avg_output_tokens_per_task": 19934.788,
      "last_run_timestamp": "2026-11-01T13:16:57.769Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Tenstorrent",
      "accuracy": 0.67833325,
      "accuracy_std_dev": 0.019075602511781895,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.01068972,
      "avg_duration_per_task_ms": 147233.725,
      "avg_output_tokens_per_task": 15363.5,
      "last_run_timestamp": "2026-11-04T18:22:16.954Z"
    },
    {
      "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-6.1-terra-20261026",
      "model_slug": "openai/gpt-6.1-terra",
      "model_name": "OpenAI: GPT-6.1 Terra",
      "provider_name": "Azure",
      "accuracy": 0.6777776666666667,
      "accuracy_std_dev": 0.017497908167041606,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.06315626,
      "avg_duration_per_task_ms": 88978.8,
      "avg_output_tokens_per_task": 4762.613333333334,
      "last_run_timestamp": "2026-11-03T15:08:23.560Z"
    },
    {
      "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": 0.6767563901345292,
      "accuracy_std_dev": 0.029269691852423077,
      "run_count": 9,
      "total_questions": 446,
      "avg_cost_per_task_usd": 0.019482901345291483,
      "avg_duration_per_task_ms": 358550.09865470853,
      "avg_output_tokens_per_task": 16851.491031390135,
      "last_run_timestamp": "2026-11-11T08:13:05.865Z"
    },
    {
      "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": 0.676667,
      "accuracy_std_dev": 0.030000000000000027,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.02188514,
      "avg_duration_per_task_ms": 255892.84,
      "avg_output_tokens_per_task": 10205.01,
      "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-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": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "DigitalOcean",
      "accuracy": 0.6755556666666667,
      "accuracy_std_dev": 0.00831485358994507,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.014157986666666667,
      "avg_duration_per_task_ms": 117427.1,
      "avg_output_tokens_per_task": 13587.66,
      "last_run_timestamp": "2026-10-31T11:44:53.544Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-astra-20260903",
      "model_slug": "openai/gpt-6-astra",
      "model_name": "OpenAI: GPT-6 Astra",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.6755556666666667,
      "accuracy_std_dev": 0.006285236475706844,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.63842362,
      "avg_duration_per_task_ms": 88201.35333333333,
      "avg_output_tokens_per_task": 3898.48,
      "last_run_timestamp": "2026-10-30T00:17:00.667Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "thinkingmachines/inkling-20260715",
      "model_slug": "thinkingmachines/inkling",
      "model_name": "Thinking Machines: Inkling",
      "provider_name": "DeepInfra",
      "accuracy": 0.6755556666666667,
      "accuracy_std_dev": 0.006285236475706844,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.07120372000000001,
      "avg_duration_per_task_ms": 144319.38666666666,
      "avg_output_tokens_per_task": 2561.846666666667,
      "last_run_timestamp": "2026-10-08T22:25:52.178Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "GMICloud",
      "accuracy": 0.6738889166666666,
      "accuracy_std_dev": 0.14166895499276846,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.11663828999999999,
      "avg_duration_per_task_ms": 354430.585,
      "avg_output_tokens_per_task": 8766.405,
      "last_run_timestamp": "2026-11-10T06:25:16.940Z"
    },
    {
      "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": 0.6734008686868687,
      "accuracy_std_dev": 0.01340150000000001,
      "run_count": 2,
      "total_questions": 99,
      "avg_cost_per_task_usd": 0.006025121212121212,
      "avg_duration_per_task_ms": 197802.63636363635,
      "avg_output_tokens_per_task": 8819.454545454546,
      "last_run_timestamp": "2026-08-30T00:01:09.439Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5-2025-08-07",
      "model_slug": "openai/gpt-5",
      "model_name": "OpenAI: GPT-5",
      "provider_name": "auto-routing",
      "accuracy": 0.6733334999999999,
      "accuracy_std_dev": 0.026666499999999982,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.6292479799999999,
      "avg_duration_per_task_ms": 729264.79,
      "avg_output_tokens_per_task": 54931.75,
      "last_run_timestamp": "2026-10-01T16:01:02.610Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Decart",
      "accuracy": 0.6733333333333333,
      "accuracy_std_dev": 0.14616051044960432,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.07790980666666668,
      "avg_duration_per_task_ms": 178517.31333333332,
      "avg_output_tokens_per_task": 18244.56,
      "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": "openai/gpt-5.6-sol-pro-20260709",
      "model_slug": "openai/gpt-5.6-sol-pro",
      "model_name": "OpenAI: GPT-5.6 Sol Pro",
      "provider_name": "Azure",
      "accuracy": 0.6733330000000001,
      "accuracy_std_dev": 0.019999999999999962,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 1.5819717300000002,
      "avg_duration_per_task_ms": 314318.37,
      "avg_output_tokens_per_task": 23725.55,
      "last_run_timestamp": "2026-10-01T16:01:13.647Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-luna-20260922",
      "model_slug": "openai/gpt-6-luna",
      "model_name": "OpenAI: GPT-6 Luna",
      "provider_name": "auto-routing",
      "accuracy": 0.673333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.00928508,
      "avg_duration_per_task_ms": 121838.98,
      "avg_output_tokens_per_task": 8898.44,
      "last_run_timestamp": "2026-10-01T08:00:22.469Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Io Net",
      "accuracy": 0.673333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.0059145000000000005,
      "avg_duration_per_task_ms": 456667.98,
      "avg_output_tokens_per_task": 17189.68,
      "last_run_timestamp": "2026-11-11T15:53:31.309Z"
    },
    {
      "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": 0.6725294120603015,
      "accuracy_std_dev": 0.11099160244860588,
      "run_count": 8,
      "total_questions": 398,
      "avg_cost_per_task_usd": 0.03270493216080402,
      "avg_duration_per_task_ms": 1961063.7788944724,
      "avg_output_tokens_per_task": 69160.1582914573,
      "last_run_timestamp": "2026-10-31T11:55:35.800Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Minimax",
      "accuracy": 0.67166675,
      "accuracy_std_dev": 0.012802094015726486,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.035549775,
      "avg_duration_per_task_ms": 154518.065,
      "avg_output_tokens_per_task": 9811.9,
      "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": "Cloudflare",
      "accuracy": 0.671111,
      "accuracy_std_dev": 0.016629528997138363,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.013801693333333332,
      "avg_duration_per_task_ms": 193265.92,
      "avg_output_tokens_per_task": 8641.586666666666,
      "last_run_timestamp": "2026-10-13T22:01:15.953Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-sol-20260922",
      "model_slug": "openai/gpt-6-sol",
      "model_name": "OpenAI: GPT-6 Sol",
      "provider_name": "Azure",
      "accuracy": 0.6706668,
      "accuracy_std_dev": 0.015548970672041286,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.142799016,
      "avg_duration_per_task_ms": 115124.684,
      "avg_output_tokens_per_task": 5068.812,
      "last_run_timestamp": "2026-10-01T08:04:38.837Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "Google",
      "accuracy": 0.6702638057553957,
      "accuracy_std_dev": 0.01542992310055007,
      "run_count": 3,
      "total_questions": 139,
      "avg_cost_per_task_usd": 0.11313071942446042,
      "avg_duration_per_task_ms": 363635,
      "avg_output_tokens_per_task": 8911.287769784172,
      "last_run_timestamp": "2026-10-17T19:51:53.622Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.4-mini-20260317",
      "model_slug": "openai/gpt-5.4-mini",
      "model_name": "OpenAI: GPT-5.4 Mini",
      "provider_name": "Azure",
      "accuracy": 0.6683335000000001,
      "accuracy_std_dev": 0.01190259075789804,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.097154565,
      "avg_duration_per_task_ms": 138323.315,
      "avg_output_tokens_per_task": 13048.88,
      "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-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "GMICloud",
      "accuracy": 0.668,
      "accuracy_std_dev": 0.08103784026983935,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.021783728,
      "avg_duration_per_task_ms": 280094.124,
      "avg_output_tokens_per_task": 19729.836,
      "last_run_timestamp": "2026-11-10T06:31:54.090Z"
    },
    {
      "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": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.6666666666666666,
      "accuracy_std_dev": 0.0524933858267454,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.03491572666666667,
      "avg_duration_per_task_ms": 274075.64666666667,
      "avg_output_tokens_per_task": 23860.97333333333,
      "last_run_timestamp": "2026-10-31T19:43:40.322Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "DigitalOcean",
      "accuracy": 0.6666665,
      "accuracy_std_dev": 0.006666499999999964,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.02906171,
      "avg_duration_per_task_ms": 246294.85,
      "avg_output_tokens_per_task": 9968.3,
      "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": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.6666665,
      "accuracy_std_dev": 0.006666499999999964,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.013174700000000001,
      "avg_duration_per_task_ms": 181950.26,
      "avg_output_tokens_per_task": 13269.31,
      "last_run_timestamp": "2026-10-31T12:42:39.997Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.4-mini-20260317",
      "model_slug": "openai/gpt-5.4-mini",
      "model_name": "OpenAI: GPT-5.4 Mini",
      "provider_name": "auto-routing",
      "accuracy": 0.6666665,
      "accuracy_std_dev": 0.013333499999999998,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.08889954,
      "avg_duration_per_task_ms": 108433.06,
      "avg_output_tokens_per_task": 12861.73,
      "last_run_timestamp": "2026-10-01T19:04:42.468Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.6666665,
      "accuracy_std_dev": 0.013333499999999998,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.00821814,
      "avg_duration_per_task_ms": 219118.91,
      "avg_output_tokens_per_task": 8875.03,
      "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Near AI",
      "accuracy": 0.66583325,
      "accuracy_std_dev": 0.03639120582211448,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.006095729999999999,
      "avg_duration_per_task_ms": 325369.79,
      "avg_output_tokens_per_task": 18643.27,
      "last_run_timestamp": "2026-10-31T19:46:03.854Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "Darkbloom",
      "accuracy": 0.6633329999999998,
      "accuracy_std_dev": 0.021602468994692835,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.007094233333333332,
      "avg_duration_per_task_ms": 456871.32,
      "avg_output_tokens_per_task": 20116.046666666665,
      "last_run_timestamp": "2026-11-13T01:53:22.455Z"
    },
    {
      "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": 0.6622221666666667,
      "accuracy_std_dev": 0.030323393570073177,
      "run_count": 30,
      "total_questions": 1500,
      "avg_cost_per_task_usd": 0.12413147533333334,
      "avg_duration_per_task_ms": 118698.298,
      "avg_output_tokens_per_task": 12588.784666666666,
      "last_run_timestamp": "2026-10-31T12:26:00.225Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "DekaLLM",
      "accuracy": 0.6604760000000001,
      "accuracy_std_dev": 0.025350253033394797,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.008717225714285714,
      "avg_duration_per_task_ms": 574345.4257142857,
      "avg_output_tokens_per_task": 17634.762857142858,
      "last_run_timestamp": "2026-10-31T20:07:05.983Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "SiliconFlow",
      "accuracy": 0.66,
      "accuracy_std_dev": 0.020000000000000018,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.022123540000000004,
      "avg_duration_per_task_ms": 232117.3,
      "avg_output_tokens_per_task": 9442.58,
      "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-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Morph",
      "accuracy": 0.66,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.06369964,
      "avg_duration_per_task_ms": 706011.28,
      "avg_output_tokens_per_task": 8805.34,
      "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": "upstage/solar-pro4-20260810",
      "model_slug": "upstage/solar-pro4",
      "model_name": "Upstage: Solar Pro 4",
      "provider_name": "KT",
      "accuracy": 0.66,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.0747839,
      "avg_duration_per_task_ms": 223467.62,
      "avg_output_tokens_per_task": 14969.76,
      "last_run_timestamp": "2026-11-11T02:34:15.627Z"
    },
    {
      "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": 0.6580952857142857,
      "accuracy_std_dev": 0.07221803406691293,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.056178177142857146,
      "avg_duration_per_task_ms": 795605.24,
      "avg_output_tokens_per_task": 17815.014285714286,
      "last_run_timestamp": "2026-10-05T20:43:11.175Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.5-haiku-20251001",
      "model_slug": "anthropic/claude-haiku-4.5",
      "model_name": "Anthropic: Claude Haiku 4.5",
      "provider_name": "Anthropic",
      "accuracy": 0.6580001,
      "accuracy_std_dev": 0.027656551565406706,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.11552560199999999,
      "avg_duration_per_task_ms": 129975.578,
      "avg_output_tokens_per_task": 12464.086,
      "last_run_timestamp": "2026-10-31T11:44:30.576Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "thinkingmachines/inkling-small-20260730",
      "model_slug": "thinkingmachines/inkling-small",
      "model_name": "Thinking Machines: Inkling Small",
      "provider_name": "Together",
      "accuracy": 0.657778,
      "accuracy_std_dev": 0.0505769576190581,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.046507033333333336,
      "avg_duration_per_task_ms": 270981.30666666664,
      "avg_output_tokens_per_task": 3230.153333333333,
      "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Parasail",
      "accuracy": 0.6574998750000001,
      "accuracy_std_dev": 0.012555363509646976,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0252459125,
      "avg_duration_per_task_ms": 148223.2925,
      "avg_output_tokens_per_task": 17051.0175,
      "last_run_timestamp": "2026-11-05T00:31:57.460Z"
    },
    {
      "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": 0.6573334,
      "accuracy_std_dev": 0.1373785957776538,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.06366923599999999,
      "avg_duration_per_task_ms": 573529.892,
      "avg_output_tokens_per_task": 10164.08,
      "last_run_timestamp": "2026-10-31T19:32:40.676Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "AkashML",
      "accuracy": 0.6573334,
      "accuracy_std_dev": 0.02284235636939995,
      "run_count": 15,
      "total_questions": 750,
      "avg_cost_per_task_usd": 0.009446806666666667,
      "avg_duration_per_task_ms": 308475.5133333333,
      "avg_output_tokens_per_task": 17229.645333333334,
      "last_run_timestamp": "2026-10-31T19:05:31.409Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.6566666,
      "accuracy_std_dev": 0.02720294101820609,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.11512897400000001,
      "avg_duration_per_task_ms": 122868.958,
      "avg_output_tokens_per_task": 12575.444,
      "last_run_timestamp": "2026-10-31T12:42:18.189Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Friendli",
      "accuracy": 0.6566665,
      "accuracy_std_dev": 0.0033334999999999892,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.029754,
      "avg_duration_per_task_ms": 250765.5,
      "avg_output_tokens_per_task": 11169.75,
      "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": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "SiliconFlow",
      "accuracy": 0.656492308139535,
      "accuracy_std_dev": 0.13396399619194066,
      "run_count": 7,
      "total_questions": 344,
      "avg_cost_per_task_usd": 0.02718404941860465,
      "avg_duration_per_task_ms": 1177043.0639534884,
      "avg_output_tokens_per_task": 35003.26453488372,
      "last_run_timestamp": "2026-10-17T21:28:32.796Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Crusoe",
      "accuracy": 0.6563219310344829,
      "accuracy_std_dev": 0.17320901072467976,
      "run_count": 3,
      "total_questions": 145,
      "avg_cost_per_task_usd": 0.02278490344827586,
      "avg_duration_per_task_ms": 272417.2068965517,
      "avg_output_tokens_per_task": 8725.041379310345,
      "last_run_timestamp": "2026-10-31T19:23:32.973Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "auto-routing",
      "accuracy": 0.6562239298245615,
      "accuracy_std_dev": 0.018946289616967236,
      "run_count": 8,
      "total_questions": 399,
      "avg_cost_per_task_usd": 0.01692722807017544,
      "avg_duration_per_task_ms": 312804.7944862155,
      "avg_output_tokens_per_task": 22811.754385964912,
      "last_run_timestamp": "2026-10-31T12:28:49.050Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.5-haiku-20251001",
      "model_slug": "anthropic/claude-haiku-4.5",
      "model_name": "Anthropic: Claude Haiku 4.5",
      "provider_name": "Azure",
      "accuracy": 0.6560001,
      "accuracy_std_dev": 0.022745126952602394,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.113423644,
      "avg_duration_per_task_ms": 134974.778,
      "avg_output_tokens_per_task": 12396.082,
      "last_run_timestamp": "2026-10-31T11:42:13.882Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "Phala",
      "accuracy": 0.6557823102040816,
      "accuracy_std_dev": 0.009318408855593308,
      "run_count": 5,
      "total_questions": 245,
      "avg_cost_per_task_usd": 0.014797057142857142,
      "avg_duration_per_task_ms": 277907.2775510204,
      "avg_output_tokens_per_task": 20245.579591836733,
      "last_run_timestamp": "2026-10-10T11:18:54.565Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "DeepInfra",
      "accuracy": 0.6555556666666666,
      "accuracy_std_dev": 0.03807065880409451,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.021039175555555557,
      "avg_duration_per_task_ms": 248791.5977777778,
      "avg_output_tokens_per_task": 26269.908888888887,
      "last_run_timestamp": "2026-10-31T12:25:21.635Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.4-nano-20260317",
      "model_slug": "openai/gpt-5.4-nano",
      "model_name": "OpenAI: GPT-5.4 Nano",
      "provider_name": "OpenAI",
      "accuracy": 0.6555556666666666,
      "accuracy_std_dev": 0.012570472951413636,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.030086579999999998,
      "avg_duration_per_task_ms": 125366.48,
      "avg_output_tokens_per_task": 16133.046666666667,
      "last_run_timestamp": "2026-10-02T23:09:54.584Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.5-haiku-20251001",
      "model_slug": "anthropic/claude-haiku-4.5",
      "model_name": "Anthropic: Claude Haiku 4.5",
      "provider_name": "Google",
      "accuracy": 0.6546666666666666,
      "accuracy_std_dev": 0.025785448094268634,
      "run_count": 30,
      "total_questions": 1500,
      "avg_cost_per_task_usd": 0.11997013933333336,
      "avg_duration_per_task_ms": 121089.03933333333,
      "avg_output_tokens_per_task": 12528.349333333334,
      "last_run_timestamp": "2026-10-31T12:17:41.489Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Crusoe",
      "accuracy": 0.6546664,
      "accuracy_std_dev": 0.03512514515614133,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.01491196,
      "avg_duration_per_task_ms": 86567.944,
      "avg_output_tokens_per_task": 19737.508,
      "last_run_timestamp": "2026-10-31T19:08:06.670Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.6535409120879121,
      "accuracy_std_dev": 0.03689102528724027,
      "run_count": 11,
      "total_questions": 546,
      "avg_cost_per_task_usd": 0.012232007326007327,
      "avg_duration_per_task_ms": 1066692.706959707,
      "avg_output_tokens_per_task": 22783.287545787545,
      "last_run_timestamp": "2026-11-04T19:03:50.422Z"
    },
    {
      "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-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Sail Research",
      "accuracy": 0.6533335,
      "accuracy_std_dev": 0.013333499999999998,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01241216,
      "avg_duration_per_task_ms": 170776.85,
      "avg_output_tokens_per_task": 17557.16,
      "last_run_timestamp": "2026-11-06T22:04:56.505Z"
    },
    {
      "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": 0.6533335,
      "accuracy_std_dev": 0.013333499999999998,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.00878474,
      "avg_duration_per_task_ms": 423822.14,
      "avg_output_tokens_per_task": 14588.95,
      "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": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "Venice",
      "accuracy": 0.653333375,
      "accuracy_std_dev": 0.07817377366472963,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0266300325,
      "avg_duration_per_task_ms": 292622.5475,
      "avg_output_tokens_per_task": 25080.435,
      "last_run_timestamp": "2026-10-17T19:25:29.604Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4-sonnet-20250522",
      "model_slug": "anthropic/claude-sonnet-4",
      "model_name": "Anthropic: Claude Sonnet 4",
      "provider_name": "Google",
      "accuracy": 0.6533333333333333,
      "accuracy_std_dev": 0.014401440259764123,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.2856746933333333,
      "avg_duration_per_task_ms": 183725.1,
      "avg_output_tokens_per_task": 8628.813333333334,
      "last_run_timestamp": "2026-09-01T13:04:20.137Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "thinkingmachines/inkling-2-small-20261107",
      "model_slug": "thinkingmachines/inkling-2-small",
      "model_name": "Thinking Machines: Inkling 2 Small",
      "provider_name": "Nebius",
      "accuracy": 0.653333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.032076099999999996,
      "avg_duration_per_task_ms": 129950.52,
      "avg_output_tokens_per_task": 8232.88,
      "last_run_timestamp": "2026-11-13T00:42:42.289Z"
    },
    {
      "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": 0.652381,
      "accuracy_std_dev": 0.035711227806231656,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.01053958857142857,
      "avg_duration_per_task_ms": 546054.9028571428,
      "avg_output_tokens_per_task": 18131.937142857143,
      "last_run_timestamp": "2026-10-31T19:41:05.491Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "Darkbloom",
      "accuracy": 0.6519705410821643,
      "accuracy_std_dev": 0.024087275055721835,
      "run_count": 10,
      "total_questions": 499,
      "avg_cost_per_task_usd": 0.012394563126252504,
      "avg_duration_per_task_ms": 729028.881763527,
      "avg_output_tokens_per_task": 25595.735470941883,
      "last_run_timestamp": "2026-11-13T02:07:14.326Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "AkashML",
      "accuracy": 0.65166675,
      "accuracy_std_dev": 0.028819327042585528,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.01676374,
      "avg_duration_per_task_ms": 406825.29,
      "avg_output_tokens_per_task": 27732.185,
      "last_run_timestamp": "2026-10-31T19:20:36.815Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Groq",
      "accuracy": 0.6490477142857143,
      "accuracy_std_dev": 0.03593273004504106,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.033029654285714284,
      "avg_duration_per_task_ms": 56698.462857142855,
      "avg_output_tokens_per_task": 14694.748571428572,
      "last_run_timestamp": "2026-10-31T18:54:40.763Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.4-mini-20260317",
      "model_slug": "openai/gpt-5.4-mini",
      "model_name": "OpenAI: GPT-5.4 Mini",
      "provider_name": "OpenAI",
      "accuracy": 0.6488889999999999,
      "accuracy_std_dev": 0.019115994210782406,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.08501352666666667,
      "avg_duration_per_task_ms": 97755.09333333334,
      "avg_output_tokens_per_task": 12128.533333333333,
      "last_run_timestamp": "2026-10-02T23:10:27.235Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.4-nano-20260317",
      "model_slug": "openai/gpt-5.4-nano",
      "model_name": "OpenAI: GPT-5.4 Nano",
      "provider_name": "auto-routing",
      "accuracy": 0.6466665,
      "accuracy_std_dev": 0.026666499999999982,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.0295618,
      "avg_duration_per_task_ms": 121397.75,
      "avg_output_tokens_per_task": 15953.93,
      "last_run_timestamp": "2026-10-01T19:04:28.503Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "DeepInfra",
      "accuracy": 0.6464647229437229,
      "accuracy_std_dev": 0.03922157828554563,
      "run_count": 14,
      "total_questions": 693,
      "avg_cost_per_task_usd": 0.02244592207792208,
      "avg_duration_per_task_ms": 314148.4891774892,
      "avg_output_tokens_per_task": 17992.321789321788,
      "last_run_timestamp": "2026-10-31T20:19:17.569Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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-5.4-nano-20260317",
      "model_slug": "openai/gpt-5.4-nano",
      "model_name": "OpenAI: GPT-5.4 Nano",
      "provider_name": "Azure",
      "accuracy": 0.6416670000000001,
      "accuracy_std_dev": 0.016583123951777013,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.029366035,
      "avg_duration_per_task_ms": 220039.595,
      "avg_output_tokens_per_task": 15350.045,
      "last_run_timestamp": "2026-10-01T19:05:16.543Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4-sonnet-20250522",
      "model_slug": "anthropic/claude-sonnet-4",
      "model_name": "Anthropic: Claude Sonnet 4",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.64,
      "accuracy_std_dev": 0.020000000000000018,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.29234208,
      "avg_duration_per_task_ms": 227202.07,
      "avg_output_tokens_per_task": 8702.01,
      "last_run_timestamp": "2026-09-01T13:00:51.629Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "ibm-granite/granite-4.2-8b-20260831",
      "model_slug": "ibm-granite/granite-4.2-8b",
      "model_name": "IBM: Granite 4.2 8B",
      "provider_name": "DeepInfra",
      "accuracy": 0.64,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.01684376,
      "avg_duration_per_task_ms": 1178898.4,
      "avg_output_tokens_per_task": 43776.78,
      "last_run_timestamp": "2026-10-09T00:39:20.903Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "thinkingmachines/inkling-2-small-20261107",
      "model_slug": "thinkingmachines/inkling-2-small",
      "model_name": "Thinking Machines: Inkling 2 Small",
      "provider_name": "Together",
      "accuracy": 0.64,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.026583100000000002,
      "avg_duration_per_task_ms": 49494.66,
      "avg_output_tokens_per_task": 8435.96,
      "last_run_timestamp": "2026-11-13T00:16:41.107Z"
    },
    {
      "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": 0.6388886666666667,
      "accuracy_std_dev": 0.09896984825805394,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.04814041333333333,
      "avg_duration_per_task_ms": 256985.87333333332,
      "avg_output_tokens_per_task": 7413.113333333334,
      "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": "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": "Venice",
      "accuracy": 0.6355553333333334,
      "accuracy_std_dev": 0.003142853940113798,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.008414553333333331,
      "avg_duration_per_task_ms": 1275373.5733333332,
      "avg_output_tokens_per_task": 17754.266666666666,
      "last_run_timestamp": "2026-10-31T21:55:54.854Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.633333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.21351704000000002,
      "avg_duration_per_task_ms": 877395.36,
      "avg_output_tokens_per_task": 27870.24,
      "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": "thinkingmachines/inkling-2-small-20261107",
      "model_slug": "thinkingmachines/inkling-2-small",
      "model_name": "Thinking Machines: Inkling 2 Small",
      "provider_name": "Modal",
      "accuracy": 0.633333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.042546600000000004,
      "avg_duration_per_task_ms": 200101.72,
      "avg_output_tokens_per_task": 26828.8,
      "last_run_timestamp": "2026-11-13T06:54:36.403Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "thinkingmachines/inkling-2-small-20261107",
      "model_slug": "thinkingmachines/inkling-2-small",
      "model_name": "Thinking Machines: Inkling 2 Small",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.633333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.01754236,
      "avg_duration_per_task_ms": 112671.48,
      "avg_output_tokens_per_task": 8415.52,
      "last_run_timestamp": "2026-11-13T03:43:05.942Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Phala",
      "accuracy": 0.6331658693467337,
      "accuracy_std_dev": 0.03188449882289982,
      "run_count": 4,
      "total_questions": 199,
      "avg_cost_per_task_usd": 0.03419022110552764,
      "avg_duration_per_task_ms": 256833.6432160804,
      "avg_output_tokens_per_task": 16666.94974874372,
      "last_run_timestamp": "2026-10-31T19:04:42.700Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Mara",
      "accuracy": 0.6326291619718309,
      "accuracy_std_dev": 0.04710741386094833,
      "run_count": 3,
      "total_questions": 142,
      "avg_cost_per_task_usd": 0.022965091549295776,
      "avg_duration_per_task_ms": 52073.43661971831,
      "avg_output_tokens_per_task": 11681.894366197183,
      "last_run_timestamp": "2026-10-24T18:40:18.531Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-nano-banana-2.1-20261006",
      "model_slug": "google/gemini-nano-banana-2.1",
      "model_name": "Google: Nano Banana 2.1",
      "provider_name": "Google",
      "accuracy": 0.631944,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 48,
      "avg_cost_per_task_usd": 0.9708985208333334,
      "avg_duration_per_task_ms": 253667.60416666666,
      "avg_output_tokens_per_task": 72982.125,
      "last_run_timestamp": "2026-11-12T21:07:31.630Z"
    },
    {
      "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": 0.6299079172932331,
      "accuracy_std_dev": 0.02774706876297926,
      "run_count": 8,
      "total_questions": 399,
      "avg_cost_per_task_usd": 0.04092825313283208,
      "avg_duration_per_task_ms": 87144.26065162907,
      "avg_output_tokens_per_task": 19555.696741854637,
      "last_run_timestamp": "2026-11-11T19:01:52.112Z"
    },
    {
      "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": 0.627619,
      "accuracy_std_dev": 0.022090270379254585,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.06924706571428572,
      "avg_duration_per_task_ms": 33379.60857142857,
      "avg_output_tokens_per_task": 21963.825714285715,
      "last_run_timestamp": "2026-10-31T19:45:42.530Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.626667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.12019484,
      "avg_duration_per_task_ms": 635855.02,
      "avg_output_tokens_per_task": 26775.66,
      "last_run_timestamp": "2026-08-22T16:01:40.197Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "thinkingmachines/inkling-2-small-20261107",
      "model_slug": "thinkingmachines/inkling-2-small",
      "model_name": "Thinking Machines: Inkling 2 Small",
      "provider_name": "Cloudflare",
      "accuracy": 0.626667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.0507923,
      "avg_duration_per_task_ms": 58773.58,
      "avg_output_tokens_per_task": 11296.32,
      "last_run_timestamp": "2026-11-13T00:15:45.631Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "Crusoe",
      "accuracy": 0.6266666666666667,
      "accuracy_std_dev": 0.14676766724392065,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.03268331333333334,
      "avg_duration_per_task_ms": 135271.32666666666,
      "avg_output_tokens_per_task": 12787.166666666666,
      "last_run_timestamp": "2026-10-31T11:51:01.680Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.6238095714285714,
      "accuracy_std_dev": 0.058538454520712026,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.023625271428571425,
      "avg_duration_per_task_ms": 375289.0742857143,
      "avg_output_tokens_per_task": 15378.111428571428,
      "last_run_timestamp": "2026-10-31T19:07:04.114Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Novita",
      "accuracy": 0.6233335,
      "accuracy_std_dev": 0.01666650000000003,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.02295258,
      "avg_duration_per_task_ms": 131170.99,
      "avg_output_tokens_per_task": 9602.23,
      "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": "AtlasCloud",
      "accuracy": 0.62,
      "accuracy_std_dev": 0.026666999999999996,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.02806645,
      "avg_duration_per_task_ms": 246864.07,
      "avg_output_tokens_per_task": 9796.43,
      "last_run_timestamp": "2026-08-29T16:13:54.380Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.62,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.006431559999999999,
      "avg_duration_per_task_ms": 321502.32,
      "avg_output_tokens_per_task": 15491.98,
      "last_run_timestamp": "2026-11-01T06:36:42.081Z"
    },
    {
      "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": 0.6173370344827587,
      "accuracy_std_dev": 0.07436853642639527,
      "run_count": 7,
      "total_questions": 348,
      "avg_cost_per_task_usd": 0.027499977011494252,
      "avg_duration_per_task_ms": 1452543.2729885057,
      "avg_output_tokens_per_task": 49684.43103448276,
      "last_run_timestamp": "2026-10-31T19:40:44.417Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-next-2025-02-03",
      "model_slug": "qwen/qwen3-coder-next",
      "model_name": "Qwen: Qwen3 Coder Next",
      "provider_name": "auto-routing",
      "accuracy": 0.6133333333333333,
      "accuracy_std_dev": 0.024944382578492966,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.04715752,
      "avg_duration_per_task_ms": 100008.15333333334,
      "avg_output_tokens_per_task": 4739.7266666666665,
      "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-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "AtlasCloud",
      "accuracy": 0.6133332857142857,
      "accuracy_std_dev": 0.12608878454226982,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.051117977142857145,
      "avg_duration_per_task_ms": 942298.3057142857,
      "avg_output_tokens_per_task": 28768.634285714284,
      "last_run_timestamp": "2026-10-31T19:22:14.979Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "thinkingmachines/inkling-small-20260730",
      "model_slug": "thinkingmachines/inkling-small",
      "model_name": "Thinking Machines: Inkling Small",
      "provider_name": "DeepInfra",
      "accuracy": 0.6133331666666667,
      "accuracy_std_dev": 0.031972140650972335,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.04608945666666667,
      "avg_duration_per_task_ms": 205329.73,
      "avg_output_tokens_per_task": 3416.34,
      "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": "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": 0.613333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.12290606,
      "avg_duration_per_task_ms": 597000.58,
      "avg_output_tokens_per_task": 27061.54,
      "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": "qwen/qwen3-coder-next-2025-02-03",
      "model_slug": "qwen/qwen3-coder-next",
      "model_name": "Qwen: Qwen3 Coder Next",
      "provider_name": "StreamLake",
      "accuracy": 0.6091664999999999,
      "accuracy_std_dev": 0.023018682026779924,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.048386775,
      "avg_duration_per_task_ms": 95869.55,
      "avg_output_tokens_per_task": 4703.465,
      "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": "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": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "StreamLake",
      "accuracy": 0.6066665,
      "accuracy_std_dev": 0.026666500000000037,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.02093522,
      "avg_duration_per_task_ms": 280645.76,
      "avg_output_tokens_per_task": 9568.11,
      "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": "google/gemini-2.5-pro",
      "model_slug": "google/gemini-2.5-pro",
      "model_name": "Google: Gemini 2.5 Pro",
      "provider_name": "Google",
      "accuracy": 0.6000001666666667,
      "accuracy_std_dev": 0.03771239112642186,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.19008634000000002,
      "avg_duration_per_task_ms": 186975.22,
      "avg_output_tokens_per_task": 11452.75,
      "last_run_timestamp": "2026-10-01T19:04:55.725Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "BaseTen",
      "accuracy": 0.593333,
      "accuracy_std_dev": 0.020000000000000018,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.04552234,
      "avg_duration_per_task_ms": 32685.39,
      "avg_output_tokens_per_task": 3125.98,
      "last_run_timestamp": "2026-09-26T08:22:32.737Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.6-20251208",
      "model_slug": "z-ai/glm-4.6v",
      "model_name": "Z.ai: GLM 4.6V",
      "provider_name": "Z.AI",
      "accuracy": 0.593333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.01896666,
      "avg_duration_per_task_ms": 306972.98,
      "avg_output_tokens_per_task": 7436.64,
      "last_run_timestamp": "2026-10-01T08:02:59.889Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.5914285714285714,
      "accuracy_std_dev": 0.03558384433120956,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.0944001457142857,
      "avg_duration_per_task_ms": 57605.94,
      "avg_output_tokens_per_task": 3609.52,
      "last_run_timestamp": "2026-10-31T19:06:25.403Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-2.5-flash",
      "model_slug": "google/gemini-2.5-flash",
      "model_name": "Google: Gemini 2.5 Flash",
      "provider_name": "Google",
      "accuracy": 0.591111,
      "accuracy_std_dev": 0.006285295405415193,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.029060479999999996,
      "avg_duration_per_task_ms": 77755.87,
      "avg_output_tokens_per_task": 7763.98,
      "last_run_timestamp": "2026-10-01T19:05:46.198Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-next-2025-02-03",
      "model_slug": "qwen/qwen3-coder-next",
      "model_name": "Qwen: Qwen3 Coder Next",
      "provider_name": "Parasail",
      "accuracy": 0.59,
      "accuracy_std_dev": 0.0196263239723252,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.020891519999999997,
      "avg_duration_per_task_ms": 97200.98666666666,
      "avg_output_tokens_per_task": 4540.886666666666,
      "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": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "DeepInfra",
      "accuracy": 0.5850695,
      "accuracy_std_dev": 0.12673650000000003,
      "run_count": 2,
      "total_questions": 96,
      "avg_cost_per_task_usd": 0.052565635416666666,
      "avg_duration_per_task_ms": 523237.0729166667,
      "avg_output_tokens_per_task": 32134.21875,
      "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": "deepseek/deepseek-r1-0528",
      "model_slug": "deepseek/deepseek-r1-0528",
      "model_name": "DeepSeek: R1 0528",
      "provider_name": "Novita",
      "accuracy": 0.5839998000000001,
      "accuracy_std_dev": 0.06031032194707637,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.21038078000000002,
      "avg_duration_per_task_ms": 2412162.416,
      "avg_output_tokens_per_task": 47038.516,
      "last_run_timestamp": "2026-10-17T19:42:10.129Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "ibm-granite/granite-4.2-8b-20260831",
      "model_slug": "ibm-granite/granite-4.2-8b",
      "model_name": "IBM: Granite 4.2 8B",
      "provider_name": "auto-routing",
      "accuracy": 0.583333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.015297799999999999,
      "avg_duration_per_task_ms": 833284.24,
      "avg_output_tokens_per_task": 38596.74,
      "last_run_timestamp": "2026-10-01T08:04:13.086Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-2.5-pro",
      "model_slug": "google/gemini-2.5-pro",
      "model_name": "Google: Gemini 2.5 Pro",
      "provider_name": "auto-routing",
      "accuracy": 0.58,
      "accuracy_std_dev": 0.026666999999999996,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.20638665,
      "avg_duration_per_task_ms": 179102.02,
      "avg_output_tokens_per_task": 12109.41,
      "last_run_timestamp": "2026-10-01T19:04:42.461Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.58,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.14568854,
      "avg_duration_per_task_ms": 545436.22,
      "avg_output_tokens_per_task": 23072.82,
      "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": "auto-routing",
      "accuracy": 0.5666665000000001,
      "accuracy_std_dev": 0.006666499999999964,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.028862480000000003,
      "avg_duration_per_task_ms": 75034.33,
      "avg_output_tokens_per_task": 7701.06,
      "last_run_timestamp": "2026-10-01T19:04:40.340Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-next-2025-02-03",
      "model_slug": "qwen/qwen3-coder-next",
      "model_name": "Qwen: Qwen3 Coder Next",
      "provider_name": "Alibaba",
      "accuracy": 0.5567129305555556,
      "accuracy_std_dev": 0.0359478068964807,
      "run_count": 3,
      "total_questions": 144,
      "avg_cost_per_task_usd": 0.06927373611111111,
      "avg_duration_per_task_ms": 83137.22916666667,
      "avg_output_tokens_per_task": 4080.9305555555557,
      "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": "Cerebras",
      "accuracy": 0.556667,
      "accuracy_std_dev": 0.04999999999999999,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.16220164,
      "avg_duration_per_task_ms": 20560.07,
      "avg_output_tokens_per_task": 2975.77,
      "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": 0.55,
      "accuracy_std_dev": 0.036667000000000005,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.2373302,
      "avg_duration_per_task_ms": 164429.2,
      "avg_output_tokens_per_task": 12508.34,
      "last_run_timestamp": "2026-10-01T19:04:52.649Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-2.5-flash",
      "model_slug": "google/gemini-2.5-flash",
      "model_name": "Google: Gemini 2.5 Flash",
      "provider_name": "Google AI Studio",
      "accuracy": 0.546667,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.03543453,
      "avg_duration_per_task_ms": 61133.82,
      "avg_output_tokens_per_task": 6656.99,
      "last_run_timestamp": "2026-10-01T19:04:42.483Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Novita",
      "accuracy": 0.5401338361204013,
      "accuracy_std_dev": 0.0534017755585482,
      "run_count": 6,
      "total_questions": 299,
      "avg_cost_per_task_usd": 0.010190789297658862,
      "avg_duration_per_task_ms": 167765.66220735785,
      "avg_output_tokens_per_task": 16034.110367892976,
      "last_run_timestamp": "2026-10-31T19:08:29.434Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-2.5-flash-lite",
      "model_slug": "google/gemini-2.5-flash-lite",
      "model_name": "Google: Gemini 2.5 Flash Lite",
      "provider_name": "Google AI Studio",
      "accuracy": 0.5370371414141414,
      "accuracy_std_dev": 0.020577999999999985,
      "run_count": 2,
      "total_questions": 99,
      "avg_cost_per_task_usd": 0.02047550505050505,
      "avg_duration_per_task_ms": 144559.1616161616,
      "avg_output_tokens_per_task": 38541.121212121216,
      "last_run_timestamp": "2026-10-01T19:05:00.102Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "auto-routing",
      "accuracy": 0.5366089075144508,
      "accuracy_std_dev": 0.018572598267559737,
      "run_count": 7,
      "total_questions": 346,
      "avg_cost_per_task_usd": 0.015467890173410406,
      "avg_duration_per_task_ms": 1042194.598265896,
      "avg_output_tokens_per_task": 62203.8063583815,
      "last_run_timestamp": "2026-10-31T19:53:39.071Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-2.5-flash-lite",
      "model_slug": "google/gemini-2.5-flash-lite",
      "model_name": "Google: Gemini 2.5 Flash Lite",
      "provider_name": "auto-routing",
      "accuracy": 0.534014,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 49,
      "avg_cost_per_task_usd": 0.01161230612244898,
      "avg_duration_per_task_ms": 94666.5306122449,
      "avg_output_tokens_per_task": 16582.326530612245,
      "last_run_timestamp": "2026-10-01T19:04:42.466Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "SiliconFlow",
      "accuracy": 0.5333334285714285,
      "accuracy_std_dev": 0.019518099050128128,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.06718807142857143,
      "avg_duration_per_task_ms": 288835.63714285713,
      "avg_output_tokens_per_task": 4054.082857142857,
      "last_run_timestamp": "2026-10-31T19:28:47.220Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Darkbloom",
      "accuracy": 0.5246082751677852,
      "accuracy_std_dev": 0.030890869335792208,
      "run_count": 9,
      "total_questions": 447,
      "avg_cost_per_task_usd": 0.004759183445190157,
      "avg_duration_per_task_ms": 649707.9328859061,
      "avg_output_tokens_per_task": 27601.37360178971,
      "last_run_timestamp": "2026-11-13T03:45:48.095Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4.1-2025-04-14",
      "model_slug": "openai/gpt-4.1",
      "model_name": "OpenAI: GPT-4.1",
      "provider_name": "Azure",
      "accuracy": 0.522831205479452,
      "accuracy_std_dev": 0.027543964303555765,
      "run_count": 3,
      "total_questions": 146,
      "avg_cost_per_task_usd": 0.11947870547945207,
      "avg_duration_per_task_ms": 69921.19178082192,
      "avg_output_tokens_per_task": 2362.404109589041,
      "last_run_timestamp": "2026-10-01T19:05:18.003Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.5211111666666667,
      "accuracy_std_dev": 0.05297332778677922,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.029030346666666665,
      "avg_duration_per_task_ms": 408571.6,
      "avg_output_tokens_per_task": 15151.166666666666,
      "last_run_timestamp": "2026-10-24T19:54:36.551Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.517778,
      "accuracy_std_dev": 0.027125670043951618,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.02834386,
      "avg_duration_per_task_ms": 441858.92333333334,
      "avg_output_tokens_per_task": 58278.28333333333,
      "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": "openai/gpt-4.1-2025-04-14",
      "model_slug": "openai/gpt-4.1",
      "model_name": "OpenAI: GPT-4.1",
      "provider_name": "OpenAI",
      "accuracy": 0.5133333333333333,
      "accuracy_std_dev": 0.028803034820811636,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.12443134666666665,
      "avg_duration_per_task_ms": 52357.68,
      "avg_output_tokens_per_task": 2755.9866666666667,
      "last_run_timestamp": "2026-10-02T23:07:04.603Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.6-20251208",
      "model_slug": "z-ai/glm-4.6v",
      "model_name": "Z.ai: GLM 4.6V",
      "provider_name": "Novita",
      "accuracy": 0.513333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.03238348,
      "avg_duration_per_task_ms": 271942.88,
      "avg_output_tokens_per_task": 7084.96,
      "last_run_timestamp": "2026-10-01T08:05:50.069Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "AtlasCloud",
      "accuracy": 0.5019102607449856,
      "accuracy_std_dev": 0.2614901575142291,
      "run_count": 7,
      "total_questions": 349,
      "avg_cost_per_task_usd": 0.013760968481375356,
      "avg_duration_per_task_ms": 262073.68481375359,
      "avg_output_tokens_per_task": 6586.329512893983,
      "last_run_timestamp": "2026-10-17T11:22:12.284Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-r1-0528",
      "model_slug": "deepseek/deepseek-r1-0528",
      "model_name": "DeepSeek: R1 0528",
      "provider_name": "auto-routing",
      "accuracy": 0.4953334,
      "accuracy_std_dev": 0.03792387345512057,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.17013926399999998,
      "avg_duration_per_task_ms": 2120647.312,
      "avg_output_tokens_per_task": 40036.992,
      "last_run_timestamp": "2026-10-17T19:35:49.791Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3-nano-30b-a3b",
      "model_slug": "nvidia/nemotron-3-nano-30b-a3b",
      "model_name": "NVIDIA: Nemotron 3 Nano 30B A3B",
      "provider_name": "DeepInfra",
      "accuracy": 0.492,
      "accuracy_std_dev": 0.02578540500360622,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.026663912,
      "avg_duration_per_task_ms": 680695.888,
      "avg_output_tokens_per_task": 52711.976,
      "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": 0.4885847808219178,
      "accuracy_std_dev": 0.019611823395311527,
      "run_count": 3,
      "total_questions": 146,
      "avg_cost_per_task_usd": 0.023597294520547946,
      "avg_duration_per_task_ms": 1384055.6575342466,
      "avg_output_tokens_per_task": 116134.45205479451,
      "last_run_timestamp": "2026-10-31T19:48:14.501Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3-nano-30b-a3b",
      "model_slug": "nvidia/nemotron-3-nano-30b-a3b",
      "model_name": "NVIDIA: Nemotron 3 Nano 30B A3B",
      "provider_name": "Nebius",
      "accuracy": 0.48850570114942526,
      "accuracy_std_dev": 0.017439968302957548,
      "run_count": 7,
      "total_questions": 348,
      "avg_cost_per_task_usd": 0.031982068965517245,
      "avg_duration_per_task_ms": 642915.867816092,
      "avg_output_tokens_per_task": 59401.63505747126,
      "last_run_timestamp": "2026-10-23T09:26:25.528Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5-nano-2025-08-07",
      "model_slug": "openai/gpt-5-nano",
      "model_name": "OpenAI: GPT-5 Nano",
      "provider_name": "auto-routing",
      "accuracy": 0.48333349999999997,
      "accuracy_std_dev": 0.03666650000000002,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.04059527,
      "avg_duration_per_task_ms": 669013.45,
      "avg_output_tokens_per_task": 93900.79,
      "last_run_timestamp": "2026-10-01T19:04:24.509Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.0282842712474619,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.0346484,
      "avg_duration_per_task_ms": 39535.24,
      "avg_output_tokens_per_task": 3626.5266666666666,
      "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/qwen3-32b-04-28",
      "model_slug": "qwen/qwen3-32b",
      "model_name": "Qwen: Qwen3 32B",
      "provider_name": "Nebius",
      "accuracy": 0.476667,
      "accuracy_std_dev": 0.010000000000000009,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.023033520000000002,
      "avg_duration_per_task_ms": 894803.72,
      "avg_output_tokens_per_task": 22315.68,
      "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": "openai/gpt-4.1-2025-04-14",
      "model_slug": "openai/gpt-4.1",
      "model_name": "OpenAI: GPT-4.1",
      "provider_name": "auto-routing",
      "accuracy": 0.47666650000000005,
      "accuracy_std_dev": 0.0166665,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.12559784000000002,
      "avg_duration_per_task_ms": 52928.96,
      "avg_output_tokens_per_task": 2668.55,
      "last_run_timestamp": "2026-10-01T19:04:26.703Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-2.5-flash-lite",
      "model_slug": "google/gemini-2.5-flash-lite",
      "model_name": "Google: Gemini 2.5 Flash Lite",
      "provider_name": "Google",
      "accuracy": 0.4754382631578947,
      "accuracy_std_dev": 0.054999999999999966,
      "run_count": 2,
      "total_questions": 95,
      "avg_cost_per_task_usd": 0.01757382105263158,
      "avg_duration_per_task_ms": 167109.45263157896,
      "avg_output_tokens_per_task": 28512.663157894738,
      "last_run_timestamp": "2026-10-01T19:06:10.500Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.475,
      "accuracy_std_dev": 0.01960262643116988,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.029653983333333335,
      "avg_duration_per_task_ms": 359829.5333333333,
      "avg_output_tokens_per_task": 63168.53,
      "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": "deepseek/deepseek-r1-0528",
      "model_slug": "deepseek/deepseek-r1-0528",
      "model_name": "DeepSeek: R1 0528",
      "provider_name": "SiliconFlow",
      "accuracy": 0.47499975000000005,
      "accuracy_std_dev": 0.002886895683515417,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.171962235,
      "avg_duration_per_task_ms": 2391474.385,
      "avg_output_tokens_per_task": 45557.86,
      "last_run_timestamp": "2026-10-17T19:38:30.603Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-32b-04-28",
      "model_slug": "qwen/qwen3-32b",
      "model_name": "Qwen: Qwen3 32B",
      "provider_name": "DeepInfra",
      "accuracy": 0.47166650000000004,
      "accuracy_std_dev": 0.011666499999999996,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.02012662,
      "avg_duration_per_task_ms": 866874.6,
      "avg_output_tokens_per_task": 18771.04,
      "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": 0.47,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.03551402,
      "avg_duration_per_task_ms": 437816.58,
      "avg_output_tokens_per_task": 20922.04,
      "last_run_timestamp": "2026-10-08T23:09:09.727Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-32b-04-28",
      "model_slug": "qwen/qwen3-32b",
      "model_name": "Qwen: Qwen3 32B",
      "provider_name": "auto-routing",
      "accuracy": 0.47,
      "accuracy_std_dev": 0.009999999999999981,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.02032128,
      "avg_duration_per_task_ms": 826135.39,
      "avg_output_tokens_per_task": 18360.23,
      "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": "deepseek/deepseek-v3.1-terminus",
      "model_slug": "deepseek/deepseek-v3.1-terminus",
      "model_name": "DeepSeek: DeepSeek V3.1 Terminus",
      "provider_name": "SiliconFlow",
      "accuracy": 0.466667,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.06821334999999999,
      "avg_duration_per_task_ms": 285666.19,
      "avg_output_tokens_per_task": 4113.59,
      "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": "openai/gpt-5-nano-2025-08-07",
      "model_slug": "openai/gpt-5-nano",
      "model_name": "OpenAI: GPT-5 Nano",
      "provider_name": "OpenAI",
      "accuracy": 0.4644446666666666,
      "accuracy_std_dev": 0.029979285029648195,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.04325750666666667,
      "avg_duration_per_task_ms": 738509.7133333334,
      "avg_output_tokens_per_task": 100007.46,
      "last_run_timestamp": "2026-10-02T23:22:06.591Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "auto-routing",
      "accuracy": 0.4622223333333333,
      "accuracy_std_dev": 0.01911642045177101,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.03776963333333334,
      "avg_duration_per_task_ms": 60587.16,
      "avg_output_tokens_per_task": 2732.9733333333334,
      "last_run_timestamp": "2026-10-03T16:07:50.332Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-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": 0.4622220000000001,
      "accuracy_std_dev": 0.02266219999617573,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.023695320000000002,
      "avg_duration_per_task_ms": 107307.77333333333,
      "avg_output_tokens_per_task": 3751.38,
      "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-nano-2025-08-07",
      "model_slug": "openai/gpt-5-nano",
      "model_name": "OpenAI: GPT-5 Nano",
      "provider_name": "Azure",
      "accuracy": 0.46166674999999996,
      "accuracy_std_dev": 0.04010416645047619,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.04448468,
      "avg_duration_per_task_ms": 738816.635,
      "avg_output_tokens_per_task": 95390.315,
      "last_run_timestamp": "2026-10-01T19:06:01.717Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "mistralai/mistral-small-2603",
      "model_slug": "mistralai/mistral-small-2603",
      "model_name": "Mistral: Mistral Small 4",
      "provider_name": "auto-routing",
      "accuracy": 0.46,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.011152899999999999,
      "avg_duration_per_task_ms": 87357.46,
      "avg_output_tokens_per_task": 9087.9,
      "last_run_timestamp": "2026-09-01T13:02:23.335Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4.1-2025-04-14",
      "model_slug": "openai/gpt-4.1",
      "model_name": "OpenAI: GPT-4.1",
      "provider_name": "FakeProvider",
      "accuracy": 0.46,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": null,
      "avg_duration_per_task_ms": 158124.84,
      "avg_output_tokens_per_task": 4515,
      "last_run_timestamp": "2026-10-01T19:06:51.244Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Parasail",
      "accuracy": 0.45720697972972973,
      "accuracy_std_dev": 0.0338422225896225,
      "run_count": 6,
      "total_questions": 296,
      "avg_cost_per_task_usd": 0.027552510135135135,
      "avg_duration_per_task_ms": 2287617.1587837837,
      "avg_output_tokens_per_task": 135257.9527027027,
      "last_run_timestamp": "2026-10-24T19:05:04.624Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Together",
      "accuracy": 0.45666671428571426,
      "accuracy_std_dev": 0.154292962939907,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.027580119999999996,
      "avg_duration_per_task_ms": 194202.16571428571,
      "avg_output_tokens_per_task": 18360.19714285714,
      "last_run_timestamp": "2026-10-31T19:46:58.356Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.45333349999999994,
      "accuracy_std_dev": 0.031001890623799048,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.021677975,
      "avg_duration_per_task_ms": 234042.77,
      "avg_output_tokens_per_task": 3500.315,
      "last_run_timestamp": "2026-10-08T22:47:27.753Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4.1-mini-2025-04-14",
      "model_slug": "openai/gpt-4.1-mini",
      "model_name": "OpenAI: GPT-4.1 Mini",
      "provider_name": "OpenAI",
      "accuracy": 0.45111100000000004,
      "accuracy_std_dev": 0.01571332689152746,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.023389446666666668,
      "avg_duration_per_task_ms": 54689.44666666666,
      "avg_output_tokens_per_task": 2398.8933333333334,
      "last_run_timestamp": "2026-10-02T23:07:02.872Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-chat-v3-0324",
      "model_slug": "deepseek/deepseek-chat-v3-0324",
      "model_name": "DeepSeek: DeepSeek V3 0324",
      "provider_name": "auto-routing",
      "accuracy": 0.45,
      "accuracy_std_dev": 0.0033330000000000026,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.07979226,
      "avg_duration_per_task_ms": 840715.05,
      "avg_output_tokens_per_task": 21258.48,
      "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-v3-0324",
      "model_slug": "deepseek/deepseek-chat-v3-0324",
      "model_name": "DeepSeek: DeepSeek V3 0324",
      "provider_name": "SiliconFlow",
      "accuracy": 0.45,
      "accuracy_std_dev": 0.010000000000000009,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.09892811,
      "avg_duration_per_task_ms": 969037.56,
      "avg_output_tokens_per_task": 23498.89,
      "last_run_timestamp": "2026-08-29T16:10:09.206Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-32b-04-28",
      "model_slug": "qwen/qwen3-32b",
      "model_name": "Qwen: Qwen3 32B",
      "provider_name": "Tenstorrent",
      "accuracy": 0.4491300441767068,
      "accuracy_std_dev": 0.023957999896485522,
      "run_count": 5,
      "total_questions": 249,
      "avg_cost_per_task_usd": 0.02306757831325301,
      "avg_duration_per_task_ms": 1875501.032128514,
      "avg_output_tokens_per_task": 14213.674698795181,
      "last_run_timestamp": "2026-11-07T04:00:08.022Z"
    },
    {
      "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": 0.4487706885245901,
      "accuracy_std_dev": 0.024486006562116254,
      "run_count": 5,
      "total_questions": 244,
      "avg_cost_per_task_usd": 0.05463102459016393,
      "avg_duration_per_task_ms": 111409.63114754099,
      "avg_output_tokens_per_task": 62778.315573770495,
      "last_run_timestamp": "2026-10-31T19:03:34.962Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Parasail",
      "accuracy": 0.4458811867816092,
      "accuracy_std_dev": 0.045214591965110434,
      "run_count": 7,
      "total_questions": 348,
      "avg_cost_per_task_usd": 0.015823612068965518,
      "avg_duration_per_task_ms": 133666.3620689655,
      "avg_output_tokens_per_task": 3370.8706896551726,
      "last_run_timestamp": "2026-10-31T18:58:15.016Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.4455556666666666,
      "accuracy_std_dev": 0.03359162888114966,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.019044513333333336,
      "avg_duration_per_task_ms": 90751.12,
      "avg_output_tokens_per_task": 3574.44,
      "last_run_timestamp": "2026-10-03T16:12:20.062Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4.1-mini-2025-04-14",
      "model_slug": "openai/gpt-4.1-mini",
      "model_name": "OpenAI: GPT-4.1 Mini",
      "provider_name": "auto-routing",
      "accuracy": 0.44333300000000003,
      "accuracy_std_dev": 0.009999999999999981,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.023192820000000003,
      "avg_duration_per_task_ms": 58714.73,
      "avg_output_tokens_per_task": 2330.08,
      "last_run_timestamp": "2026-10-01T19:04:38.295Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.44222233333333333,
      "accuracy_std_dev": 0.03281073296279061,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.026376873333333332,
      "avg_duration_per_task_ms": 59340.306666666664,
      "avg_output_tokens_per_task": 3752.213333333333,
      "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/qwen3-14b-04-28",
      "model_slug": "qwen/qwen3-14b",
      "model_name": "Qwen: Qwen3 14B",
      "provider_name": "auto-routing",
      "accuracy": 0.44,
      "accuracy_std_dev": 0.0066669999999999785,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.03130282,
      "avg_duration_per_task_ms": 680376.23,
      "avg_output_tokens_per_task": 16688.33,
      "last_run_timestamp": "2026-10-01T16:00:49.878Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Google",
      "accuracy": 0.43949773972602735,
      "accuracy_std_dev": 0.06335403547437912,
      "run_count": 3,
      "total_questions": 146,
      "avg_cost_per_task_usd": 0.053683828767123284,
      "avg_duration_per_task_ms": 81785.06849315068,
      "avg_output_tokens_per_task": 3261.6164383561645,
      "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": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Parasail",
      "accuracy": 0.43555550000000004,
      "accuracy_std_dev": 0.03136793204186934,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.00797013,
      "avg_duration_per_task_ms": 203443.01,
      "avg_output_tokens_per_task": 4634.386666666666,
      "last_run_timestamp": "2026-10-31T19:46:05.351Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Nebius",
      "accuracy": 0.434,
      "accuracy_std_dev": 0.03129784515266187,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.053281564000000003,
      "avg_duration_per_task_ms": 160868.104,
      "avg_output_tokens_per_task": 3661.612,
      "last_run_timestamp": "2026-10-17T19:11:29.194Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Venice",
      "accuracy": 0.43277766666666667,
      "accuracy_std_dev": 0.021638103680518977,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.03916930333333333,
      "avg_duration_per_task_ms": 287824.24666666664,
      "avg_output_tokens_per_task": 3498.32,
      "last_run_timestamp": "2026-10-31T19:56:10.591Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4o",
      "model_slug": "openai/gpt-4o",
      "model_name": "OpenAI: GPT-4o",
      "provider_name": "OpenAI",
      "accuracy": 0.4322221666666667,
      "accuracy_std_dev": 0.01651776073823432,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.21270556333333332,
      "avg_duration_per_task_ms": 46305.06,
      "avg_output_tokens_per_task": 2266.36,
      "last_run_timestamp": "2026-10-02T23:10:05.271Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4o-2024-08-06",
      "model_slug": "openai/gpt-4o-2024-08-06",
      "model_name": "OpenAI: GPT-4o (2024-08-06)",
      "provider_name": "OpenAI",
      "accuracy": 0.431111,
      "accuracy_std_dev": 0.00314238253559302,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.24174756666666666,
      "avg_duration_per_task_ms": 46576.24,
      "avg_output_tokens_per_task": 2473.9733333333334,
      "last_run_timestamp": "2026-10-02T23:09:02.629Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-chat-v3",
      "model_slug": "deepseek/deepseek-chat",
      "model_name": "DeepSeek: DeepSeek V3",
      "provider_name": "DeepInfra",
      "accuracy": 0.4301802162162162,
      "accuracy_std_dev": 0.02412263108922131,
      "run_count": 3,
      "total_questions": 148,
      "avg_cost_per_task_usd": 0.06743897297297298,
      "avg_duration_per_task_ms": 360892.5135135135,
      "avg_output_tokens_per_task": 3318.9054054054054,
      "last_run_timestamp": "2026-10-08T22:54:33.845Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "auto-routing",
      "accuracy": 0.43,
      "accuracy_std_dev": 0.17237997649475806,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.07492743333333333,
      "avg_duration_per_task_ms": 588774.0866666667,
      "avg_output_tokens_per_task": 9586.053333333333,
      "last_run_timestamp": "2026-10-31T19:59:00.038Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Novita",
      "accuracy": 0.42888899999999996,
      "accuracy_std_dev": 0.01498961780700229,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.025741633333333337,
      "avg_duration_per_task_ms": 239678.91333333333,
      "avg_output_tokens_per_task": 3503.96,
      "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-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": 0.4288886666666667,
      "accuracy_std_dev": 0.011331263428624758,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.04006192,
      "avg_duration_per_task_ms": 171579.76,
      "avg_output_tokens_per_task": 3385.233333333333,
      "last_run_timestamp": "2026-10-03T16:18:20.377Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Chutes",
      "accuracy": 0.42822731213872833,
      "accuracy_std_dev": 0.02920419045134392,
      "run_count": 7,
      "total_questions": 346,
      "avg_cost_per_task_usd": 0.0052024595375722545,
      "avg_duration_per_task_ms": 337191.0375722543,
      "avg_output_tokens_per_task": 4868.939306358381,
      "last_run_timestamp": "2026-10-31T19:55:29.787Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4o-2024-08-06",
      "model_slug": "openai/gpt-4o-2024-08-06",
      "model_name": "OpenAI: GPT-4o (2024-08-06)",
      "provider_name": "auto-routing",
      "accuracy": 0.42666650000000006,
      "accuracy_std_dev": 0.00666650000000002,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.21223042999999997,
      "avg_duration_per_task_ms": 51685.1,
      "avg_output_tokens_per_task": 2189.48,
      "last_run_timestamp": "2026-10-01T16:00:49.880Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "DeepInfra",
      "accuracy": 0.42555566666666667,
      "accuracy_std_dev": 0.02910172028515764,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.027489039999999996,
      "avg_duration_per_task_ms": 68221.56666666667,
      "avg_output_tokens_per_task": 2795.5733333333333,
      "last_run_timestamp": "2026-10-03T16:12:20.034Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-14b-04-28",
      "model_slug": "qwen/qwen3-14b",
      "model_name": "Qwen: Qwen3 14B",
      "provider_name": "DeepInfra",
      "accuracy": 0.4255553333333334,
      "accuracy_std_dev": 0.034462310778910675,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.029985173333333334,
      "avg_duration_per_task_ms": 822675.42,
      "avg_output_tokens_per_task": 18074.12,
      "last_run_timestamp": "2026-10-08T22:41:10.664Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "DeepInfra",
      "accuracy": 0.423750125,
      "accuracy_std_dev": 0.03212117519424492,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.023984480000000002,
      "avg_duration_per_task_ms": 452779.21,
      "avg_output_tokens_per_task": 3544.9625,
      "last_run_timestamp": "2026-10-31T19:22:37.207Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.4233335,
      "accuracy_std_dev": 0.0166665,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.07010642,
      "avg_duration_per_task_ms": 253218.68,
      "avg_output_tokens_per_task": 2643.24,
      "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": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Modal",
      "accuracy": 0.423333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.00707156,
      "avg_duration_per_task_ms": 19040.86,
      "avg_output_tokens_per_task": 3450.68,
      "last_run_timestamp": "2026-11-04T09:13:34.415Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "mistralai/mistral-small-2603",
      "model_slug": "mistralai/mistral-small-2603",
      "model_name": "Mistral: Mistral Small 4",
      "provider_name": "Mistral",
      "accuracy": 0.41888899999999996,
      "accuracy_std_dev": 0.014614444065147779,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.01463076,
      "avg_duration_per_task_ms": 76549.69333333333,
      "avg_output_tokens_per_task": 8814.213333333333,
      "last_run_timestamp": "2026-09-17T02:10:06.482Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "Google",
      "accuracy": 0.4188886666666667,
      "accuracy_std_dev": 0.03424661828885039,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.052051466666666664,
      "avg_duration_per_task_ms": 38040.333333333336,
      "avg_output_tokens_per_task": 2546.826666666667,
      "last_run_timestamp": "2026-10-03T16:12:29.903Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta-llama/llama-4-maverick-17b-128e-instruct",
      "model_slug": "meta-llama/llama-4-maverick",
      "model_name": "Meta: Llama 4 Maverick",
      "provider_name": "Parasail",
      "accuracy": 0.4183672857142857,
      "accuracy_std_dev": 0.0459725,
      "run_count": 2,
      "total_questions": 98,
      "avg_cost_per_task_usd": 0.06366735714285714,
      "avg_duration_per_task_ms": 88350.82653061225,
      "avg_output_tokens_per_task": 2676.5102040816328,
      "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-vl-235b-a22b-instruct",
      "model_slug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_name": "Qwen: Qwen3 VL 235B A22B Instruct",
      "provider_name": "Venice",
      "accuracy": 0.4177776666666667,
      "accuracy_std_dev": 0.03325923616414275,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.07143351333333332,
      "avg_duration_per_task_ms": 333242.26,
      "avg_output_tokens_per_task": 14143.333333333334,
      "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": "openai/gpt-4o",
      "model_slug": "openai/gpt-4o",
      "model_name": "OpenAI: GPT-4o",
      "provider_name": "auto-routing",
      "accuracy": 0.416667,
      "accuracy_std_dev": 0.03,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.19647352999999998,
      "avg_duration_per_task_ms": 53166.02,
      "avg_output_tokens_per_task": 2086.91,
      "last_run_timestamp": "2026-10-01T16:01:13.623Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "Venice",
      "accuracy": 0.41555566666666666,
      "accuracy_std_dev": 0.015713798296048267,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.01727216,
      "avg_duration_per_task_ms": 121607.27333333333,
      "avg_output_tokens_per_task": 2838.9866666666667,
      "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/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "StreamLake",
      "accuracy": 0.415,
      "accuracy_std_dev": 0.02763845951749121,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.051504745,
      "avg_duration_per_task_ms": 126446.71,
      "avg_output_tokens_per_task": 3278.695,
      "last_run_timestamp": "2026-10-03T16:19:54.632Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta-llama/llama-4-maverick-17b-128e-instruct",
      "model_slug": "meta-llama/llama-4-maverick",
      "model_name": "Meta: Llama 4 Maverick",
      "provider_name": "DigitalOcean",
      "accuracy": 0.4144140810810811,
      "accuracy_std_dev": 0.03472111109333275,
      "run_count": 3,
      "total_questions": 148,
      "avg_cost_per_task_usd": 0.07118006081081081,
      "avg_duration_per_task_ms": 303449.4054054054,
      "avg_output_tokens_per_task": 2651.4324324324325,
      "last_run_timestamp": "2026-10-05T20:09:54.588Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "StreamLake",
      "accuracy": 0.4136123821989529,
      "accuracy_std_dev": 0.024976865859981302,
      "run_count": 4,
      "total_questions": 191,
      "avg_cost_per_task_usd": 0.009809837696335078,
      "avg_duration_per_task_ms": 75939.89528795812,
      "avg_output_tokens_per_task": 2670.895287958115,
      "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-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "Novita",
      "accuracy": 0.40888899999999995,
      "accuracy_std_dev": 0.08368091364622321,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.2754822733333333,
      "avg_duration_per_task_ms": 135119.14,
      "avg_output_tokens_per_task": 4063.2466666666664,
      "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-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "Alibaba",
      "accuracy": 0.40888866666666673,
      "accuracy_std_dev": 0.0599587430173634,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.6742300533333333,
      "avg_duration_per_task_ms": 128242.34,
      "avg_output_tokens_per_task": 3782.766666666667,
      "last_run_timestamp": "2026-10-03T16:14:53.189Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_slug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_name": "Qwen: Qwen3 VL 235B A22B Instruct",
      "provider_name": "Novita",
      "accuracy": 0.40888866666666673,
      "accuracy_std_dev": 0.03189478845969806,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.07531134,
      "avg_duration_per_task_ms": 109862.38666666667,
      "avg_output_tokens_per_task": 3035.5333333333333,
      "last_run_timestamp": "2026-10-03T16:12:07.495Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 VL 30B A3B Instruct",
      "provider_name": "Novita",
      "accuracy": 0.40878389864864867,
      "accuracy_std_dev": 0.025276434244942244,
      "run_count": 6,
      "total_questions": 296,
      "avg_cost_per_task_usd": 0.06894762837837838,
      "avg_duration_per_task_ms": 264205.23986486485,
      "avg_output_tokens_per_task": 5484.4628378378375,
      "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": "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": 0.4069373604060914,
      "accuracy_std_dev": 0.01904818344849452,
      "run_count": 4,
      "total_questions": 197,
      "avg_cost_per_task_usd": 0.059585842639593906,
      "avg_duration_per_task_ms": 733094.8121827411,
      "avg_output_tokens_per_task": 26202.4923857868,
      "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": "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": 0.406667,
      "accuracy_std_dev": 0.016329931618554536,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.06185437333333333,
      "avg_duration_per_task_ms": 96669.76666666666,
      "avg_output_tokens_per_task": 2924.06,
      "last_run_timestamp": "2026-10-03T16:12:33.500Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "SiliconFlow",
      "accuracy": 0.4066666666666667,
      "accuracy_std_dev": 0.02494438257849294,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.028486626666666664,
      "avg_duration_per_task_ms": 245942.52666666667,
      "avg_output_tokens_per_task": 4212.966666666666,
      "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": "Parasail",
      "accuracy": 0.40444433333333335,
      "accuracy_std_dev": 0.008314853589945066,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.034242013333333335,
      "avg_duration_per_task_ms": 165678.99333333335,
      "avg_output_tokens_per_task": 3408.8533333333335,
      "last_run_timestamp": "2026-10-03T16:08:50.061Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-chat-v3",
      "model_slug": "deepseek/deepseek-chat",
      "model_name": "DeepSeek: DeepSeek V3",
      "provider_name": "StreamLake",
      "accuracy": 0.40380291275167784,
      "accuracy_std_dev": 0.008840018715415086,
      "run_count": 3,
      "total_questions": 149,
      "avg_cost_per_task_usd": 0.07104265771812081,
      "avg_duration_per_task_ms": 235544.13422818793,
      "avg_output_tokens_per_task": 8668.979865771813,
      "last_run_timestamp": "2026-10-02T15:44:48.231Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4.1-mini-2025-04-14",
      "model_slug": "openai/gpt-4.1-mini",
      "model_name": "OpenAI: GPT-4.1 Mini",
      "provider_name": "Azure",
      "accuracy": 0.40333325000000003,
      "accuracy_std_dev": 0.01914841159698371,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.023807599999999998,
      "avg_duration_per_task_ms": 77145.375,
      "avg_output_tokens_per_task": 2775.135,
      "last_run_timestamp": "2026-10-01T19:05:00.133Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "AkashML",
      "accuracy": 0.401667,
      "accuracy_std_dev": 0.025000000000000022,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01796061,
      "avg_duration_per_task_ms": 1329879.6,
      "avg_output_tokens_per_task": 125682.59,
      "last_run_timestamp": "2026-10-31T20:05:39.534Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.4,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.016500189999999998,
      "avg_duration_per_task_ms": 114969.84,
      "avg_output_tokens_per_task": 649.31,
      "last_run_timestamp": "2026-08-29T17:06:47.617Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_slug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_name": "Qwen: Qwen3 VL 235B A22B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.4,
      "accuracy_std_dev": 0.014401594564491808,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.08217413333333333,
      "avg_duration_per_task_ms": 498824.76,
      "avg_output_tokens_per_task": 14205.28,
      "last_run_timestamp": "2026-10-03T16:05:51.296Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-chat-v3",
      "model_slug": "deepseek/deepseek-chat",
      "model_name": "DeepSeek: DeepSeek V3",
      "provider_name": "auto-routing",
      "accuracy": 0.39833349999999995,
      "accuracy_std_dev": 0.018333500000000003,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.06903712000000001,
      "avg_duration_per_task_ms": 315593.59,
      "avg_output_tokens_per_task": 7308.28,
      "last_run_timestamp": "2026-10-01T19:04:42.463Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-chat-v3-0324",
      "model_slug": "deepseek/deepseek-chat-v3-0324",
      "model_name": "DeepSeek: DeepSeek V3 0324",
      "provider_name": "DeepInfra",
      "accuracy": 0.393333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.03258216,
      "avg_duration_per_task_ms": 687273.92,
      "avg_output_tokens_per_task": 4792.38,
      "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-vl-235b-a22b-instruct",
      "model_slug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_name": "Qwen: Qwen3 VL 235B A22B Instruct",
      "provider_name": "DeepInfra",
      "accuracy": 0.3872055252525253,
      "accuracy_std_dev": 0.01879637043919651,
      "run_count": 4,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.05699022222222223,
      "avg_duration_per_task_ms": 384913.5606060606,
      "avg_output_tokens_per_task": 3819.5858585858587,
      "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/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "AtlasCloud",
      "accuracy": 0.386667,
      "accuracy_std_dev": 0.06,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.052157499999999996,
      "avg_duration_per_task_ms": 111017.7,
      "avg_output_tokens_per_task": 2942.27,
      "last_run_timestamp": "2026-08-29T16:12:10.467Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "Ambient",
      "accuracy": 0.386667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.014510460000000001,
      "avg_duration_per_task_ms": 64332.54,
      "avg_output_tokens_per_task": 4106.96,
      "last_run_timestamp": "2026-10-09T15:37:38.473Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 VL 30B A3B Instruct",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.386667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.0592903,
      "avg_duration_per_task_ms": 203536.16,
      "avg_output_tokens_per_task": 5170.96,
      "last_run_timestamp": "2026-10-04T03:00:16.402Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "auto-routing",
      "accuracy": 0.3844443333333334,
      "accuracy_std_dev": 0.008314853589945041,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.02693198,
      "avg_duration_per_task_ms": 219333.54666666666,
      "avg_output_tokens_per_task": 4471.74,
      "last_run_timestamp": "2026-10-03T16:03:16.197Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "Novita",
      "accuracy": 0.38148099999999996,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 45,
      "avg_cost_per_task_usd": 0.008997444444444445,
      "avg_duration_per_task_ms": 232957.2888888889,
      "avg_output_tokens_per_task": 3561.511111111111,
      "last_run_timestamp": "2026-10-03T16:15:51.302Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "CoreWeave",
      "accuracy": 0.38,
      "accuracy_std_dev": 0.0066669999999999785,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.02802411,
      "avg_duration_per_task_ms": 79737.77,
      "avg_output_tokens_per_task": 4093.08,
      "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/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "auto-routing",
      "accuracy": 0.3780952857142857,
      "accuracy_std_dev": 0.13062556690864188,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.019071360000000002,
      "avg_duration_per_task_ms": 234782.5542857143,
      "avg_output_tokens_per_task": 3254.2371428571428,
      "last_run_timestamp": "2026-10-31T19:18:14.606Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.3742767863818424,
      "accuracy_std_dev": 0.02350411518067989,
      "run_count": 15,
      "total_questions": 749,
      "avg_cost_per_task_usd": 0.013144802403204274,
      "avg_duration_per_task_ms": 41061.73831775701,
      "avg_output_tokens_per_task": 781.1562082777036,
      "last_run_timestamp": "2026-10-31T19:10:53.562Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "DeepInfra",
      "accuracy": 0.3694728520408163,
      "accuracy_std_dev": 0.1273174900715113,
      "run_count": 8,
      "total_questions": 392,
      "avg_cost_per_task_usd": 0.05665731377551021,
      "avg_duration_per_task_ms": 831256.9948979592,
      "avg_output_tokens_per_task": 9740.272959183674,
      "last_run_timestamp": "2026-10-31T19:55:08.491Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "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": "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": 0.36666650000000006,
      "accuracy_std_dev": 0.021602468996621663,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.033141745,
      "avg_duration_per_task_ms": 134753.71,
      "avg_output_tokens_per_task": 4403.68,
      "last_run_timestamp": "2026-10-23T08:14:36.871Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.36,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.041000339999999996,
      "avg_duration_per_task_ms": 322345.3,
      "avg_output_tokens_per_task": 4212.74,
      "last_run_timestamp": "2026-10-04T03:00:10.237Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 VL 30B A3B Instruct",
      "provider_name": "DeepInfra",
      "accuracy": 0.3568007183908046,
      "accuracy_std_dev": 0.033975072562581136,
      "run_count": 7,
      "total_questions": 348,
      "avg_cost_per_task_usd": 0.06755664942528736,
      "avg_duration_per_task_ms": 409777.1005747126,
      "avg_output_tokens_per_task": 7849.508620689655,
      "last_run_timestamp": "2026-10-08T23:34:52.306Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 VL 30B A3B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.35570486577181204,
      "accuracy_std_dev": 0.024452362985677545,
      "run_count": 6,
      "total_questions": 298,
      "avg_cost_per_task_usd": 0.07792019798657716,
      "avg_duration_per_task_ms": 210172.6476510067,
      "avg_output_tokens_per_task": 7299.98322147651,
      "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/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Alibaba",
      "accuracy": 0.35333333333333333,
      "accuracy_std_dev": 0.04455559988997518,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.06266134666666666,
      "avg_duration_per_task_ms": 100248.80666666667,
      "avg_output_tokens_per_task": 3931.2533333333336,
      "last_run_timestamp": "2026-10-03T16:05:51.296Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "Alibaba",
      "accuracy": 0.351111,
      "accuracy_std_dev": 0.013698950130089059,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.057490099999999995,
      "avg_duration_per_task_ms": 70349.37333333334,
      "avg_output_tokens_per_task": 4393.28,
      "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": "CoreWeave",
      "accuracy": 0.35047614285714285,
      "accuracy_std_dev": 0.03596402460566301,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.09883995142857144,
      "avg_duration_per_task_ms": 29505.717142857142,
      "avg_output_tokens_per_task": 1132.6742857142858,
      "last_run_timestamp": "2026-10-31T18:48:48.352Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "CoreWeave",
      "accuracy": 0.34013619387755106,
      "accuracy_std_dev": 0.08215978685882466,
      "run_count": 4,
      "total_questions": 196,
      "avg_cost_per_task_usd": 0.5213178979591837,
      "avg_duration_per_task_ms": 540508.3367346938,
      "avg_output_tokens_per_task": 21094.061224489797,
      "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": "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": 0.33622356069364157,
      "accuracy_std_dev": 0.046445988932944285,
      "run_count": 7,
      "total_questions": 346,
      "avg_cost_per_task_usd": 0.057168858381502895,
      "avg_duration_per_task_ms": 73268.67919075144,
      "avg_output_tokens_per_task": 1325.586705202312,
      "last_run_timestamp": "2026-10-31T19:07:35.559Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.333333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.051354699999999996,
      "avg_duration_per_task_ms": 358714.54,
      "avg_output_tokens_per_task": 69285.14,
      "last_run_timestamp": "2026-10-04T03:00:10.240Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-vl-8b-instruct",
      "model_slug": "qwen/qwen3-vl-8b-instruct",
      "model_name": "Qwen: Qwen3 VL 8B Instruct",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.326667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.02872976,
      "avg_duration_per_task_ms": 133695.86,
      "avg_output_tokens_per_task": 3408.56,
      "last_run_timestamp": "2026-10-04T03:00:16.402Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 VL 30B A3B Instruct",
      "provider_name": "Alibaba",
      "accuracy": 0.3233333333333333,
      "accuracy_std_dev": 0.027419971095697552,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.07416080666666668,
      "avg_duration_per_task_ms": 130738.53,
      "avg_output_tokens_per_task": 7429.926666666666,
      "last_run_timestamp": "2026-10-03T08:23:09.856Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.313333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.15186704,
      "avg_duration_per_task_ms": 32635.52,
      "avg_output_tokens_per_task": 1294.54,
      "last_run_timestamp": "2026-10-04T03:00:16.406Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.31,
      "accuracy_std_dev": 0.12415428957205972,
      "run_count": 14,
      "total_questions": 700,
      "avg_cost_per_task_usd": 0.037997529999999995,
      "avg_duration_per_task_ms": 470949.48285714287,
      "avg_output_tokens_per_task": 124103.05857142857,
      "last_run_timestamp": "2026-10-31T20:07:06.512Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4o-mini",
      "model_slug": "openai/gpt-4o-mini",
      "model_name": "OpenAI: GPT-4o-mini",
      "provider_name": "auto-routing",
      "accuracy": 0.30666649999999995,
      "accuracy_std_dev": 0.006666499999999992,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01779966,
      "avg_duration_per_task_ms": 55498.41,
      "avg_output_tokens_per_task": 3406.81,
      "last_run_timestamp": "2026-10-01T19:04:58.075Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "DeepInfra",
      "accuracy": 0.3055554583333333,
      "accuracy_std_dev": 0.054493000000000014,
      "run_count": 2,
      "total_questions": 96,
      "avg_cost_per_task_usd": 0.024076770833333334,
      "avg_duration_per_task_ms": 191488.85416666666,
      "avg_output_tokens_per_task": 2060.5833333333335,
      "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": 0.28533339999999996,
      "accuracy_std_dev": 0.02872100586400135,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.174218524,
      "avg_duration_per_task_ms": 145567.084,
      "avg_output_tokens_per_task": 1889.128,
      "last_run_timestamp": "2026-10-17T19:19:08.724Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4o-mini",
      "model_slug": "openai/gpt-4o-mini",
      "model_name": "OpenAI: GPT-4o-mini",
      "provider_name": "OpenAI",
      "accuracy": 0.2844446666666667,
      "accuracy_std_dev": 0.02997950740459592,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.025635820000000004,
      "avg_duration_per_task_ms": 67806.75333333333,
      "avg_output_tokens_per_task": 4107.166666666667,
      "last_run_timestamp": "2026-10-02T23:13:47.090Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen-2.5-72b-instruct",
      "model_slug": "qwen/qwen-2.5-72b-instruct",
      "model_name": "Qwen2.5 72B Instruct",
      "provider_name": "Novita",
      "accuracy": 0.2733335,
      "accuracy_std_dev": 0.013333499999999998,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.07945186,
      "avg_duration_per_task_ms": 632916.72,
      "avg_output_tokens_per_task": 5086.04,
      "last_run_timestamp": "2026-10-01T16:00:45.063Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.25632182758620686,
      "accuracy_std_dev": 0.05543966898249743,
      "run_count": 3,
      "total_questions": 145,
      "avg_cost_per_task_usd": 0.005220103448275862,
      "avg_duration_per_task_ms": 18594.572413793103,
      "avg_output_tokens_per_task": 2050.062068965517,
      "last_run_timestamp": "2026-10-24T19:30:38.210Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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/qwen3-coder-next-2025-02-03",
      "model_slug": "qwen/qwen3-coder-next",
      "model_name": "Qwen: Qwen3 Coder Next",
      "provider_name": "Novita",
      "accuracy": 0.24666666666666667,
      "accuracy_std_dev": 0.08018487896660371,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.013351073333333331,
      "avg_duration_per_task_ms": 21559.86,
      "avg_output_tokens_per_task": 1099.86,
      "last_run_timestamp": "2026-10-03T16:13:42.255Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen-2.5-72b-instruct",
      "model_slug": "qwen/qwen-2.5-72b-instruct",
      "model_name": "Qwen2.5 72B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.24,
      "accuracy_std_dev": 0.033333,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.07327475,
      "avg_duration_per_task_ms": 662655.63,
      "avg_output_tokens_per_task": 5322.46,
      "last_run_timestamp": "2026-10-01T16:00:49.856Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "AtlasCloud",
      "accuracy": 0.2366665,
      "accuracy_std_dev": 0.1233335,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.10252787000000002,
      "avg_duration_per_task_ms": 324411.25,
      "avg_output_tokens_per_task": 8077.93,
      "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "CoreWeave",
      "accuracy": 0.22000033333333335,
      "accuracy_std_dev": 0.30169889330626026,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.003046386666666667,
      "avg_duration_per_task_ms": 163087.96666666667,
      "avg_output_tokens_per_task": 5509.633333333333,
      "last_run_timestamp": "2026-10-24T19:32:27.020Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-next-2025-02-03",
      "model_slug": "qwen/qwen3-coder-next",
      "model_name": "Qwen: Qwen3 Coder Next",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.186667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.01693508,
      "avg_duration_per_task_ms": 26941.4,
      "avg_output_tokens_per_task": 1041.58,
      "last_run_timestamp": "2026-10-04T03:00:16.405Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "amazon/nova-micro-v1",
      "model_slug": "amazon/nova-micro-v1",
      "model_name": "Amazon: Nova Micro 1.0",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.17942213775510205,
      "accuracy_std_dev": 0.009739601557558703,
      "run_count": 4,
      "total_questions": 196,
      "avg_cost_per_task_usd": 0.010118459183673469,
      "avg_duration_per_task_ms": 155344.85714285713,
      "avg_output_tokens_per_task": 4996.790816326531,
      "last_run_timestamp": "2026-10-01T19:04:52.650Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 VL 30B A3B Instruct",
      "provider_name": "SiliconFlow",
      "accuracy": 0.1466666,
      "accuracy_std_dev": 0.1569147341005299,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.26473028,
      "avg_duration_per_task_ms": 1632124.184,
      "avg_output_tokens_per_task": 10106.508,
      "last_run_timestamp": "2026-10-03T08:09:23.857Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4.1-nano-2025-04-14",
      "model_slug": "openai/gpt-4.1-nano",
      "model_name": "OpenAI: GPT-4.1 Nano",
      "provider_name": "Azure",
      "accuracy": 0.14,
      "accuracy_std_dev": 0.014142135623730954,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.0072234,
      "avg_duration_per_task_ms": 76455.195,
      "avg_output_tokens_per_task": 2322.375,
      "last_run_timestamp": "2026-10-01T19:04:58.093Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "BaseTen",
      "accuracy": 0.1366668,
      "accuracy_std_dev": 0.2566795840906713,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.004989168,
      "avg_duration_per_task_ms": 18430.86,
      "avg_output_tokens_per_task": 5212.152,
      "last_run_timestamp": "2026-10-31T19:36:09.009Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Nebius",
      "accuracy": 0.132499875,
      "accuracy_std_dev": 0.20457100455064345,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.04783906,
      "avg_duration_per_task_ms": 64368.49,
      "avg_output_tokens_per_task": 12062.3875,
      "last_run_timestamp": "2026-10-31T19:46:05.356Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4.1-nano-2025-04-14",
      "model_slug": "openai/gpt-4.1-nano",
      "model_name": "OpenAI: GPT-4.1 Nano",
      "provider_name": "OpenAI",
      "accuracy": 0.111111,
      "accuracy_std_dev": 0.01571332689152746,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.007342933333333333,
      "avg_duration_per_task_ms": 57269.24,
      "avg_output_tokens_per_task": 2382.0866666666666,
      "last_run_timestamp": "2026-10-02T23:07:59.112Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4.1-nano-2025-04-14",
      "model_slug": "openai/gpt-4.1-nano",
      "model_name": "OpenAI: GPT-4.1 Nano",
      "provider_name": "auto-routing",
      "accuracy": 0.11,
      "accuracy_std_dev": 0.009999999999999995,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.006762440000000001,
      "avg_duration_per_task_ms": 64306.02,
      "avg_output_tokens_per_task": 2312.51,
      "last_run_timestamp": "2026-10-01T19:04:36.194Z"
    },
    {
      "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": 0.09172277852348994,
      "accuracy_std_dev": 0.1783848110580669,
      "run_count": 6,
      "total_questions": 298,
      "avg_cost_per_task_usd": 0.003983040268456375,
      "avg_duration_per_task_ms": 283449.6208053691,
      "avg_output_tokens_per_task": 25704.31879194631,
      "last_run_timestamp": "2026-10-31T19:13:59.240Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.1-terminus",
      "model_slug": "deepseek/deepseek-v3.1-terminus",
      "model_name": "DeepSeek: DeepSeek V3.1 Terminus",
      "provider_name": "AtlasCloud",
      "accuracy": 0.08666650000000001,
      "accuracy_std_dev": 0.013333500000000005,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.00166133,
      "avg_duration_per_task_ms": 9310,
      "avg_output_tokens_per_task": 202.06,
      "last_run_timestamp": "2026-08-29T16:17:08.780Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "Venice",
      "accuracy": 0.0633335,
      "accuracy_std_dev": 0.14742210372515377,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.001561505,
      "avg_duration_per_task_ms": 4223.18,
      "avg_output_tokens_per_task": 528.385,
      "last_run_timestamp": "2026-10-17T19:48:24.252Z"
    }
  ]
}
