{
  "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": "2026-11-07",
  "retrieved_at": "2026-11-07T01:45:34.869Z",
  "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": 2263,
  "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": "google/gemini-3.1-pro-preview-20260219",
      "model_slug": "google/gemini-3.1-pro-preview",
      "model_name": "Google: Gemini 3.1 Pro Preview",
      "provider_name": "Google AI Studio",
      "accuracy": 0.9511779999999999,
      "accuracy_std_dev": 0.003367000000000009,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.20247136363636364,
      "avg_duration_per_task_ms": 137578.30555555556,
      "avg_output_tokens_per_task": 16748.56818181818,
      "last_run_timestamp": "2026-10-01T19:12:42.155Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-astra-pro-20260903",
      "model_slug": "openai/gpt-6-astra-pro",
      "model_name": "OpenAI: GPT-6 Astra Pro",
      "provider_name": "Azure",
      "accuracy": 0.949495,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.35495546717171716,
      "avg_duration_per_task_ms": 72102.17676767676,
      "avg_output_tokens_per_task": 4622.60101010101,
      "last_run_timestamp": "2026-10-01T08:29:03.480Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.8-flash-20260902",
      "model_slug": "google/gemini-3.8-flash",
      "model_name": "Google: Gemini 3.8 Flash",
      "provider_name": "Google",
      "accuracy": 0.9482614885496183,
      "accuracy_std_dev": 0.0012429999999999941,
      "run_count": 2,
      "total_questions": 393,
      "avg_cost_per_task_usd": 0.07536056743002545,
      "avg_duration_per_task_ms": 132814.94147582698,
      "avg_output_tokens_per_task": 18975.51145038168,
      "last_run_timestamp": "2026-10-15T21:00:04.244Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "inclusionai/ling-3.1-flash-20261002",
      "model_slug": "inclusionai/ling-3.1-flash",
      "model_name": "inclusionAI: Ling 3.1 Flash",
      "provider_name": "DeepInfra",
      "accuracy": 0.947917,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 96,
      "avg_cost_per_task_usd": 0.009331666666666667,
      "avg_duration_per_task_ms": 232878.71875,
      "avg_output_tokens_per_task": 10278.75,
      "last_run_timestamp": "2026-10-14T18:00:04.673Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6.1-sol-20260929",
      "model_slug": "openai/gpt-6.1-sol",
      "model_name": "OpenAI: GPT-6.1 Sol",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.947811,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.021484267676767678,
      "avg_duration_per_task_ms": 55623.856060606064,
      "avg_output_tokens_per_task": 1890.328282828283,
      "last_run_timestamp": "2026-10-30T00:03:33.414Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-astra-20260903",
      "model_slug": "openai/gpt-6-astra",
      "model_name": "OpenAI: GPT-6 Astra",
      "provider_name": "OpenAI",
      "accuracy": 0.9472503333333333,
      "accuracy_std_dev": 0.0015872190215034077,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.11162585016835018,
      "avg_duration_per_task_ms": 48234.978114478115,
      "avg_output_tokens_per_task": 2076.191919191919,
      "last_run_timestamp": "2026-10-02T23:16:05.967Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-astra-20260903",
      "model_slug": "openai/gpt-6-astra",
      "model_name": "OpenAI: GPT-6 Astra",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.9461276666666666,
      "accuracy_std_dev": 0.001374572014200141,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.1181814292929293,
      "avg_duration_per_task_ms": 36670.56565656565,
      "avg_output_tokens_per_task": 2059.811447811448,
      "last_run_timestamp": "2026-10-30T00:02:55.917Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.1-pro-preview-20260219",
      "model_slug": "google/gemini-3.1-pro-preview",
      "model_name": "Google: Gemini 3.1 Pro Preview",
      "provider_name": "auto-routing",
      "accuracy": 0.945286,
      "accuracy_std_dev": 0.0008420000000000094,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.20203182828282829,
      "avg_duration_per_task_ms": 133595.55303030304,
      "avg_output_tokens_per_task": 16712.858585858587,
      "last_run_timestamp": "2026-10-01T19:16:14.342Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-astra-20260903",
      "model_slug": "openai/gpt-6-astra",
      "model_name": "OpenAI: GPT-6 Astra",
      "provider_name": "auto-routing",
      "accuracy": 0.944444,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.11333201010101009,
      "avg_duration_per_task_ms": 57056.792929292926,
      "avg_output_tokens_per_task": 2108.121212121212,
      "last_run_timestamp": "2026-10-01T08:25:16.514Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6.1-sol-20260929",
      "model_slug": "openai/gpt-6.1-sol",
      "model_name": "OpenAI: GPT-6.1 Sol",
      "provider_name": "OpenAI",
      "accuracy": 0.944444,
      "accuracy_std_dev": 0.0027491439879836154,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.020607563973063972,
      "avg_duration_per_task_ms": 47519.37542087542,
      "avg_output_tokens_per_task": 1902.2373737373737,
      "last_run_timestamp": "2026-10-02T23:20:22.253Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6.1-sol-20260929",
      "model_slug": "openai/gpt-6.1-sol",
      "model_name": "OpenAI: GPT-6.1 Sol",
      "provider_name": "auto-routing",
      "accuracy": 0.944444,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.02136650505050505,
      "avg_duration_per_task_ms": 52411.32323232323,
      "avg_output_tokens_per_task": 1978.1313131313132,
      "last_run_timestamp": "2026-10-01T08:24:51.870Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-astra-20260903",
      "model_slug": "openai/gpt-6-astra",
      "model_name": "OpenAI: GPT-6 Astra",
      "provider_name": "Azure",
      "accuracy": 0.9431713315363881,
      "accuracy_std_dev": 0.006147978549897494,
      "run_count": 4,
      "total_questions": 742,
      "avg_cost_per_task_usd": 0.10982969407008086,
      "avg_duration_per_task_ms": 117613.10242587602,
      "avg_output_tokens_per_task": 1949.2371967654988,
      "last_run_timestamp": "2026-10-01T08:26:36.876Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.7-flash-20260813",
      "model_slug": "google/gemini-3.7-flash",
      "model_name": "Google: Gemini 3.7 Flash",
      "provider_name": "Google AI Studio",
      "accuracy": 0.9427606666666666,
      "accuracy_std_dev": 0.0036369299629708685,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.029872336700336698,
      "avg_duration_per_task_ms": 28492.575757575756,
      "avg_output_tokens_per_task": 7819.474747474747,
      "last_run_timestamp": "2026-10-01T16:16:03.631Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-sol-pro-20260709",
      "model_slug": "openai/gpt-5.6-sol-pro",
      "model_name": "OpenAI: GPT-5.6 Sol Pro",
      "provider_name": "auto-routing",
      "accuracy": 0.941077,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.15125092424242426,
      "avg_duration_per_task_ms": 97603.77272727272,
      "avg_output_tokens_per_task": 11559.585858585859,
      "last_run_timestamp": "2026-10-01T16:10:33.208Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": "google/gemini-3.7-flash-20260813",
      "model_slug": "google/gemini-3.7-flash",
      "model_name": "Google: Gemini 3.7 Flash",
      "provider_name": "Google",
      "accuracy": 0.9391951653543308,
      "accuracy_std_dev": 0.009884880879403648,
      "run_count": 4,
      "total_questions": 762,
      "avg_cost_per_task_usd": 0.026313035433070862,
      "avg_duration_per_task_ms": 56131.107611548556,
      "avg_output_tokens_per_task": 7729.320209973754,
      "last_run_timestamp": "2026-10-15T21:01:11.190Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": "OpenAI",
      "accuracy": 0.9360266666666668,
      "accuracy_std_dev": 0.0036369299629708854,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.3141816919191919,
      "avg_duration_per_task_ms": 142474.1195286195,
      "avg_output_tokens_per_task": 10342.412457912458,
      "last_run_timestamp": "2026-10-02T23:25:41.614Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.5-20260423",
      "model_slug": "openai/gpt-5.5",
      "model_name": "OpenAI: GPT-5.5",
      "provider_name": "Azure",
      "accuracy": 0.9336215714285715,
      "accuracy_std_dev": 0.008810277924222438,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.3281882835497835,
      "avg_duration_per_task_ms": 164668.6948051948,
      "avg_output_tokens_per_task": 10091.254689754689,
      "last_run_timestamp": "2026-10-01T19:14:07.472Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Io Net",
      "accuracy": 0.933333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 195,
      "avg_cost_per_task_usd": 0.10421853333333334,
      "avg_duration_per_task_ms": 136414.0358974359,
      "avg_output_tokens_per_task": 6800.2,
      "last_run_timestamp": "2026-11-04T16:38:30.598Z"
    },
    {
      "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": "google/gemini-3.6-flash-20260721",
      "model_slug": "google/gemini-3.6-flash",
      "model_name": "Google: Gemini 3.6 Flash",
      "provider_name": "Google",
      "accuracy": 0.9309762,
      "accuracy_std_dev": 0.005106374267520954,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.057293524242424246,
      "avg_duration_per_task_ms": 66775.28888888888,
      "avg_output_tokens_per_task": 11798.8,
      "last_run_timestamp": "2026-10-01T19:12:42.153Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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-5.5-20260423",
      "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": "openai/gpt-5.6-sol-20260709",
      "model_slug": "openai/gpt-5.6-sol",
      "model_name": "OpenAI: GPT-5.6 Sol",
      "provider_name": "Azure",
      "accuracy": 0.9276091428571428,
      "accuracy_std_dev": 0.005010400338398178,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.10054259307359308,
      "avg_duration_per_task_ms": 55310.95093795094,
      "avg_output_tokens_per_task": 3496.1031746031745,
      "last_run_timestamp": "2026-10-01T19:42:48.407Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3.1-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": 0.9264542826738137,
      "accuracy_std_dev": 0.007871563573392863,
      "run_count": 34,
      "total_questions": 6343,
      "avg_cost_per_task_usd": 0.1726583693835724,
      "avg_duration_per_task_ms": 167044.34431656945,
      "avg_output_tokens_per_task": 9360.229386725525,
      "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": "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": 0.9244886751858735,
      "accuracy_std_dev": 0.010749601793658319,
      "run_count": 12,
      "total_questions": 2152,
      "avg_cost_per_task_usd": 0.14484657992565056,
      "avg_duration_per_task_ms": 111589.46514869889,
      "avg_output_tokens_per_task": 9509.516728624536,
      "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": "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": "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": "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": 0.9226093316749586,
      "accuracy_std_dev": 0.014318661477247108,
      "run_count": 4,
      "total_questions": 603,
      "avg_cost_per_task_usd": 0.01542157711442786,
      "avg_duration_per_task_ms": 482101.75787728024,
      "avg_output_tokens_per_task": 30680.03648424544,
      "last_run_timestamp": "2026-11-04T18:05:51.266Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "ParetoAlpha",
      "accuracy": 0.92256925,
      "accuracy_std_dev": 0.005618592528382903,
      "run_count": 5,
      "total_questions": 960,
      "avg_cost_per_task_usd": 0.148641165625,
      "avg_duration_per_task_ms": 385822.7,
      "avg_output_tokens_per_task": 9697.144791666668,
      "last_run_timestamp": "2026-10-31T08:38:36.234Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.5-flash-20260519",
      "model_slug": "google/gemini-3.5-flash",
      "model_name": "Google: Gemini 3.5 Flash",
      "provider_name": "Google AI Studio",
      "accuracy": 0.9225589999999999,
      "accuracy_std_dev": 0.003367000000000009,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.14588472727272728,
      "avg_duration_per_task_ms": 63933.833333333336,
      "avg_output_tokens_per_task": 16085.368686868687,
      "last_run_timestamp": "2026-10-01T19:12:44.398Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Sail Research",
      "accuracy": 0.9224701578947367,
      "accuracy_std_dev": 0.007214706969744515,
      "run_count": 19,
      "total_questions": 3762,
      "avg_cost_per_task_usd": 0.13204979505582137,
      "avg_duration_per_task_ms": 175028.0927698033,
      "avg_output_tokens_per_task": 11401.640882509304,
      "last_run_timestamp": "2026-11-06T22:04:48.119Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-sol-20260709",
      "model_slug": "openai/gpt-5.6-sol",
      "model_name": "OpenAI: GPT-5.6 Sol",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.9219976666666666,
      "accuracy_std_dev": 0.0028614651647170097,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.08309612457912459,
      "avg_duration_per_task_ms": 49364.952861952865,
      "avg_output_tokens_per_task": 3748.658249158249,
      "last_run_timestamp": "2026-10-30T00:00:06.279Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "DigitalOcean",
      "accuracy": 0.9208753333333334,
      "accuracy_std_dev": 0.003636621356271357,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.019656516835016836,
      "avg_duration_per_task_ms": 90437.9696969697,
      "avg_output_tokens_per_task": 20830.57239057239,
      "last_run_timestamp": "2026-10-31T08:09:30.885Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-sol-20260922",
      "model_slug": "openai/gpt-6-sol",
      "model_name": "OpenAI: GPT-6 Sol",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.9208750000000001,
      "accuracy_std_dev": 0.004761657064510223,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.026268675084175085,
      "avg_duration_per_task_ms": 25466.75420875421,
      "avg_output_tokens_per_task": 2308.39898989899,
      "last_run_timestamp": "2026-10-30T00:00:05.373Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6-sol-20260922",
      "model_slug": "openai/gpt-6-sol",
      "model_name": "OpenAI: GPT-6 Sol",
      "provider_name": "OpenAI",
      "accuracy": 0.9208749999999999,
      "accuracy_std_dev": 0.002749143987983661,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.025590255892255893,
      "avg_duration_per_task_ms": 33502.22222222222,
      "avg_output_tokens_per_task": 2400.506734006734,
      "last_run_timestamp": "2026-10-02T23:17:42.236Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.9205386,
      "accuracy_std_dev": 0.004178401589124734,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.14947429494949496,
      "avg_duration_per_task_ms": 120588.33434343434,
      "avg_output_tokens_per_task": 9732.809090909092,
      "last_run_timestamp": "2026-10-31T08:09:30.327Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "BaseTen",
      "accuracy": 0.9200085693112466,
      "accuracy_std_dev": 0.012700463146629303,
      "run_count": 12,
      "total_questions": 2294,
      "avg_cost_per_task_usd": 0.13002592894507412,
      "avg_duration_per_task_ms": 114105.1320836966,
      "avg_output_tokens_per_task": 8645,
      "last_run_timestamp": "2026-10-31T08:14:19.970Z"
    },
    {
      "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": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "auto-routing",
      "accuracy": 0.9187325454545454,
      "accuracy_std_dev": 0.01067374820732572,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.12858450137741045,
      "avg_duration_per_task_ms": 254711.14095500458,
      "avg_output_tokens_per_task": 10409.111570247935,
      "last_run_timestamp": "2026-10-31T08:21:51.280Z"
    },
    {
      "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": "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": "Moonshot AI",
      "accuracy": 0.916859788863109,
      "accuracy_std_dev": 0.00696582562310962,
      "run_count": 11,
      "total_questions": 2155,
      "avg_cost_per_task_usd": 0.1413055136890951,
      "avg_duration_per_task_ms": 281514.6403712297,
      "avg_output_tokens_per_task": 9248.70162412993,
      "last_run_timestamp": "2026-10-24T08:44:42.687Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Baidu",
      "accuracy": 0.9162757342365815,
      "accuracy_std_dev": 0.00785699694603481,
      "run_count": 10,
      "total_questions": 1919,
      "avg_cost_per_task_usd": 0.09128974934861907,
      "avg_duration_per_task_ms": 645077.3731109953,
      "avg_output_tokens_per_task": 39631.607087024495,
      "last_run_timestamp": "2026-10-31T08:24:03.784Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "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": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Morph",
      "accuracy": 0.9154666900732507,
      "accuracy_std_dev": 0.026006001110512943,
      "run_count": 20,
      "total_questions": 3959,
      "avg_cost_per_task_usd": 0.14730550719878757,
      "avg_duration_per_task_ms": 390507.92093963124,
      "avg_output_tokens_per_task": 9397.29047739328,
      "last_run_timestamp": "2026-10-31T09:12:10.294Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "AtlasCloud",
      "accuracy": 0.9149830000000001,
      "accuracy_std_dev": 0.009335323856192691,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.15949898674242427,
      "avg_duration_per_task_ms": 579778.1357323233,
      "avg_output_tokens_per_task": 39960.98863636364,
      "last_run_timestamp": "2026-10-31T08:49:58.581Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "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": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Chutes",
      "accuracy": 0.9148920332741792,
      "accuracy_std_dev": 0.014548429093080425,
      "run_count": 12,
      "total_questions": 2254,
      "avg_cost_per_task_usd": 0.16241754968944103,
      "avg_duration_per_task_ms": 496269.2985803017,
      "avg_output_tokens_per_task": 11219.274622892635,
      "last_run_timestamp": "2026-10-31T08:37:30.440Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Baidu",
      "accuracy": 0.914552902173913,
      "accuracy_std_dev": 0.010001261731512806,
      "run_count": 3,
      "total_questions": 552,
      "avg_cost_per_task_usd": 0.01860343115942029,
      "avg_duration_per_task_ms": 136525.615942029,
      "avg_output_tokens_per_task": 19948.84963768116,
      "last_run_timestamp": "2026-10-31T08:12:46.334Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Minimax",
      "accuracy": 0.9134399166666667,
      "accuracy_std_dev": 0.013804946332494593,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.033723330808080806,
      "avg_duration_per_task_ms": 222484.18686868687,
      "avg_output_tokens_per_task": 27941.286616161615,
      "last_run_timestamp": "2026-10-31T16:05:09.530Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Wafer",
      "accuracy": 0.9129648851570964,
      "accuracy_std_dev": 0.008195106272424223,
      "run_count": 14,
      "total_questions": 2769,
      "avg_cost_per_task_usd": 0.0985225175153485,
      "avg_duration_per_task_ms": 283575.7085590466,
      "avg_output_tokens_per_task": 25018.89021307331,
      "last_run_timestamp": "2026-10-31T08:40:32.804Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Together",
      "accuracy": 0.9119986363636363,
      "accuracy_std_dev": 0.011140273942059605,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.03546824609733701,
      "avg_duration_per_task_ms": 297176.681359045,
      "avg_output_tokens_per_task": 29467.73370064279,
      "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-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Alibaba",
      "accuracy": 0.911926105263158,
      "accuracy_std_dev": 0.013578898854258508,
      "run_count": 19,
      "total_questions": 3762,
      "avg_cost_per_task_usd": 0.1411968139287613,
      "avg_duration_per_task_ms": 616719.149122807,
      "avg_output_tokens_per_task": 40116.86735778841,
      "last_run_timestamp": "2026-10-31T08:31:11.942Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Venice",
      "accuracy": 0.911616,
      "accuracy_std_dev": 0.004209000000000018,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.024379722222222223,
      "avg_duration_per_task_ms": 102982.82575757576,
      "avg_output_tokens_per_task": 20129.815656565657,
      "last_run_timestamp": "2026-10-31T08:34:49.804Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-6.1-terra-20261026",
      "model_slug": "openai/gpt-6.1-terra",
      "model_name": "OpenAI: GPT-6.1 Terra",
      "provider_name": "OpenAI",
      "accuracy": 0.910774,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.011462444444444444,
      "avg_duration_per_task_ms": 27557.444444444445,
      "avg_output_tokens_per_task": 2133.969696969697,
      "last_run_timestamp": "2026-10-28T00:00:04.598Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "DekaLLM",
      "accuracy": 0.9106891772151898,
      "accuracy_std_dev": 0.008562817793226714,
      "run_count": 5,
      "total_questions": 948,
      "avg_cost_per_task_usd": 0.14278316033755276,
      "avg_duration_per_task_ms": 333972.91455696203,
      "avg_output_tokens_per_task": 37406.36181434599,
      "last_run_timestamp": "2026-10-31T08:31:11.854Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "GMICloud",
      "accuracy": 0.9103534999999999,
      "accuracy_std_dev": 0.003644869990822727,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.019048060606060607,
      "avg_duration_per_task_ms": 135272.13636363635,
      "avg_output_tokens_per_task": 22610.929292929293,
      "last_run_timestamp": "2026-10-31T08:28:22.473Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "GMICloud",
      "accuracy": 0.9090908333333333,
      "accuracy_std_dev": 0.008189992133017543,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.03826549410774411,
      "avg_duration_per_task_ms": 134585.28703703705,
      "avg_output_tokens_per_task": 28750.484848484848,
      "last_run_timestamp": "2026-10-27T00:36:05.210Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Phala",
      "accuracy": 0.9086667275553827,
      "accuracy_std_dev": 0.04751984064982348,
      "run_count": 13,
      "total_questions": 2573,
      "avg_cost_per_task_usd": 0.14066514496696458,
      "avg_duration_per_task_ms": 326890.27283326857,
      "avg_output_tokens_per_task": 10307.279440342014,
      "last_run_timestamp": "2026-10-31T09:37:23.390Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Venice",
      "accuracy": 0.9086299041331802,
      "accuracy_std_dev": 0.031330168891428464,
      "run_count": 9,
      "total_questions": 1742,
      "avg_cost_per_task_usd": 0.046619725602755444,
      "avg_duration_per_task_ms": 419278.7634902411,
      "avg_output_tokens_per_task": 14257.997129735935,
      "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-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": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Mancer 2",
      "accuracy": 0.9078013574468085,
      "accuracy_std_dev": 0.012928926018815361,
      "run_count": 3,
      "total_questions": 470,
      "avg_cost_per_task_usd": 0.01807536170212766,
      "avg_duration_per_task_ms": 357786.7425531915,
      "avg_output_tokens_per_task": 17947.382978723403,
      "last_run_timestamp": "2026-10-31T09:05:48.501Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "SiliconFlow",
      "accuracy": 0.9076662307692308,
      "accuracy_std_dev": 0.02122131075178129,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.15455806177156176,
      "avg_duration_per_task_ms": 600133.6076146077,
      "avg_output_tokens_per_task": 38964.08663558664,
      "last_run_timestamp": "2026-10-31T08:28:10.440Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Baidu",
      "accuracy": 0.9076534680851064,
      "accuracy_std_dev": 0.005694785441280666,
      "run_count": 6,
      "total_questions": 1128,
      "avg_cost_per_task_usd": 0.025873300531914892,
      "avg_duration_per_task_ms": 176651.64184397162,
      "avg_output_tokens_per_task": 28451.522163120568,
      "last_run_timestamp": "2026-10-31T08:42:09.092Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "CoreWeave",
      "accuracy": 0.9074649821150856,
      "accuracy_std_dev": 0.007911648334003893,
      "run_count": 13,
      "total_questions": 2572,
      "avg_cost_per_task_usd": 0.030564498833592534,
      "avg_duration_per_task_ms": 392326.3246500778,
      "avg_output_tokens_per_task": 31615.309875583203,
      "last_run_timestamp": "2026-10-31T16:32:36.327Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Cohere",
      "accuracy": 0.9074074999999998,
      "accuracy_std_dev": 0.008417500000000022,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.014508287878787877,
      "avg_duration_per_task_ms": 90086.95707070707,
      "avg_output_tokens_per_task": 21790.61111111111,
      "last_run_timestamp": "2026-10-31T08:13:11.036Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Nebius",
      "accuracy": 0.9074072499999999,
      "accuracy_std_dev": 0.004610545920766882,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.025605553030303027,
      "avg_duration_per_task_ms": 96285.60858585859,
      "avg_output_tokens_per_task": 21160.45580808081,
      "last_run_timestamp": "2026-10-31T09:42:32.702Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": "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": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Together",
      "accuracy": 0.9061883101265825,
      "accuracy_std_dev": 0.03992056619739797,
      "run_count": 12,
      "total_questions": 2370,
      "avg_cost_per_task_usd": 0.1382923181434599,
      "avg_duration_per_task_ms": 209596.56919831224,
      "avg_output_tokens_per_task": 9286.707594936708,
      "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-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Morph",
      "accuracy": 0.9061447499999998,
      "accuracy_std_dev": 0.005102889738912664,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.007892248737373738,
      "avg_duration_per_task_ms": 123987.29040404041,
      "avg_output_tokens_per_task": 16306.166666666666,
      "last_run_timestamp": "2026-10-31T08:53:36.415Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "Novita",
      "accuracy": 0.9060041666666666,
      "accuracy_std_dev": 0.006179339377222198,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.032878036195286196,
      "avg_duration_per_task_ms": 809018.2592592592,
      "avg_output_tokens_per_task": 37448.21212121212,
      "last_run_timestamp": "2026-10-31T08:23:27.402Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "StreamLake",
      "accuracy": 0.9059643571428573,
      "accuracy_std_dev": 0.02530737982319202,
      "run_count": 14,
      "total_questions": 2772,
      "avg_cost_per_task_usd": 0.039628380591630596,
      "avg_duration_per_task_ms": 275120.5137085137,
      "avg_output_tokens_per_task": 30935.2702020202,
      "last_run_timestamp": "2026-10-31T16:23:41.132Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Decart",
      "accuracy": 0.9059342499999998,
      "accuracy_std_dev": 0.010910172663874526,
      "run_count": 3,
      "total_questions": 528,
      "avg_cost_per_task_usd": 0.008819143939393941,
      "avg_duration_per_task_ms": 163195.54924242425,
      "avg_output_tokens_per_task": 22179.03787878788,
      "last_run_timestamp": "2026-10-31T08:49:09.016Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "SiliconFlow",
      "accuracy": 0.9057238888888889,
      "accuracy_std_dev": 0.009356618102764875,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.03432713860830528,
      "avg_duration_per_task_ms": 196757.96408529743,
      "avg_output_tokens_per_task": 28444.87317620651,
      "last_run_timestamp": "2026-10-31T08:29:56.453Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Makora",
      "accuracy": 0.9054373465045593,
      "accuracy_std_dev": 0.0037554479359991244,
      "run_count": 5,
      "total_questions": 987,
      "avg_cost_per_task_usd": 0.026207787234042552,
      "avg_duration_per_task_ms": 127141.36068895644,
      "avg_output_tokens_per_task": 21741.29989868288,
      "last_run_timestamp": "2026-10-31T08:31:11.848Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "InferenceNet",
      "accuracy": 0.905376,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 155,
      "avg_cost_per_task_usd": 0.09219292258064517,
      "avg_duration_per_task_ms": 892432.9419354838,
      "avg_output_tokens_per_task": 36790.02580645162,
      "last_run_timestamp": "2026-10-02T06:44:44.722Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Venice",
      "accuracy": 0.9051541212505279,
      "accuracy_std_dev": 0.011635351751833228,
      "run_count": 12,
      "total_questions": 2367,
      "avg_cost_per_task_usd": 0.0273217351077313,
      "avg_duration_per_task_ms": 321380.98436839884,
      "avg_output_tokens_per_task": 22616.132234896493,
      "last_run_timestamp": "2026-10-31T16:12:28.227Z"
    },
    {
      "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": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "GMICloud",
      "accuracy": 0.9049466923076923,
      "accuracy_std_dev": 0.025579447099289023,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.03460525097125097,
      "avg_duration_per_task_ms": 225805.94988344988,
      "avg_output_tokens_per_task": 28771.273892773894,
      "last_run_timestamp": "2026-10-31T16:04:48.154Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-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": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Novita",
      "accuracy": 0.9047416666666667,
      "accuracy_std_dev": 0.021088018854843207,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.03692853914141414,
      "avg_duration_per_task_ms": 252506.9945286195,
      "avg_output_tokens_per_task": 30920.941498316497,
      "last_run_timestamp": "2026-10-31T16:09:19.241Z"
    },
    {
      "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": 0.904724560327997,
      "accuracy_std_dev": 0.021378149302998656,
      "run_count": 13,
      "total_questions": 2561,
      "avg_cost_per_task_usd": 0.12218845060523233,
      "avg_duration_per_task_ms": 327538.06130417803,
      "avg_output_tokens_per_task": 8367.566966028895,
      "last_run_timestamp": "2026-10-31T08:25:00.194Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Parasail",
      "accuracy": 0.9046013333333334,
      "accuracy_std_dev": 0.009152126540986068,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.02685950505050505,
      "avg_duration_per_task_ms": 136981.2441077441,
      "avg_output_tokens_per_task": 22236.84175084175,
      "last_run_timestamp": "2026-10-31T08:31:11.499Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Ionstream",
      "accuracy": 0.9046013333333334,
      "accuracy_std_dev": 0.003967811851498118,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.016546668350168353,
      "avg_duration_per_task_ms": 140240.16835016836,
      "avg_output_tokens_per_task": 21954.457912457914,
      "last_run_timestamp": "2026-10-31T08:27:53.304Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "AtlasCloud",
      "accuracy": 0.9046013333333334,
      "accuracy_std_dev": 0.008285549120138134,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.01933321717171717,
      "avg_duration_per_task_ms": 139538.02861952863,
      "avg_output_tokens_per_task": 20900.63973063973,
      "last_run_timestamp": "2026-10-31T08:20:08.070Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "DeepInfra",
      "accuracy": 0.9044004522003033,
      "accuracy_std_dev": 0.00856689234728672,
      "run_count": 10,
      "total_questions": 1977,
      "avg_cost_per_task_usd": 0.014848743045017704,
      "avg_duration_per_task_ms": 370376.5558927668,
      "avg_output_tokens_per_task": 28855.582195245323,
      "last_run_timestamp": "2026-10-31T08:03:09.643Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Relace",
      "accuracy": 0.9042316703786192,
      "accuracy_std_dev": 0.0031656583896560934,
      "run_count": 5,
      "total_questions": 898,
      "avg_cost_per_task_usd": 0.22350682405345215,
      "avg_duration_per_task_ms": 441723.66035634745,
      "avg_output_tokens_per_task": 42286.14922048998,
      "last_run_timestamp": "2026-10-31T08:44:55.396Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Alibaba",
      "accuracy": 0.9041981924485772,
      "accuracy_std_dev": 0.01113877499359892,
      "run_count": 18,
      "total_questions": 3549,
      "avg_cost_per_task_usd": 0.027435362073823615,
      "avg_duration_per_task_ms": 185052.6246830093,
      "avg_output_tokens_per_task": 25932.813750352212,
      "last_run_timestamp": "2026-10-31T08:28:13.141Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3-flash-preview-20251217",
      "model_slug": "google/gemini-3-flash-preview",
      "model_name": "Google: Gemini 3 Flash Preview",
      "provider_name": "Google",
      "accuracy": 0.9040405,
      "accuracy_std_dev": 0.011784499999999976,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.13174327272727274,
      "avg_duration_per_task_ms": 249918.9898989899,
      "avg_output_tokens_per_task": 43791.378787878784,
      "last_run_timestamp": "2026-10-01T19:18:35.588Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "x-ai/grok-4.6-20260810",
      "model_slug": "x-ai/grok-4.6",
      "model_name": "SpaceXAI: Grok 4.6",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.903508802631579,
      "accuracy_std_dev": 0.0984941779678147,
      "run_count": 3,
      "total_questions": 456,
      "avg_cost_per_task_usd": 0.11285829166666667,
      "avg_duration_per_task_ms": 423954.7916666667,
      "avg_output_tokens_per_task": 16846.184210526317,
      "last_run_timestamp": "2026-10-01T16:04:59.818Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Mistral",
      "accuracy": 0.9034793333333332,
      "accuracy_std_dev": 0.005722799567422317,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.025755799663299665,
      "avg_duration_per_task_ms": 89661.28282828283,
      "avg_output_tokens_per_task": 21298.74074074074,
      "last_run_timestamp": "2026-10-31T09:51:08.575Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.9033668,
      "accuracy_std_dev": 0.005061737504059227,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.026716864646464646,
      "avg_duration_per_task_ms": 106368.8292929293,
      "avg_output_tokens_per_task": 22085.938383838384,
      "last_run_timestamp": "2026-10-31T08:13:57.548Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "auto-routing",
      "accuracy": 0.9030760960809103,
      "accuracy_std_dev": 0.01701828861577222,
      "run_count": 12,
      "total_questions": 2373,
      "avg_cost_per_task_usd": 0.10602527054361567,
      "avg_duration_per_task_ms": 544239.2713864307,
      "avg_output_tokens_per_task": 37272.33965444585,
      "last_run_timestamp": "2026-10-31T08:11:08.576Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Together",
      "accuracy": 0.9029179999999999,
      "accuracy_std_dev": 0.0028613344206273234,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.023974107744107742,
      "avg_duration_per_task_ms": 50011.877104377105,
      "avg_output_tokens_per_task": 19809.430976430976,
      "last_run_timestamp": "2026-10-31T08:51:04.442Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "auto-routing",
      "accuracy": 0.9029179166666667,
      "accuracy_std_dev": 0.01824016717055307,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.03467677272727272,
      "avg_duration_per_task_ms": 316187.70707070705,
      "avg_output_tokens_per_task": 30484.299663299662,
      "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-v4.1-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": "openai/gpt-5.4-20260305",
      "model_slug": "openai/gpt-5.4",
      "model_name": "OpenAI: GPT-5.4",
      "provider_name": "auto-routing",
      "accuracy": 0.902357,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.1437863712121212,
      "avg_duration_per_task_ms": 149585.63636363635,
      "avg_output_tokens_per_task": 9456.126262626263,
      "last_run_timestamp": "2026-10-01T19:14:23.097Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta/muse-spark-open-20261015",
      "model_slug": "meta/muse-spark-open",
      "model_name": "Meta: Muse Spark Open",
      "provider_name": "Fireworks",
      "accuracy": 0.9023568,
      "accuracy_std_dev": 0.003838970664123373,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.0695745484848485,
      "avg_duration_per_task_ms": 344238.6484848485,
      "avg_output_tokens_per_task": 16158.315151515151,
      "last_run_timestamp": "2026-10-31T08:52:05.708Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-opus-5.5-20260921",
      "model_slug": "anthropic/claude-opus-5.5",
      "model_name": "Anthropic: Claude Opus 5.5",
      "provider_name": "auto-routing",
      "accuracy": 0.9020761666666667,
      "accuracy_std_dev": 0.005271695534223824,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.05407119865319866,
      "avg_duration_per_task_ms": 30482.292087542086,
      "avg_output_tokens_per_task": 2495.6936026936028,
      "last_run_timestamp": "2026-10-31T08:27:47.532Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-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": "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": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "SiliconFlow",
      "accuracy": 0.901515,
      "accuracy_std_dev": 0.004532823209656408,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.13843164835858585,
      "avg_duration_per_task_ms": 1759646.8768939395,
      "avg_output_tokens_per_task": 58099.75189393939,
      "last_run_timestamp": "2026-10-31T16:13:09.866Z"
    },
    {
      "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": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Parasail",
      "accuracy": 0.9014423347902099,
      "accuracy_std_dev": 0.008872458280591035,
      "run_count": 13,
      "total_questions": 2288,
      "avg_cost_per_task_usd": 0.02873364816433567,
      "avg_duration_per_task_ms": 177751.28103146853,
      "avg_output_tokens_per_task": 22378.798513986014,
      "last_run_timestamp": "2026-11-06T06:01:51.918Z"
    },
    {
      "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": 0.9012793199999999,
      "accuracy_std_dev": 0.006151174747769718,
      "run_count": 25,
      "total_questions": 4950,
      "avg_cost_per_task_usd": 0.20707172282828287,
      "avg_duration_per_task_ms": 98196.02,
      "avg_output_tokens_per_task": 7866.467070707071,
      "last_run_timestamp": "2026-10-31T08:46:46.997Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "Xiaomi",
      "accuracy": 0.9012343333333332,
      "accuracy_std_dev": 0.0037221826780652403,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.03380553956228956,
      "avg_duration_per_task_ms": 826187.4343434343,
      "avg_output_tokens_per_task": 38504.42255892256,
      "last_run_timestamp": "2026-10-31T08:14:20.398Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-opus-5.5-20260921",
      "model_slug": "anthropic/claude-opus-5.5",
      "model_name": "Anthropic: Claude Opus 5.5",
      "provider_name": "Anthropic",
      "accuracy": 0.9011541428571429,
      "accuracy_std_dev": 0.004473489479416358,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.053932944444444436,
      "avg_duration_per_task_ms": 26662.532467532466,
      "avg_output_tokens_per_task": 2480.9805194805194,
      "last_run_timestamp": "2026-10-31T08:25:18.221Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Baidu",
      "accuracy": 0.9011222848914487,
      "accuracy_std_dev": 0.007337523255215825,
      "run_count": 12,
      "total_questions": 2257,
      "avg_cost_per_task_usd": 0.013226397430217102,
      "avg_duration_per_task_ms": 143480.65086397872,
      "avg_output_tokens_per_task": 26118.721754541428,
      "last_run_timestamp": "2026-10-31T08:07:19.235Z"
    },
    {
      "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-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Novita",
      "accuracy": 0.9009462380396732,
      "accuracy_std_dev": 0.019684329717769646,
      "run_count": 13,
      "total_questions": 2571,
      "avg_cost_per_task_usd": 0.12647589887203423,
      "avg_duration_per_task_ms": 518480.37456242705,
      "avg_output_tokens_per_task": 38463.91637495138,
      "last_run_timestamp": "2026-10-31T08:17:28.484Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "auto-routing",
      "accuracy": 0.9006730000000001,
      "accuracy_std_dev": 0.006733999999999962,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.010972851010101009,
      "avg_duration_per_task_ms": 117479.04797979798,
      "avg_output_tokens_per_task": 19415.13888888889,
      "last_run_timestamp": "2026-10-31T08:11:46.712Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Cloudflare",
      "accuracy": 0.9005650372881355,
      "accuracy_std_dev": 0.0005644881654108368,
      "run_count": 3,
      "total_questions": 590,
      "avg_cost_per_task_usd": 0.013123520338983051,
      "avg_duration_per_task_ms": 181977.62711864407,
      "avg_output_tokens_per_task": 21777.491525423728,
      "last_run_timestamp": "2026-10-31T08:31:11.333Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-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": 0.9002637837837837,
      "accuracy_std_dev": 0.005876451733850735,
      "run_count": 37,
      "total_questions": 7326,
      "avg_cost_per_task_usd": 0.2161343588588589,
      "avg_duration_per_task_ms": 93366.365001365,
      "avg_output_tokens_per_task": 7877.386704886705,
      "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": "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": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "GMICloud",
      "accuracy": 0.8998314999999999,
      "accuracy_std_dev": 0.000841499999999995,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.11599434848484849,
      "avg_duration_per_task_ms": 1589231.9873737374,
      "avg_output_tokens_per_task": 46211.58585858586,
      "last_run_timestamp": "2026-10-31T08:31:10.881Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "Cohere",
      "accuracy": 0.8996625716694773,
      "accuracy_std_dev": 0.0020759481367959394,
      "run_count": 3,
      "total_questions": 593,
      "avg_cost_per_task_usd": 0.11324433389544686,
      "avg_duration_per_task_ms": 858665.3389544688,
      "avg_output_tokens_per_task": 45077.612141652615,
      "last_run_timestamp": "2026-10-31T08:28:18.045Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "DigitalOcean",
      "accuracy": 0.8995687234401349,
      "accuracy_std_dev": 0.007447984730328704,
      "run_count": 9,
      "total_questions": 1779,
      "avg_cost_per_task_usd": 0.0197627712197864,
      "avg_duration_per_task_ms": 382016.3063518831,
      "avg_output_tokens_per_task": 22859.64811691962,
      "last_run_timestamp": "2026-10-31T08:33:02.133Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.8-opus-20260528",
      "model_slug": "anthropic/claude-opus-4.8",
      "model_name": "Anthropic: Claude Opus 4.8",
      "provider_name": "Google",
      "accuracy": 0.8995041666666669,
      "accuracy_std_dev": 0.005842819508213075,
      "run_count": 36,
      "total_questions": 7128,
      "avg_cost_per_task_usd": 0.22327578970258133,
      "avg_duration_per_task_ms": 94284.44626823794,
      "avg_output_tokens_per_task": 8148.210998877666,
      "last_run_timestamp": "2026-10-31T08:31:12.840Z"
    },
    {
      "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": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Wafer",
      "accuracy": 0.8994017354790972,
      "accuracy_std_dev": 0.12179597930511851,
      "run_count": 28,
      "total_questions": 5406,
      "avg_cost_per_task_usd": 0.16268426008139103,
      "avg_duration_per_task_ms": 193647.5,
      "avg_output_tokens_per_task": 10722.489456159823,
      "last_run_timestamp": "2026-10-31T09:24:16.791Z"
    },
    {
      "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": 0.8993883798165137,
      "accuracy_std_dev": 0.03327796223455359,
      "run_count": 8,
      "total_questions": 1090,
      "avg_cost_per_task_usd": 0.02275113119266055,
      "avg_duration_per_task_ms": 820057.9403669725,
      "avg_output_tokens_per_task": 21074.469724770643,
      "last_run_timestamp": "2026-10-31T16:03:21.470Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "ModelRun",
      "accuracy": 0.8993857647058824,
      "accuracy_std_dev": 0.018671156495670343,
      "run_count": 17,
      "total_questions": 3366,
      "avg_cost_per_task_usd": 0.07737615121806299,
      "avg_duration_per_task_ms": 217682.72222222222,
      "avg_output_tokens_per_task": 30235.886809269163,
      "last_run_timestamp": "2026-10-31T16:10:09.223Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "DeepInfra",
      "accuracy": 0.8993782307692306,
      "accuracy_std_dev": 0.014693466141603143,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.004332889665889667,
      "avg_duration_per_task_ms": 367485.18026418024,
      "avg_output_tokens_per_task": 23771.594405594406,
      "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-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "NextBit",
      "accuracy": 0.8993638888888887,
      "accuracy_std_dev": 0.004029367042778795,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.030411617283950615,
      "avg_duration_per_task_ms": 237553.6509539843,
      "avg_output_tokens_per_task": 26633.1290684624,
      "last_run_timestamp": "2026-10-03T10:19:34.320Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-fable-5.1-20260831",
      "model_slug": "anthropic/claude-fable-5.1",
      "model_name": "Anthropic: Claude Fable 5.1",
      "provider_name": "auto-routing",
      "accuracy": 0.8993264,
      "accuracy_std_dev": 0.006511664598242117,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.13052546969696968,
      "avg_duration_per_task_ms": 30874.333333333332,
      "avg_output_tokens_per_task": 2403.130303030303,
      "last_run_timestamp": "2026-10-01T08:24:30.121Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "GMICloud",
      "accuracy": 0.8992487692307694,
      "accuracy_std_dev": 0.016131982491330434,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.020707853146853153,
      "avg_duration_per_task_ms": 203007.47940947942,
      "avg_output_tokens_per_task": 26050.9560994561,
      "last_run_timestamp": "2026-10-31T09:18:59.078Z"
    },
    {
      "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": 0.8989896800000002,
      "accuracy_std_dev": 0.007736742160470359,
      "run_count": 25,
      "total_questions": 4950,
      "avg_cost_per_task_usd": 0.12731747616161618,
      "avg_duration_per_task_ms": 58520.54424242424,
      "avg_output_tokens_per_task": 4742.417575757576,
      "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-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Cloudflare",
      "accuracy": 0.8987081781963057,
      "accuracy_std_dev": 0.03249433928983099,
      "run_count": 14,
      "total_questions": 2761,
      "avg_cost_per_task_usd": 0.14709861970300617,
      "avg_duration_per_task_ms": 823480.3976819993,
      "avg_output_tokens_per_task": 36846.00507062658,
      "last_run_timestamp": "2026-10-31T08:15:03.301Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": "anthropic/claude-4.7-opus-20260416",
      "model_slug": "anthropic/claude-opus-4.7",
      "model_name": "Anthropic: Claude Opus 4.7",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.8980075833333333,
      "accuracy_std_dev": 0.004784274352123693,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.23032522264309765,
      "avg_duration_per_task_ms": 110445.48653198653,
      "avg_output_tokens_per_task": 8998.862373737375,
      "last_run_timestamp": "2026-10-31T08:24:37.252Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "StreamLake",
      "accuracy": 0.8978673333333335,
      "accuracy_std_dev": 0.007570651065499923,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.03724634175084175,
      "avg_duration_per_task_ms": 745311.5892255892,
      "avg_output_tokens_per_task": 37370.69696969697,
      "last_run_timestamp": "2026-10-31T08:12:46.423Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.4-20260305",
      "model_slug": "openai/gpt-5.4",
      "model_name": "OpenAI: GPT-5.4",
      "provider_name": "Azure",
      "accuracy": 0.8977871428571428,
      "accuracy_std_dev": 0.005599042607665047,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.15678494083694083,
      "avg_duration_per_task_ms": 138524.9898989899,
      "avg_output_tokens_per_task": 9631.916305916306,
      "last_run_timestamp": "2026-10-01T19:18:24.536Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "GMICloud",
      "accuracy": 0.8976429,
      "accuracy_std_dev": 0.019280368027867086,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.024536105555555552,
      "avg_duration_per_task_ms": 164306.0883838384,
      "avg_output_tokens_per_task": 28203.131818181817,
      "last_run_timestamp": "2026-10-31T08:31:11.869Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8973428130434782,
      "accuracy_std_dev": 0.010729397445850358,
      "run_count": 7,
      "total_questions": 1380,
      "avg_cost_per_task_usd": 0.19345281014492755,
      "avg_duration_per_task_ms": 1610739.8637681159,
      "avg_output_tokens_per_task": 56727.897826086955,
      "last_run_timestamp": "2026-10-24T16:20:08.208Z"
    },
    {
      "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-4.7-opus-20260416",
      "model_slug": "anthropic/claude-opus-4.7",
      "model_name": "Anthropic: Claude Opus 4.7",
      "provider_name": "Google",
      "accuracy": 0.8972595277777778,
      "accuracy_std_dev": 0.008117361043145963,
      "run_count": 36,
      "total_questions": 7128,
      "avg_cost_per_task_usd": 0.20771573681257013,
      "avg_duration_per_task_ms": 71213.86826599327,
      "avg_output_tokens_per_task": 7567.590488215488,
      "last_run_timestamp": "2026-10-31T08:51:35.523Z"
    },
    {
      "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": "Novita",
      "accuracy": 0.8968853333333332,
      "accuracy_std_dev": 0.018366417657114165,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.032345546296296296,
      "avg_duration_per_task_ms": 248464.89772727274,
      "avg_output_tokens_per_task": 27393.215909090908,
      "last_run_timestamp": "2026-10-31T08:36:32.971Z"
    },
    {
      "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": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "GMICloud",
      "accuracy": 0.8966514444444446,
      "accuracy_std_dev": 0.022056702876959412,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.06759996408529742,
      "avg_duration_per_task_ms": 533319.7861952862,
      "avg_output_tokens_per_task": 36554.644219977556,
      "last_run_timestamp": "2026-10-31T16:05:14.632Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8966242936708859,
      "accuracy_std_dev": 0.010248237855304126,
      "run_count": 12,
      "total_questions": 2370,
      "avg_cost_per_task_usd": 0.017466570042194093,
      "avg_duration_per_task_ms": 356256.63333333336,
      "avg_output_tokens_per_task": 28695.1776371308,
      "last_run_timestamp": "2026-10-31T09:16:30.257Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Wafer",
      "accuracy": 0.8965486500000001,
      "accuracy_std_dev": 0.014777080321480958,
      "run_count": 20,
      "total_questions": 3960,
      "avg_cost_per_task_usd": 0.007944752525252527,
      "avg_duration_per_task_ms": 153300.83762626263,
      "avg_output_tokens_per_task": 23808.83282828283,
      "last_run_timestamp": "2026-10-31T09:24:59.899Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-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": "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": 0.8961839166666666,
      "accuracy_std_dev": 0.011424969966833842,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.2001774431818182,
      "avg_duration_per_task_ms": 95856.68686868687,
      "avg_output_tokens_per_task": 7788.865740740741,
      "last_run_timestamp": "2026-10-31T08:32:06.814Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-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": "anthropic/claude-4.8-opus-20260528",
      "model_slug": "anthropic/claude-opus-4.8",
      "model_name": "Anthropic: Claude Opus 4.8",
      "provider_name": "Anthropic",
      "accuracy": 0.8958817692307695,
      "accuracy_std_dev": 0.006430888982138585,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.20706321833721833,
      "avg_duration_per_task_ms": 100110.66355866355,
      "avg_output_tokens_per_task": 8069.975524475524,
      "last_run_timestamp": "2026-10-31T08:27:47.283Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-luna-pro-20260709",
      "model_slug": "openai/gpt-5.6-luna-pro",
      "model_name": "OpenAI: GPT-5.6 Luna Pro",
      "provider_name": "auto-routing",
      "accuracy": 0.895623,
      "accuracy_std_dev": 0.0033669999999999534,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.03841230555555556,
      "avg_duration_per_task_ms": 144059.0782828283,
      "avg_output_tokens_per_task": 27039.489898989897,
      "last_run_timestamp": "2026-10-01T16:12:34.647Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "AtlasCloud",
      "accuracy": 0.895412375,
      "accuracy_std_dev": 0.020051295631065218,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.02414115151515151,
      "avg_duration_per_task_ms": 210274.94823232322,
      "avg_output_tokens_per_task": 27810.135101010103,
      "last_run_timestamp": "2026-10-31T08:27:49.932Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Phala",
      "accuracy": 0.8952905937499999,
      "accuracy_std_dev": 0.016969367452854572,
      "run_count": 8,
      "total_questions": 1536,
      "avg_cost_per_task_usd": 0.032397834635416664,
      "avg_duration_per_task_ms": 236218.794921875,
      "avg_output_tokens_per_task": 32195.716796875,
      "last_run_timestamp": "2026-10-31T09:17:17.877Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-opus-5-20260723",
      "model_slug": "anthropic/claude-opus-5",
      "model_name": "Anthropic: Claude Opus 5",
      "provider_name": "Anthropic",
      "accuracy": 0.89520175,
      "accuracy_std_dev": 0.0077796281522640644,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.12088184385521886,
      "avg_duration_per_task_ms": 58635.83712121212,
      "avg_output_tokens_per_task": 4614.950757575758,
      "last_run_timestamp": "2026-10-31T08:31:11.765Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "BaseTen",
      "accuracy": 0.8951904072144289,
      "accuracy_std_dev": 0.019820245214898276,
      "run_count": 13,
      "total_questions": 2495,
      "avg_cost_per_task_usd": 0.159686151503006,
      "avg_duration_per_task_ms": 368440.1851703407,
      "avg_output_tokens_per_task": 40060.66533066132,
      "last_run_timestamp": "2026-10-31T08:09:30.320Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Cohere",
      "accuracy": 0.8951417142857143,
      "accuracy_std_dev": 0.009679824699036726,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.008688588023088023,
      "avg_duration_per_task_ms": 171107.75108225108,
      "avg_output_tokens_per_task": 30552.396103896102,
      "last_run_timestamp": "2026-10-31T09:09:17.234Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "GMICloud",
      "accuracy": 0.8950615555555554,
      "accuracy_std_dev": 0.010765037370127736,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.005121220538720539,
      "avg_duration_per_task_ms": 221349.66778900113,
      "avg_output_tokens_per_task": 28186.213804713803,
      "last_run_timestamp": "2026-10-31T16:31:13.440Z"
    },
    {
      "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": 0.8950615000000002,
      "accuracy_std_dev": 0.009654758735980914,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.2321890096801347,
      "avg_duration_per_task_ms": 110673.35858585859,
      "avg_output_tokens_per_task": 9063.56186868687,
      "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-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Novita",
      "accuracy": 0.894781,
      "accuracy_std_dev": 0.028308178049814503,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.1154155827020202,
      "avg_duration_per_task_ms": 425629.0599747475,
      "avg_output_tokens_per_task": 36939.916666666664,
      "last_run_timestamp": "2026-10-31T16:10:36.257Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-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": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "Makora",
      "accuracy": 0.8934709999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 194,
      "avg_cost_per_task_usd": 0.11686864948453608,
      "avg_duration_per_task_ms": 605687.7474226804,
      "avg_output_tokens_per_task": 46496.701030927834,
      "last_run_timestamp": "2026-10-13T03:00:07.016Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "reflection/beam-1.0-501b-a23b-20261016",
      "model_slug": "reflection/beam-1.0-501b-a23b",
      "model_name": "Reflection: Beam 1.0 501B A23B",
      "provider_name": "BaseTen",
      "accuracy": 0.8933780000000001,
      "accuracy_std_dev": 0.0028613344206273234,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.03135734848484849,
      "avg_duration_per_task_ms": 31339.205387205388,
      "avg_output_tokens_per_task": 13928.784511784512,
      "last_run_timestamp": "2026-10-31T08:28:17.755Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "reflection/beam-1.0-501b-a23b-20261016",
      "model_slug": "reflection/beam-1.0-501b-a23b",
      "model_name": "Reflection: Beam 1.0 501B A23B",
      "provider_name": "DigitalOcean",
      "accuracy": 0.8933779999999999,
      "accuracy_std_dev": 0.0034590473640392213,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.01648433501683502,
      "avg_duration_per_task_ms": 93266.93602693603,
      "avg_output_tokens_per_task": 13659.838383838383,
      "last_run_timestamp": "2026-10-31T08:40:23.037Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": "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-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Fireworks",
      "accuracy": 0.8928570861678005,
      "accuracy_std_dev": 0.015340394005370288,
      "run_count": 5,
      "total_questions": 882,
      "avg_cost_per_task_usd": 0.12683532766439912,
      "avg_duration_per_task_ms": 285532.1031746032,
      "avg_output_tokens_per_task": 28366.77097505669,
      "last_run_timestamp": "2026-09-19T09:14:09.704Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Novita",
      "accuracy": 0.892817,
      "accuracy_std_dev": 0.00915257617650169,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.007166784511784513,
      "avg_duration_per_task_ms": 121588.36531986532,
      "avg_output_tokens_per_task": 11764.878787878788,
      "last_run_timestamp": "2026-10-31T08:24:44.390Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "Novita",
      "accuracy": 0.892817,
      "accuracy_std_dev": 0.0067805750493597225,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.11382627104377104,
      "avg_duration_per_task_ms": 836209.1372053872,
      "avg_output_tokens_per_task": 45359.85101010101,
      "last_run_timestamp": "2026-10-31T09:36:23.668Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.7-opus-20260416",
      "model_slug": "anthropic/claude-opus-4.7",
      "model_name": "Anthropic: Claude Opus 4.7",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.8928169166666666,
      "accuracy_std_dev": 0.00928060290748336,
      "run_count": 24,
      "total_questions": 4752,
      "avg_cost_per_task_usd": 0.2432147763047138,
      "avg_duration_per_task_ms": 85093.25778619529,
      "avg_output_tokens_per_task": 9036.06712962963,
      "last_run_timestamp": "2026-10-31T08:12:47.061Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-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": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Inceptron",
      "accuracy": 0.8925062583961616,
      "accuracy_std_dev": 0.02075156285341843,
      "run_count": 8,
      "total_questions": 1459,
      "avg_cost_per_task_usd": 0.1553590150788211,
      "avg_duration_per_task_ms": 645745.0089102124,
      "avg_output_tokens_per_task": 50031.70390678547,
      "last_run_timestamp": "2026-10-31T16:34:03.928Z"
    },
    {
      "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": 0.8925003799071338,
      "accuracy_std_dev": 0.009652319598878228,
      "run_count": 12,
      "total_questions": 2369,
      "avg_cost_per_task_usd": 0.007965678767412412,
      "avg_duration_per_task_ms": 323776.0540312368,
      "avg_output_tokens_per_task": 28074.01392992824,
      "last_run_timestamp": "2026-10-31T08:16:11.302Z"
    },
    {
      "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": 0.8917748571428571,
      "accuracy_std_dev": 0.009721593908534165,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.04607931240981241,
      "avg_duration_per_task_ms": 43786.40764790765,
      "avg_output_tokens_per_task": 3453.090909090909,
      "last_run_timestamp": "2026-10-01T19:13:06.514Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "OpenInference",
      "accuracy": 0.8916946666666665,
      "accuracy_std_dev": 0.009944030347678707,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.011122356902356903,
      "avg_duration_per_task_ms": 302424.50673400675,
      "avg_output_tokens_per_task": 25444.905723905726,
      "last_run_timestamp": "2026-10-31T09:01:06.311Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.7-opus-20260416",
      "model_slug": "anthropic/claude-opus-4.7",
      "model_name": "Anthropic: Claude Opus 4.7",
      "provider_name": "Anthropic",
      "accuracy": 0.8913493076923077,
      "accuracy_std_dev": 0.011970151958322219,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.20830912587412587,
      "avg_duration_per_task_ms": 99479.31274281275,
      "avg_output_tokens_per_task": 8103.869463869464,
      "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-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8912736666666665,
      "accuracy_std_dev": 0.018072691158823626,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.03149746801346801,
      "avg_duration_per_task_ms": 189533.00841750842,
      "avg_output_tokens_per_task": 25443.46254208754,
      "last_run_timestamp": "2026-10-31T08:53:51.020Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Relace",
      "accuracy": 0.8910890642064206,
      "accuracy_std_dev": 0.023037670395819873,
      "run_count": 17,
      "total_questions": 3333,
      "avg_cost_per_task_usd": 0.007215581758175818,
      "avg_duration_per_task_ms": 268643.4326432643,
      "avg_output_tokens_per_task": 27966.94419441944,
      "last_run_timestamp": "2026-11-02T17:25:00.911Z"
    },
    {
      "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": 0.8910866666666668,
      "accuracy_std_dev": 0.009472170005395327,
      "run_count": 36,
      "total_questions": 7128,
      "avg_cost_per_task_usd": 0.1333184758698092,
      "avg_duration_per_task_ms": 59365.143237934906,
      "avg_output_tokens_per_task": 4764.966189674523,
      "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": "openai/gpt-6-luna-pro-20260922",
      "model_slug": "openai/gpt-6-luna-pro",
      "model_name": "OpenAI: GPT-6 Luna Pro",
      "provider_name": "OpenAI",
      "accuracy": 0.8900113333333334,
      "accuracy_std_dev": 0.0015872190215034077,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.008230473063973063,
      "avg_duration_per_task_ms": 80128.80303030302,
      "avg_output_tokens_per_task": 12935.998316498317,
      "last_run_timestamp": "2026-10-02T23:23:43.005Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "GMICloud",
      "accuracy": 0.8899710714285712,
      "accuracy_std_dev": 0.03484950356372031,
      "run_count": 14,
      "total_questions": 2772,
      "avg_cost_per_task_usd": 0.11476333874458873,
      "avg_duration_per_task_ms": 499182.5613275613,
      "avg_output_tokens_per_task": 36700.38997113997,
      "last_run_timestamp": "2026-10-31T08:09:42.543Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "DeepInfra",
      "accuracy": 0.889876224111283,
      "accuracy_std_dev": 0.005998909101508626,
      "run_count": 7,
      "total_questions": 1294,
      "avg_cost_per_task_usd": 0.030587522411128287,
      "avg_duration_per_task_ms": 772887.7256568779,
      "avg_output_tokens_per_task": 36034.0115919629,
      "last_run_timestamp": "2026-10-31T08:09:29.858Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Fireworks",
      "accuracy": 0.8897900085877862,
      "accuracy_std_dev": 0.006183862216824232,
      "run_count": 6,
      "total_questions": 1048,
      "avg_cost_per_task_usd": 0.01576835591603053,
      "avg_duration_per_task_ms": 219138.39885496182,
      "avg_output_tokens_per_task": 25718.326335877864,
      "last_run_timestamp": "2026-09-19T08:54:58.652Z"
    },
    {
      "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": 0.8897305833333333,
      "accuracy_std_dev": 0.008597912444874173,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.20647175841750842,
      "avg_duration_per_task_ms": 99405.93939393939,
      "avg_output_tokens_per_task": 8030.848063973064,
      "last_run_timestamp": "2026-10-31T09:23:04.665Z"
    },
    {
      "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": "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": "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-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": "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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "DeepInfra",
      "accuracy": 0.887631592274678,
      "accuracy_std_dev": 0.04755846530025938,
      "run_count": 13,
      "total_questions": 2563,
      "avg_cost_per_task_usd": 0.07255826414358174,
      "avg_duration_per_task_ms": 586062.0109246976,
      "avg_output_tokens_per_task": 39208.97737026922,
      "last_run_timestamp": "2026-10-31T16:51:33.999Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Nebius",
      "accuracy": 0.8875500348763478,
      "accuracy_std_dev": 0.013802209238356553,
      "run_count": 8,
      "total_questions": 1577,
      "avg_cost_per_task_usd": 0.1294895707038681,
      "avg_duration_per_task_ms": 1720933.2999365884,
      "avg_output_tokens_per_task": 32195.95688015219,
      "last_run_timestamp": "2026-10-31T16:35:08.639Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "NextBit",
      "accuracy": 0.887537924012158,
      "accuracy_std_dev": 0.020250680116973837,
      "run_count": 5,
      "total_questions": 987,
      "avg_cost_per_task_usd": 0.12033997061803446,
      "avg_duration_per_task_ms": 580755.4113475177,
      "avg_output_tokens_per_task": 34357.5663627153,
      "last_run_timestamp": "2026-10-03T16:44:56.199Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Phala",
      "accuracy": 0.8874858333333334,
      "accuracy_std_dev": 0.011999648139517349,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.028057361531986534,
      "avg_duration_per_task_ms": 346451.00589225587,
      "avg_output_tokens_per_task": 27164.929713804715,
      "last_run_timestamp": "2026-10-31T09:11:16.308Z"
    },
    {
      "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": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Moonshot AI",
      "accuracy": 0.8872051428571429,
      "accuracy_std_dev": 0.02235225083858491,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.23087303607503606,
      "avg_duration_per_task_ms": 953449.9184704184,
      "avg_output_tokens_per_task": 57894.41414141414,
      "last_run_timestamp": "2026-10-24T16:07:10.471Z"
    },
    {
      "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": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Mancer 2",
      "accuracy": 0.8870650833333332,
      "accuracy_std_dev": 0.018422131673878617,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.013124187710437711,
      "avg_duration_per_task_ms": 345264.73526936024,
      "avg_output_tokens_per_task": 23251.8265993266,
      "last_run_timestamp": "2026-10-31T08:24:31.558Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-luna-20260709",
      "model_slug": "openai/gpt-5.6-luna",
      "model_name": "OpenAI: GPT-5.6 Luna",
      "provider_name": "Azure",
      "accuracy": 0.887042338150289,
      "accuracy_std_dev": 0.010371066166434932,
      "run_count": 7,
      "total_questions": 1384,
      "avg_cost_per_task_usd": 0.010374203757225434,
      "avg_duration_per_task_ms": 78284.89884393064,
      "avg_output_tokens_per_task": 7947.398121387283,
      "last_run_timestamp": "2026-10-01T19:12:42.994Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Fireworks",
      "accuracy": 0.8867051248554914,
      "accuracy_std_dev": 0.021208571037200977,
      "run_count": 5,
      "total_questions": 865,
      "avg_cost_per_task_usd": 0.16137210289017342,
      "avg_duration_per_task_ms": 430828.78150289017,
      "avg_output_tokens_per_task": 40500.787283237,
      "last_run_timestamp": "2026-09-19T08:53:57.162Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Nebius",
      "accuracy": 0.8865318000000001,
      "accuracy_std_dev": 0.007110736653821441,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.030907244444444443,
      "avg_duration_per_task_ms": 199105.52626262626,
      "avg_output_tokens_per_task": 25534.58484848485,
      "last_run_timestamp": "2026-10-31T08:28:17.984Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "amazon/nova-micro-v1",
      "model_slug": "amazon/nova-micro-v1",
      "model_name": "Amazon: Nova Micro 1.0",
      "provider_name": "auto-routing",
      "accuracy": 0.8863635000000001,
      "accuracy_std_dev": 0.002525500000000014,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.22157512626262627,
      "avg_duration_per_task_ms": 106335.13383838384,
      "avg_output_tokens_per_task": 8638.217171717171,
      "last_run_timestamp": "2026-10-01T19:10:42.213Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Novita",
      "accuracy": 0.8863634999999997,
      "accuracy_std_dev": 0.014636111679108384,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.11741310185185187,
      "avg_duration_per_task_ms": 565931.4103535353,
      "avg_output_tokens_per_task": 30650.19318181818,
      "last_run_timestamp": "2026-10-31T16:23:41.134Z"
    },
    {
      "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": 0.8863451596434361,
      "accuracy_std_dev": 0.0171093277870711,
      "run_count": 13,
      "total_questions": 2468,
      "avg_cost_per_task_usd": 0.035033690842787686,
      "avg_duration_per_task_ms": 334174.35372771474,
      "avg_output_tokens_per_task": 27531.184764991896,
      "last_run_timestamp": "2026-10-31T08:10:35.431Z"
    },
    {
      "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": 0.886153,
      "accuracy_std_dev": 0.029061101979278085,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.20035622727272726,
      "avg_duration_per_task_ms": 1000635.4015151515,
      "avg_output_tokens_per_task": 59277.57196969697,
      "last_run_timestamp": "2026-10-31T16:24:42.659Z"
    },
    {
      "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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "DigitalOcean",
      "accuracy": 0.8860398461538462,
      "accuracy_std_dev": 0.032945951927935986,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.08817068492618493,
      "avg_duration_per_task_ms": 564559.2735042735,
      "avg_output_tokens_per_task": 40601.23853923854,
      "last_run_timestamp": "2026-10-31T16:16:10.979Z"
    },
    {
      "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": 0.8857262179115301,
      "accuracy_std_dev": 0.016856765875324656,
      "run_count": 14,
      "total_questions": 2758,
      "avg_cost_per_task_usd": 0.09745979006526469,
      "avg_duration_per_task_ms": 455192.95939086296,
      "avg_output_tokens_per_task": 33925.17730239304,
      "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-v4.1-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": "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": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Inceptron",
      "accuracy": 0.885161357171555,
      "accuracy_std_dev": 0.008788568152862834,
      "run_count": 13,
      "total_questions": 2489,
      "avg_cost_per_task_usd": 0.09624794174367216,
      "avg_duration_per_task_ms": 466564.90518280433,
      "avg_output_tokens_per_task": 28784.21052631579,
      "last_run_timestamp": "2026-10-31T16:12:28.231Z"
    },
    {
      "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": 0.8851303585776956,
      "accuracy_std_dev": 0.013660790573422498,
      "run_count": 22,
      "total_questions": 4331,
      "avg_cost_per_task_usd": 0.1910637587162318,
      "avg_duration_per_task_ms": 367733.6695913184,
      "avg_output_tokens_per_task": 31574.25467559455,
      "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-v4.1-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": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8844695,
      "accuracy_std_dev": 0.019793231336747402,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.008258296717171718,
      "avg_duration_per_task_ms": 405954.99242424243,
      "avg_output_tokens_per_task": 29077.366161616163,
      "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": "Ambient",
      "accuracy": 0.8843993333333334,
      "accuracy_std_dev": 0.0067804647005611455,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.004449752525252526,
      "avg_duration_per_task_ms": 221818.0185185185,
      "avg_output_tokens_per_task": 21318.58754208754,
      "last_run_timestamp": "2026-08-29T10:57:26.955Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Cloudflare",
      "accuracy": 0.8842879351276745,
      "accuracy_std_dev": 0.03053132246153121,
      "run_count": 8,
      "total_questions": 1449,
      "avg_cost_per_task_usd": 0.20776448999309868,
      "avg_duration_per_task_ms": 1449823.9765355417,
      "avg_output_tokens_per_task": 51794.29537612147,
      "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-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "DigitalOcean",
      "accuracy": 0.8842513271500845,
      "accuracy_std_dev": 0.03154031064539513,
      "run_count": 18,
      "total_questions": 3558,
      "avg_cost_per_task_usd": 0.1480000303541315,
      "avg_duration_per_task_ms": 355671.73243395163,
      "avg_output_tokens_per_task": 11013.858909499719,
      "last_run_timestamp": "2026-10-31T09:14:35.378Z"
    },
    {
      "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": "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-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "ModelRun",
      "accuracy": 0.8833011934235977,
      "accuracy_std_dev": 0.027231369868529694,
      "run_count": 6,
      "total_questions": 1034,
      "avg_cost_per_task_usd": 0.09606284816247583,
      "avg_duration_per_task_ms": 191330.07156673115,
      "avg_output_tokens_per_task": 25799.25918762089,
      "last_run_timestamp": "2026-09-26T09:19:13.012Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-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": "anthropic/claude-opus-5-20260723",
      "model_slug": "anthropic/claude-opus-5",
      "model_name": "Anthropic: Claude Opus 5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.8830720146096483,
      "accuracy_std_dev": 0.009833326482613501,
      "run_count": 34,
      "total_questions": 6571,
      "avg_cost_per_task_usd": 0.14151279226906105,
      "avg_duration_per_task_ms": 62922.72804748136,
      "avg_output_tokens_per_task": 5091.605843859382,
      "last_run_timestamp": "2026-10-31T08:31:11.082Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Minimax",
      "accuracy": 0.8829965,
      "accuracy_std_dev": 0.002525500000000014,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.006819977272727273,
      "avg_duration_per_task_ms": 114306.47727272728,
      "avg_output_tokens_per_task": 11182.35606060606,
      "last_run_timestamp": "2026-10-31T08:07:19.817Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Reka",
      "accuracy": 0.8829290000000001,
      "accuracy_std_dev": 0.012699110785799135,
      "run_count": 4,
      "total_questions": 783,
      "avg_cost_per_task_usd": 0.027103435504469985,
      "avg_duration_per_task_ms": 200899.3601532567,
      "avg_output_tokens_per_task": 30539.469987228607,
      "last_run_timestamp": "2026-10-31T08:15:03.390Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Inceptron",
      "accuracy": 0.8825948758169935,
      "accuracy_std_dev": 0.010166708210464918,
      "run_count": 14,
      "total_questions": 2754,
      "avg_cost_per_task_usd": 0.007344207697893974,
      "avg_duration_per_task_ms": 375236.5925925926,
      "avg_output_tokens_per_task": 20279.856209150326,
      "last_run_timestamp": "2026-10-31T08:07:18.580Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Together",
      "accuracy": 0.882365125,
      "accuracy_std_dev": 0.0043279540904883765,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.029606650252525252,
      "avg_duration_per_task_ms": 78746.81123737374,
      "avg_output_tokens_per_task": 24519.371843434343,
      "last_run_timestamp": "2026-10-31T09:02:59.407Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.8821789256756756,
      "accuracy_std_dev": 0.01706691482241463,
      "run_count": 6,
      "total_questions": 1184,
      "avg_cost_per_task_usd": 0.04157157263513513,
      "avg_duration_per_task_ms": 164106.9873310811,
      "avg_output_tokens_per_task": 34476.61317567567,
      "last_run_timestamp": "2026-10-31T08:31:11.691Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "ModelRun",
      "accuracy": 0.8821550000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.026309151515151515,
      "avg_duration_per_task_ms": 202247.18686868687,
      "avg_output_tokens_per_task": 39119.5505050505,
      "last_run_timestamp": "2026-10-29T03:00:03.996Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Phala",
      "accuracy": 0.8821550000000001,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.129083601010101,
      "avg_duration_per_task_ms": 388751.6590909091,
      "avg_output_tokens_per_task": 32112.176767676767,
      "last_run_timestamp": "2026-10-17T16:07:32.047Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Novita",
      "accuracy": 0.88215475,
      "accuracy_std_dev": 0.013094556929789563,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.007754531565656565,
      "avg_duration_per_task_ms": 220861.55050505052,
      "avg_output_tokens_per_task": 27364.97095959596,
      "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": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "GMICloud",
      "accuracy": 0.8817339999999999,
      "accuracy_std_dev": 0.016681744617994862,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.008588069444444443,
      "avg_duration_per_task_ms": 150888.15656565657,
      "avg_output_tokens_per_task": 16172.003787878788,
      "last_run_timestamp": "2026-10-31T08:07:19.698Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "Parasail",
      "accuracy": 0.8809202687667559,
      "accuracy_std_dev": 0.012345881812749338,
      "run_count": 8,
      "total_questions": 1492,
      "avg_cost_per_task_usd": 0.0663811581769437,
      "avg_duration_per_task_ms": 315790.1621983914,
      "avg_output_tokens_per_task": 18342.8572386059,
      "last_run_timestamp": "2026-10-31T16:03:21.450Z"
    },
    {
      "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": 0.8808921666666667,
      "accuracy_std_dev": 0.019159369526132352,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.12913603156565656,
      "avg_duration_per_task_ms": 63290.436026936026,
      "avg_output_tokens_per_task": 4931.569023569024,
      "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-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Parasail",
      "accuracy": 0.8805721763140385,
      "accuracy_std_dev": 0.01575367618208524,
      "run_count": 8,
      "total_questions": 1503,
      "avg_cost_per_task_usd": 0.18305890685296072,
      "avg_duration_per_task_ms": 788943.5455755156,
      "avg_output_tokens_per_task": 52764.211576846305,
      "last_run_timestamp": "2026-10-31T16:35:50.033Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "StreamLake",
      "accuracy": 0.8804710769230769,
      "accuracy_std_dev": 0.0116074195466755,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.006037170163170164,
      "avg_duration_per_task_ms": 408315.95765345765,
      "avg_output_tokens_per_task": 30238.239316239316,
      "last_run_timestamp": "2026-10-31T08:19:56.454Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "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-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Decart",
      "accuracy": 0.8798644568527919,
      "accuracy_std_dev": 0.004527248931623837,
      "run_count": 3,
      "total_questions": 591,
      "avg_cost_per_task_usd": 0.0029938477157360403,
      "avg_duration_per_task_ms": 316818.90016920475,
      "avg_output_tokens_per_task": 18321.88663282572,
      "last_run_timestamp": "2026-08-22T10:20:26.941Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Venice",
      "accuracy": 0.87984,
      "accuracy_std_dev": 0.018052164925570553,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.036141941919191914,
      "avg_duration_per_task_ms": 139656.41224747474,
      "avg_output_tokens_per_task": 28081.721590909092,
      "last_run_timestamp": "2026-10-31T09:14:35.460Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "OpenInference",
      "accuracy": 0.879386,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 76,
      "avg_cost_per_task_usd": 0.03120878947368421,
      "avg_duration_per_task_ms": 274583.63157894736,
      "avg_output_tokens_per_task": 9741.526315789473,
      "last_run_timestamp": "2026-10-09T00:00:06.206Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Cloudflare",
      "accuracy": 0.8793139256044638,
      "accuracy_std_dev": 0.0299771488382257,
      "run_count": 9,
      "total_questions": 1613,
      "avg_cost_per_task_usd": 0.03197218970861749,
      "avg_duration_per_task_ms": 352833.73217606946,
      "avg_output_tokens_per_task": 26068.50278983261,
      "last_run_timestamp": "2026-10-31T16:15:26.636Z"
    },
    {
      "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": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "CoreWeave",
      "accuracy": 0.8792495990621337,
      "accuracy_std_dev": 0.015176521845450746,
      "run_count": 13,
      "total_questions": 2559,
      "avg_cost_per_task_usd": 0.10830853575615476,
      "avg_duration_per_task_ms": 503979.6647127784,
      "avg_output_tokens_per_task": 31069.37202032044,
      "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-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "CoreWeave",
      "accuracy": 0.8790466153846153,
      "accuracy_std_dev": 0.013384300197438356,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.008804418026418025,
      "avg_duration_per_task_ms": 213500.61344211345,
      "avg_output_tokens_per_task": 31022.890054390053,
      "last_run_timestamp": "2026-10-31T08:31:11.360Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "AkashML",
      "accuracy": 0.8790284807947019,
      "accuracy_std_dev": 0.024140287647157404,
      "run_count": 4,
      "total_questions": 755,
      "avg_cost_per_task_usd": 0.004138932450331125,
      "avg_duration_per_task_ms": 381646.0185430464,
      "avg_output_tokens_per_task": 15960.21986754967,
      "last_run_timestamp": "2026-09-05T08:24:32.148Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-opus-5-20260723",
      "model_slug": "anthropic/claude-opus-5",
      "model_name": "Anthropic: Claude Opus 5",
      "provider_name": "auto-routing",
      "accuracy": 0.8789280833333333,
      "accuracy_std_dev": 0.018215839684453818,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.12799191666666665,
      "avg_duration_per_task_ms": 62532.678872053875,
      "avg_output_tokens_per_task": 4895.166245791246,
      "last_run_timestamp": "2026-10-31T08:12:45.681Z"
    },
    {
      "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": "StreamLake",
      "accuracy": 0.878787777777778,
      "accuracy_std_dev": 0.016872358550651074,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.023912988215488217,
      "avg_duration_per_task_ms": 730084.7261503928,
      "avg_output_tokens_per_task": 36009.7519640853,
      "last_run_timestamp": "2026-10-31T16:46:00.416Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Cloudflare",
      "accuracy": 0.8787235883100379,
      "accuracy_std_dev": 0.03310670771944613,
      "run_count": 12,
      "total_questions": 2361,
      "avg_cost_per_task_usd": 0.10775736086404068,
      "avg_duration_per_task_ms": 649248.8576874206,
      "avg_output_tokens_per_task": 26784.826768318508,
      "last_run_timestamp": "2026-10-31T16:08:08.536Z"
    },
    {
      "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": 0.878634635903919,
      "accuracy_std_dev": 0.006072744766577977,
      "run_count": 12,
      "total_questions": 2373,
      "avg_cost_per_task_usd": 0.11808843784239359,
      "avg_duration_per_task_ms": 774332.4475347662,
      "avg_output_tokens_per_task": 30907.657395701644,
      "last_run_timestamp": "2026-10-31T16:19:28.502Z"
    },
    {
      "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": "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": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "CoreWeave",
      "accuracy": 0.8773146249999999,
      "accuracy_std_dev": 0.019550283047423515,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.20197871338383838,
      "avg_duration_per_task_ms": 597313.053030303,
      "avg_output_tokens_per_task": 59103.44823232323,
      "last_run_timestamp": "2026-10-31T16:09:16.278Z"
    },
    {
      "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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Venice",
      "accuracy": 0.8771411280539175,
      "accuracy_std_dev": 0.030213255709644882,
      "run_count": 12,
      "total_questions": 2374,
      "avg_cost_per_task_usd": 0.17571395745577084,
      "avg_duration_per_task_ms": 362774.2093513058,
      "avg_output_tokens_per_task": 39884.13732097725,
      "last_run_timestamp": "2026-10-31T16:32:36.327Z"
    },
    {
      "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": 0.8771043749999999,
      "accuracy_std_dev": 0.026498402465702973,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.13848938320707072,
      "avg_duration_per_task_ms": 860393.4109848485,
      "avg_output_tokens_per_task": 62250.43244949495,
      "last_run_timestamp": "2026-10-31T16:04:29.123Z"
    },
    {
      "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": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "Novita",
      "accuracy": 0.8765656184971099,
      "accuracy_std_dev": 0.007664911544146613,
      "run_count": 7,
      "total_questions": 1384,
      "avg_cost_per_task_usd": 0.12350613511560693,
      "avg_duration_per_task_ms": 893845.7933526011,
      "avg_output_tokens_per_task": 43364.7362716763,
      "last_run_timestamp": "2026-10-24T16:19:29.598Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-fable-5.1-20260831",
      "model_slug": "anthropic/claude-fable-5.1",
      "model_name": "Anthropic: Claude Fable 5.1",
      "provider_name": "Google",
      "accuracy": 0.8765431666666668,
      "accuracy_std_dev": 0.03217293038884639,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.11818556060606061,
      "avg_duration_per_task_ms": 28260.9031986532,
      "avg_output_tokens_per_task": 2161.995791245791,
      "last_run_timestamp": "2026-10-01T08:25:56.890Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-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": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "Together",
      "accuracy": 0.8755709520547945,
      "accuracy_std_dev": 0.007533567739715949,
      "run_count": 3,
      "total_questions": 584,
      "avg_cost_per_task_usd": 0.031521625,
      "avg_duration_per_task_ms": 247947.07705479453,
      "avg_output_tokens_per_task": 62802.890410958906,
      "last_run_timestamp": "2026-10-31T08:14:43.030Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "Nebius",
      "accuracy": 0.875421,
      "accuracy_std_dev": 0.010101000000000027,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.05678013888888889,
      "avg_duration_per_task_ms": 453449.72222222225,
      "avg_output_tokens_per_task": 62821.7904040404,
      "last_run_timestamp": "2026-10-31T08:23:57.761Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Together",
      "accuracy": 0.8754206666666668,
      "accuracy_std_dev": 0.004955863351178644,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.01787970538720539,
      "avg_duration_per_task_ms": 181946.53367003368,
      "avg_output_tokens_per_task": 29504.771043771045,
      "last_run_timestamp": "2026-10-31T08:09:31.246Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Cloudflare",
      "accuracy": 0.8754205,
      "accuracy_std_dev": 0.0016835000000000044,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.04156198232323233,
      "avg_duration_per_task_ms": 309570.8888888889,
      "avg_output_tokens_per_task": 39159.9494949495,
      "last_run_timestamp": "2026-10-17T08:26:18.265Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Venice",
      "accuracy": 0.8752804166666668,
      "accuracy_std_dev": 0.013423450409999739,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.005922073653198653,
      "avg_duration_per_task_ms": 268671.5,
      "avg_output_tokens_per_task": 18183.949074074073,
      "last_run_timestamp": "2026-10-31T08:36:26.194Z"
    },
    {
      "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": 0.87521025,
      "accuracy_std_dev": 0.005483586138422573,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.08232593308080809,
      "avg_duration_per_task_ms": 423311.8996212121,
      "avg_output_tokens_per_task": 23409.34974747475,
      "last_run_timestamp": "2026-10-31T16:15:54.664Z"
    },
    {
      "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": 0.8749371894736842,
      "accuracy_std_dev": 0.025253942161324032,
      "run_count": 4,
      "total_questions": 665,
      "avg_cost_per_task_usd": 0.11522267669172932,
      "avg_duration_per_task_ms": 531594.2751879699,
      "avg_output_tokens_per_task": 48718.98195488722,
      "last_run_timestamp": "2026-10-03T17:01:24.374Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Ambient",
      "accuracy": 0.8748596666666667,
      "accuracy_std_dev": 0.004827144451490511,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.09818961279461279,
      "avg_duration_per_task_ms": 617620.867003367,
      "avg_output_tokens_per_task": 28483.28787878788,
      "last_run_timestamp": "2026-08-29T11:31:24.094Z"
    },
    {
      "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-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Together",
      "accuracy": 0.874579,
      "accuracy_std_dev": 0.009259000000000017,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.11236396717171716,
      "avg_duration_per_task_ms": 447332.6186868687,
      "avg_output_tokens_per_task": 32120.56313131313,
      "last_run_timestamp": "2026-08-22T16:49:15.579Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Makora",
      "accuracy": 0.8745317312013828,
      "accuracy_std_dev": 0.011186512454241005,
      "run_count": 6,
      "total_questions": 1157,
      "avg_cost_per_task_usd": 0.0043274416594641305,
      "avg_duration_per_task_ms": 185028.52031114954,
      "avg_output_tokens_per_task": 21870.51080380294,
      "last_run_timestamp": "2026-09-26T08:53:59.525Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "auto-routing",
      "accuracy": 0.8744105999999999,
      "accuracy_std_dev": 0.014149500063253112,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.09050255303030302,
      "avg_duration_per_task_ms": 196310.64444444445,
      "avg_output_tokens_per_task": 14775.43686868687,
      "last_run_timestamp": "2026-10-31T08:07:27.863Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "BaseTen",
      "accuracy": 0.8743250351006383,
      "accuracy_std_dev": 0.026296453097608783,
      "run_count": 21,
      "total_questions": 4074,
      "avg_cost_per_task_usd": 0.03975532817869414,
      "avg_duration_per_task_ms": 80921.18360333824,
      "avg_output_tokens_per_task": 25880.549828178693,
      "last_run_timestamp": "2026-10-31T09:03:20.938Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "SambaNova",
      "accuracy": 0.8742983333333334,
      "accuracy_std_dev": 0.0055553002519115975,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.02925406565656566,
      "avg_duration_per_task_ms": 153006.37373737374,
      "avg_output_tokens_per_task": 28825.781144781144,
      "last_run_timestamp": "2026-10-31T08:50:42.011Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Makora",
      "accuracy": 0.8740494840344654,
      "accuracy_std_dev": 0.0038017229265163467,
      "run_count": 10,
      "total_questions": 1973,
      "avg_cost_per_task_usd": 0.04434169640141917,
      "avg_duration_per_task_ms": 202197.7790167258,
      "avg_output_tokens_per_task": 38209.482513938165,
      "last_run_timestamp": "2026-10-31T08:10:37.345Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-5-fable-20260609",
      "model_slug": "anthropic/claude-fable-5",
      "model_name": "Anthropic: Claude Fable 5",
      "provider_name": "Anthropic",
      "accuracy": 0.8740176666666666,
      "accuracy_std_dev": 0.03308150602510245,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.19026189393939397,
      "avg_duration_per_task_ms": 46363.66582491583,
      "avg_output_tokens_per_task": 3596.4814814814813,
      "last_run_timestamp": "2026-10-31T08:26:18.906Z"
    },
    {
      "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": 0.8739667650531286,
      "accuracy_std_dev": 0.012755233259575166,
      "run_count": 9,
      "total_questions": 1694,
      "avg_cost_per_task_usd": 0.008466527744982291,
      "avg_duration_per_task_ms": 226306.17237308147,
      "avg_output_tokens_per_task": 29408.253837072018,
      "last_run_timestamp": "2026-10-28T03:34:12.001Z"
    },
    {
      "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-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Baidu",
      "accuracy": 0.8735266250000001,
      "accuracy_std_dev": 0.01725979185808378,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.03874042929292929,
      "avg_duration_per_task_ms": 506055.12563131313,
      "avg_output_tokens_per_task": 35547.941287878784,
      "last_run_timestamp": "2026-10-31T16:06:00.097Z"
    },
    {
      "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": 0.8727552499999999,
      "accuracy_std_dev": 0.03887536655159452,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.18207171254208754,
      "avg_duration_per_task_ms": 44211.63804713805,
      "avg_output_tokens_per_task": 3431.4545454545455,
      "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-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "AtlasCloud",
      "accuracy": 0.872685,
      "accuracy_std_dev": 0.023807447553234246,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.11981438194444444,
      "avg_duration_per_task_ms": 617486.5606060605,
      "avg_output_tokens_per_task": 35339.65404040404,
      "last_run_timestamp": "2026-10-31T16:25:53.821Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "DigitalOcean",
      "accuracy": 0.8726149999999999,
      "accuracy_std_dev": 0.0020998715833752102,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.017651676767676767,
      "avg_duration_per_task_ms": 95564.90235690236,
      "avg_output_tokens_per_task": 29293.20202020202,
      "last_run_timestamp": "2026-10-31T08:33:05.096Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "DigitalOcean",
      "accuracy": 0.8725128181818181,
      "accuracy_std_dev": 0.06935719834676424,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.00775320890725436,
      "avg_duration_per_task_ms": 564125.6129476584,
      "avg_output_tokens_per_task": 31933.148301193756,
      "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-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Together",
      "accuracy": 0.8723242701325178,
      "accuracy_std_dev": 0.009812683936620001,
      "run_count": 10,
      "total_questions": 1962,
      "avg_cost_per_task_usd": 0.008813283384301733,
      "avg_duration_per_task_ms": 356040.2660550459,
      "avg_output_tokens_per_task": 31140.901121304792,
      "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": "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": "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": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "Alibaba",
      "accuracy": 0.8717947243121842,
      "accuracy_std_dev": 0.00772939649952788,
      "run_count": 9,
      "total_questions": 1781,
      "avg_cost_per_task_usd": 0.04942068893879843,
      "avg_duration_per_task_ms": 291097.89219539583,
      "avg_output_tokens_per_task": 20991.978663672093,
      "last_run_timestamp": "2026-10-31T16:14:42.903Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Crusoe",
      "accuracy": 0.8717391442028984,
      "accuracy_std_dev": 0.05316368413010626,
      "run_count": 7,
      "total_questions": 1380,
      "avg_cost_per_task_usd": 0.18806824130434782,
      "avg_duration_per_task_ms": 1566628.5471014492,
      "avg_output_tokens_per_task": 54424.17246376812,
      "last_run_timestamp": "2026-10-31T16:04:29.123Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-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": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "auto-routing",
      "accuracy": 0.871422375,
      "accuracy_std_dev": 0.01757503885442008,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.0447298547979798,
      "avg_duration_per_task_ms": 688258.494949495,
      "avg_output_tokens_per_task": 35524.118055555555,
      "last_run_timestamp": "2026-10-31T16:16:55.624Z"
    },
    {
      "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": "thinkingmachines/inkling-small-20260730",
      "model_slug": "thinkingmachines/inkling-small",
      "model_name": "Thinking Machines: Inkling Small",
      "provider_name": "Together",
      "accuracy": 0.8710116425661915,
      "accuracy_std_dev": 0.01025225770257459,
      "run_count": 5,
      "total_questions": 982,
      "avg_cost_per_task_usd": 0.025059682281059068,
      "avg_duration_per_task_ms": 328330.6405295316,
      "avg_output_tokens_per_task": 20642.19348268839,
      "last_run_timestamp": "2026-09-12T08:54:29.103Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.8708191333333333,
      "accuracy_std_dev": 0.005011630059061523,
      "run_count": 15,
      "total_questions": 2970,
      "avg_cost_per_task_usd": 0.004255378451178452,
      "avg_duration_per_task_ms": 39486.23198653199,
      "avg_output_tokens_per_task": 7745.265656565656,
      "last_run_timestamp": "2026-10-31T08:49:45.152Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Parasail",
      "accuracy": 0.8701296753246752,
      "accuracy_std_dev": 0.0026430000000000065,
      "run_count": 2,
      "total_questions": 385,
      "avg_cost_per_task_usd": 0.09188206233766234,
      "avg_duration_per_task_ms": 599880.7792207792,
      "avg_output_tokens_per_task": 27189.84935064935,
      "last_run_timestamp": "2026-08-22T09:51:24.721Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "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": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Z.AI",
      "accuracy": 0.8695517762430939,
      "accuracy_std_dev": 0.027759107412948937,
      "run_count": 11,
      "total_questions": 2172,
      "avg_cost_per_task_usd": 0.10362995395948434,
      "avg_duration_per_task_ms": 356439.69521178637,
      "avg_output_tokens_per_task": 23309.89779005525,
      "last_run_timestamp": "2026-10-31T08:27:45.609Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Parasail",
      "accuracy": 0.8695285000000001,
      "accuracy_std_dev": 0.00624258508151873,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.008183505681818182,
      "avg_duration_per_task_ms": 364666.8503787879,
      "avg_output_tokens_per_task": 28881.530303030304,
      "last_run_timestamp": "2026-10-31T16:18:48.433Z"
    },
    {
      "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": "anthropic/claude-5-fable-20260609",
      "model_slug": "anthropic/claude-fable-5",
      "model_name": "Anthropic: Claude Fable 5",
      "provider_name": "Azure",
      "accuracy": 0.8693882500000001,
      "accuracy_std_dev": 0.037552927930954993,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.18755991456228957,
      "avg_duration_per_task_ms": 46608.63804713805,
      "avg_output_tokens_per_task": 3539.8329124579122,
      "last_run_timestamp": "2026-10-31T08:09:25.606Z"
    },
    {
      "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": 0.8693602,
      "accuracy_std_dev": 0.027670849909607036,
      "run_count": 25,
      "total_questions": 4950,
      "avg_cost_per_task_usd": 0.15936079595959599,
      "avg_duration_per_task_ms": 153687.5486868687,
      "avg_output_tokens_per_task": 15356.345858585859,
      "last_run_timestamp": "2026-10-31T08:57:30.925Z"
    },
    {
      "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": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "AtlasCloud",
      "accuracy": 0.868965475862069,
      "accuracy_std_dev": 0.09920724349312471,
      "run_count": 11,
      "total_questions": 2175,
      "avg_cost_per_task_usd": 0.026564598160919543,
      "avg_duration_per_task_ms": 193083.65563218392,
      "avg_output_tokens_per_task": 22703.94482758621,
      "last_run_timestamp": "2026-10-31T16:04:29.094Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.1-20251113",
      "model_slug": "openai/gpt-5.1",
      "model_name": "OpenAI: GPT-5.1",
      "provider_name": "Azure",
      "accuracy": 0.8686867500000001,
      "accuracy_std_dev": 0.009143975622643609,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.2541581641414141,
      "avg_duration_per_task_ms": 251149.7840909091,
      "avg_output_tokens_per_task": 24116.972222222223,
      "last_run_timestamp": "2026-10-01T16:14:35.237Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-sonnet-5-20260630",
      "model_slug": "anthropic/claude-sonnet-5",
      "model_name": "Anthropic: Claude Sonnet 5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.8686372647058822,
      "accuracy_std_dev": 0.02658550229012359,
      "run_count": 34,
      "total_questions": 6732,
      "avg_cost_per_task_usd": 0.17476259848484849,
      "avg_duration_per_task_ms": 159604.14185977422,
      "avg_output_tokens_per_task": 16236.420083184788,
      "last_run_timestamp": "2026-10-31T08:33:01.888Z"
    },
    {
      "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": "thinkingmachines/inkling-small-20260730",
      "model_slug": "thinkingmachines/inkling-small",
      "model_name": "Thinking Machines: Inkling Small",
      "provider_name": "DeepInfra",
      "accuracy": 0.867634625,
      "accuracy_std_dev": 0.007431114989311851,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.025694838383838382,
      "avg_duration_per_task_ms": 248942.36553030304,
      "avg_output_tokens_per_task": 21167.008207070707,
      "last_run_timestamp": "2026-10-08T22:14:16.087Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "Novita",
      "accuracy": 0.8676153636363637,
      "accuracy_std_dev": 0.004014278793432891,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.0702307419651056,
      "avg_duration_per_task_ms": 231760.75160697888,
      "avg_output_tokens_per_task": 11437.757575757576,
      "last_run_timestamp": "2026-10-31T08:17:29.729Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "Google",
      "accuracy": 0.8675645333333333,
      "accuracy_std_dev": 0.00826267276262484,
      "run_count": 15,
      "total_questions": 2970,
      "avg_cost_per_task_usd": 0.004299042424242424,
      "avg_duration_per_task_ms": 36217.24208754209,
      "avg_output_tokens_per_task": 7829.657575757576,
      "last_run_timestamp": "2026-10-31T08:38:41.553Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.6-luna-20260709",
      "model_slug": "openai/gpt-5.6-luna",
      "model_name": "OpenAI: GPT-5.6 Luna",
      "provider_name": "OpenAI",
      "accuracy": 0.8675643333333334,
      "accuracy_std_dev": 0.0028614651647170092,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.009448984848484849,
      "avg_duration_per_task_ms": 79377.02525252526,
      "avg_output_tokens_per_task": 7744.79797979798,
      "last_run_timestamp": "2026-10-02T23:19:10.742Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "auto-routing",
      "accuracy": 0.8674240000000001,
      "accuracy_std_dev": 0.0048169925783625705,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.004064253787878788,
      "avg_duration_per_task_ms": 38624.37373737374,
      "avg_output_tokens_per_task": 7893.444444444444,
      "last_run_timestamp": "2026-10-31T08:27:45.642Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": "anthropic/claude-5-fable-20260609",
      "model_slug": "anthropic/claude-fable-5",
      "model_name": "Anthropic: Claude Fable 5",
      "provider_name": "Google",
      "accuracy": 0.8670734166666667,
      "accuracy_std_dev": 0.04494368828036987,
      "run_count": 24,
      "total_questions": 4752,
      "avg_cost_per_task_usd": 0.1912950467171717,
      "avg_duration_per_task_ms": 41633.452230639734,
      "avg_output_tokens_per_task": 3438.911616161616,
      "last_run_timestamp": "2026-10-31T08:27:45.697Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.1-20251113",
      "model_slug": "openai/gpt-5.1",
      "model_name": "OpenAI: GPT-5.1",
      "provider_name": "auto-routing",
      "accuracy": 0.867003,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.22265377272727274,
      "avg_duration_per_task_ms": 265427.9646464646,
      "avg_output_tokens_per_task": 22169.171717171717,
      "last_run_timestamp": "2026-10-01T16:12:54.241Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "Chutes",
      "accuracy": 0.867003,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.06361645454545455,
      "avg_duration_per_task_ms": 248018.0404040404,
      "avg_output_tokens_per_task": 21121.30808080808,
      "last_run_timestamp": "2026-08-15T15:55:02.752Z"
    },
    {
      "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": 0.8666972727272727,
      "accuracy_std_dev": 0.01677101145153263,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.10232939393939393,
      "avg_duration_per_task_ms": 485607.88200183655,
      "avg_output_tokens_per_task": 27236.82139577594,
      "last_run_timestamp": "2026-10-31T16:03:21.454Z"
    },
    {
      "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": 0.866391090909091,
      "accuracy_std_dev": 0.02614258303804019,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.007744717630853995,
      "avg_duration_per_task_ms": 429168.4632690542,
      "avg_output_tokens_per_task": 26301.81772268136,
      "last_run_timestamp": "2026-10-31T08:17:29.008Z"
    },
    {
      "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": 0.8663721249999999,
      "accuracy_std_dev": 0.06286092990768888,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.2573122026515151,
      "avg_duration_per_task_ms": 1217529.4318181819,
      "avg_output_tokens_per_task": 57063.51452020202,
      "last_run_timestamp": "2026-10-31T16:10:09.220Z"
    },
    {
      "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": 0.8663051040204212,
      "accuracy_std_dev": 0.009200990020610552,
      "run_count": 8,
      "total_questions": 1567,
      "avg_cost_per_task_usd": 0.07569640906190171,
      "avg_duration_per_task_ms": 283298.48819400126,
      "avg_output_tokens_per_task": 20901.779195915762,
      "last_run_timestamp": "2026-10-31T16:05:36.103Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "GMICloud",
      "accuracy": 0.8658810000000001,
      "accuracy_std_dev": 0.04916542176842041,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.11255429335016834,
      "avg_duration_per_task_ms": 478692.68855218857,
      "avg_output_tokens_per_task": 28354.992845117846,
      "last_run_timestamp": "2026-10-31T16:01:53.651Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "DeepInfra",
      "accuracy": 0.8656255430210325,
      "accuracy_std_dev": 0.009966828669561601,
      "run_count": 9,
      "total_questions": 1569,
      "avg_cost_per_task_usd": 0.057410206500956025,
      "avg_duration_per_task_ms": 438181.7106437221,
      "avg_output_tokens_per_task": 19027.14722753346,
      "last_run_timestamp": "2026-10-31T16:05:14.702Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "thinkingmachines/inkling-small-20260730",
      "model_slug": "thinkingmachines/inkling-small",
      "model_name": "Thinking Machines: Inkling Small",
      "provider_name": "auto-routing",
      "accuracy": 0.8656005,
      "accuracy_std_dev": 0.00803956708515244,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.02525451262626263,
      "avg_duration_per_task_ms": 233486.3627946128,
      "avg_output_tokens_per_task": 20811.0765993266,
      "last_run_timestamp": "2026-10-01T16:08:04.290Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-sonnet-5-20260630",
      "model_slug": "anthropic/claude-sonnet-5",
      "model_name": "Anthropic: Claude Sonnet 5",
      "provider_name": "Anthropic",
      "accuracy": 0.86560025,
      "accuracy_std_dev": 0.030176816372078863,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.1543242760942761,
      "avg_duration_per_task_ms": 151975.23611111112,
      "avg_output_tokens_per_task": 15215.097222222223,
      "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": "AtlasCloud",
      "accuracy": 0.865530375,
      "accuracy_std_dev": 0.017904480611969035,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.07363626578282828,
      "avg_duration_per_task_ms": 288838.01957070705,
      "avg_output_tokens_per_task": 20955.541666666668,
      "last_run_timestamp": "2026-10-31T16:09:19.242Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Relace",
      "accuracy": 0.86532,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.01495640909090909,
      "avg_duration_per_task_ms": 120441.0101010101,
      "avg_output_tokens_per_task": 29795.70707070707,
      "last_run_timestamp": "2026-10-24T15:00:02.228Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "NextBit",
      "accuracy": 0.8653195,
      "accuracy_std_dev": 0.0016835000000000044,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.008587179292929294,
      "avg_duration_per_task_ms": 240899.30303030304,
      "avg_output_tokens_per_task": 30387.573232323233,
      "last_run_timestamp": "2026-08-29T18:38:15.907Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "StreamLake",
      "accuracy": 0.8652866062099558,
      "accuracy_std_dev": 0.03145510165404201,
      "run_count": 11,
      "total_questions": 2029,
      "avg_cost_per_task_usd": 0.023189925086249383,
      "avg_duration_per_task_ms": 458930.39822572697,
      "avg_output_tokens_per_task": 22064.744701823558,
      "last_run_timestamp": "2026-10-17T08:18:04.346Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Azure",
      "accuracy": 0.8649456666666666,
      "accuracy_std_dev": 0.026980251460816472,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.13166756565656565,
      "avg_duration_per_task_ms": 468303.6307519641,
      "avg_output_tokens_per_task": 34458.837261503926,
      "last_run_timestamp": "2026-10-31T16:30:53.502Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Sail Research",
      "accuracy": 0.8647586666666667,
      "accuracy_std_dev": 0.0028614651647170097,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.005299535353535353,
      "avg_duration_per_task_ms": 492201.62457912456,
      "avg_output_tokens_per_task": 29273.70707070707,
      "last_run_timestamp": "2026-08-22T16:48:48.690Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-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": "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": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Morph",
      "accuracy": 0.8646462,
      "accuracy_std_dev": 0.03469156072822321,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.005541933333333334,
      "avg_duration_per_task_ms": 433520.95454545453,
      "avg_output_tokens_per_task": 17519.044444444444,
      "last_run_timestamp": "2026-11-05T09:00:05.070Z"
    },
    {
      "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": 0.8644780000000001,
      "accuracy_std_dev": 0.014143861990276906,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.7500333691077442,
      "avg_duration_per_task_ms": 503555.07028619526,
      "avg_output_tokens_per_task": 29837.847643097644,
      "last_run_timestamp": "2026-10-31T08:23:25.857Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8642675,
      "accuracy_std_dev": 0.01795386030217457,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.16045302777777776,
      "avg_duration_per_task_ms": 1018899.9987373737,
      "avg_output_tokens_per_task": 50975.2380050505,
      "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": "DeepInfra",
      "accuracy": 0.8641672550805886,
      "accuracy_std_dev": 0.016481482149957736,
      "run_count": 8,
      "total_questions": 1427,
      "avg_cost_per_task_usd": 0.00518134197617379,
      "avg_duration_per_task_ms": 698929.2803083392,
      "avg_output_tokens_per_task": 28479.621583742115,
      "last_run_timestamp": "2026-10-31T16:22:14.129Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.5-opus-20251124",
      "model_slug": "anthropic/claude-opus-4.5",
      "model_name": "Anthropic: Claude Opus 4.5",
      "provider_name": "auto-routing",
      "accuracy": 0.8640571666666669,
      "accuracy_std_dev": 0.010173704494376127,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.8425685521885522,
      "avg_duration_per_task_ms": 410851.17003367003,
      "avg_output_tokens_per_task": 33511.65193602694,
      "last_run_timestamp": "2026-10-31T08:13:57.628Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Phala",
      "accuracy": 0.863852179419525,
      "accuracy_std_dev": 0.02754414592685712,
      "run_count": 10,
      "total_questions": 1895,
      "avg_cost_per_task_usd": 0.012207302902374668,
      "avg_duration_per_task_ms": 440151.53984168865,
      "avg_output_tokens_per_task": 28225.067018469657,
      "last_run_timestamp": "2026-10-31T08:36:30.903Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "GMICloud",
      "accuracy": 0.8633403579175706,
      "accuracy_std_dev": 0.03134350405895737,
      "run_count": 7,
      "total_questions": 1383,
      "avg_cost_per_task_usd": 0.03313610050614606,
      "avg_duration_per_task_ms": 634565.6637744035,
      "avg_output_tokens_per_task": 38353.89515545915,
      "last_run_timestamp": "2026-10-31T08:09:30.864Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "Fireworks",
      "accuracy": 0.86321525,
      "accuracy_std_dev": 0.010127171825712258,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.022006194444444443,
      "avg_duration_per_task_ms": 265798.13636363635,
      "avg_output_tokens_per_task": 33108.22601010101,
      "last_run_timestamp": "2026-09-19T08:51:58.825Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "Alibaba",
      "accuracy": 0.8630751666666665,
      "accuracy_std_dev": 0.009048483766478349,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.0646409191919192,
      "avg_duration_per_task_ms": 213655.1308922559,
      "avg_output_tokens_per_task": 10505.142255892257,
      "last_run_timestamp": "2026-10-31T08:13:57.551Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "auto-routing",
      "accuracy": 0.8629184270372711,
      "accuracy_std_dev": 0.04098321044018362,
      "run_count": 8,
      "total_questions": 1583,
      "avg_cost_per_task_usd": 0.1642263594440935,
      "avg_duration_per_task_ms": 968291.7119393557,
      "avg_output_tokens_per_task": 55352.77511054959,
      "last_run_timestamp": "2026-10-31T16:11:41.349Z"
    },
    {
      "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": 0.8627945000000001,
      "accuracy_std_dev": 0.00840343408474575,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.7843768287037037,
      "avg_duration_per_task_ms": 448345.18644781143,
      "avg_output_tokens_per_task": 31202.670875420874,
      "last_run_timestamp": "2026-10-31T08:15:24.290Z"
    },
    {
      "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": 0.8627944166666667,
      "accuracy_std_dev": 0.018753065569209103,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.7605652819865321,
      "avg_duration_per_task_ms": 513193.70033670036,
      "avg_output_tokens_per_task": 30293.478535353537,
      "last_run_timestamp": "2026-10-31T08:30:25.667Z"
    },
    {
      "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": "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": "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": "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": 0.8617424166666667,
      "accuracy_std_dev": 0.0076116939579650746,
      "run_count": 24,
      "total_questions": 4752,
      "avg_cost_per_task_usd": 0.8836652169612793,
      "avg_duration_per_task_ms": 413694.60353535356,
      "avg_output_tokens_per_task": 33474.18686868687,
      "last_run_timestamp": "2026-10-31T08:24:08.173Z"
    },
    {
      "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": 0.8615641538461539,
      "accuracy_std_dev": 0.011629020620487162,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.8434282964257965,
      "avg_duration_per_task_ms": 459445.4759129759,
      "avg_output_tokens_per_task": 33558.62315462316,
      "last_run_timestamp": "2026-10-31T08:07:51.358Z"
    },
    {
      "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-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Baidu",
      "accuracy": 0.8613828571428571,
      "accuracy_std_dev": 0.010406829149049206,
      "run_count": 8,
      "total_questions": 1533,
      "avg_cost_per_task_usd": 0.004015036529680365,
      "avg_duration_per_task_ms": 245941.22635355513,
      "avg_output_tokens_per_task": 28674.964122635356,
      "last_run_timestamp": "2026-10-31T16:05:36.103Z"
    },
    {
      "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": 0.8613568027812896,
      "accuracy_std_dev": 0.010662573677940023,
      "run_count": 8,
      "total_questions": 1582,
      "avg_cost_per_task_usd": 0.0375921409608091,
      "avg_duration_per_task_ms": 554775.7035398231,
      "avg_output_tokens_per_task": 34882.013274336285,
      "last_run_timestamp": "2026-10-31T16:11:17.542Z"
    },
    {
      "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": 0.8608771111111111,
      "accuracy_std_dev": 0.03464756893169482,
      "run_count": 36,
      "total_questions": 7128,
      "avg_cost_per_task_usd": 0.16595036826599324,
      "avg_duration_per_task_ms": 152410.04938271604,
      "avg_output_tokens_per_task": 15337.304854096521,
      "last_run_timestamp": "2026-10-31T08:51:08.192Z"
    },
    {
      "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": "anthropic/claude-4.6-opus-20260205",
      "model_slug": "anthropic/claude-opus-4.6",
      "model_name": "Anthropic: Claude Opus 4.6",
      "provider_name": "Anthropic",
      "accuracy": 0.8603986923076923,
      "accuracy_std_dev": 0.01707241294461563,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.7520348795648795,
      "avg_duration_per_task_ms": 507732.5505050505,
      "avg_output_tokens_per_task": 29977.13014763015,
      "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": "Together",
      "accuracy": 0.860269,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.2318445,
      "avg_duration_per_task_ms": 364508.0101010101,
      "avg_output_tokens_per_task": 51949.767676767675,
      "last_run_timestamp": "2026-08-15T08:57:45.830Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "reflection/beam-1.0-501b-a23b-20261016",
      "model_slug": "reflection/beam-1.0-501b-a23b",
      "model_name": "Reflection: Beam 1.0 501B A23B",
      "provider_name": "Io Net",
      "accuracy": 0.860269,
      "accuracy_std_dev": 0.010100999999999971,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.00987439393939394,
      "avg_duration_per_task_ms": 74831.28787878787,
      "avg_output_tokens_per_task": 9997.058080808081,
      "last_run_timestamp": "2026-10-24T08:14:36.264Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Chutes",
      "accuracy": 0.859848375,
      "accuracy_std_dev": 0.022266733645381724,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.13791987941919195,
      "avg_duration_per_task_ms": 1281636.821338384,
      "avg_output_tokens_per_task": 52297.90909090909,
      "last_run_timestamp": "2026-10-31T16:08:31.067Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.6-opus-20260205",
      "model_slug": "anthropic/claude-opus-4.6",
      "model_name": "Anthropic: Claude Opus 4.6",
      "provider_name": "Google",
      "accuracy": 0.8598483333333333,
      "accuracy_std_dev": 0.018068234858139545,
      "run_count": 24,
      "total_questions": 4752,
      "avg_cost_per_task_usd": 0.8039200313552188,
      "avg_duration_per_task_ms": 487583.53682659933,
      "avg_output_tokens_per_task": 30465.617424242424,
      "last_run_timestamp": "2026-10-31T09:25:22.406Z"
    },
    {
      "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": 0.8597737453957996,
      "accuracy_std_dev": 0.00946268128798041,
      "run_count": 19,
      "total_questions": 3095,
      "avg_cost_per_task_usd": 0.17348523166397414,
      "avg_duration_per_task_ms": 253011.81744749597,
      "avg_output_tokens_per_task": 31870.79353796446,
      "last_run_timestamp": "2026-09-26T08:54:22.466Z"
    },
    {
      "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": 0.859708,
      "accuracy_std_dev": 0.01589205338106649,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.09237588832772166,
      "avg_duration_per_task_ms": 179383.0931537598,
      "avg_output_tokens_per_task": 25530.111672278337,
      "last_run_timestamp": "2026-10-31T16:04:29.122Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "Io Net",
      "accuracy": 0.8596746149425287,
      "accuracy_std_dev": 0.000831499999999985,
      "run_count": 2,
      "total_questions": 348,
      "avg_cost_per_task_usd": 0.0899417959770115,
      "avg_duration_per_task_ms": 1024745.4281609196,
      "avg_output_tokens_per_task": 35825.26724137931,
      "last_run_timestamp": "2026-10-27T13:57:40.548Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Inceptron",
      "accuracy": 0.8594970187441424,
      "accuracy_std_dev": 0.03146274773954612,
      "run_count": 11,
      "total_questions": 2134,
      "avg_cost_per_task_usd": 0.1325802572633552,
      "avg_duration_per_task_ms": 467358.20337394567,
      "avg_output_tokens_per_task": 43781.12417994377,
      "last_run_timestamp": "2026-10-31T16:16:35.161Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "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-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Mancer 2",
      "accuracy": 0.859217125,
      "accuracy_std_dev": 0.010875888888241494,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.014054850378787878,
      "avg_duration_per_task_ms": 809652.5277777778,
      "avg_output_tokens_per_task": 28625.664772727272,
      "last_run_timestamp": "2026-10-31T16:17:15.719Z"
    },
    {
      "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": 0.8590065833333335,
      "accuracy_std_dev": 0.008723943751331079,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.8530048674242424,
      "avg_duration_per_task_ms": 481951.62794612796,
      "avg_output_tokens_per_task": 33929.37457912458,
      "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": "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": "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": 0.858165,
      "accuracy_std_dev": 0.026280391530568943,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.07870610921717172,
      "avg_duration_per_task_ms": 466337.7960858586,
      "avg_output_tokens_per_task": 23353.214646464647,
      "last_run_timestamp": "2026-10-31T16:12:07.201Z"
    },
    {
      "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": "anthropic/claude-4.5-opus-20251124",
      "model_slug": "anthropic/claude-opus-4.5",
      "model_name": "Anthropic: Claude Opus 4.5",
      "provider_name": "Azure",
      "accuracy": 0.8580245833333333,
      "accuracy_std_dev": 0.017007195744048704,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.8381309111952863,
      "avg_duration_per_task_ms": 457642.2281144781,
      "avg_output_tokens_per_task": 33349.60101010101,
      "last_run_timestamp": "2026-10-31T08:40:44.982Z"
    },
    {
      "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": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "Minimax",
      "accuracy": 0.8572705625000001,
      "accuracy_std_dev": 0.007924119438214786,
      "run_count": 16,
      "total_questions": 3168,
      "avg_cost_per_task_usd": 0.06404825631313131,
      "avg_duration_per_task_ms": 595272.8756313132,
      "avg_output_tokens_per_task": 35321.874684343435,
      "last_run_timestamp": "2026-10-31T16:08:56.835Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Azure",
      "accuracy": 0.8570892222222223,
      "accuracy_std_dev": 0.031137791974732713,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.015986119528619527,
      "avg_duration_per_task_ms": 333449.29124579125,
      "avg_output_tokens_per_task": 28587.83613916947,
      "last_run_timestamp": "2026-10-31T16:04:48.120Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Wafer",
      "accuracy": 0.8569025,
      "accuracy_std_dev": 0.005050499999999958,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.018898722222222223,
      "avg_duration_per_task_ms": 164757.63383838383,
      "avg_output_tokens_per_task": 31337.035353535353,
      "last_run_timestamp": "2026-10-24T08:17:56.393Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5-2025-08-07",
      "model_slug": "openai/gpt-5",
      "model_name": "OpenAI: GPT-5",
      "provider_name": "auto-routing",
      "accuracy": 0.8569019999999999,
      "accuracy_std_dev": 0.0033669999999999534,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.20989276262626264,
      "avg_duration_per_task_ms": 263850.72222222225,
      "avg_output_tokens_per_task": 20893.070707070707,
      "last_run_timestamp": "2026-10-01T16:31:36.833Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Sail Research",
      "accuracy": 0.8568698741301679,
      "accuracy_std_dev": 0.08960692857100729,
      "run_count": 25,
      "total_questions": 4886,
      "avg_cost_per_task_usd": 0.00782735918952108,
      "avg_duration_per_task_ms": 558967.4410560786,
      "avg_output_tokens_per_task": 27058.752762996315,
      "last_run_timestamp": "2026-11-06T22:28:22.713Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "BaseTen",
      "accuracy": 0.8566210871819038,
      "accuracy_std_dev": 0.02083134388139214,
      "run_count": 43,
      "total_questions": 8488,
      "avg_cost_per_task_usd": 0.2924742340951932,
      "avg_duration_per_task_ms": 342907.9387370405,
      "avg_output_tokens_per_task": 52593.62170122526,
      "last_run_timestamp": "2026-10-31T16:26:37.892Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Novita",
      "accuracy": 0.8564403552465234,
      "accuracy_std_dev": 0.02676663000516378,
      "run_count": 12,
      "total_questions": 2373,
      "avg_cost_per_task_usd": 0.08082342730720607,
      "avg_duration_per_task_ms": 485283.54403708386,
      "avg_output_tokens_per_task": 41227.2022756005,
      "last_run_timestamp": "2026-10-31T16:30:10.027Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "auto-routing",
      "accuracy": 0.8563411666666667,
      "accuracy_std_dev": 0.009991279087562064,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.005837240740740741,
      "avg_duration_per_task_ms": 544489.4410774411,
      "avg_output_tokens_per_task": 31556.968855218856,
      "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": "Venice",
      "accuracy": 0.855850125,
      "accuracy_std_dev": 0.029146120652830884,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.09797882512626263,
      "avg_duration_per_task_ms": 344142.03661616164,
      "avg_output_tokens_per_task": 21821.31313131313,
      "last_run_timestamp": "2026-10-31T16:15:29.279Z"
    },
    {
      "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": 0.8558501249999999,
      "accuracy_std_dev": 0.0195049467087038,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.04610065845959596,
      "avg_duration_per_task_ms": 601194.1792929292,
      "avg_output_tokens_per_task": 37602.47222222222,
      "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": "GMICloud",
      "accuracy": 0.8556823130287649,
      "accuracy_std_dev": 0.021169598748854452,
      "run_count": 12,
      "total_questions": 2364,
      "avg_cost_per_task_usd": 0.08525841285956007,
      "avg_duration_per_task_ms": 295845.3159898477,
      "avg_output_tokens_per_task": 25006.826565143823,
      "last_run_timestamp": "2026-10-31T08:34:08.441Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "BaseTen",
      "accuracy": 0.8556049172108208,
      "accuracy_std_dev": 0.03188071644780284,
      "run_count": 22,
      "total_questions": 4288,
      "avg_cost_per_task_usd": 0.003744977378731343,
      "avg_duration_per_task_ms": 73667.2943097015,
      "avg_output_tokens_per_task": 13976.44776119403,
      "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-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Baidu",
      "accuracy": 0.8554592857142856,
      "accuracy_std_dev": 0.010776903155946642,
      "run_count": 21,
      "total_questions": 4158,
      "avg_cost_per_task_usd": 0.14148054425204423,
      "avg_duration_per_task_ms": 433829.56758056756,
      "avg_output_tokens_per_task": 44977.72101972102,
      "last_run_timestamp": "2026-10-31T16:26:59.503Z"
    },
    {
      "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": 0.8552186666666667,
      "accuracy_std_dev": 0.0148680429631998,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.007792587542087542,
      "avg_duration_per_task_ms": 536143.4478114478,
      "avg_output_tokens_per_task": 28061.833333333332,
      "last_run_timestamp": "2026-08-29T18:49:16.416Z"
    },
    {
      "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": 0.8550504000000001,
      "accuracy_std_dev": 0.012697892748011365,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.029575816161616166,
      "avg_duration_per_task_ms": 786613.2691919192,
      "avg_output_tokens_per_task": 33730.74898989899,
      "last_run_timestamp": "2026-10-17T08:19:33.830Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "DigitalOcean",
      "accuracy": 0.8549781428571429,
      "accuracy_std_dev": 0.06322451474914856,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.16852446320346318,
      "avg_duration_per_task_ms": 1962682.6176046175,
      "avg_output_tokens_per_task": 59425.0505050505,
      "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": "AtlasCloud",
      "accuracy": 0.8546926249999999,
      "accuracy_std_dev": 0.022345683453060337,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.04371308901515151,
      "avg_duration_per_task_ms": 649064.4501262626,
      "avg_output_tokens_per_task": 36365.37247474748,
      "last_run_timestamp": "2026-10-31T16:02:14.360Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Inceptron",
      "accuracy": 0.8545531474226805,
      "accuracy_std_dev": 0.018117679996070137,
      "run_count": 10,
      "total_questions": 1940,
      "avg_cost_per_task_usd": 0.08358154742268041,
      "avg_duration_per_task_ms": 325785.14793814434,
      "avg_output_tokens_per_task": 28959.811340206186,
      "last_run_timestamp": "2026-10-31T08:56:36.753Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "GMICloud",
      "accuracy": 0.8540963333333333,
      "accuracy_std_dev": 0.019358062075872505,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.03021066049382716,
      "avg_duration_per_task_ms": 606023.7901234567,
      "avg_output_tokens_per_task": 34216.56677890011,
      "last_run_timestamp": "2026-10-31T16:02:14.339Z"
    },
    {
      "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": 0.8540087088607595,
      "accuracy_std_dev": 0.011340000000000017,
      "run_count": 2,
      "total_questions": 395,
      "avg_cost_per_task_usd": 0.023069045569620254,
      "avg_duration_per_task_ms": 780937.6962025317,
      "avg_output_tokens_per_task": 22650.07341772152,
      "last_run_timestamp": "2026-08-22T10:03:34.994Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Ambient",
      "accuracy": 0.8535090068846816,
      "accuracy_std_dev": 0.005804304535067259,
      "run_count": 9,
      "total_questions": 1743,
      "avg_cost_per_task_usd": 0.09211812679288582,
      "avg_duration_per_task_ms": 680955.4825014343,
      "avg_output_tokens_per_task": 42013.00631095812,
      "last_run_timestamp": "2026-10-31T16:11:17.543Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "DekaLLM",
      "accuracy": 0.8530642646239555,
      "accuracy_std_dev": 0.022329477337983097,
      "run_count": 4,
      "total_questions": 718,
      "avg_cost_per_task_usd": 0.02860370194986072,
      "avg_duration_per_task_ms": 460751.2409470752,
      "avg_output_tokens_per_task": 34791.85933147632,
      "last_run_timestamp": "2026-10-10T08:22:41.628Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Together",
      "accuracy": 0.852836585106383,
      "accuracy_std_dev": 0.00992599999999999,
      "run_count": 2,
      "total_questions": 376,
      "avg_cost_per_task_usd": 0.10881258244680853,
      "avg_duration_per_task_ms": 389377.2340425532,
      "avg_output_tokens_per_task": 27704.662234042553,
      "last_run_timestamp": "2026-08-22T10:19:36.962Z"
    },
    {
      "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": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Baidu",
      "accuracy": 0.8521639072512648,
      "accuracy_std_dev": 0.012873132633608224,
      "run_count": 9,
      "total_questions": 1779,
      "avg_cost_per_task_usd": 0.08519417144463183,
      "avg_duration_per_task_ms": 288650.1905564924,
      "avg_output_tokens_per_task": 29819.736368746486,
      "last_run_timestamp": "2026-10-31T08:22:52.355Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8521324166666666,
      "accuracy_std_dev": 0.01894507738375299,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.12555288047138044,
      "avg_duration_per_task_ms": 374701.57112794614,
      "avg_output_tokens_per_task": 39360.16035353536,
      "last_run_timestamp": "2026-10-31T16:08:08.543Z"
    },
    {
      "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": 0.8521323333333334,
      "accuracy_std_dev": 0.03225728546993512,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.005255346801346801,
      "avg_duration_per_task_ms": 1011045.1414141414,
      "avg_output_tokens_per_task": 28586.09259259259,
      "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-sonnet-5-20260630",
      "model_slug": "anthropic/claude-sonnet-5",
      "model_name": "Anthropic: Claude Sonnet 5",
      "provider_name": "auto-routing",
      "accuracy": 0.8521322499999999,
      "accuracy_std_dev": 0.047036847074261895,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.16922138846801346,
      "avg_duration_per_task_ms": 167377.74537037036,
      "avg_output_tokens_per_task": 16702.02441077441,
      "last_run_timestamp": "2026-10-31T08:12:45.468Z"
    },
    {
      "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": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8520922857142857,
      "accuracy_std_dev": 0.010237499481448224,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.014528669552669555,
      "avg_duration_per_task_ms": 664432.2056277057,
      "avg_output_tokens_per_task": 19049.7316017316,
      "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-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": "openai/gpt-5-2025-08-07",
      "model_slug": "openai/gpt-5",
      "model_name": "OpenAI: GPT-5",
      "provider_name": "Azure",
      "accuracy": 0.8518517499999999,
      "accuracy_std_dev": 0.0020620621201845155,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.21063902146464644,
      "avg_duration_per_task_ms": 251655.29545454544,
      "avg_output_tokens_per_task": 19971.159090909092,
      "last_run_timestamp": "2026-10-01T16:27:58.636Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "Parasail",
      "accuracy": 0.8514385818947369,
      "accuracy_std_dev": 0.00836367476510108,
      "run_count": 12,
      "total_questions": 2375,
      "avg_cost_per_task_usd": 0.03818178821052631,
      "avg_duration_per_task_ms": 384700.4193684211,
      "avg_output_tokens_per_task": 38833.43747368421,
      "last_run_timestamp": "2026-10-31T16:26:59.494Z"
    },
    {
      "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": 0.8513607500000001,
      "accuracy_std_dev": 0.010092092937253725,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.09869350589225588,
      "avg_duration_per_task_ms": 554753.9515993266,
      "avg_output_tokens_per_task": 30538.262205387204,
      "last_run_timestamp": "2026-10-31T16:01:53.623Z"
    },
    {
      "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-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Decart",
      "accuracy": 0.8512828567483604,
      "accuracy_std_dev": 0.02197685353720036,
      "run_count": 15,
      "total_questions": 2897,
      "avg_cost_per_task_usd": 0.21227700207110806,
      "avg_duration_per_task_ms": 253125.91128753882,
      "avg_output_tokens_per_task": 43881.1304798067,
      "last_run_timestamp": "2026-10-24T16:29:51.093Z"
    },
    {
      "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": 0.850783975720789,
      "accuracy_std_dev": 0.012623895335434301,
      "run_count": 10,
      "total_questions": 1977,
      "avg_cost_per_task_usd": 0.03144315275670207,
      "avg_duration_per_task_ms": 755960.0136570561,
      "avg_output_tokens_per_task": 32811.26707132018,
      "last_run_timestamp": "2026-10-17T08:19:39.242Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Crusoe",
      "accuracy": 0.8502175226400613,
      "accuracy_std_dev": 0.012397473946332766,
      "run_count": 8,
      "total_questions": 1303,
      "avg_cost_per_task_usd": 0.00643011281657713,
      "avg_duration_per_task_ms": 366999.48887183424,
      "avg_output_tokens_per_task": 15845.35226400614,
      "last_run_timestamp": "2026-10-31T16:08:33.701Z"
    },
    {
      "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": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Minimax",
      "accuracy": 0.8501683333333333,
      "accuracy_std_dev": 0.004956089811759068,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.05497935774410773,
      "avg_duration_per_task_ms": 419820.80387205386,
      "avg_output_tokens_per_task": 28107.88973063973,
      "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-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Sail Research",
      "accuracy": 0.850168,
      "accuracy_std_dev": 0.004761657064510223,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.09743340740740741,
      "avg_duration_per_task_ms": 424537.84175084176,
      "avg_output_tokens_per_task": 31131.93771043771,
      "last_run_timestamp": "2026-08-29T11:19:17.684Z"
    },
    {
      "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": 0.84974725,
      "accuracy_std_dev": 0.054530591026390865,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.005316606060606061,
      "avg_duration_per_task_ms": 207511.0707070707,
      "avg_output_tokens_per_task": 18629.959595959597,
      "last_run_timestamp": "2026-11-05T15:00:03.331Z"
    },
    {
      "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-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "auto-routing",
      "accuracy": 0.8492564166666666,
      "accuracy_std_dev": 0.015991238316957257,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.1367386346801347,
      "avg_duration_per_task_ms": 450843.9692760943,
      "avg_output_tokens_per_task": 49203.65193602694,
      "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": "Parasail",
      "accuracy": 0.8492264405594405,
      "accuracy_std_dev": 0.009705494209950108,
      "run_count": 8,
      "total_questions": 1573,
      "avg_cost_per_task_usd": 0.007261696757787667,
      "avg_duration_per_task_ms": 592695.9872854418,
      "avg_output_tokens_per_task": 17964.459631277812,
      "last_run_timestamp": "2026-10-31T16:12:28.235Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-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": "anthropic/claude-sonnet-5-20260630",
      "model_slug": "anthropic/claude-sonnet-5",
      "model_name": "Anthropic: Claude Sonnet 5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.84890575,
      "accuracy_std_dev": 0.05035112340045156,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.1629483771043771,
      "avg_duration_per_task_ms": 162891.44065656565,
      "avg_output_tokens_per_task": 16117.493265993266,
      "last_run_timestamp": "2026-10-31T08:45:37.246Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8489055833333334,
      "accuracy_std_dev": 0.01662490835793455,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.052647822390572384,
      "avg_duration_per_task_ms": 274255.4238215488,
      "avg_output_tokens_per_task": 45768.43518518518,
      "last_run_timestamp": "2026-10-31T16:53:23.847Z"
    },
    {
      "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": "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": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Inceptron",
      "accuracy": 0.848485,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.019539080808080807,
      "avg_duration_per_task_ms": 826109.8787878788,
      "avg_output_tokens_per_task": 23180.459595959597,
      "last_run_timestamp": "2026-08-15T14:57:08.773Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Alibaba",
      "accuracy": 0.84811653125,
      "accuracy_std_dev": 0.0186791945439712,
      "run_count": 32,
      "total_questions": 6336,
      "avg_cost_per_task_usd": 0.2417097290088384,
      "avg_duration_per_task_ms": 494923.91840277775,
      "avg_output_tokens_per_task": 47859.75946969697,
      "last_run_timestamp": "2026-10-31T17:01:17.198Z"
    },
    {
      "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": 0.8480038571428572,
      "accuracy_std_dev": 0.046172433270539756,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.09876766017316017,
      "avg_duration_per_task_ms": 1234895.2164502165,
      "avg_output_tokens_per_task": 39606.40476190476,
      "last_run_timestamp": "2026-10-31T16:01:53.650Z"
    },
    {
      "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": 0.847949384816754,
      "accuracy_std_dev": 0.009393024418270302,
      "run_count": 12,
      "total_questions": 2292,
      "avg_cost_per_task_usd": 0.06595389223385689,
      "avg_duration_per_task_ms": 190454.0772251309,
      "avg_output_tokens_per_task": 20514.38787085515,
      "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-5.3-20260816",
      "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": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "GMICloud",
      "accuracy": 0.8473217418810629,
      "accuracy_std_dev": 0.031381320315484754,
      "run_count": 12,
      "total_questions": 2371,
      "avg_cost_per_task_usd": 0.009025258118937158,
      "avg_duration_per_task_ms": 479873.75579924084,
      "avg_output_tokens_per_task": 32127.526360185577,
      "last_run_timestamp": "2026-10-31T08:31:37.876Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-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": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "AtlasCloud",
      "accuracy": 0.84680125,
      "accuracy_std_dev": 0.03303257248440542,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.011049029040404038,
      "avg_duration_per_task_ms": 454169.0126262626,
      "avg_output_tokens_per_task": 39139.167297979795,
      "last_run_timestamp": "2026-10-31T16:04:29.124Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "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": "ModelRun",
      "accuracy": 0.8459595000000001,
      "accuracy_std_dev": 0.010636232717304867,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.019541941077441075,
      "avg_duration_per_task_ms": 139818.19360269362,
      "avg_output_tokens_per_task": 19064.3265993266,
      "last_run_timestamp": "2026-10-31T16:05:14.635Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-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-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8446468343465044,
      "accuracy_std_dev": 0.01783018725336332,
      "run_count": 10,
      "total_questions": 1974,
      "avg_cost_per_task_usd": 0.0660811013171226,
      "avg_duration_per_task_ms": 245353.75430597772,
      "avg_output_tokens_per_task": 20336.286220871327,
      "last_run_timestamp": "2026-10-31T08:32:53.731Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "SambaNova",
      "accuracy": 0.8446020893155258,
      "accuracy_std_dev": 0.020816297010995962,
      "run_count": 8,
      "total_questions": 1198,
      "avg_cost_per_task_usd": 0.018853567612687815,
      "avg_duration_per_task_ms": 104380.7020033389,
      "avg_output_tokens_per_task": 16184.987479131887,
      "last_run_timestamp": "2026-10-31T16:06:00.067Z"
    },
    {
      "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/gemini-3.5-flash-lite-20260721",
      "model_slug": "google/gemini-3.5-flash-lite",
      "model_name": "Google: Gemini 3.5 Flash Lite",
      "provider_name": "auto-routing",
      "accuracy": 0.844276,
      "accuracy_std_dev": 0.009259000000000017,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.036391522727272724,
      "avg_duration_per_task_ms": 53072.40404040404,
      "avg_output_tokens_per_task": 14468.015151515152,
      "last_run_timestamp": "2026-10-01T19:12:26.032Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.8439077604166666,
      "accuracy_std_dev": 0.009070260174817954,
      "run_count": 12,
      "total_questions": 2112,
      "avg_cost_per_task_usd": 0.12450989630681818,
      "avg_duration_per_task_ms": 153233.97348484848,
      "avg_output_tokens_per_task": 31747.522727272728,
      "last_run_timestamp": "2026-10-31T08:45:40.383Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Mistral",
      "accuracy": 0.8438831999999998,
      "accuracy_std_dev": 0.010796604899689534,
      "run_count": 30,
      "total_questions": 5940,
      "avg_cost_per_task_usd": 0.1486385685185185,
      "avg_duration_per_task_ms": 173235.29764309764,
      "avg_output_tokens_per_task": 32642.098316498315,
      "last_run_timestamp": "2026-10-31T08:24:42.872Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "Moonshot AI",
      "accuracy": 0.8434343333333334,
      "accuracy_std_dev": 0.0404743207505972,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.12289456902356903,
      "avg_duration_per_task_ms": 641373.1111111111,
      "avg_output_tokens_per_task": 41514.85353535353,
      "last_run_timestamp": "2026-08-29T18:25:54.650Z"
    },
    {
      "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": "qwen/qwen3.5-122b-a10b-20260224",
      "model_slug": "qwen/qwen3.5-122b-a10b",
      "model_name": "Qwen: Qwen3.5-122B-A10B",
      "provider_name": "DeepInfra",
      "accuracy": 0.8433156148597423,
      "accuracy_std_dev": 0.007093117054653797,
      "run_count": 7,
      "total_questions": 1319,
      "avg_cost_per_task_usd": 0.05370390523123579,
      "avg_duration_per_task_ms": 267310.77255496586,
      "avg_output_tokens_per_task": 22285.294920394237,
      "last_run_timestamp": "2026-09-26T09:50:39.842Z"
    },
    {
      "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": 0.843159054532056,
      "accuracy_std_dev": 0.012250910211082282,
      "run_count": 8,
      "total_questions": 1357,
      "avg_cost_per_task_usd": 0.0069029462048636695,
      "avg_duration_per_task_ms": 626843.2291820191,
      "avg_output_tokens_per_task": 17018.759027266027,
      "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-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "Z.AI",
      "accuracy": 0.8430962619246861,
      "accuracy_std_dev": 0.029385860831636818,
      "run_count": 7,
      "total_questions": 1195,
      "avg_cost_per_task_usd": 0.08830847447698745,
      "avg_duration_per_task_ms": 1349078.5882845188,
      "avg_output_tokens_per_task": 40536.07280334728,
      "last_run_timestamp": "2026-10-31T16:04:23.355Z"
    },
    {
      "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": 0.8430133333333335,
      "accuracy_std_dev": 0.021951449114402964,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.05437653409090909,
      "avg_duration_per_task_ms": 247748.2095959596,
      "avg_output_tokens_per_task": 23043.897727272728,
      "last_run_timestamp": "2026-10-31T16:08:08.536Z"
    },
    {
      "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": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Io Net",
      "accuracy": 0.8425924999999999,
      "accuracy_std_dev": 0.015993499999999994,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.006982,
      "avg_duration_per_task_ms": 216832.21464646465,
      "avg_output_tokens_per_task": 14749.073232323231,
      "last_run_timestamp": "2026-08-22T09:56:05.550Z"
    },
    {
      "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": 0.8425923750000001,
      "accuracy_std_dev": 0.07901703473450755,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.06442433396464646,
      "avg_duration_per_task_ms": 1182218.1875,
      "avg_output_tokens_per_task": 33113.868055555555,
      "last_run_timestamp": "2026-10-31T16:05:14.716Z"
    },
    {
      "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": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "StreamLake",
      "accuracy": 0.8423118888888889,
      "accuracy_std_dev": 0.04281090341242117,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.0038540690235690234,
      "avg_duration_per_task_ms": 561172.7755331089,
      "avg_output_tokens_per_task": 41338.6531986532,
      "last_run_timestamp": "2026-10-31T16:18:02.140Z"
    },
    {
      "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": "Decart",
      "accuracy": 0.8422337612565445,
      "accuracy_std_dev": 0.023212455418589382,
      "run_count": 10,
      "total_questions": 1910,
      "avg_cost_per_task_usd": 0.10300291204188483,
      "avg_duration_per_task_ms": 203070.60157068062,
      "avg_output_tokens_per_task": 30819.346073298428,
      "last_run_timestamp": "2026-10-31T08:52:06.333Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Together",
      "accuracy": 0.842080958563536,
      "accuracy_std_dev": 0.01924201603413352,
      "run_count": 11,
      "total_questions": 2172,
      "avg_cost_per_task_usd": 0.02133896546961326,
      "avg_duration_per_task_ms": 337269.2900552486,
      "avg_output_tokens_per_task": 42459.25598526704,
      "last_run_timestamp": "2026-10-31T08:44:51.872Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "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": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Together",
      "accuracy": 0.8416756038338659,
      "accuracy_std_dev": 0.009989855274226949,
      "run_count": 5,
      "total_questions": 939,
      "avg_cost_per_task_usd": 0.01604648136315229,
      "avg_duration_per_task_ms": 662324.1448349308,
      "avg_output_tokens_per_task": 17011.09478168264,
      "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-5.3-20260816",
      "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": "qwen/qwen3.5-122b-a10b-20260224",
      "model_slug": "qwen/qwen3.5-122b-a10b",
      "model_name": "Qwen: Qwen3.5-122B-A10B",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8414702500000001,
      "accuracy_std_dev": 0.0327729719792723,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.05281460311447811,
      "avg_duration_per_task_ms": 205567.1611952862,
      "avg_output_tokens_per_task": 21904.65446127946,
      "last_run_timestamp": "2026-10-31T16:28:22.255Z"
    },
    {
      "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": 0.8413764444444444,
      "accuracy_std_dev": 0.014280020589085009,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.13749752525252526,
      "avg_duration_per_task_ms": 888166.4298540965,
      "avg_output_tokens_per_task": 54880.480359147026,
      "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": "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": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "StreamLake",
      "accuracy": 0.8411896666666668,
      "accuracy_std_dev": 0.025531658844135698,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.16060966610549945,
      "avg_duration_per_task_ms": 1022232.3557800225,
      "avg_output_tokens_per_task": 52797.9152637486,
      "last_run_timestamp": "2026-10-31T16:39:29.347Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.5-flash-lite-20260721",
      "model_slug": "google/gemini-3.5-flash-lite",
      "model_name": "Google: Gemini 3.5 Flash Lite",
      "provider_name": "Google AI Studio",
      "accuracy": 0.8409089999999999,
      "accuracy_std_dev": 0.010943000000000036,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.03708839646464646,
      "avg_duration_per_task_ms": 47515.843434343435,
      "avg_output_tokens_per_task": 14746.045454545454,
      "last_run_timestamp": "2026-10-01T19:13:29.958Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Fireworks",
      "accuracy": 0.840891889212828,
      "accuracy_std_dev": 0.01539818649719325,
      "run_count": 16,
      "total_questions": 3087,
      "avg_cost_per_task_usd": 0.022577127632005183,
      "avg_duration_per_task_ms": 250189.5425979916,
      "avg_output_tokens_per_task": 38449.29964366699,
      "last_run_timestamp": "2026-10-31T08:54:15.638Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Together",
      "accuracy": 0.8408359820854386,
      "accuracy_std_dev": 0.011558364030339736,
      "run_count": 11,
      "total_questions": 2177,
      "avg_cost_per_task_usd": 0.14478948920532844,
      "avg_duration_per_task_ms": 203069.29581993568,
      "avg_output_tokens_per_task": 32673.29122645843,
      "last_run_timestamp": "2026-10-31T09:21:48.528Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-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": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "Venice",
      "accuracy": 0.8404879999999999,
      "accuracy_std_dev": 0.020003728864889173,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.037026556818181816,
      "avg_duration_per_task_ms": 113840.68308080808,
      "avg_output_tokens_per_task": 34919.68308080808,
      "last_run_timestamp": "2026-10-31T09:18:03.023Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Sail Research",
      "accuracy": 0.8403390039479797,
      "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": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Venice",
      "accuracy": 0.8398567499999999,
      "accuracy_std_dev": 0.010281237665646122,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.005851238636363636,
      "avg_duration_per_task_ms": 457462.80303030304,
      "avg_output_tokens_per_task": 25951.3125,
      "last_run_timestamp": "2026-10-31T16:45:40.019Z"
    },
    {
      "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": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "DigitalOcean",
      "accuracy": 0.839812126197183,
      "accuracy_std_dev": 0.07223149496882991,
      "run_count": 9,
      "total_questions": 1775,
      "avg_cost_per_task_usd": 0.06668295830985914,
      "avg_duration_per_task_ms": 1704327.0591549296,
      "avg_output_tokens_per_task": 20074.104225352112,
      "last_run_timestamp": "2026-10-31T16:13:09.868Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Io Net",
      "accuracy": 0.8397870000000001,
      "accuracy_std_dev": 0.06305207943174174,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.013501392255892255,
      "avg_duration_per_task_ms": 269245.1767676768,
      "avg_output_tokens_per_task": 24293.242424242424,
      "last_run_timestamp": "2026-10-01T17:44:47.331Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Together",
      "accuracy": 0.8393658333333334,
      "accuracy_std_dev": 0.01474792674600588,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.23697320664983168,
      "avg_duration_per_task_ms": 282494.2058080808,
      "avg_output_tokens_per_task": 53697.90698653198,
      "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-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Mistral",
      "accuracy": 0.8391730814863104,
      "accuracy_std_dev": 0.011927781960856079,
      "run_count": 31,
      "total_questions": 6136,
      "avg_cost_per_task_usd": 0.26252152737940027,
      "avg_duration_per_task_ms": 275082.67649934813,
      "avg_output_tokens_per_task": 57033.21740547588,
      "last_run_timestamp": "2026-10-31T16:16:55.693Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "StreamLake",
      "accuracy": 0.8387440275454993,
      "accuracy_std_dev": 0.019690853286246784,
      "run_count": 11,
      "total_questions": 2033,
      "avg_cost_per_task_usd": 0.07400631874077719,
      "avg_duration_per_task_ms": 403244.14658140676,
      "avg_output_tokens_per_task": 24425.756025577964,
      "last_run_timestamp": "2026-10-31T16:18:04.837Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Chutes",
      "accuracy": 0.8387141903143586,
      "accuracy_std_dev": 0.029140860908219966,
      "run_count": 12,
      "total_questions": 2354,
      "avg_cost_per_task_usd": 0.0665123759558199,
      "avg_duration_per_task_ms": 1085367.209855565,
      "avg_output_tokens_per_task": 33157.14528462192,
      "last_run_timestamp": "2026-10-31T16:05:36.103Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "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-5.3-20260816",
      "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": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8383836666666667,
      "accuracy_std_dev": 0.006299258386685094,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.02860740067340067,
      "avg_duration_per_task_ms": 515900.1077441077,
      "avg_output_tokens_per_task": 23652.510101010103,
      "last_run_timestamp": "2026-08-29T18:16:39.840Z"
    },
    {
      "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": 0.838173375,
      "accuracy_std_dev": 0.011212592975506409,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.09136887752525254,
      "avg_duration_per_task_ms": 1122835.0227272727,
      "avg_output_tokens_per_task": 46187.42108585859,
      "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": "InferenceNet",
      "accuracy": 0.8378226666666666,
      "accuracy_std_dev": 0.00975199857579064,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.15828378956228956,
      "avg_duration_per_task_ms": 446608.7861952862,
      "avg_output_tokens_per_task": 56298.93097643097,
      "last_run_timestamp": "2026-10-10T08:30:28.864Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Z.AI",
      "accuracy": 0.8376822500000001,
      "accuracy_std_dev": 0.06783075684024788,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.20651400042087545,
      "avg_duration_per_task_ms": 551769.9718013467,
      "avg_output_tokens_per_task": 46912.232323232325,
      "last_run_timestamp": "2026-10-31T16:18:48.433Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.4-mini-20260317",
      "model_slug": "openai/gpt-5.4-mini",
      "model_name": "OpenAI: GPT-5.4 Mini",
      "provider_name": "auto-routing",
      "accuracy": 0.837542,
      "accuracy_std_dev": 0.005892000000000008,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.06186689898989899,
      "avg_duration_per_task_ms": 75089.61363636363,
      "avg_output_tokens_per_task": 13620.507575757576,
      "last_run_timestamp": "2026-10-01T19:12:44.401Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8375419999999999,
      "accuracy_std_dev": 0.012551288592544344,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.08632508417508418,
      "avg_duration_per_task_ms": 711440.0951178451,
      "avg_output_tokens_per_task": 49539.13468013468,
      "last_run_timestamp": "2026-10-31T16:46:10.481Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Modal",
      "accuracy": 0.8371593636363637,
      "accuracy_std_dev": 0.013027195241792026,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.141865238292011,
      "avg_duration_per_task_ms": 181084.47015610652,
      "avg_output_tokens_per_task": 32015.05968778696,
      "last_run_timestamp": "2026-10-31T08:16:11.213Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8371268528072838,
      "accuracy_std_dev": 0.033679846992081194,
      "run_count": 10,
      "total_questions": 1977,
      "avg_cost_per_task_usd": 0.1096482544258978,
      "avg_duration_per_task_ms": 1061118.531613556,
      "avg_output_tokens_per_task": 34187.799696509865,
      "last_run_timestamp": "2026-10-24T16:17:57.470Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "Groq",
      "accuracy": 0.8369882061403507,
      "accuracy_std_dev": 0.09929590339099784,
      "run_count": 8,
      "total_questions": 1368,
      "avg_cost_per_task_usd": 0.042823769005847945,
      "avg_duration_per_task_ms": 127078.64692982456,
      "avg_output_tokens_per_task": 28056.13523391813,
      "last_run_timestamp": "2026-10-31T16:08:08.538Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "DeepInfra",
      "accuracy": 0.8367623703349283,
      "accuracy_std_dev": 0.019945105490341078,
      "run_count": 12,
      "total_questions": 2090,
      "avg_cost_per_task_usd": 0.09667845789473684,
      "avg_duration_per_task_ms": 610571.3789473685,
      "avg_output_tokens_per_task": 30136.36889952153,
      "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-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Alibaba",
      "accuracy": 0.83672117730096,
      "accuracy_std_dev": 0.04633394919974385,
      "run_count": 9,
      "total_questions": 1771,
      "avg_cost_per_task_usd": 0.192393172783738,
      "avg_duration_per_task_ms": 603566.0841332581,
      "avg_output_tokens_per_task": 45791.96103896104,
      "last_run_timestamp": "2026-10-31T16:25:21.093Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "DeepInfra",
      "accuracy": 0.8367003333333334,
      "accuracy_std_dev": 0.04194124922899757,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.09553346127946127,
      "avg_duration_per_task_ms": 1453001.978114478,
      "avg_output_tokens_per_task": 42378.70538720539,
      "last_run_timestamp": "2026-08-29T18:36:10.818Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8365599166666667,
      "accuracy_std_dev": 0.02438133608746088,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.1378553611111111,
      "avg_duration_per_task_ms": 542424.4175084175,
      "avg_output_tokens_per_task": 43872.40151515151,
      "last_run_timestamp": "2026-10-31T16:18:28.926Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-35b-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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Cloudflare",
      "accuracy": 0.8356791761071061,
      "accuracy_std_dev": 0.01579187012442372,
      "run_count": 15,
      "total_questions": 2913,
      "avg_cost_per_task_usd": 0.29531686096807414,
      "avg_duration_per_task_ms": 404294.6398901476,
      "avg_output_tokens_per_task": 50028.10332990045,
      "last_run_timestamp": "2026-10-31T16:56:08.871Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "inclusionai/ling-3.0-flash-vl-20260910",
      "model_slug": "inclusionai/ling-3.0-flash-vl",
      "model_name": "inclusionAI: Ling 3.0 Flash VL",
      "provider_name": "DeepInfra",
      "accuracy": 0.8351849999999998,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 180,
      "avg_cost_per_task_usd": 0.0030483999999999997,
      "avg_duration_per_task_ms": 288802.22222222225,
      "avg_output_tokens_per_task": 16703.5,
      "last_run_timestamp": "2026-10-01T08:43:11.415Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "AkashML",
      "accuracy": 0.8348581735005084,
      "accuracy_std_dev": 0.011926219294758369,
      "run_count": 18,
      "total_questions": 2951,
      "avg_cost_per_task_usd": 0.1190305462555066,
      "avg_duration_per_task_ms": 291312.74483226024,
      "avg_output_tokens_per_task": 28021.837004405286,
      "last_run_timestamp": "2026-10-31T08:09:30.690Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "DigitalOcean",
      "accuracy": 0.8347361666666666,
      "accuracy_std_dev": 0.01234876949897801,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.04779885185185185,
      "avg_duration_per_task_ms": 978213.5824915825,
      "avg_output_tokens_per_task": 29118.793771043773,
      "last_run_timestamp": "2026-10-17T08:20:47.476Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Chutes",
      "accuracy": 0.8346745432595574,
      "accuracy_std_dev": 0.02838986518520959,
      "run_count": 8,
      "total_questions": 1491,
      "avg_cost_per_task_usd": 0.006750653923541246,
      "avg_duration_per_task_ms": 767094.1871227365,
      "avg_output_tokens_per_task": 18119.238095238095,
      "last_run_timestamp": "2026-10-31T16:04:29.121Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Mara",
      "accuracy": 0.8346120933503837,
      "accuracy_std_dev": 0.029860364128538167,
      "run_count": 6,
      "total_questions": 782,
      "avg_cost_per_task_usd": 0.09530405242966751,
      "avg_duration_per_task_ms": 330065.4283887468,
      "avg_output_tokens_per_task": 39472.66240409207,
      "last_run_timestamp": "2026-10-31T16:08:31.063Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Modal",
      "accuracy": 0.8344576188870152,
      "accuracy_std_dev": 0.007843600547934322,
      "run_count": 3,
      "total_questions": 593,
      "avg_cost_per_task_usd": 0.033416617200674535,
      "avg_duration_per_task_ms": 208747.37942664418,
      "avg_output_tokens_per_task": 14517.00168634064,
      "last_run_timestamp": "2026-10-31T08:47:23.899Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Crusoe",
      "accuracy": 0.8344556666666667,
      "accuracy_std_dev": 0.02104175490674567,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.21689809427609427,
      "avg_duration_per_task_ms": 815662.5723905724,
      "avg_output_tokens_per_task": 49349.5,
      "last_run_timestamp": "2026-08-29T18:53:23.077Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Phala",
      "accuracy": 0.8341984152249134,
      "accuracy_std_dev": 0.029002188381032377,
      "run_count": 3,
      "total_questions": 578,
      "avg_cost_per_task_usd": 0.005492470588235294,
      "avg_duration_per_task_ms": 956260.6712802768,
      "avg_output_tokens_per_task": 15404.778546712803,
      "last_run_timestamp": "2026-08-29T18:08:44.008Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "DekaLLM",
      "accuracy": 0.8341543388067869,
      "accuracy_std_dev": 0.008301927538228698,
      "run_count": 10,
      "total_questions": 1827,
      "avg_cost_per_task_usd": 0.0431522140120416,
      "avg_duration_per_task_ms": 94743.94636015326,
      "avg_output_tokens_per_task": 15061.821018062397,
      "last_run_timestamp": "2026-10-31T08:22:53.415Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Alibaba",
      "accuracy": 0.8341254705882353,
      "accuracy_std_dev": 0.010649185432078013,
      "run_count": 17,
      "total_questions": 3366,
      "avg_cost_per_task_usd": 0.25025737195484254,
      "avg_duration_per_task_ms": 384709.65567439096,
      "avg_output_tokens_per_task": 40253.002970885325,
      "last_run_timestamp": "2026-10-31T08:41:02.403Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "thinkingmachines/inkling-20260715",
      "model_slug": "thinkingmachines/inkling",
      "model_name": "Thinking Machines: Inkling",
      "provider_name": "auto-routing",
      "accuracy": 0.8340364430379746,
      "accuracy_std_dev": 0.01069546220626094,
      "run_count": 6,
      "total_questions": 1185,
      "avg_cost_per_task_usd": 0.0954248565400844,
      "avg_duration_per_task_ms": 293770.05400843883,
      "avg_output_tokens_per_task": 23403.932489451476,
      "last_run_timestamp": "2026-10-01T16:11:05.832Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.8339644999999999,
      "accuracy_std_dev": 0.02083967769064579,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.09829379103535353,
      "avg_duration_per_task_ms": 1158045.8926767677,
      "avg_output_tokens_per_task": 45010.31502525252,
      "last_run_timestamp": "2026-10-31T16:08:56.903Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8338943333333333,
      "accuracy_std_dev": 0.011690806112307073,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.03773407239057239,
      "avg_duration_per_task_ms": 276396.1228956229,
      "avg_output_tokens_per_task": 28290.345117845118,
      "last_run_timestamp": "2026-09-12T09:56:44.384Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "CoreWeave",
      "accuracy": 0.833754,
      "accuracy_std_dev": 0.016681719388300487,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.14335061994949497,
      "avg_duration_per_task_ms": 311316.28472222225,
      "avg_output_tokens_per_task": 59097.76136363636,
      "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": "Parasail",
      "accuracy": 0.8335494788062284,
      "accuracy_std_dev": 0.010937881698736537,
      "run_count": 12,
      "total_questions": 2312,
      "avg_cost_per_task_usd": 0.23824533304498272,
      "avg_duration_per_task_ms": 303110.6955017301,
      "avg_output_tokens_per_task": 53975.97967128028,
      "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-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-5.3-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": "StreamLake",
      "accuracy": 0.8328153846153846,
      "accuracy_std_dev": 0.007481944466909486,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.08116497474747475,
      "avg_duration_per_task_ms": 696165.8154623155,
      "avg_output_tokens_per_task": 59217.08663558664,
      "last_run_timestamp": "2026-10-31T16:05:36.105Z"
    },
    {
      "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-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Crusoe",
      "accuracy": 0.8326600000000001,
      "accuracy_std_dev": 0.017376865045226092,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.21192213939393936,
      "avg_duration_per_task_ms": 312080.35656565655,
      "avg_output_tokens_per_task": 48262.278787878786,
      "last_run_timestamp": "2026-09-12T08:53:53.147Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "AkashML",
      "accuracy": 0.832407,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 180,
      "avg_cost_per_task_usd": 0.05314901111111111,
      "avg_duration_per_task_ms": 479967.9111111111,
      "avg_output_tokens_per_task": 22823.577777777777,
      "last_run_timestamp": "2026-08-15T08:47:51.369Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.4-mini-20260317",
      "model_slug": "openai/gpt-5.4-mini",
      "model_name": "OpenAI: GPT-5.4 Mini",
      "provider_name": "Azure",
      "accuracy": 0.8319864,
      "accuracy_std_dev": 0.01402072559605957,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.06441977171717173,
      "avg_duration_per_task_ms": 117114.99292929293,
      "avg_output_tokens_per_task": 13401.907070707071,
      "last_run_timestamp": "2026-10-01T19:14:05.661Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-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": "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": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Friendli",
      "accuracy": 0.8316432832699621,
      "accuracy_std_dev": 0.013089827300825316,
      "run_count": 9,
      "total_questions": 1578,
      "avg_cost_per_task_usd": 0.2110187237008872,
      "avg_duration_per_task_ms": 469662.1115335868,
      "avg_output_tokens_per_task": 47863.309252217994,
      "last_run_timestamp": "2026-10-31T16:35:08.642Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "AkashML",
      "accuracy": 0.8315871731272918,
      "accuracy_std_dev": 0.020923668294302502,
      "run_count": 20,
      "total_questions": 3818,
      "avg_cost_per_task_usd": 0.04136215741225774,
      "avg_duration_per_task_ms": 440931.9447354636,
      "avg_output_tokens_per_task": 45362.3198009429,
      "last_run_timestamp": "2026-10-31T16:27:38.037Z"
    },
    {
      "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/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "CoreWeave",
      "accuracy": 0.8310010869565219,
      "accuracy_std_dev": 0.01538004204178586,
      "run_count": 13,
      "total_questions": 2001,
      "avg_cost_per_task_usd": 0.04013055072463768,
      "avg_duration_per_task_ms": 299275.6486756622,
      "avg_output_tokens_per_task": 32699.943028485755,
      "last_run_timestamp": "2026-10-31T16:19:08.399Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "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/qwen3.5-122b-a10b-20260224",
      "model_slug": "qwen/qwen3.5-122b-a10b",
      "model_name": "Qwen: Qwen3.5-122B-A10B",
      "provider_name": "Alibaba",
      "accuracy": 0.8303971644295303,
      "accuracy_std_dev": 0.07664602971803865,
      "run_count": 13,
      "total_questions": 2384,
      "avg_cost_per_task_usd": 0.040612590604026846,
      "avg_duration_per_task_ms": 150268.53691275168,
      "avg_output_tokens_per_task": 19434.7072147651,
      "last_run_timestamp": "2026-10-31T16:17:39.770Z"
    },
    {
      "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": 0.830387,
      "accuracy_std_dev": 0.012044722890129136,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.007987552398989899,
      "avg_duration_per_task_ms": 519162.2380050505,
      "avg_output_tokens_per_task": 20553.491792929293,
      "last_run_timestamp": "2026-10-31T16:01:53.649Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Cloudflare",
      "accuracy": 0.8302074978687126,
      "accuracy_std_dev": 0.018798024809619108,
      "run_count": 12,
      "total_questions": 2346,
      "avg_cost_per_task_usd": 0.14303728772378516,
      "avg_duration_per_task_ms": 291115.02642796247,
      "avg_output_tokens_per_task": 32292.15643648764,
      "last_run_timestamp": "2026-10-31T08:14:43.196Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "Phala",
      "accuracy": 0.8301065833333333,
      "accuracy_std_dev": 0.009803426675218673,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.06415462542087542,
      "avg_duration_per_task_ms": 540961.1531986533,
      "avg_output_tokens_per_task": 50401.976851851854,
      "last_run_timestamp": "2026-10-31T16:08:56.901Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "Venice",
      "accuracy": 0.830042909090909,
      "accuracy_std_dev": 0.0271229167426385,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.03839932506887053,
      "avg_duration_per_task_ms": 364543.0197428834,
      "avg_output_tokens_per_task": 38863.25987144169,
      "last_run_timestamp": "2026-10-24T16:01:48.224Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Venice",
      "accuracy": 0.8298383174258858,
      "accuracy_std_dev": 0.09656172773698192,
      "run_count": 7,
      "total_questions": 1383,
      "avg_cost_per_task_usd": 0.08489557556037598,
      "avg_duration_per_task_ms": 562990.3246565438,
      "avg_output_tokens_per_task": 23490.788864786697,
      "last_run_timestamp": "2026-10-31T16:16:30.410Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "AtlasCloud",
      "accuracy": 0.829787,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 141,
      "avg_cost_per_task_usd": 0.06982787943262411,
      "avg_duration_per_task_ms": 171508.63120567377,
      "avg_output_tokens_per_task": 17479.241134751774,
      "last_run_timestamp": "2026-08-15T08:47:41.523Z"
    },
    {
      "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": 0.8295282648401827,
      "accuracy_std_dev": 0.02709499999999998,
      "run_count": 2,
      "total_questions": 219,
      "avg_cost_per_task_usd": 0.11806798630136987,
      "avg_duration_per_task_ms": 301952.6621004566,
      "avg_output_tokens_per_task": 26947.013698630137,
      "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": "DeepInfra",
      "accuracy": 0.8290707212276216,
      "accuracy_std_dev": 0.048577604072121076,
      "run_count": 7,
      "total_questions": 1173,
      "avg_cost_per_task_usd": 0.04622304262574595,
      "avg_duration_per_task_ms": 675082.569479966,
      "avg_output_tokens_per_task": 30812.56862745098,
      "last_run_timestamp": "2026-09-26T09:56:00.145Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Friendli",
      "accuracy": 0.8290008259157149,
      "accuracy_std_dev": 0.02019136648766136,
      "run_count": 13,
      "total_questions": 2539,
      "avg_cost_per_task_usd": 0.25050771012209533,
      "avg_duration_per_task_ms": 410910.65537613234,
      "avg_output_tokens_per_task": 56752.72193777077,
      "last_run_timestamp": "2026-10-31T16:06:46.433Z"
    },
    {
      "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": "CoreWeave",
      "accuracy": 0.8288806316793894,
      "accuracy_std_dev": 0.011093878321949571,
      "run_count": 3,
      "total_questions": 524,
      "avg_cost_per_task_usd": 0.02635685687022901,
      "avg_duration_per_task_ms": 109838.5858778626,
      "avg_output_tokens_per_task": 20953.05534351145,
      "last_run_timestamp": "2026-08-29T18:29:13.946Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "auto-routing",
      "accuracy": 0.828487042983565,
      "accuracy_std_dev": 0.04343161920830901,
      "run_count": 12,
      "total_questions": 2373,
      "avg_cost_per_task_usd": 0.08886351622418878,
      "avg_duration_per_task_ms": 761181.4357353561,
      "avg_output_tokens_per_task": 31575.004214075012,
      "last_run_timestamp": "2026-10-31T16:01:53.624Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "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/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8280725000000001,
      "accuracy_std_dev": 0.008005017801354371,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.047915450757575764,
      "avg_duration_per_task_ms": 452440.26767676766,
      "avg_output_tokens_per_task": 26534.96148989899,
      "last_run_timestamp": "2026-10-31T16:10:31.145Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "CoreWeave",
      "accuracy": 0.8280023333333334,
      "accuracy_std_dev": 0.01954028429225691,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.11466727609427607,
      "avg_duration_per_task_ms": 496633.7373737374,
      "avg_output_tokens_per_task": 31920.87037037037,
      "last_run_timestamp": "2026-08-29T10:53:09.444Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "Google",
      "accuracy": 0.8280021666666666,
      "accuracy_std_dev": 0.014721942482981734,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.09966979545454545,
      "avg_duration_per_task_ms": 354652.75336700334,
      "avg_output_tokens_per_task": 45738.045454545456,
      "last_run_timestamp": "2026-10-17T16:06:16.399Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Alibaba",
      "accuracy": 0.8273763076923075,
      "accuracy_std_dev": 0.07054214326135352,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.09363684304584305,
      "avg_duration_per_task_ms": 516457.86635586637,
      "avg_output_tokens_per_task": 34542.806915306915,
      "last_run_timestamp": "2026-10-31T16:05:36.297Z"
    },
    {
      "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": "qwen/qwen3.5-122b-a10b-20260224",
      "model_slug": "qwen/qwen3.5-122b-a10b",
      "model_name": "Qwen: Qwen3.5-122B-A10B",
      "provider_name": "SiliconFlow",
      "accuracy": 0.82695,
      "accuracy_std_dev": 0.05592365383568327,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.05177901725589225,
      "avg_duration_per_task_ms": 522861.78745791246,
      "avg_output_tokens_per_task": 24790.76473063973,
      "last_run_timestamp": "2026-10-31T16:01:53.625Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "BaseTen",
      "accuracy": 0.8269358,
      "accuracy_std_dev": 0.006767606146932625,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.04578402121212121,
      "avg_duration_per_task_ms": 60291.652525252524,
      "avg_output_tokens_per_task": 11285.754545454545,
      "last_run_timestamp": "2026-09-26T09:19:03.901Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Friendli",
      "accuracy": 0.8266521780172412,
      "accuracy_std_dev": 0.022372414615840108,
      "run_count": 12,
      "total_questions": 2320,
      "avg_cost_per_task_usd": 0.02234062025862069,
      "avg_duration_per_task_ms": 235753.16637931034,
      "avg_output_tokens_per_task": 44456.286637931036,
      "last_run_timestamp": "2026-10-31T08:28:04.955Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "Darkbloom",
      "accuracy": 0.826417643153527,
      "accuracy_std_dev": 0.01028202813164795,
      "run_count": 4,
      "total_questions": 723,
      "avg_cost_per_task_usd": 0.010974759336099586,
      "avg_duration_per_task_ms": 1541635.907330567,
      "avg_output_tokens_per_task": 31233.395573997233,
      "last_run_timestamp": "2026-10-31T08:24:48.465Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "CoreWeave",
      "accuracy": 0.8264159143968872,
      "accuracy_std_dev": 0.0040531378816294926,
      "run_count": 4,
      "total_questions": 771,
      "avg_cost_per_task_usd": 0.13822659403372245,
      "avg_duration_per_task_ms": 272716.0363164721,
      "avg_output_tokens_per_task": 34382.090791180286,
      "last_run_timestamp": "2026-10-31T08:38:15.143Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "CoreWeave",
      "accuracy": 0.8264111325878595,
      "accuracy_std_dev": 0.01425677691801343,
      "run_count": 10,
      "total_questions": 1878,
      "avg_cost_per_task_usd": 0.02454355484558041,
      "avg_duration_per_task_ms": 264967.17305644305,
      "avg_output_tokens_per_task": 50263.019169329076,
      "last_run_timestamp": "2026-10-31T08:17:29.105Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Io Net",
      "accuracy": 0.8263836795903439,
      "accuracy_std_dev": 0.029166866420929173,
      "run_count": 7,
      "total_questions": 1367,
      "avg_cost_per_task_usd": 0.07172207900512069,
      "avg_duration_per_task_ms": 248028.12801755668,
      "avg_output_tokens_per_task": 21803.532553035846,
      "last_run_timestamp": "2026-10-31T09:17:12.929Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Ionstream",
      "accuracy": 0.8259871818181819,
      "accuracy_std_dev": 0.010774204132752242,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.03615010422405877,
      "avg_duration_per_task_ms": 305458.8122130395,
      "avg_output_tokens_per_task": 15557.78971533517,
      "last_run_timestamp": "2026-10-31T08:32:33.985Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "Novita",
      "accuracy": 0.8259259000000001,
      "accuracy_std_dev": 0.018259649079048597,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.012357968686868687,
      "avg_duration_per_task_ms": 759517.2151515152,
      "avg_output_tokens_per_task": 36563.7601010101,
      "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": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "Xiaomi",
      "accuracy": 0.8255892,
      "accuracy_std_dev": 0.023390445078279278,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.01099114494949495,
      "avg_duration_per_task_ms": 757385.2308080808,
      "avg_output_tokens_per_task": 39012.57828282828,
      "last_run_timestamp": "2026-10-17T08:02:36.531Z"
    },
    {
      "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": 0.8254769999999999,
      "accuracy_std_dev": 0.021834969445059152,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.025108806397306397,
      "avg_duration_per_task_ms": 234041.265993266,
      "avg_output_tokens_per_task": 21052.969696969696,
      "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": "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": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "Venice",
      "accuracy": 0.8253217093954844,
      "accuracy_std_dev": 0.12574835204649074,
      "run_count": 7,
      "total_questions": 1373,
      "avg_cost_per_task_usd": 0.12495739184268023,
      "avg_duration_per_task_ms": 776035.8018936635,
      "avg_output_tokens_per_task": 37930.34668608886,
      "last_run_timestamp": "2026-10-31T16:02:14.362Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "SiliconFlow",
      "accuracy": 0.82491575,
      "accuracy_std_dev": 0.013148603345507848,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.19637407828282827,
      "avg_duration_per_task_ms": 945885.1603535353,
      "avg_output_tokens_per_task": 52273.792929292926,
      "last_run_timestamp": "2026-10-31T16:50:02.353Z"
    },
    {
      "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": 0.8249156666666667,
      "accuracy_std_dev": 0.016206137938722134,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.024119345117845116,
      "avg_duration_per_task_ms": 456487.2491582492,
      "avg_output_tokens_per_task": 22656.301346801345,
      "last_run_timestamp": "2026-08-29T18:09:03.379Z"
    },
    {
      "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": 0.8249156666666667,
      "accuracy_std_dev": 0.0254575838907693,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.07950963804713804,
      "avg_duration_per_task_ms": 692043.8686868687,
      "avg_output_tokens_per_task": 43016.01683501683,
      "last_run_timestamp": "2026-10-04T00:07:08.217Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "Novita",
      "accuracy": 0.8249156249999999,
      "accuracy_std_dev": 0.030431425779683337,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.09934535101010099,
      "avg_duration_per_task_ms": 1389882.110479798,
      "avg_output_tokens_per_task": 45786.19065656565,
      "last_run_timestamp": "2026-10-31T16:08:33.697Z"
    },
    {
      "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": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "DigitalOcean",
      "accuracy": 0.8241641919770775,
      "accuracy_std_dev": 0.013440726394979233,
      "run_count": 9,
      "total_questions": 1745,
      "avg_cost_per_task_usd": 0.01836141260744986,
      "avg_duration_per_task_ms": 609134.4246418338,
      "avg_output_tokens_per_task": 20155.189684813755,
      "last_run_timestamp": "2026-10-31T16:13:07.320Z"
    },
    {
      "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": 0.8240738888888891,
      "accuracy_std_dev": 0.06341916550564268,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.02869584455667789,
      "avg_duration_per_task_ms": 721909.4079685747,
      "avg_output_tokens_per_task": 30019.711560044892,
      "last_run_timestamp": "2026-10-17T08:18:04.911Z"
    },
    {
      "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": 0.8240561561100449,
      "accuracy_std_dev": 0.00960574900775572,
      "run_count": 8,
      "total_questions": 1563,
      "avg_cost_per_task_usd": 0.008232310300703776,
      "avg_duration_per_task_ms": 344296.5265515035,
      "avg_output_tokens_per_task": 19189.83877159309,
      "last_run_timestamp": "2026-10-31T16:08:31.066Z"
    },
    {
      "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": 0.8237933333333333,
      "accuracy_std_dev": 0.003967811851498118,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.01872473063973064,
      "avg_duration_per_task_ms": 507797.69696969696,
      "avg_output_tokens_per_task": 18599.063973063974,
      "last_run_timestamp": "2026-08-29T18:35:57.732Z"
    },
    {
      "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": 0.8236530000000001,
      "accuracy_std_dev": 0.010804411622110653,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.02350044696969697,
      "avg_duration_per_task_ms": 359036.5871212121,
      "avg_output_tokens_per_task": 20960.843434343435,
      "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-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Z.AI",
      "accuracy": 0.82323225,
      "accuracy_std_dev": 0.029134592633800457,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.2227477537878788,
      "avg_duration_per_task_ms": 1217960.3686868686,
      "avg_output_tokens_per_task": 50933.75315656565,
      "last_run_timestamp": "2026-10-31T16:30:07.021Z"
    },
    {
      "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": 0.8228234324639938,
      "accuracy_std_dev": 0.0330962900700723,
      "run_count": 13,
      "total_questions": 2569,
      "avg_cost_per_task_usd": 0.08174266757493187,
      "avg_duration_per_task_ms": 338894.74659400544,
      "avg_output_tokens_per_task": 20300.639159205915,
      "last_run_timestamp": "2026-10-31T16:02:14.359Z"
    },
    {
      "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": 0.8226709999999999,
      "accuracy_std_dev": 0.023502234801538922,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.02574937037037037,
      "avg_duration_per_task_ms": 439979.5740740741,
      "avg_output_tokens_per_task": 23407.67845117845,
      "last_run_timestamp": "2026-08-29T18:16:26.161Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "DeepInfra",
      "accuracy": 0.8226336142278774,
      "accuracy_std_dev": 0.01129717437085189,
      "run_count": 9,
      "total_questions": 1729,
      "avg_cost_per_task_usd": 0.02482087044534413,
      "avg_duration_per_task_ms": 433623.25274725276,
      "avg_output_tokens_per_task": 24707.683632157317,
      "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-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Morph",
      "accuracy": 0.8225450489150423,
      "accuracy_std_dev": 0.023532795309234442,
      "run_count": 14,
      "total_questions": 2719,
      "avg_cost_per_task_usd": 0.026525477749172497,
      "avg_duration_per_task_ms": 850874.7859507172,
      "avg_output_tokens_per_task": 49891.477013607946,
      "last_run_timestamp": "2026-10-31T08:42:18.327Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-2.5-pro",
      "model_slug": "google/gemini-2.5-pro",
      "model_name": "Google: Gemini 2.5 Pro",
      "provider_name": "Google AI Studio",
      "accuracy": 0.8223905000000001,
      "accuracy_std_dev": 0.0244105,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.2415952904040404,
      "avg_duration_per_task_ms": 181953.0101010101,
      "avg_output_tokens_per_task": 24066.49494949495,
      "last_run_timestamp": "2026-10-01T19:12:31.045Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Chutes",
      "accuracy": 0.8223905,
      "accuracy_std_dev": 0.019262412208322532,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.03192426473063974,
      "avg_duration_per_task_ms": 260830.34259259258,
      "avg_output_tokens_per_task": 14583.287878787878,
      "last_run_timestamp": "2026-10-31T08:17:29.187Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-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-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Makora",
      "accuracy": 0.8213506405228759,
      "accuracy_std_dev": 0.02248216924587128,
      "run_count": 3,
      "total_questions": 459,
      "avg_cost_per_task_usd": 0.08758356862745097,
      "avg_duration_per_task_ms": 304618.9760348584,
      "avg_output_tokens_per_task": 34720.03050108933,
      "last_run_timestamp": "2026-08-29T10:57:24.797Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta/muse-glimmer-30b-20260810",
      "model_slug": "meta/muse-glimmer-30b",
      "model_name": "Meta: Muse Glimmer 30B",
      "provider_name": "Parasail",
      "accuracy": 0.8212070710659898,
      "accuracy_std_dev": 0.011946281848340899,
      "run_count": 3,
      "total_questions": 591,
      "avg_cost_per_task_usd": 0.01807649915397631,
      "avg_duration_per_task_ms": 302940.0592216582,
      "avg_output_tokens_per_task": 16413.382402707277,
      "last_run_timestamp": "2026-08-29T09:59:35.603Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Io Net",
      "accuracy": 0.8210525603238865,
      "accuracy_std_dev": 0.07599126782917703,
      "run_count": 7,
      "total_questions": 1235,
      "avg_cost_per_task_usd": 0.007044106882591094,
      "avg_duration_per_task_ms": 525854.9595141701,
      "avg_output_tokens_per_task": 14201.662348178137,
      "last_run_timestamp": "2026-10-03T10:22:02.091Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.1-flash-lite-20260507",
      "model_slug": "google/gemini-3.1-flash-lite",
      "model_name": "Google: Gemini 3.1 Flash Lite",
      "provider_name": "auto-routing",
      "accuracy": 0.8207070000000001,
      "accuracy_std_dev": 0.024411000000000016,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.04063828030303031,
      "avg_duration_per_task_ms": 83296.42424242424,
      "avg_output_tokens_per_task": 26969.131313131315,
      "last_run_timestamp": "2026-10-01T19:23:01.327Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "CoreWeave",
      "accuracy": 0.8204965000000001,
      "accuracy_std_dev": 0.019212136906133062,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.0063976704545454545,
      "avg_duration_per_task_ms": 423192.5814393939,
      "avg_output_tokens_per_task": 18733.522727272728,
      "last_run_timestamp": "2026-10-31T16:12:28.233Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "CoreWeave",
      "accuracy": 0.8203833076923077,
      "accuracy_std_dev": 0.018076297388844354,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.042996285159285164,
      "avg_duration_per_task_ms": 241748.8038073038,
      "avg_output_tokens_per_task": 14318.661227661227,
      "last_run_timestamp": "2026-10-31T08:09:30.241Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "auto-routing",
      "accuracy": 0.820286125,
      "accuracy_std_dev": 0.014129067020131756,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.007522392676767676,
      "avg_duration_per_task_ms": 484868.283459596,
      "avg_output_tokens_per_task": 18453.40404040404,
      "last_run_timestamp": "2026-10-31T16:13:09.863Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "thinkingmachines/inkling-20260715",
      "model_slug": "thinkingmachines/inkling",
      "model_name": "Thinking Machines: Inkling",
      "provider_name": "BaseTen",
      "accuracy": 0.8202743410372041,
      "accuracy_std_dev": 0.011484132116496031,
      "run_count": 9,
      "total_questions": 1774,
      "avg_cost_per_task_usd": 0.09063781341600902,
      "avg_duration_per_task_ms": 168767.6234498309,
      "avg_output_tokens_per_task": 22236.227170236754,
      "last_run_timestamp": "2026-09-19T10:14:07.966Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Mancer 2",
      "accuracy": 0.8200275393258427,
      "accuracy_std_dev": 0.011794769394173345,
      "run_count": 9,
      "total_questions": 1691,
      "avg_cost_per_task_usd": 0.03818077646363098,
      "avg_duration_per_task_ms": 215799.96629213484,
      "avg_output_tokens_per_task": 14692.68894145476,
      "last_run_timestamp": "2026-10-31T08:55:11.136Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "StreamLake",
      "accuracy": 0.819865,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.13854771717171718,
      "avg_duration_per_task_ms": 540824.1212121212,
      "avg_output_tokens_per_task": 31240.651515151516,
      "last_run_timestamp": "2026-09-07T06:00:57.184Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "auto-routing",
      "accuracy": 0.8194060909090909,
      "accuracy_std_dev": 0.013884218840464785,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.033594433884297524,
      "avg_duration_per_task_ms": 211994.46005509642,
      "avg_output_tokens_per_task": 13897.44214876033,
      "last_run_timestamp": "2026-10-31T08:09:30.951Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "SambaNova",
      "accuracy": 0.8192149207920791,
      "accuracy_std_dev": 0.018823182524040232,
      "run_count": 6,
      "total_questions": 909,
      "avg_cost_per_task_usd": 0.11224348184818483,
      "avg_duration_per_task_ms": 401079.6094609461,
      "avg_output_tokens_per_task": 46532.80198019802,
      "last_run_timestamp": "2026-10-17T16:06:46.637Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "Parasail",
      "accuracy": 0.8188040208464941,
      "accuracy_std_dev": 0.005162697447798004,
      "run_count": 8,
      "total_questions": 1583,
      "avg_cost_per_task_usd": 0.02620403032217309,
      "avg_duration_per_task_ms": 318291.60138976626,
      "avg_output_tokens_per_task": 26132.884396715097,
      "last_run_timestamp": "2026-10-31T16:04:48.178Z"
    },
    {
      "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": 0.8186413984710942,
      "accuracy_std_dev": 0.062429697723794066,
      "run_count": 22,
      "total_questions": 4186,
      "avg_cost_per_task_usd": 0.0077469773053033915,
      "avg_duration_per_task_ms": 504549.9739608218,
      "avg_output_tokens_per_task": 17360.977544194935,
      "last_run_timestamp": "2026-10-31T16:45:45.611Z"
    },
    {
      "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": 0.8180615714285715,
      "accuracy_std_dev": 0.014803209766390367,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.015330031024531026,
      "avg_duration_per_task_ms": 451421.6443001443,
      "avg_output_tokens_per_task": 12902.48051948052,
      "last_run_timestamp": "2026-11-04T17:56:05.778Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Near AI",
      "accuracy": 0.8178727706422019,
      "accuracy_std_dev": 0.03174962566960436,
      "run_count": 5,
      "total_questions": 981,
      "avg_cost_per_task_usd": 0.023857419979612642,
      "avg_duration_per_task_ms": 166236.06523955148,
      "avg_output_tokens_per_task": 14523.781855249745,
      "last_run_timestamp": "2026-10-31T08:07:51.399Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta/muse-glimmer-30b-20260810",
      "model_slug": "meta/muse-glimmer-30b",
      "model_name": "Meta: Muse Glimmer 30B",
      "provider_name": "Together",
      "accuracy": 0.8178077777777777,
      "accuracy_std_dev": 0.009999885885768644,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.024910318742985414,
      "avg_duration_per_task_ms": 190389.1571268238,
      "avg_output_tokens_per_task": 16514.1734006734,
      "last_run_timestamp": "2026-10-03T10:10:35.343Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.81776075,
      "accuracy_std_dev": 0.030405091048827666,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.08916797159090908,
      "avg_duration_per_task_ms": 1052180.1603535353,
      "avg_output_tokens_per_task": 43838.132575757576,
      "last_run_timestamp": "2026-10-31T16:08:08.535Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "StreamLake",
      "accuracy": 0.8177376144306652,
      "accuracy_std_dev": 0.03151916867911054,
      "run_count": 9,
      "total_questions": 1774,
      "avg_cost_per_task_usd": 0.013423886696730551,
      "avg_duration_per_task_ms": 855848.4718151072,
      "avg_output_tokens_per_task": 39679.556933483655,
      "last_run_timestamp": "2026-10-17T08:19:03.202Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Phala",
      "accuracy": 0.8173077377622378,
      "accuracy_std_dev": 0.04183386114142466,
      "run_count": 3,
      "total_questions": 572,
      "avg_cost_per_task_usd": 0.00918055944055944,
      "avg_duration_per_task_ms": 556054.2482517483,
      "avg_output_tokens_per_task": 20397.216783216783,
      "last_run_timestamp": "2026-08-29T19:08:46.030Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Reka",
      "accuracy": 0.8172424786585366,
      "accuracy_std_dev": 0.032381379178472315,
      "run_count": 8,
      "total_questions": 984,
      "avg_cost_per_task_usd": 0.0567359319105691,
      "avg_duration_per_task_ms": 328244.918699187,
      "avg_output_tokens_per_task": 21710.94918699187,
      "last_run_timestamp": "2026-10-31T16:26:17.224Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "OpenInference",
      "accuracy": 0.8171817336142323,
      "accuracy_std_dev": 0.04982830145510509,
      "run_count": 11,
      "total_questions": 2136,
      "avg_cost_per_task_usd": 0.018343408239700373,
      "avg_duration_per_task_ms": 793559.7439138577,
      "avg_output_tokens_per_task": 48776.84925093633,
      "last_run_timestamp": "2026-10-31T08:31:53.887Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Darkbloom",
      "accuracy": 0.8171795172255397,
      "accuracy_std_dev": 0.008683223783407119,
      "run_count": 11,
      "total_questions": 2177,
      "avg_cost_per_task_usd": 0.03297541662838769,
      "avg_duration_per_task_ms": 545292.2503445108,
      "avg_output_tokens_per_task": 15758.780431786863,
      "last_run_timestamp": "2026-10-31T08:29:48.679Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "Novita",
      "accuracy": 0.8170593333333332,
      "accuracy_std_dev": 0.013208478749988154,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.018857220538720538,
      "avg_duration_per_task_ms": 518487.1228956229,
      "avg_output_tokens_per_task": 18659.164983164985,
      "last_run_timestamp": "2026-08-30T00:16:38.394Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "StreamLake",
      "accuracy": 0.8170593333333332,
      "accuracy_std_dev": 0.004418546090690392,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.006106776094276093,
      "avg_duration_per_task_ms": 508955.7373737374,
      "avg_output_tokens_per_task": 18545.604377104377,
      "last_run_timestamp": "2026-08-29T19:09:30.030Z"
    },
    {
      "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": 0.8170593333333332,
      "accuracy_std_dev": 0.006780464700561124,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.08866408754208754,
      "avg_duration_per_task_ms": 189591.79966329967,
      "avg_output_tokens_per_task": 25298.508417508416,
      "last_run_timestamp": "2026-08-29T18:58:14.477Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta/muse-glimmer-30b-20260810",
      "model_slug": "meta/muse-glimmer-30b",
      "model_name": "Meta: Muse Glimmer 30B",
      "provider_name": "Phala",
      "accuracy": 0.816967559819413,
      "accuracy_std_dev": 0.022532849378432182,
      "run_count": 9,
      "total_questions": 1772,
      "avg_cost_per_task_usd": 0.021389000564334085,
      "avg_duration_per_task_ms": 297135.54796839727,
      "avg_output_tokens_per_task": 17946.336343115123,
      "last_run_timestamp": "2026-10-03T09:25:02.390Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.5-sonnet-20250929",
      "model_slug": "anthropic/claude-sonnet-4.5",
      "model_name": "Anthropic: Claude Sonnet 4.5",
      "provider_name": "auto-routing",
      "accuracy": 0.8169191666666666,
      "accuracy_std_dev": 0.011358349086562799,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.3136177163299663,
      "avg_duration_per_task_ms": 289623.9978956229,
      "avg_output_tokens_per_task": 20717.387205387204,
      "last_run_timestamp": "2026-10-31T08:15:24.404Z"
    },
    {
      "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": 0.8169191250000001,
      "accuracy_std_dev": 0.047316166086332226,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.20553683143939394,
      "avg_duration_per_task_ms": 898204.4223484849,
      "avg_output_tokens_per_task": 51369.868055555555,
      "last_run_timestamp": "2026-10-31T16:08:08.537Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "Google",
      "accuracy": 0.8168020459873759,
      "accuracy_std_dev": 0.03036523422023943,
      "run_count": 6,
      "total_questions": 1109,
      "avg_cost_per_task_usd": 0.028931809738503157,
      "avg_duration_per_task_ms": 508171.53381424706,
      "avg_output_tokens_per_task": 17143.577998196575,
      "last_run_timestamp": "2026-10-17T16:03:44.233Z"
    },
    {
      "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": 0.81649825,
      "accuracy_std_dev": 0.007896181684048294,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.026513486111111113,
      "avg_duration_per_task_ms": 167722.74494949495,
      "avg_output_tokens_per_task": 23391.1047979798,
      "last_run_timestamp": "2026-10-31T16:09:44.821Z"
    },
    {
      "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": 0.815937,
      "accuracy_std_dev": 0.00915257617650169,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.06866670033670035,
      "avg_duration_per_task_ms": 839163.7121212122,
      "avg_output_tokens_per_task": 39474.887205387204,
      "last_run_timestamp": "2026-08-30T00:37:59.094Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "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": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "SiliconFlow",
      "accuracy": 0.8150252499999998,
      "accuracy_std_dev": 0.004075243451316723,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.0115347702020202,
      "avg_duration_per_task_ms": 842527.0176767677,
      "avg_output_tokens_per_task": 18589.04419191919,
      "last_run_timestamp": "2026-10-31T16:13:09.867Z"
    },
    {
      "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": 0.8148633940983607,
      "accuracy_std_dev": 0.008422212201041645,
      "run_count": 8,
      "total_questions": 1525,
      "avg_cost_per_task_usd": 0.042689679344262293,
      "avg_duration_per_task_ms": 185447.63540983608,
      "avg_output_tokens_per_task": 16898.064918032785,
      "last_run_timestamp": "2026-11-04T18:19:31.563Z"
    },
    {
      "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": 0.8148148,
      "accuracy_std_dev": 0.009606302751839527,
      "run_count": 25,
      "total_questions": 4950,
      "avg_cost_per_task_usd": 0.33180516727272724,
      "avg_duration_per_task_ms": 293289.88505050505,
      "avg_output_tokens_per_task": 20920.7098989899,
      "last_run_timestamp": "2026-10-31T08:27:57.306Z"
    },
    {
      "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": "Google",
      "accuracy": 0.8146744166666667,
      "accuracy_std_dev": 0.010590848930864255,
      "run_count": 24,
      "total_questions": 4752,
      "avg_cost_per_task_usd": 0.36361028135521894,
      "avg_duration_per_task_ms": 357888.61384680134,
      "avg_output_tokens_per_task": 22899.654882154882,
      "last_run_timestamp": "2026-10-31T08:33:07.626Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta/muse-glimmer-30b-20260810",
      "model_slug": "meta/muse-glimmer-30b",
      "model_name": "Meta: Muse Glimmer 30B",
      "provider_name": "Fireworks",
      "accuracy": 0.8143338571428571,
      "accuracy_std_dev": 0.014248444761532711,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.026212479076479073,
      "avg_duration_per_task_ms": 329590.34992784995,
      "avg_output_tokens_per_task": 17511.52380952381,
      "last_run_timestamp": "2026-09-19T09:05:52.633Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.1-flash-lite-20260507",
      "model_slug": "google/gemini-3.1-flash-lite",
      "model_name": "Google: Gemini 3.1 Flash Lite",
      "provider_name": "Google AI Studio",
      "accuracy": 0.813973,
      "accuracy_std_dev": 0.01262599999999997,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.04186610353535354,
      "avg_duration_per_task_ms": 76135.87878787878,
      "avg_output_tokens_per_task": 27786.689393939392,
      "last_run_timestamp": "2026-10-01T19:16:58.595Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Chutes",
      "accuracy": 0.8136124285714287,
      "accuracy_std_dev": 0.04335004526231662,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.14883965584415582,
      "avg_duration_per_task_ms": 1660514.1673881675,
      "avg_output_tokens_per_task": 48710.61111111111,
      "last_run_timestamp": "2026-10-31T16:08:56.833Z"
    },
    {
      "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": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Morph",
      "accuracy": 0.813131,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0701835909090909,
      "avg_duration_per_task_ms": 493979.91414141416,
      "avg_output_tokens_per_task": 29227.76767676768,
      "last_run_timestamp": "2026-08-15T10:07:37.662Z"
    },
    {
      "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-5-fable-20260609",
      "model_slug": "anthropic/claude-fable-5",
      "model_name": "Anthropic: Claude Fable 5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.811448,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.1522642575757576,
      "avg_duration_per_task_ms": 39952.71717171717,
      "avg_output_tokens_per_task": 2844.2373737373737,
      "last_run_timestamp": "2026-08-15T09:15:35.213Z"
    },
    {
      "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": "anthropic/claude-4.5-sonnet-20250929",
      "model_slug": "anthropic/claude-sonnet-4.5",
      "model_name": "Anthropic: Claude Sonnet 4.5",
      "provider_name": "Anthropic",
      "accuracy": 0.8104117692307693,
      "accuracy_std_dev": 0.017452206376509655,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.3396733146853147,
      "avg_duration_per_task_ms": 352646.4836829837,
      "avg_output_tokens_per_task": 22475.542735042734,
      "last_run_timestamp": "2026-10-31T08:46:10.648Z"
    },
    {
      "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": 0.8103954999999999,
      "accuracy_std_dev": 0.025710588553162297,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.009251354166666666,
      "avg_duration_per_task_ms": 588683.1515151515,
      "avg_output_tokens_per_task": 19069.81186868687,
      "last_run_timestamp": "2026-10-31T16:08:08.535Z"
    },
    {
      "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": 0.8101851666666666,
      "accuracy_std_dev": 0.020562625427513432,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.3433783051346802,
      "avg_duration_per_task_ms": 356918.0021043771,
      "avg_output_tokens_per_task": 22728.853956228955,
      "last_run_timestamp": "2026-10-31T08:17:29.306Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "StreamLake",
      "accuracy": 0.8099512222222223,
      "accuracy_std_dev": 0.024479317441363887,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.11924308024691359,
      "avg_duration_per_task_ms": 781765.5089786757,
      "avg_output_tokens_per_task": 62033.351851851854,
      "last_run_timestamp": "2026-10-31T16:10:09.226Z"
    },
    {
      "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-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Nebius",
      "accuracy": 0.8089722206946454,
      "accuracy_std_dev": 0.07286502949135142,
      "run_count": 7,
      "total_questions": 1382,
      "avg_cost_per_task_usd": 0.1989648010130246,
      "avg_duration_per_task_ms": 1711843.8125904487,
      "avg_output_tokens_per_task": 46110.4507959479,
      "last_run_timestamp": "2026-10-31T16:22:14.129Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "AtlasCloud",
      "accuracy": 0.8089225,
      "accuracy_std_dev": 0.06617775297824792,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.09753928787878788,
      "avg_duration_per_task_ms": 791029.2310606061,
      "avg_output_tokens_per_task": 39011.65909090909,
      "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-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "auto-routing",
      "accuracy": 0.8081858749999998,
      "accuracy_std_dev": 0.01700473100667502,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.23625982323232325,
      "avg_duration_per_task_ms": 1060789.2386363635,
      "avg_output_tokens_per_task": 63306.07323232323,
      "last_run_timestamp": "2026-10-31T16:12:28.230Z"
    },
    {
      "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": 0.8080808,
      "accuracy_std_dev": 0.053692861534472135,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.2018329616161616,
      "avg_duration_per_task_ms": 1561759.3434343433,
      "avg_output_tokens_per_task": 48030.74141414141,
      "last_run_timestamp": "2026-10-31T16:31:54.109Z"
    },
    {
      "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": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Friendli",
      "accuracy": 0.8069586666666666,
      "accuracy_std_dev": 0.03732496637491973,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.026121195286195286,
      "avg_duration_per_task_ms": 474388.5303030303,
      "avg_output_tokens_per_task": 21767.089225589225,
      "last_run_timestamp": "2026-08-29T19:19:06.625Z"
    },
    {
      "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": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Morph",
      "accuracy": 0.806397,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.006770893939393939,
      "avg_duration_per_task_ms": 1110169.2828282828,
      "avg_output_tokens_per_task": 17470.782828282827,
      "last_run_timestamp": "2026-08-15T10:05:32.709Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5-mini-2025-08-07",
      "model_slug": "openai/gpt-5-mini",
      "model_name": "OpenAI: GPT-5 Mini",
      "provider_name": "OpenAI",
      "accuracy": 0.806397,
      "accuracy_std_dev": 0.003367000000000009,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.04417995707070707,
      "avg_duration_per_task_ms": 210130.46464646465,
      "avg_output_tokens_per_task": 21993.772727272728,
      "last_run_timestamp": "2026-10-01T19:26:25.385Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta/muse-glimmer-30b-20260810",
      "model_slug": "meta/muse-glimmer-30b",
      "model_name": "Meta: Muse Glimmer 30B",
      "provider_name": "auto-routing",
      "accuracy": 0.8059763749999999,
      "accuracy_std_dev": 0.01367028747811745,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.022531703914141412,
      "avg_duration_per_task_ms": 251092.985479798,
      "avg_output_tokens_per_task": 17458.415404040403,
      "last_run_timestamp": "2026-10-03T09:13:36.277Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "StreamLake",
      "accuracy": 0.805275,
      "accuracy_std_dev": 0.018459119444509446,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.10121038888888889,
      "avg_duration_per_task_ms": 584553.9983164984,
      "avg_output_tokens_per_task": 37644.86531986532,
      "last_run_timestamp": "2026-08-29T18:26:14.942Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Novita",
      "accuracy": 0.80492425,
      "accuracy_std_dev": 0.025086207414384096,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.2853950606060606,
      "avg_duration_per_task_ms": 994351.720959596,
      "avg_output_tokens_per_task": 64899.24684343435,
      "last_run_timestamp": "2026-10-31T16:12:28.234Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "SambaNova",
      "accuracy": 0.804281,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 109,
      "avg_cost_per_task_usd": 0.08876922935779817,
      "avg_duration_per_task_ms": 124213.02752293578,
      "avg_output_tokens_per_task": 36876.293577981654,
      "last_run_timestamp": "2026-08-15T14:54:40.470Z"
    },
    {
      "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": 0.8041526666666666,
      "accuracy_std_dev": 0.02864721697167495,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.028751296296296296,
      "avg_duration_per_task_ms": 459929.2138047138,
      "avg_output_tokens_per_task": 25017.868686868685,
      "last_run_timestamp": "2026-08-29T18:05:36.579Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "Friendli",
      "accuracy": 0.8040123333333333,
      "accuracy_std_dev": 0.026646777658012516,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.029773004208754207,
      "avg_duration_per_task_ms": 321073.38468013465,
      "avg_output_tokens_per_task": 19724.395622895623,
      "last_run_timestamp": "2026-10-17T16:09:25.569Z"
    },
    {
      "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": "z-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": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "Parasail",
      "accuracy": 0.8015174199706314,
      "accuracy_std_dev": 0.034082802627718296,
      "run_count": 10,
      "total_questions": 1362,
      "avg_cost_per_task_usd": 0.006941986784140968,
      "avg_duration_per_task_ms": 316770.08076358296,
      "avg_output_tokens_per_task": 27549.08516886931,
      "last_run_timestamp": "2026-10-31T16:04:29.125Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "auto-routing",
      "accuracy": 0.801346625,
      "accuracy_std_dev": 0.04197794992593582,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.026733222222222217,
      "avg_duration_per_task_ms": 301102.7582070707,
      "avg_output_tokens_per_task": 25848.362373737375,
      "last_run_timestamp": "2026-10-31T16:05:36.303Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Mistral",
      "accuracy": 0.8006073724696356,
      "accuracy_std_dev": 0.018676167604730915,
      "run_count": 5,
      "total_questions": 988,
      "avg_cost_per_task_usd": 0.026427639676113363,
      "avg_duration_per_task_ms": 228773.67307692306,
      "avg_output_tokens_per_task": 52620.16093117409,
      "last_run_timestamp": "2026-10-31T08:22:47.424Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.800084125,
      "accuracy_std_dev": 0.03890081182455419,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.18029773989898992,
      "avg_duration_per_task_ms": 560473.1414141414,
      "avg_output_tokens_per_task": 56128.871212121216,
      "last_run_timestamp": "2026-10-31T16:15:54.668Z"
    },
    {
      "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": 0.7999998975308643,
      "accuracy_std_dev": 0.02657019067436704,
      "run_count": 6,
      "total_questions": 810,
      "avg_cost_per_task_usd": 0.06976524074074074,
      "avg_duration_per_task_ms": 1694220.9679012345,
      "avg_output_tokens_per_task": 39748.25802469136,
      "last_run_timestamp": "2026-10-31T16:05:14.633Z"
    },
    {
      "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": 0.79966325,
      "accuracy_std_dev": 0.00518913881944781,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.013944108585858586,
      "avg_duration_per_task_ms": 338268.0871212121,
      "avg_output_tokens_per_task": 14269.474747474747,
      "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-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Ionstream",
      "accuracy": 0.7979796666666668,
      "accuracy_std_dev": 0.06576415746353702,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.041038570707070704,
      "avg_duration_per_task_ms": 1120543.265993266,
      "avg_output_tokens_per_task": 23360.329966329966,
      "last_run_timestamp": "2026-08-29T19:26:41.546Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Io Net",
      "accuracy": 0.7976190000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 98,
      "avg_cost_per_task_usd": 0.012265704081632654,
      "avg_duration_per_task_ms": 285854.5612244898,
      "avg_output_tokens_per_task": 10545.989795918367,
      "last_run_timestamp": "2026-10-03T17:16:45.102Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.1-terminus",
      "model_slug": "deepseek/deepseek-v3.1-terminus",
      "model_name": "DeepSeek: DeepSeek V3.1 Terminus",
      "provider_name": "StreamLake",
      "accuracy": 0.7974186666666667,
      "accuracy_std_dev": 0.009356358206529358,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.014677102693602694,
      "avg_duration_per_task_ms": 310917.8265993266,
      "avg_output_tokens_per_task": 14041.452861952863,
      "last_run_timestamp": "2026-08-29T16:10:33.070Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "auto-routing",
      "accuracy": 0.79706775,
      "accuracy_std_dev": 0.041385108447212035,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.043133531565656566,
      "avg_duration_per_task_ms": 497051.0765993266,
      "avg_output_tokens_per_task": 42719.71127946128,
      "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-r1-0528",
      "model_slug": "deepseek/deepseek-r1-0528",
      "model_name": "DeepSeek: R1 0528",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7965251890459364,
      "accuracy_std_dev": 0.006229073430740378,
      "run_count": 6,
      "total_questions": 1132,
      "avg_cost_per_task_usd": 0.05014427738515901,
      "avg_duration_per_task_ms": 1091376.4240282685,
      "avg_output_tokens_per_task": 22832.588339222613,
      "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": "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": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7940516666666667,
      "accuracy_std_dev": 0.017513670362192944,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.17459135353535354,
      "avg_duration_per_task_ms": 780357.1144781145,
      "avg_output_tokens_per_task": 55892.003367003366,
      "last_run_timestamp": "2026-08-29T19:16:33.097Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.1-terminus",
      "model_slug": "deepseek/deepseek-v3.1-terminus",
      "model_name": "DeepSeek: DeepSeek V3.1 Terminus",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7940513333333334,
      "accuracy_std_dev": 0.009654676391377534,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.013367496632996632,
      "avg_duration_per_task_ms": 287892.37373737374,
      "avg_output_tokens_per_task": 13942.378787878788,
      "last_run_timestamp": "2026-08-29T19:12:57.906Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "Venice",
      "accuracy": 0.7937709000000001,
      "accuracy_std_dev": 0.011766387869265602,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.004587615656565657,
      "avg_duration_per_task_ms": 253050.72525252524,
      "avg_output_tokens_per_task": 30034.337373737373,
      "last_run_timestamp": "2026-10-17T16:03:34.733Z"
    },
    {
      "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": 0.7931395,
      "accuracy_std_dev": 0.031746818911506706,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.03944074053030303,
      "avg_duration_per_task_ms": 195523.23863636365,
      "avg_output_tokens_per_task": 21811.546085858587,
      "last_run_timestamp": "2026-10-31T16:08:31.063Z"
    },
    {
      "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": 0.7926977023289665,
      "accuracy_std_dev": 0.008002668144348565,
      "run_count": 7,
      "total_questions": 1374,
      "avg_cost_per_task_usd": 0.0120994730713246,
      "avg_duration_per_task_ms": 824903.4272197962,
      "avg_output_tokens_per_task": 39852.10480349345,
      "last_run_timestamp": "2026-11-04T19:16:42.901Z"
    },
    {
      "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": 0.7925247381378694,
      "accuracy_std_dev": 0.015422197277112174,
      "run_count": 12,
      "total_questions": 2234,
      "avg_cost_per_task_usd": 0.004671950760966876,
      "avg_duration_per_task_ms": 1479604.6521933752,
      "avg_output_tokens_per_task": 30665.252014324084,
      "last_run_timestamp": "2026-10-31T16:08:08.537Z"
    },
    {
      "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": 0.7923760646992054,
      "accuracy_std_dev": 0.150042241777375,
      "run_count": 9,
      "total_questions": 1762,
      "avg_cost_per_task_usd": 0.015620138479001138,
      "avg_duration_per_task_ms": 554789.2985244041,
      "avg_output_tokens_per_task": 26201.363223609536,
      "last_run_timestamp": "2026-10-17T08:10:43.825Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta/muse-glimmer-30b-20260810",
      "model_slug": "meta/muse-glimmer-30b",
      "model_name": "Meta: Muse Glimmer 30B",
      "provider_name": "DeepInfra",
      "accuracy": 0.7920900579268293,
      "accuracy_std_dev": 0.011512971285033247,
      "run_count": 10,
      "total_questions": 1968,
      "avg_cost_per_task_usd": 0.01824498475609756,
      "avg_duration_per_task_ms": 181543.2469512195,
      "avg_output_tokens_per_task": 15051.83943089431,
      "last_run_timestamp": "2026-10-08T22:29:11.504Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "Alibaba",
      "accuracy": 0.7916197777777776,
      "accuracy_std_dev": 0.10449398651148174,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.029375244107744107,
      "avg_duration_per_task_ms": 135177.0364758698,
      "avg_output_tokens_per_task": 22499.014029180697,
      "last_run_timestamp": "2026-10-31T16:08:08.538Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-thinking-2507",
      "model_slug": "qwen/qwen3-235b-a22b-thinking-2507",
      "model_name": "Qwen: Qwen3 235B A22B Thinking 2507",
      "provider_name": "Alibaba",
      "accuracy": 0.7912454999999999,
      "accuracy_std_dev": 0.011784500000000031,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.05368662373737374,
      "avg_duration_per_task_ms": 408830.87373737374,
      "avg_output_tokens_per_task": 23323.171717171717,
      "last_run_timestamp": "2026-08-22T17:18:01.115Z"
    },
    {
      "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": 0.79103525,
      "accuracy_std_dev": 0.01161593222421257,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.01797352335858586,
      "avg_duration_per_task_ms": 60686.5625,
      "avg_output_tokens_per_task": 5836.820707070707,
      "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": "Together",
      "accuracy": 0.7907960444642064,
      "accuracy_std_dev": 0.011335818096181775,
      "run_count": 12,
      "total_questions": 2249,
      "avg_cost_per_task_usd": 0.008774971098265897,
      "avg_duration_per_task_ms": 377108.3983992886,
      "avg_output_tokens_per_task": 34581.72787905736,
      "last_run_timestamp": "2026-10-31T16:15:26.662Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.1-terminus",
      "model_slug": "deepseek/deepseek-v3.1-terminus",
      "model_name": "DeepSeek: DeepSeek V3.1 Terminus",
      "provider_name": "DeepInfra",
      "accuracy": 0.787879,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.01368820202020202,
      "avg_duration_per_task_ms": 414867.4494949495,
      "avg_output_tokens_per_task": 14233.494949494949,
      "last_run_timestamp": "2026-08-15T14:56:55.269Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "GMICloud",
      "accuracy": 0.7876197692307692,
      "accuracy_std_dev": 0.03230095147242056,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.10465196542346543,
      "avg_duration_per_task_ms": 367907.6076146076,
      "avg_output_tokens_per_task": 27503.11655011655,
      "last_run_timestamp": "2026-10-31T16:09:41.959Z"
    },
    {
      "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": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "CoreWeave",
      "accuracy": 0.7868265000000001,
      "accuracy_std_dev": 0.010545056341717679,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.021032600378787878,
      "avg_duration_per_task_ms": 242572.1452020202,
      "avg_output_tokens_per_task": 12501.890782828283,
      "last_run_timestamp": "2026-10-31T16:10:36.260Z"
    },
    {
      "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": 0.7865083595251555,
      "accuracy_std_dev": 0.02939069554623296,
      "run_count": 9,
      "total_questions": 1769,
      "avg_cost_per_task_usd": 0.011274476540418316,
      "avg_duration_per_task_ms": 619482.3962690786,
      "avg_output_tokens_per_task": 37337.42283776145,
      "last_run_timestamp": "2026-10-31T16:03:21.448Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Morph",
      "accuracy": 0.7864661884870098,
      "accuracy_std_dev": 0.10678677266478277,
      "run_count": 10,
      "total_questions": 1963,
      "avg_cost_per_task_usd": 0.12194484309730003,
      "avg_duration_per_task_ms": 745901.2404482934,
      "avg_output_tokens_per_task": 33847.68874172185,
      "last_run_timestamp": "2026-10-31T17:00:44.290Z"
    },
    {
      "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": "deepseek/deepseek-r1-0528",
      "model_slug": "deepseek/deepseek-r1-0528",
      "model_name": "DeepSeek: R1 0528",
      "provider_name": "DeepInfra",
      "accuracy": 0.7861940813859789,
      "accuracy_std_dev": 0.020955855485826815,
      "run_count": 7,
      "total_questions": 1241,
      "avg_cost_per_task_usd": 0.04030405640612409,
      "avg_duration_per_task_ms": 737558.6027397261,
      "avg_output_tokens_per_task": 18622.764705882353,
      "last_run_timestamp": "2026-10-17T16:06:15.029Z"
    },
    {
      "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": 0.7854701461538461,
      "accuracy_std_dev": 0.11062032174825517,
      "run_count": 4,
      "total_questions": 780,
      "avg_cost_per_task_usd": 0.10532577051282052,
      "avg_duration_per_task_ms": 1185336.45,
      "avg_output_tokens_per_task": 37894.10512820513,
      "last_run_timestamp": "2026-10-24T16:13:05.627Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-thinking-2507",
      "model_slug": "qwen/qwen3-235b-a22b-thinking-2507",
      "model_name": "Qwen: Qwen3 235B A22B Thinking 2507",
      "provider_name": "DeepInfra",
      "accuracy": 0.7854607925531915,
      "accuracy_std_dev": 0.004311000000000009,
      "run_count": 2,
      "total_questions": 376,
      "avg_cost_per_task_usd": 0.05007911170212766,
      "avg_duration_per_task_ms": 706928.4574468085,
      "avg_output_tokens_per_task": 21703.061170212764,
      "last_run_timestamp": "2026-08-22T16:06:14.894Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Nebius",
      "accuracy": 0.78493275,
      "accuracy_std_dev": 0.010934561842501992,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.03407076388888889,
      "avg_duration_per_task_ms": 483005.77398989897,
      "avg_output_tokens_per_task": 67906.58964646465,
      "last_run_timestamp": "2026-10-31T08:31:13.285Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "Venice",
      "accuracy": 0.7847687086124402,
      "accuracy_std_dev": 0.12803182926015888,
      "run_count": 11,
      "total_questions": 2090,
      "avg_cost_per_task_usd": 0.10580474545454546,
      "avg_duration_per_task_ms": 663710.6894736842,
      "avg_output_tokens_per_task": 30101.89138755981,
      "last_run_timestamp": "2026-10-31T16:40:50.323Z"
    },
    {
      "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": 0.7846955057471265,
      "accuracy_std_dev": 0.10720142439753014,
      "run_count": 8,
      "total_questions": 1566,
      "avg_cost_per_task_usd": 0.07534132247765006,
      "avg_duration_per_task_ms": 702935.9335887611,
      "avg_output_tokens_per_task": 27301.47062579821,
      "last_run_timestamp": "2026-10-10T16:29:40.552Z"
    },
    {
      "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": 0.7828283333333333,
      "accuracy_std_dev": 0.014481820841623765,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.02915533333333333,
      "avg_duration_per_task_ms": 454449.42087542085,
      "avg_output_tokens_per_task": 24277.760942760942,
      "last_run_timestamp": "2026-08-29T18:09:57.600Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-r1-0528",
      "model_slug": "deepseek/deepseek-r1-0528",
      "model_name": "DeepSeek: R1 0528",
      "provider_name": "StreamLake",
      "accuracy": 0.7825877142857144,
      "accuracy_std_dev": 0.026746482528225146,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.053754077200577205,
      "avg_duration_per_task_ms": 455374.8766233766,
      "avg_output_tokens_per_task": 23369.723665223664,
      "last_run_timestamp": "2026-10-17T16:03:34.700Z"
    },
    {
      "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": 0.7807207329462988,
      "accuracy_std_dev": 0.06482919923639853,
      "run_count": 7,
      "total_questions": 1378,
      "avg_cost_per_task_usd": 0.00722692815674891,
      "avg_duration_per_task_ms": 850153.4637155298,
      "avg_output_tokens_per_task": 18320.05587808418,
      "last_run_timestamp": "2026-10-24T16:21:54.634Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "DeepInfra",
      "accuracy": 0.7804782183266932,
      "accuracy_std_dev": 0.017482235460356838,
      "run_count": 9,
      "total_questions": 1255,
      "avg_cost_per_task_usd": 0.04653852270916335,
      "avg_duration_per_task_ms": 966873.6972111554,
      "avg_output_tokens_per_task": 23672.268525896416,
      "last_run_timestamp": "2026-10-31T16:04:29.123Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "Novita",
      "accuracy": 0.7800223333333333,
      "accuracy_std_dev": 0.011195005920300127,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.006137688552188552,
      "avg_duration_per_task_ms": 345084.8249158249,
      "avg_output_tokens_per_task": 14881.631313131313,
      "last_run_timestamp": "2026-08-29T19:20:50.265Z"
    },
    {
      "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": 0.77988225,
      "accuracy_std_dev": 0.01598762236661534,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.3256477702020202,
      "avg_duration_per_task_ms": 1001085.1660353536,
      "avg_output_tokens_per_task": 82083.53661616161,
      "last_run_timestamp": "2026-10-31T16:04:02.366Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "Google",
      "accuracy": 0.7797533759071117,
      "accuracy_std_dev": 0.019631960574532543,
      "run_count": 7,
      "total_questions": 1378,
      "avg_cost_per_task_usd": 0.02252375471698113,
      "avg_duration_per_task_ms": 854891.6850507982,
      "avg_output_tokens_per_task": 38423.74455732946,
      "last_run_timestamp": "2026-10-24T16:37:21.981Z"
    },
    {
      "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": 0.779461111111111,
      "accuracy_std_dev": 0.13454197858284844,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.008024448933782269,
      "avg_duration_per_task_ms": 403581.7491582492,
      "avg_output_tokens_per_task": 25395.133557800225,
      "last_run_timestamp": "2026-10-31T16:19:28.505Z"
    },
    {
      "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": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "Darkbloom",
      "accuracy": 0.77925075,
      "accuracy_std_dev": 0.01980242038710168,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.02103724936868687,
      "avg_duration_per_task_ms": 475607.264520202,
      "avg_output_tokens_per_task": 27969.91224747475,
      "last_run_timestamp": "2026-10-31T16:16:35.158Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "auto-routing",
      "accuracy": 0.7784991428571428,
      "accuracy_std_dev": 0.04711966006252765,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.013313156565656565,
      "avg_duration_per_task_ms": 825301.5577200577,
      "avg_output_tokens_per_task": 39367.342712842714,
      "last_run_timestamp": "2026-10-17T08:18:06.210Z"
    },
    {
      "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": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "auto-routing",
      "accuracy": 0.7749718333333332,
      "accuracy_std_dev": 0.019112577010236546,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.005948839646464647,
      "avg_duration_per_task_ms": 683117.8017676767,
      "avg_output_tokens_per_task": 37109.85984848485,
      "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-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/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Crusoe",
      "accuracy": 0.77381,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 140,
      "avg_cost_per_task_usd": 0.00325335,
      "avg_duration_per_task_ms": 70786.31428571428,
      "avg_output_tokens_per_task": 3951.1785714285716,
      "last_run_timestamp": "2026-08-15T14:54:48.654Z"
    },
    {
      "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": 0.7737795,
      "accuracy_std_dev": 0.054440706080101486,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.15974002967171716,
      "avg_duration_per_task_ms": 914205.7670454546,
      "avg_output_tokens_per_task": 68299.18939393939,
      "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-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "Phala",
      "accuracy": 0.7736031459537571,
      "accuracy_std_dev": 0.07146005981537049,
      "run_count": 7,
      "total_questions": 1384,
      "avg_cost_per_task_usd": 0.01837689523121387,
      "avg_duration_per_task_ms": 1270075.3273121386,
      "avg_output_tokens_per_task": 17747.31430635838,
      "last_run_timestamp": "2026-10-31T16:45:39.993Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5.4-nano-20260317",
      "model_slug": "openai/gpt-5.4-nano",
      "model_name": "OpenAI: GPT-5.4 Nano",
      "provider_name": "Azure",
      "accuracy": 0.7734006,
      "accuracy_std_dev": 0.005280843894681953,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.014113778787878787,
      "avg_duration_per_task_ms": 123713.42222222222,
      "avg_output_tokens_per_task": 10542.545454545454,
      "last_run_timestamp": "2026-10-01T19:20:04.986Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Venice",
      "accuracy": 0.7732982293193716,
      "accuracy_std_dev": 0.013698592432801259,
      "run_count": 5,
      "total_questions": 955,
      "avg_cost_per_task_usd": 0.00232973612565445,
      "avg_duration_per_task_ms": 792784.167539267,
      "avg_output_tokens_per_task": 15348.367539267016,
      "last_run_timestamp": "2026-10-31T16:10:36.260Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "DeepInfra",
      "accuracy": 0.7732734397522524,
      "accuracy_std_dev": 0.007399312031750627,
      "run_count": 9,
      "total_questions": 1776,
      "avg_cost_per_task_usd": 0.011318120495495496,
      "avg_duration_per_task_ms": 924768.2730855856,
      "avg_output_tokens_per_task": 11765.394144144144,
      "last_run_timestamp": "2026-10-31T16:02:14.358Z"
    },
    {
      "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": 0.7727273,
      "accuracy_std_dev": 0.04359568158671222,
      "run_count": 10,
      "total_questions": 1980,
      "avg_cost_per_task_usd": 0.22926952626262626,
      "avg_duration_per_task_ms": 887813.6904040404,
      "avg_output_tokens_per_task": 76915.14747474747,
      "last_run_timestamp": "2026-10-17T16:06:46.266Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Mara",
      "accuracy": 0.7722982392694063,
      "accuracy_std_dev": 0.026492001321153518,
      "run_count": 8,
      "total_questions": 1095,
      "avg_cost_per_task_usd": 0.013083080365296807,
      "avg_duration_per_task_ms": 53025.53789954338,
      "avg_output_tokens_per_task": 17342.828310502282,
      "last_run_timestamp": "2026-10-31T16:39:49.876Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7721661666666666,
      "accuracy_std_dev": 0.007120472442580984,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.020572006734006738,
      "avg_duration_per_task_ms": 737897.6776094276,
      "avg_output_tokens_per_task": 51210.5968013468,
      "last_run_timestamp": "2026-10-31T16:23:19.027Z"
    },
    {
      "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": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Cerebras",
      "accuracy": 0.7714366132879046,
      "accuracy_std_dev": 0.005025261541010102,
      "run_count": 3,
      "total_questions": 587,
      "avg_cost_per_task_usd": 0.00432709880749574,
      "avg_duration_per_task_ms": 4109.833049403748,
      "avg_output_tokens_per_task": 2384.9233390119252,
      "last_run_timestamp": "2026-08-29T18:58:30.919Z"
    },
    {
      "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": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "Mara",
      "accuracy": 0.7706375705329154,
      "accuracy_std_dev": 0.012487370499829025,
      "run_count": 3,
      "total_questions": 319,
      "avg_cost_per_task_usd": 0.03633863949843261,
      "avg_duration_per_task_ms": 130589.2789968652,
      "avg_output_tokens_per_task": 37748.849529780564,
      "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": "Parasail",
      "accuracy": 0.7705974206349206,
      "accuracy_std_dev": 0.01526894876684195,
      "run_count": 9,
      "total_questions": 1764,
      "avg_cost_per_task_usd": 0.021992218253968247,
      "avg_duration_per_task_ms": 244131.76473922902,
      "avg_output_tokens_per_task": 29949.219387755104,
      "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": 0.7704452803915377,
      "accuracy_std_dev": 0.015743859747037087,
      "run_count": 16,
      "total_questions": 3167,
      "avg_cost_per_task_usd": 0.006528186927691823,
      "avg_duration_per_task_ms": 442659.0584149037,
      "avg_output_tokens_per_task": 29091.698137038205,
      "last_run_timestamp": "2026-10-31T16:02:14.375Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "BaseTen",
      "accuracy": 0.7698783076923077,
      "accuracy_std_dev": 0.010511058333485764,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.014805616938616937,
      "avg_duration_per_task_ms": 135269.72105672106,
      "avg_output_tokens_per_task": 29456.09829059829,
      "last_run_timestamp": "2026-10-31T16:51:23.396Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "SambaNova",
      "accuracy": 0.7694390259909032,
      "accuracy_std_dev": 0.01052704853876431,
      "run_count": 8,
      "total_questions": 1539,
      "avg_cost_per_task_usd": 0.02523249252761533,
      "avg_duration_per_task_ms": 74630.34632878493,
      "avg_output_tokens_per_task": 26425.37621832359,
      "last_run_timestamp": "2026-10-31T16:06:00.111Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "DeepInfra",
      "accuracy": 0.7684847733333334,
      "accuracy_std_dev": 0.03945984809060131,
      "run_count": 9,
      "total_questions": 1650,
      "avg_cost_per_task_usd": 0.007370210909090909,
      "avg_duration_per_task_ms": 876615.1224242424,
      "avg_output_tokens_per_task": 18987.695151515152,
      "last_run_timestamp": "2026-10-31T16:04:29.109Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "DekaLLM",
      "accuracy": 0.7669798106936416,
      "accuracy_std_dev": 0.011686708686296891,
      "run_count": 7,
      "total_questions": 1384,
      "avg_cost_per_task_usd": 0.019905706647398845,
      "avg_duration_per_task_ms": 1154492.3439306358,
      "avg_output_tokens_per_task": 60181.387283236996,
      "last_run_timestamp": "2026-10-31T16:03:21.445Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4-sonnet-20250522",
      "model_slug": "anthropic/claude-sonnet-4",
      "model_name": "Anthropic: Claude Sonnet 4",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7669488305084746,
      "accuracy_std_dev": 0.016365499999999977,
      "run_count": 2,
      "total_questions": 354,
      "avg_cost_per_task_usd": 0.21194037853107345,
      "avg_duration_per_task_ms": 159764.62146892655,
      "avg_output_tokens_per_task": 13982.706214689266,
      "last_run_timestamp": "2026-09-01T13:33:46.786Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-2.5-flash",
      "model_slug": "google/gemini-2.5-flash",
      "model_name": "Google: Gemini 2.5 Flash",
      "provider_name": "Google AI Studio",
      "accuracy": 0.7668349999999999,
      "accuracy_std_dev": 0.004208999999999963,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.057140755050505046,
      "avg_duration_per_task_ms": 101024.92676767676,
      "avg_output_tokens_per_task": 22766.989898989897,
      "last_run_timestamp": "2026-10-01T19:11:42.323Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": "deepseek/deepseek-r1-0528",
      "model_slug": "deepseek/deepseek-r1-0528",
      "model_name": "DeepSeek: R1 0528",
      "provider_name": "auto-routing",
      "accuracy": 0.7651515,
      "accuracy_std_dev": 0.037753760451774886,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.052040077441077436,
      "avg_duration_per_task_ms": 879769.8745791246,
      "avg_output_tokens_per_task": 23186.810606060608,
      "last_run_timestamp": "2026-10-17T16:03:24.588Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Nebius",
      "accuracy": 0.765058,
      "accuracy_std_dev": 0.01367922480747111,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.018581179573512904,
      "avg_duration_per_task_ms": 112140.46801346801,
      "avg_output_tokens_per_task": 30739.353535353537,
      "last_run_timestamp": "2026-10-31T16:52:10.938Z"
    },
    {
      "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": "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": "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": 0.7638887500000001,
      "accuracy_std_dev": 0.01180708327392928,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.02126011868686869,
      "avg_duration_per_task_ms": 1098413.3244949495,
      "avg_output_tokens_per_task": 53038.14141414141,
      "last_run_timestamp": "2026-10-31T16:51:21.148Z"
    },
    {
      "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": "deepseek/deepseek-v3.1-terminus",
      "model_slug": "deepseek/deepseek-v3.1-terminus",
      "model_name": "DeepSeek: DeepSeek V3.1 Terminus",
      "provider_name": "auto-routing",
      "accuracy": 0.7631873333333332,
      "accuracy_std_dev": 0.012472702469348371,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.017184111111111112,
      "avg_duration_per_task_ms": 455850.7845117845,
      "avg_output_tokens_per_task": 17097.883838383837,
      "last_run_timestamp": "2026-08-29T18:21:38.666Z"
    },
    {
      "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7617845000000001,
      "accuracy_std_dev": 0.014110256482431497,
      "run_count": 16,
      "total_questions": 3168,
      "avg_cost_per_task_usd": 0.01827727462121212,
      "avg_duration_per_task_ms": 146955.67613636365,
      "avg_output_tokens_per_task": 30220.935606060608,
      "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": "Together",
      "accuracy": 0.7617845,
      "accuracy_std_dev": 0.011198548678288631,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.017532584595959598,
      "avg_duration_per_task_ms": 239388.20833333334,
      "avg_output_tokens_per_task": 29069.847853535353,
      "last_run_timestamp": "2026-10-31T16:22:37.563Z"
    },
    {
      "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": 0.7617845,
      "accuracy_std_dev": 0.0244105,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.05473536363636364,
      "avg_duration_per_task_ms": 625909.4974747475,
      "avg_output_tokens_per_task": 22336.9797979798,
      "last_run_timestamp": "2026-08-22T17:09:41.940Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Groq",
      "accuracy": 0.761574,
      "accuracy_std_dev": 0.011842521785920426,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.017385299873737374,
      "avg_duration_per_task_ms": 63084.142045454544,
      "avg_output_tokens_per_task": 28736.113636363636,
      "last_run_timestamp": "2026-10-31T16:23:59.761Z"
    },
    {
      "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": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "GMICloud",
      "accuracy": 0.7613168888888889,
      "accuracy_std_dev": 0.08989112326642029,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.1361467480359147,
      "avg_duration_per_task_ms": 937300.8019079686,
      "avg_output_tokens_per_task": 71180.4657687991,
      "last_run_timestamp": "2026-10-31T16:04:29.124Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "auto-routing",
      "accuracy": 0.7607322499999999,
      "accuracy_std_dev": 0.04032017199166567,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.058705520202020196,
      "avg_duration_per_task_ms": 1045870.6470959596,
      "avg_output_tokens_per_task": 29725.0625,
      "last_run_timestamp": "2026-10-31T16:08:31.065Z"
    },
    {
      "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": 0.7598906249999999,
      "accuracy_std_dev": 0.032074197830723274,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.004828941287878788,
      "avg_duration_per_task_ms": 600231.7380050505,
      "avg_output_tokens_per_task": 28770.83648989899,
      "last_run_timestamp": "2026-10-31T16:20:12.041Z"
    },
    {
      "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": 0.7592847364130435,
      "accuracy_std_dev": 0.03960627555022057,
      "run_count": 5,
      "total_questions": 736,
      "avg_cost_per_task_usd": 0.058288675271739136,
      "avg_duration_per_task_ms": 649092.0529891305,
      "avg_output_tokens_per_task": 49476.220108695656,
      "last_run_timestamp": "2026-08-29T19:07:57.500Z"
    },
    {
      "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": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7570146666666666,
      "accuracy_std_dev": 0.017186872710169103,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.05602859764309764,
      "avg_duration_per_task_ms": 578162.5252525252,
      "avg_output_tokens_per_task": 26055.28787878788,
      "last_run_timestamp": "2026-08-29T18:26:43.144Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "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": "qwen/qwen3-coder-next-2025-02-03",
      "model_slug": "qwen/qwen3-coder-next",
      "model_name": "Qwen: Qwen3 Coder Next",
      "provider_name": "Alibaba",
      "accuracy": 0.7558922500000002,
      "accuracy_std_dev": 0.007240774798838877,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.022548712121212122,
      "avg_duration_per_task_ms": 94076.5808080808,
      "avg_output_tokens_per_task": 14873.041666666666,
      "last_run_timestamp": "2026-10-03T16:56:57.458Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "mistralai/mistral-small-2603",
      "model_slug": "mistralai/mistral-small-2603",
      "model_name": "Mistral: Mistral Small 4",
      "provider_name": "Venice",
      "accuracy": 0.755892,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.018171858585858586,
      "avg_duration_per_task_ms": 285760.0353535354,
      "avg_output_tokens_per_task": 24027.954545454544,
      "last_run_timestamp": "2026-09-01T12:59:55.618Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "Google",
      "accuracy": 0.755695906329114,
      "accuracy_std_dev": 0.006554500000000019,
      "run_count": 2,
      "total_questions": 395,
      "avg_cost_per_task_usd": 0.020412756962025318,
      "avg_duration_per_task_ms": 122185.9164556962,
      "avg_output_tokens_per_task": 11814.949367088608,
      "last_run_timestamp": "2026-08-22T17:12:08.919Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "OpenInference",
      "accuracy": 0.7547698333333334,
      "accuracy_std_dev": 0.0766049855914889,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.004288400673400673,
      "avg_duration_per_task_ms": 447675.2971380471,
      "avg_output_tokens_per_task": 15291.021885521886,
      "last_run_timestamp": "2026-10-17T08:26:03.926Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "DeepInfra",
      "accuracy": 0.7539042317380352,
      "accuracy_std_dev": 0.03122595132646972,
      "run_count": 12,
      "total_questions": 1985,
      "avg_cost_per_task_usd": 0.045778590428211585,
      "avg_duration_per_task_ms": 263292.8277078086,
      "avg_output_tokens_per_task": 17254.131486146096,
      "last_run_timestamp": "2026-10-31T08:39:08.315Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "mistralai/mistral-small-2603",
      "model_slug": "mistralai/mistral-small-2603",
      "model_name": "Mistral: Mistral Small 4",
      "provider_name": "Mistral",
      "accuracy": 0.7528056666666667,
      "accuracy_std_dev": 0.004601995207395269,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.015185508417508417,
      "avg_duration_per_task_ms": 152622.4537037037,
      "avg_output_tokens_per_task": 24359.508417508416,
      "last_run_timestamp": "2026-09-17T02:17:49.284Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.7525251818181818,
      "accuracy_std_dev": 0.013807926121801336,
      "run_count": 11,
      "total_questions": 2178,
      "avg_cost_per_task_usd": 0.01597986134067952,
      "avg_duration_per_task_ms": 1293582.3039485766,
      "avg_output_tokens_per_task": 55438.24655647383,
      "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": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "DeepInfra",
      "accuracy": 0.7516448305921052,
      "accuracy_std_dev": 0.02929436769346626,
      "run_count": 5,
      "total_questions": 608,
      "avg_cost_per_task_usd": 0.051999368421052634,
      "avg_duration_per_task_ms": 386838.6842105263,
      "avg_output_tokens_per_task": 15199.644736842105,
      "last_run_timestamp": "2026-09-19T08:53:38.031Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "Makora",
      "accuracy": 0.7514029999999999,
      "accuracy_std_dev": 0.013279676389129373,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.015289352693602695,
      "avg_duration_per_task_ms": 350945.7154882155,
      "avg_output_tokens_per_task": 62534.38636363636,
      "last_run_timestamp": "2026-10-31T16:38:47.420Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "DeepInfra",
      "accuracy": 0.7502983046594983,
      "accuracy_std_dev": 0.030897500000000022,
      "run_count": 2,
      "total_questions": 279,
      "avg_cost_per_task_usd": 0.12269224014336919,
      "avg_duration_per_task_ms": 1216119.093189964,
      "avg_output_tokens_per_task": 34943.36559139785,
      "last_run_timestamp": "2026-08-22T17:30:52.646Z"
    },
    {
      "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": 0.749495,
      "accuracy_std_dev": 0.010721542221154592,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.013542626262626261,
      "avg_duration_per_task_ms": 93554.2,
      "avg_output_tokens_per_task": 14884.874747474747,
      "last_run_timestamp": "2026-10-03T16:54:39.591Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "GMICloud",
      "accuracy": 0.7494390818228898,
      "accuracy_std_dev": 0.10270027330445619,
      "run_count": 10,
      "total_questions": 1931,
      "avg_cost_per_task_usd": 0.006055046090108752,
      "avg_duration_per_task_ms": 549414.4293112376,
      "avg_output_tokens_per_task": 26366.88865872605,
      "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": "auto-routing",
      "accuracy": 0.7491585,
      "accuracy_std_dev": 0.009959720314848229,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.014780738636363637,
      "avg_duration_per_task_ms": 104786.22727272728,
      "avg_output_tokens_per_task": 14546.537878787878,
      "last_run_timestamp": "2026-10-03T16:43:33.673Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "inclusionai/ling-3.0-flash-20260723",
      "model_slug": "inclusionai/ling-3.0-flash",
      "model_name": "inclusionAI: Ling 3.0 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.7487311624365482,
      "accuracy_std_dev": 0.019042500000000018,
      "run_count": 2,
      "total_questions": 394,
      "avg_cost_per_task_usd": 0.00260698730964467,
      "avg_duration_per_task_ms": 166191.14467005077,
      "avg_output_tokens_per_task": 30980.81725888325,
      "last_run_timestamp": "2026-10-01T16:04:24.420Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "Wafer",
      "accuracy": 0.7481013645569621,
      "accuracy_std_dev": 0.004003499999999993,
      "run_count": 2,
      "total_questions": 395,
      "avg_cost_per_task_usd": 0.010130237974683544,
      "avg_duration_per_task_ms": 298779.6734177215,
      "avg_output_tokens_per_task": 60721.07341772152,
      "last_run_timestamp": "2026-10-10T08:39:12.933Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "Baidu",
      "accuracy": 0.7455807499999999,
      "accuracy_std_dev": 0.025882017723460052,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.1971899930555556,
      "avg_duration_per_task_ms": 1087151.1325757576,
      "avg_output_tokens_per_task": 89342.12184343435,
      "last_run_timestamp": "2026-10-31T16:01:53.649Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Io Net",
      "accuracy": 0.7450601862113403,
      "accuracy_std_dev": 0.02284683246884292,
      "run_count": 8,
      "total_questions": 1552,
      "avg_cost_per_task_usd": 0.016163423324742267,
      "avg_duration_per_task_ms": 377921.4826030928,
      "avg_output_tokens_per_task": 53593.9587628866,
      "last_run_timestamp": "2026-10-31T08:07:17.099Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "StreamLake",
      "accuracy": 0.7444443999999999,
      "accuracy_std_dev": 0.008741194188439036,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.006576761616161616,
      "avg_duration_per_task_ms": 163397.33737373736,
      "avg_output_tokens_per_task": 7642.029292929293,
      "last_run_timestamp": "2026-10-03T16:58:32.590Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "GMICloud",
      "accuracy": 0.7441077777777777,
      "accuracy_std_dev": 0.02634512333864457,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.33108891245791244,
      "avg_duration_per_task_ms": 924653.0869809203,
      "avg_output_tokens_per_task": 82641.20707070707,
      "last_run_timestamp": "2026-10-31T16:39:49.878Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7421462352941177,
      "accuracy_std_dev": 0.08394561417332949,
      "run_count": 8,
      "total_questions": 1581,
      "avg_cost_per_task_usd": 0.004914519924098672,
      "avg_duration_per_task_ms": 596355.0860215054,
      "avg_output_tokens_per_task": 32270,
      "last_run_timestamp": "2026-10-17T16:03:21.011Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Nebius",
      "accuracy": 0.7421436666666666,
      "accuracy_std_dev": 0.0029824300867283098,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.0052554671717171715,
      "avg_duration_per_task_ms": 138780.79882154881,
      "avg_output_tokens_per_task": 8488.298821548822,
      "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-coder-next-2025-02-03",
      "model_slug": "qwen/qwen3-coder-next",
      "model_name": "Qwen: Qwen3 Coder Next",
      "provider_name": "Parasail",
      "accuracy": 0.741847766304348,
      "accuracy_std_dev": 0.004303026282455164,
      "run_count": 4,
      "total_questions": 736,
      "avg_cost_per_task_usd": 0.008672221467391304,
      "avg_duration_per_task_ms": 124088.47690217392,
      "avg_output_tokens_per_task": 10910.259510869566,
      "last_run_timestamp": "2026-10-03T17:03:40.665Z"
    },
    {
      "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": "qwen/qwen3-coder-next-2025-02-03",
      "model_slug": "qwen/qwen3-coder-next",
      "model_name": "Qwen: Qwen3 Coder Next",
      "provider_name": "Novita",
      "accuracy": 0.7398992499999999,
      "accuracy_std_dev": 0.0060112973797259595,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.02330285101010101,
      "avg_duration_per_task_ms": 95915.94318181818,
      "avg_output_tokens_per_task": 15426.907828282829,
      "last_run_timestamp": "2026-10-03T16:58:08.462Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Alibaba",
      "accuracy": 0.739899,
      "accuracy_std_dev": 0.007575777781851822,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.0048982929292929285,
      "avg_duration_per_task_ms": 108030.26641414141,
      "avg_output_tokens_per_task": 7988.015151515152,
      "last_run_timestamp": "2026-10-03T16:59:36.480Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "DeepInfra",
      "accuracy": 0.7385018107983389,
      "accuracy_std_dev": 0.1187447930907569,
      "run_count": 12,
      "total_questions": 2167,
      "avg_cost_per_task_usd": 0.033791137517305024,
      "avg_duration_per_task_ms": 1331765.8463313337,
      "avg_output_tokens_per_task": 35484.031379787724,
      "last_run_timestamp": "2026-10-31T16:04:48.182Z"
    },
    {
      "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-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "SiliconFlow",
      "accuracy": 0.738005,
      "accuracy_std_dev": 0.009886114694357948,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.11405846527777778,
      "avg_duration_per_task_ms": 757664.5296717172,
      "avg_output_tokens_per_task": 44516.394570707074,
      "last_run_timestamp": "2026-10-31T16:04:48.154Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "Novita",
      "accuracy": 0.7368489013657057,
      "accuracy_std_dev": 0.05193843550644497,
      "run_count": 7,
      "total_questions": 1318,
      "avg_cost_per_task_usd": 0.018259321699544768,
      "avg_duration_per_task_ms": 840685.5872534142,
      "avg_output_tokens_per_task": 45854.41881638847,
      "last_run_timestamp": "2026-10-31T16:47:37.740Z"
    },
    {
      "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": 0.7366410572519084,
      "accuracy_std_dev": 0.012366834942068835,
      "run_count": 8,
      "total_questions": 1572,
      "avg_cost_per_task_usd": 0.006647374045801527,
      "avg_duration_per_task_ms": 328667.01463104325,
      "avg_output_tokens_per_task": 26397.287531806614,
      "last_run_timestamp": "2026-10-31T16:16:13.798Z"
    },
    {
      "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": 0.736532,
      "accuracy_std_dev": 0.01262599999999997,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.07812436868686869,
      "avg_duration_per_task_ms": 803892.6035353536,
      "avg_output_tokens_per_task": 22434.27525252525,
      "last_run_timestamp": "2026-08-22T17:38:59.682Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Parasail",
      "accuracy": 0.7363656367656349,
      "accuracy_std_dev": 0.008751528723600225,
      "run_count": 8,
      "total_questions": 1583,
      "avg_cost_per_task_usd": 0.007021190145293747,
      "avg_duration_per_task_ms": 251648.8812381554,
      "avg_output_tokens_per_task": 8683.42135186355,
      "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-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "auto-routing",
      "accuracy": 0.735900625,
      "accuracy_std_dev": 0.023012779036534772,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.006714852904040405,
      "avg_duration_per_task_ms": 396625.5202020202,
      "avg_output_tokens_per_task": 22776.6452020202,
      "last_run_timestamp": "2026-10-31T16:08:31.072Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Venice",
      "accuracy": 0.7328042857142857,
      "accuracy_std_dev": 0.006328935223339255,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.006012205627705627,
      "avg_duration_per_task_ms": 432462.59451659454,
      "avg_output_tokens_per_task": 7938.487734487734,
      "last_run_timestamp": "2026-10-31T16:08:56.903Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Relace",
      "accuracy": 0.732440575,
      "accuracy_std_dev": 0.057246596407316816,
      "run_count": 3,
      "total_questions": 560,
      "avg_cost_per_task_usd": 0.014524144642857144,
      "avg_duration_per_task_ms": 428029.5660714286,
      "avg_output_tokens_per_task": 58878.205357142855,
      "last_run_timestamp": "2026-10-31T08:03:10.090Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.732155,
      "accuracy_std_dev": 0.02424727451075689,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.016461675757575758,
      "avg_duration_per_task_ms": 673272.2666666667,
      "avg_output_tokens_per_task": 60883.92525252525,
      "last_run_timestamp": "2026-10-31T08:31:11.846Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "Mara",
      "accuracy": 0.7306920882647944,
      "accuracy_std_dev": 0.018533662569629808,
      "run_count": 7,
      "total_questions": 997,
      "avg_cost_per_task_usd": 0.016062558676028082,
      "avg_duration_per_task_ms": 67044.65897693079,
      "avg_output_tokens_per_task": 9235.0962888666,
      "last_run_timestamp": "2026-10-31T16:30:53.500Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7305295887850467,
      "accuracy_std_dev": 0.01062550000000001,
      "run_count": 2,
      "total_questions": 321,
      "avg_cost_per_task_usd": 0.08696355451713396,
      "avg_duration_per_task_ms": 1922851.7601246105,
      "avg_output_tokens_per_task": 28235.09968847352,
      "last_run_timestamp": "2026-08-22T17:09:42.218Z"
    },
    {
      "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": 0.7304292499999999,
      "accuracy_std_dev": 0.021067858379234947,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.3208809797979798,
      "avg_duration_per_task_ms": 1030227.9185606061,
      "avg_output_tokens_per_task": 94983.5441919192,
      "last_run_timestamp": "2026-10-31T16:11:15.423Z"
    },
    {
      "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": 0.7303591666666667,
      "accuracy_std_dev": 0.00844092435531926,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.1903736430976431,
      "avg_duration_per_task_ms": 238390.79924242425,
      "avg_output_tokens_per_task": 37883.638468013465,
      "last_run_timestamp": "2026-10-31T08:20:09.279Z"
    },
    {
      "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": 0.7303417538461537,
      "accuracy_std_dev": 0.11666421720821496,
      "run_count": 6,
      "total_questions": 1170,
      "avg_cost_per_task_usd": 0.03974357521367521,
      "avg_duration_per_task_ms": 502126.6564102564,
      "avg_output_tokens_per_task": 35593.947863247864,
      "last_run_timestamp": "2026-09-19T17:18:00.087Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Google",
      "accuracy": 0.7301687421940927,
      "accuracy_std_dev": 0.019520676718688693,
      "run_count": 12,
      "total_questions": 2370,
      "avg_cost_per_task_usd": 0.0065871569620253164,
      "avg_duration_per_task_ms": 133816.87552742616,
      "avg_output_tokens_per_task": 6899.4869198312235,
      "last_run_timestamp": "2026-10-17T16:03:44.090Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.5-haiku-20251001",
      "model_slug": "anthropic/claude-haiku-4.5",
      "model_name": "Anthropic: Claude Haiku 4.5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7299572432432432,
      "accuracy_std_dev": 0.012015924167099304,
      "run_count": 37,
      "total_questions": 7326,
      "avg_cost_per_task_usd": 0.20459208230958226,
      "avg_duration_per_task_ms": 238486.2523887524,
      "avg_output_tokens_per_task": 38144.53794703795,
      "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": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "Phala",
      "accuracy": 0.7271808509532062,
      "accuracy_std_dev": 0.03540222199908109,
      "run_count": 6,
      "total_questions": 1154,
      "avg_cost_per_task_usd": 0.011152869150779895,
      "avg_duration_per_task_ms": 504501.5831889082,
      "avg_output_tokens_per_task": 55506.62998266898,
      "last_run_timestamp": "2026-10-10T08:24:27.616Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.5-haiku-20251001",
      "model_slug": "anthropic/claude-haiku-4.5",
      "model_name": "Anthropic: Claude Haiku 4.5",
      "provider_name": "Azure",
      "accuracy": 0.7268518333333334,
      "accuracy_std_dev": 0.0122393019874047,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.18790157449494946,
      "avg_duration_per_task_ms": 344629.31313131313,
      "avg_output_tokens_per_task": 37389.22474747475,
      "last_run_timestamp": "2026-10-31T08:16:11.695Z"
    },
    {
      "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": 0.7260101666666667,
      "accuracy_std_dev": 0.010493774144965926,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.18633288762626263,
      "avg_duration_per_task_ms": 306780.9974747475,
      "avg_output_tokens_per_task": 37097.41203703704,
      "last_run_timestamp": "2026-10-31T08:09:30.774Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "auto-routing",
      "accuracy": 0.725589375,
      "accuracy_std_dev": 0.015859785220310358,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.005495951388888889,
      "avg_duration_per_task_ms": 318269.0252525252,
      "avg_output_tokens_per_task": 8055.998106060606,
      "last_run_timestamp": "2026-10-31T16:22:55.771Z"
    },
    {
      "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": 0.7249578750000001,
      "accuracy_std_dev": 0.041655408686740475,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.3217009703282828,
      "avg_duration_per_task_ms": 1291699.7613636365,
      "avg_output_tokens_per_task": 101576.43434343435,
      "last_run_timestamp": "2026-10-31T16:18:48.427Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "CoreWeave",
      "accuracy": 0.7236893527788727,
      "accuracy_std_dev": 0.03777209883376474,
      "run_count": 13,
      "total_questions": 2537,
      "avg_cost_per_task_usd": 0.010590881750098544,
      "avg_duration_per_task_ms": 263440.8253843122,
      "avg_output_tokens_per_task": 48345.98265668112,
      "last_run_timestamp": "2026-10-31T08:46:43.680Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Krea",
      "accuracy": 0.7234044255319149,
      "accuracy_std_dev": 0.1917994784299478,
      "run_count": 3,
      "total_questions": 517,
      "avg_cost_per_task_usd": 0.033565591876208896,
      "avg_duration_per_task_ms": 242772.43907156674,
      "avg_output_tokens_per_task": 37102.30560928433,
      "last_run_timestamp": "2026-09-29T20:57:20.673Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "Novita",
      "accuracy": 0.723310259001895,
      "accuracy_std_dev": 0.03972893833844543,
      "run_count": 8,
      "total_questions": 1583,
      "avg_cost_per_task_usd": 0.31632930574857865,
      "avg_duration_per_task_ms": 1267532.5773847126,
      "avg_output_tokens_per_task": 99778.44725205307,
      "last_run_timestamp": "2026-10-31T16:18:28.925Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Novita",
      "accuracy": 0.722643,
      "accuracy_std_dev": 0.0177518271735616,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.004502867424242425,
      "avg_duration_per_task_ms": 166323.601010101,
      "avg_output_tokens_per_task": 7638.002525252526,
      "last_run_timestamp": "2026-10-03T16:57:10.991Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "DeepInfra",
      "accuracy": 0.7221286666666666,
      "accuracy_std_dev": 0.010833865812554817,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.00440245847362514,
      "avg_duration_per_task_ms": 697842.308080808,
      "avg_output_tokens_per_task": 7879.701459034793,
      "last_run_timestamp": "2026-10-31T16:02:18.363Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "ModelRun",
      "accuracy": 0.7219408075949366,
      "accuracy_std_dev": 0.025598999999999983,
      "run_count": 2,
      "total_questions": 395,
      "avg_cost_per_task_usd": 0.27939689620253166,
      "avg_duration_per_task_ms": 450249.90379746835,
      "avg_output_tokens_per_task": 45941.00253164557,
      "last_run_timestamp": "2026-08-15T08:16:15.061Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.5-haiku-20251001",
      "model_slug": "anthropic/claude-haiku-4.5",
      "model_name": "Anthropic: Claude Haiku 4.5",
      "provider_name": "Google",
      "accuracy": 0.7218013888888889,
      "accuracy_std_dev": 0.013975083045425636,
      "run_count": 36,
      "total_questions": 7128,
      "avg_cost_per_task_usd": 0.19767881116722782,
      "avg_duration_per_task_ms": 272001.3790684624,
      "avg_output_tokens_per_task": 36868.61924803592,
      "last_run_timestamp": "2026-10-31T08:29:35.324Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Wafer",
      "accuracy": 0.7210140000000002,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 92,
      "avg_cost_per_task_usd": 0.08365161956521738,
      "avg_duration_per_task_ms": 673616.0108695652,
      "avg_output_tokens_per_task": 34073.82608695652,
      "last_run_timestamp": "2026-08-15T15:24:07.556Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "DeepInfra",
      "accuracy": 0.7207904490979381,
      "accuracy_std_dev": 0.05082374714542744,
      "run_count": 16,
      "total_questions": 3104,
      "avg_cost_per_task_usd": 0.008789419136597938,
      "avg_duration_per_task_ms": 550071.2590206185,
      "avg_output_tokens_per_task": 26091.442332474227,
      "last_run_timestamp": "2026-10-31T16:52:55.644Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Makora",
      "accuracy": 0.7183742857142856,
      "accuracy_std_dev": 0.012394494661678109,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.01898589466089466,
      "avg_duration_per_task_ms": 501749.6746031746,
      "avg_output_tokens_per_task": 67465.82683982684,
      "last_run_timestamp": "2026-10-31T08:06:04.074Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-next-80b-a3b-instruct-2509",
      "model_slug": "qwen/qwen3-next-80b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Next 80B A3B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.7180135,
      "accuracy_std_dev": 0.002525166677667045,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.011038688131313132,
      "avg_duration_per_task_ms": 91885.46717171717,
      "avg_output_tokens_per_task": 10364.281565656565,
      "last_run_timestamp": "2026-10-03T16:59:33.949Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-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": 0.7176046841432225,
      "accuracy_std_dev": 0.014357802800219841,
      "run_count": 4,
      "total_questions": 782,
      "avg_cost_per_task_usd": 0.011155867007672635,
      "avg_duration_per_task_ms": 100550.45524296675,
      "avg_output_tokens_per_task": 10241.003836317135,
      "last_run_timestamp": "2026-10-03T16:43:17.105Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-next-80b-a3b-instruct-2509",
      "model_slug": "qwen/qwen3-next-80b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Next 80B A3B Instruct",
      "provider_name": "Alibaba",
      "accuracy": 0.715488,
      "accuracy_std_dev": 0.008584199351133475,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.007857320707070707,
      "avg_duration_per_task_ms": 80203.69570707071,
      "avg_output_tokens_per_task": 9977.103535353535,
      "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": "Novita",
      "accuracy": 0.7112795000000001,
      "accuracy_std_dev": 0.008459507683665782,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.008834092171717172,
      "avg_duration_per_task_ms": 129721.28282828283,
      "avg_output_tokens_per_task": 5726.907828282829,
      "last_run_timestamp": "2026-10-03T16:45:33.688Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-next-80b-a3b-instruct-2509",
      "model_slug": "qwen/qwen3-next-80b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Next 80B A3B Instruct",
      "provider_name": "Novita",
      "accuracy": 0.7104377500000001,
      "accuracy_std_dev": 0.0035712428071331226,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.015488117424242423,
      "avg_duration_per_task_ms": 97440.21590909091,
      "avg_output_tokens_per_task": 10244.169191919193,
      "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": "Alibaba",
      "accuracy": 0.7074915000000002,
      "accuracy_std_dev": 0.007657142629597565,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.0059589507575757575,
      "avg_duration_per_task_ms": 108794.86363636363,
      "avg_output_tokens_per_task": 5529.719696969697,
      "last_run_timestamp": "2026-10-03T17:02:24.058Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_slug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_name": "Qwen: Qwen3 VL 235B A22B Instruct",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.707071,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.008867828282828283,
      "avg_duration_per_task_ms": 116045.65151515152,
      "avg_output_tokens_per_task": 5749.39898989899,
      "last_run_timestamp": "2026-10-04T03:05:59.018Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-a22b-thinking-2507",
      "model_slug": "qwen/qwen3-235b-a22b-thinking-2507",
      "model_name": "Qwen: Qwen3 235B A22B Thinking 2507",
      "provider_name": "Novita",
      "accuracy": 0.7070704999999999,
      "accuracy_std_dev": 0.025252499999999956,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.06934866919191919,
      "avg_duration_per_task_ms": 1041055.3939393939,
      "avg_output_tokens_per_task": 23164.866161616163,
      "last_run_timestamp": "2026-08-22T17:45:06.645Z"
    },
    {
      "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": 0.7059485833333332,
      "accuracy_std_dev": 0.13670659769524554,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.615879143097643,
      "avg_duration_per_task_ms": 638649.8851010101,
      "avg_output_tokens_per_task": 40901.45538720539,
      "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": 0.7051066666666668,
      "accuracy_std_dev": 0.14363924551362076,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.619232220959596,
      "avg_duration_per_task_ms": 639502.4095117846,
      "avg_output_tokens_per_task": 41183.20328282828,
      "last_run_timestamp": "2026-10-31T08:28:17.671Z"
    },
    {
      "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": 0.7049665,
      "accuracy_std_dev": 0.017307340935279462,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.010932695707070707,
      "avg_duration_per_task_ms": 188922.1123737374,
      "avg_output_tokens_per_task": 5690.835858585859,
      "last_run_timestamp": "2026-10-03T16:59:57.847Z"
    },
    {
      "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": 0.704655008,
      "accuracy_std_dev": 0.13819566036718844,
      "run_count": 36,
      "total_questions": 7125,
      "avg_cost_per_task_usd": 0.6637001717894737,
      "avg_duration_per_task_ms": 604263.9247719299,
      "avg_output_tokens_per_task": 41325.50835087719,
      "last_run_timestamp": "2026-10-31T08:51:32.081Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5-nano-2025-08-07",
      "model_slug": "openai/gpt-5-nano",
      "model_name": "OpenAI: GPT-5 Nano",
      "provider_name": "auto-routing",
      "accuracy": 0.7045455,
      "accuracy_std_dev": 0.004208500000000004,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.01336774494949495,
      "avg_duration_per_task_ms": 231791.09848484848,
      "avg_output_tokens_per_task": 33323.15909090909,
      "last_run_timestamp": "2026-10-01T19:27:22.032Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "auto-routing",
      "accuracy": 0.7038606610169492,
      "accuracy_std_dev": 0.025828434379814764,
      "run_count": 11,
      "total_questions": 2124,
      "avg_cost_per_task_usd": 0.009122728342749528,
      "avg_duration_per_task_ms": 365582.73399246705,
      "avg_output_tokens_per_task": 45350.33474576271,
      "last_run_timestamp": "2026-10-31T08:03:09.796Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "anthropic/claude-4.6-sonnet-20260217",
      "model_slug": "anthropic/claude-sonnet-4.6",
      "model_name": "Anthropic: Claude Sonnet 4.6",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.703844,
      "accuracy_std_dev": 0.13064484273849722,
      "run_count": 12,
      "total_questions": 2376,
      "avg_cost_per_task_usd": 0.611304741161616,
      "avg_duration_per_task_ms": 637806.0361952862,
      "avg_output_tokens_per_task": 40663.66961279461,
      "last_run_timestamp": "2026-10-31T08:38:33.095Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-235b-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": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Venice",
      "accuracy": 0.7025463750000001,
      "accuracy_std_dev": 0.0578804880549946,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.13540828724747475,
      "avg_duration_per_task_ms": 664851.7664141414,
      "avg_output_tokens_per_task": 40688.72285353536,
      "last_run_timestamp": "2026-10-31T16:56:31.817Z"
    },
    {
      "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": 0.7024950168823861,
      "accuracy_std_dev": 0.09479570165959954,
      "run_count": 9,
      "total_questions": 1777,
      "avg_cost_per_task_usd": 0.011357701181767021,
      "avg_duration_per_task_ms": 821641.8733821047,
      "avg_output_tokens_per_task": 33243.20652785594,
      "last_run_timestamp": "2026-10-31T16:08:08.535Z"
    },
    {
      "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": 0.7024297297297296,
      "accuracy_std_dev": 0.13362786647407585,
      "run_count": 37,
      "total_questions": 7326,
      "avg_cost_per_task_usd": 0.6591401015561016,
      "avg_duration_per_task_ms": 440446.8645918646,
      "avg_output_tokens_per_task": 40987.61315861316,
      "last_run_timestamp": "2026-10-31T08:31:32.893Z"
    },
    {
      "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": 0.701915125,
      "accuracy_std_dev": 0.04741224100492801,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.015016424242424244,
      "avg_duration_per_task_ms": 909166.5378787878,
      "avg_output_tokens_per_task": 50028.55808080808,
      "last_run_timestamp": "2026-10-31T16:11:46.431Z"
    },
    {
      "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": 0.7003709888776543,
      "accuracy_std_dev": 0.00943615013869532,
      "run_count": 5,
      "total_questions": 989,
      "avg_cost_per_task_usd": 0.010341509605662285,
      "avg_duration_per_task_ms": 140930.19615773507,
      "avg_output_tokens_per_task": 9335.02628918099,
      "last_run_timestamp": "2026-10-08T22:40:37.311Z"
    },
    {
      "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": 0.6998505497757848,
      "accuracy_std_dev": 0.1222266014861715,
      "run_count": 8,
      "total_questions": 1115,
      "avg_cost_per_task_usd": 0.09317169596412556,
      "avg_duration_per_task_ms": 1222512.3695067265,
      "avg_output_tokens_per_task": 26526.149775784754,
      "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-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": 0.6973907500000001,
      "accuracy_std_dev": 0.0320390349837117,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.01086133207070707,
      "avg_duration_per_task_ms": 63422.920454545456,
      "avg_output_tokens_per_task": 9015.968434343435,
      "last_run_timestamp": "2026-10-03T16:36:16.895Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "DeepInfra",
      "accuracy": 0.6970787019475021,
      "accuracy_std_dev": 0.046119854611155975,
      "run_count": 12,
      "total_questions": 2362,
      "avg_cost_per_task_usd": 0.007317635478408128,
      "avg_duration_per_task_ms": 292695.22861981374,
      "avg_output_tokens_per_task": 40004.958086367486,
      "last_run_timestamp": "2026-10-31T08:15:24.314Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.6947253333333333,
      "accuracy_std_dev": 0.06031432281712934,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0065146380471380475,
      "avg_duration_per_task_ms": 125260.52525252526,
      "avg_output_tokens_per_task": 7227.368686868687,
      "last_run_timestamp": "2026-08-29T18:01:46.940Z"
    },
    {
      "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": 0.6922558,
      "accuracy_std_dev": 0.010774381251839934,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.00555330606060606,
      "avg_duration_per_task_ms": 393393.0888888889,
      "avg_output_tokens_per_task": 6125.930303030303,
      "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": 0.6907968888888889,
      "accuracy_std_dev": 0.011334983541079488,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.006490662738496071,
      "avg_duration_per_task_ms": 110618.54152637486,
      "avg_output_tokens_per_task": 15950.401234567902,
      "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": "anthropic/claude-4.6-sonnet-20260217",
      "model_slug": "anthropic/claude-sonnet-4.6",
      "model_name": "Anthropic: Claude Sonnet 4.6",
      "provider_name": "Anthropic",
      "accuracy": 0.6885522307692307,
      "accuracy_std_dev": 0.1455903473845781,
      "run_count": 13,
      "total_questions": 2574,
      "avg_cost_per_task_usd": 0.6175294277389277,
      "avg_duration_per_task_ms": 635458.5120435121,
      "avg_output_tokens_per_task": 41042.69852369852,
      "last_run_timestamp": "2026-10-31T08:22:47.508Z"
    },
    {
      "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": 0.6876397825763216,
      "accuracy_std_dev": 0.02033882284635741,
      "run_count": 7,
      "total_questions": 1343,
      "avg_cost_per_task_usd": 0.01665832092330603,
      "avg_duration_per_task_ms": 832849.0714817572,
      "avg_output_tokens_per_task": 83099.43037974683,
      "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-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-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "Venice",
      "accuracy": 0.6858604107485604,
      "accuracy_std_dev": 0.04695142144480293,
      "run_count": 8,
      "total_questions": 1563,
      "avg_cost_per_task_usd": 0.02905995585412668,
      "avg_duration_per_task_ms": 808566.8777991043,
      "avg_output_tokens_per_task": 18265.64043506078,
      "last_run_timestamp": "2026-10-31T16:23:19.099Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-5-nano-2025-08-07",
      "model_slug": "openai/gpt-5-nano",
      "model_name": "OpenAI: GPT-5 Nano",
      "provider_name": "OpenAI",
      "accuracy": 0.6843435000000001,
      "accuracy_std_dev": 0.005892499999999967,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.01356670707070707,
      "avg_duration_per_task_ms": 229761.702020202,
      "avg_output_tokens_per_task": 33820.56313131313,
      "last_run_timestamp": "2026-10-01T19:28:34.261Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_slug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_name": "Qwen: Qwen3 VL 235B A22B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.6843435,
      "accuracy_std_dev": 0.01913932645758466,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.009632915404040404,
      "avg_duration_per_task_ms": 272984.89015151514,
      "avg_output_tokens_per_task": 6630.618686868687,
      "last_run_timestamp": "2026-10-03T17:02:48.284Z"
    },
    {
      "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": 0.6816816949806949,
      "accuracy_std_dev": 0.019977202224895615,
      "run_count": 8,
      "total_questions": 1554,
      "avg_cost_per_task_usd": 0.002334259330759331,
      "avg_duration_per_task_ms": 214658.21685971686,
      "avg_output_tokens_per_task": 4092.794723294723,
      "last_run_timestamp": "2026-10-31T16:18:48.428Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "Darkbloom",
      "accuracy": 0.681502625,
      "accuracy_std_dev": 0.06815065878246794,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.007994105429292929,
      "avg_duration_per_task_ms": 1440482.744949495,
      "avg_output_tokens_per_task": 62597.993055555555,
      "last_run_timestamp": "2026-10-31T16:08:08.531Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-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": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Venice",
      "accuracy": 0.677734640625,
      "accuracy_std_dev": 0.0033374214597500216,
      "run_count": 3,
      "total_questions": 512,
      "avg_cost_per_task_usd": 0.04002998046875,
      "avg_duration_per_task_ms": 677472.7421875,
      "avg_output_tokens_per_task": 42379.095703125,
      "last_run_timestamp": "2026-08-29T18:51:07.939Z"
    },
    {
      "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": 0.6770081428571428,
      "accuracy_std_dev": 0.13047305606076964,
      "run_count": 7,
      "total_questions": 1386,
      "avg_cost_per_task_usd": 0.007556641414141414,
      "avg_duration_per_task_ms": 737595.6847041846,
      "avg_output_tokens_per_task": 26027.38023088023,
      "last_run_timestamp": "2026-10-31T16:45:39.993Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "Venice",
      "accuracy": 0.6764644057507988,
      "accuracy_std_dev": 0.06386519178597505,
      "run_count": 8,
      "total_questions": 1565,
      "avg_cost_per_task_usd": 0.05298015335463258,
      "avg_duration_per_task_ms": 1168436.861341853,
      "avg_output_tokens_per_task": 24293.87412140575,
      "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-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "DeepInfra",
      "accuracy": 0.6691934983520105,
      "accuracy_std_dev": 0.018592824882494627,
      "run_count": 9,
      "total_questions": 1517,
      "avg_cost_per_task_usd": 0.00669224719841793,
      "avg_duration_per_task_ms": 471626.4996704021,
      "avg_output_tokens_per_task": 47630.16677653263,
      "last_run_timestamp": "2026-10-31T16:03:21.483Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-chat-v3",
      "model_slug": "deepseek/deepseek-chat",
      "model_name": "DeepSeek: DeepSeek V3",
      "provider_name": "StreamLake",
      "accuracy": 0.6689113333333334,
      "accuracy_std_dev": 0.0067804647005611455,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0026409175084175086,
      "avg_duration_per_task_ms": 47935.90572390572,
      "avg_output_tokens_per_task": 2383.6902356902356,
      "last_run_timestamp": "2026-10-02T15:53:31.869Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": "CoreWeave",
      "accuracy": 0.667718875,
      "accuracy_std_dev": 0.01755485164731891,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.007828616792929293,
      "avg_duration_per_task_ms": 514528.57007575757,
      "avg_output_tokens_per_task": 60437.62941919192,
      "last_run_timestamp": "2026-10-31T16:08:31.061Z"
    },
    {
      "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": 0.6677188749999998,
      "accuracy_std_dev": 0.006886930456260996,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.00510659595959596,
      "avg_duration_per_task_ms": 104136.78345959596,
      "avg_output_tokens_per_task": 9617.265151515152,
      "last_run_timestamp": "2026-10-03T09:23:08.256Z"
    },
    {
      "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": 0.66726417562724,
      "accuracy_std_dev": 0.001755225531567576,
      "run_count": 3,
      "total_questions": 279,
      "avg_cost_per_task_usd": 0.004358928315412186,
      "avg_duration_per_task_ms": 566517.577060932,
      "avg_output_tokens_per_task": 29219.63799283154,
      "last_run_timestamp": "2026-08-29T17:24:11.557Z"
    },
    {
      "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": 0.667228,
      "accuracy_std_dev": 0.006780740569190578,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0024873282828282825,
      "avg_duration_per_task_ms": 87761.60606060606,
      "avg_output_tokens_per_task": 2304.691919191919,
      "last_run_timestamp": "2026-08-29T20:54:16.286Z"
    },
    {
      "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": 0.6672276666666667,
      "accuracy_std_dev": 0.006918527315998866,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.002891920875420875,
      "avg_duration_per_task_ms": 83806.60269360269,
      "avg_output_tokens_per_task": 3112.5656565656564,
      "last_run_timestamp": "2026-08-29T19:49:56.943Z"
    },
    {
      "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": 0.6670875,
      "accuracy_std_dev": 0.016207809136339183,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.005778919191919192,
      "avg_duration_per_task_ms": 411871.7727272727,
      "avg_output_tokens_per_task": 15840.212121212122,
      "last_run_timestamp": "2026-10-31T16:17:15.718Z"
    },
    {
      "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": 0.6666666666666666,
      "accuracy_std_dev": 0.020248820744812664,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.005717542087542088,
      "avg_duration_per_task_ms": 446868.9478114478,
      "avg_output_tokens_per_task": 16225.250841750842,
      "last_run_timestamp": "2026-08-29T18:40:21.422Z"
    },
    {
      "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": 0.666246,
      "accuracy_std_dev": 0.052274901764613566,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.01620974368686869,
      "avg_duration_per_task_ms": 334155.6628787879,
      "avg_output_tokens_per_task": 8982.006313131313,
      "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": "DigitalOcean",
      "accuracy": 0.665825,
      "accuracy_std_dev": 0.0075755555571852436,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.002092334595959596,
      "avg_duration_per_task_ms": 237098.96464646465,
      "avg_output_tokens_per_task": 3099.743686868687,
      "last_run_timestamp": "2026-10-05T20:44:48.611Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.6643520625000001,
      "accuracy_std_dev": 0.013584297200576622,
      "run_count": 16,
      "total_questions": 3168,
      "avg_cost_per_task_usd": 0.009127087752525253,
      "avg_duration_per_task_ms": 320481.77809343435,
      "avg_output_tokens_per_task": 60446.78566919192,
      "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-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": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "Parasail",
      "accuracy": 0.6626460482456141,
      "accuracy_std_dev": 0.07409874774620387,
      "run_count": 3,
      "total_questions": 456,
      "avg_cost_per_task_usd": 0.0063980043859649125,
      "avg_duration_per_task_ms": 453491.5789473684,
      "avg_output_tokens_per_task": 34112.30263157895,
      "last_run_timestamp": "2026-08-29T12:30:09.617Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Parasail",
      "accuracy": 0.6603534999999999,
      "accuracy_std_dev": 0.021669912920914115,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.008152684974747474,
      "avg_duration_per_task_ms": 795790.2424242424,
      "avg_output_tokens_per_task": 55529.520833333336,
      "last_run_timestamp": "2026-10-31T16:08:56.902Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.6-20251208",
      "model_slug": "z-ai/glm-4.6v",
      "model_name": "Z.ai: GLM 4.6V",
      "provider_name": "Z.AI",
      "accuracy": 0.659933,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.014984520202020202,
      "avg_duration_per_task_ms": 348987.1767676768,
      "avg_output_tokens_per_task": 16461.863636363636,
      "last_run_timestamp": "2026-10-01T08:28:22.621Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4.1-2025-04-14",
      "model_slug": "openai/gpt-4.1",
      "model_name": "OpenAI: GPT-4.1",
      "provider_name": "OpenAI",
      "accuracy": 0.6599326666666666,
      "accuracy_std_dev": 0.013255807590972711,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.015927558922558923,
      "avg_duration_per_task_ms": 14095.8367003367,
      "avg_output_tokens_per_task": 1794.7289562289561,
      "last_run_timestamp": "2026-10-02T23:17:48.072Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "DigitalOcean",
      "accuracy": 0.657360497461929,
      "accuracy_std_dev": 0.009259500000000032,
      "run_count": 2,
      "total_questions": 394,
      "avg_cost_per_task_usd": 0.04522630456852792,
      "avg_duration_per_task_ms": 3082748.9060913706,
      "avg_output_tokens_per_task": 28446.494923857867,
      "last_run_timestamp": "2026-08-22T18:19:05.094Z"
    },
    {
      "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": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "auto-routing",
      "accuracy": 0.654250875,
      "accuracy_std_dev": 0.13793927785482052,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.014789621212121212,
      "avg_duration_per_task_ms": 422898.68813131313,
      "avg_output_tokens_per_task": 12693.433080808081,
      "last_run_timestamp": "2026-10-31T16:01:53.629Z"
    },
    {
      "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": 0.6531986666666667,
      "accuracy_std_dev": 0.009013858120817182,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0030076632996633,
      "avg_duration_per_task_ms": 41226.700336700334,
      "avg_output_tokens_per_task": 2842.754208754209,
      "last_run_timestamp": "2026-08-29T19:51:51.198Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4.1-mini-2025-04-14",
      "model_slug": "openai/gpt-4.1-mini",
      "model_name": "OpenAI: GPT-4.1 Mini",
      "provider_name": "auto-routing",
      "accuracy": 0.652357,
      "accuracy_std_dev": 0.01262599999999997,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.004161952020202021,
      "avg_duration_per_task_ms": 27160.11111111111,
      "avg_output_tokens_per_task": 2408.8232323232323,
      "last_run_timestamp": "2026-10-01T19:12:31.553Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "ibm-granite/granite-4.2-8b-20260831",
      "model_slug": "ibm-granite/granite-4.2-8b",
      "model_name": "IBM: Granite 4.2 8B",
      "provider_name": "DeepInfra",
      "accuracy": 0.6523569999999999,
      "accuracy_std_dev": 0.0025249999999999995,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.01652235606060606,
      "avg_duration_per_task_ms": 1917916.494949495,
      "avg_output_tokens_per_task": 65966.15909090909,
      "last_run_timestamp": "2026-10-09T00:15:37.020Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-r1-0528",
      "model_slug": "deepseek/deepseek-r1-0528",
      "model_name": "DeepSeek: R1 0528",
      "provider_name": "Novita",
      "accuracy": 0.651515,
      "accuracy_std_dev": 0.1905456270048375,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.058297676767676776,
      "avg_duration_per_task_ms": 1098090.5791245792,
      "avg_output_tokens_per_task": 23174.20202020202,
      "last_run_timestamp": "2026-10-17T16:03:44.041Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "auto-routing",
      "accuracy": 0.6508838749999999,
      "accuracy_std_dev": 0.023523002829345063,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.005865840909090909,
      "avg_duration_per_task_ms": 657820.1502525252,
      "avg_output_tokens_per_task": 42824.80997474748,
      "last_run_timestamp": "2026-10-31T16:01:53.622Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4.1-mini-2025-04-14",
      "model_slug": "openai/gpt-4.1-mini",
      "model_name": "OpenAI: GPT-4.1 Mini",
      "provider_name": "OpenAI",
      "accuracy": 0.6492706666666667,
      "accuracy_std_dev": 0.010498645193018418,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0041127255892255896,
      "avg_duration_per_task_ms": 24949.132996632998,
      "avg_output_tokens_per_task": 2375.9343434343436,
      "last_run_timestamp": "2026-10-02T23:15:53.868Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-chat-v3-0324",
      "model_slug": "deepseek/deepseek-chat-v3-0324",
      "model_name": "DeepSeek: DeepSeek V3 0324",
      "provider_name": "GMICloud",
      "accuracy": 0.6484773807106599,
      "accuracy_std_dev": 0.0054344006339245585,
      "run_count": 4,
      "total_questions": 788,
      "avg_cost_per_task_usd": 0.00290907614213198,
      "avg_duration_per_task_ms": 58072.35659898477,
      "avg_output_tokens_per_task": 2371.812182741117,
      "last_run_timestamp": "2026-10-27T00:30:02.893Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Google",
      "accuracy": 0.6479142515802782,
      "accuracy_std_dev": 0.10866298734612617,
      "run_count": 8,
      "total_questions": 1582,
      "avg_cost_per_task_usd": 0.009831766750948167,
      "avg_duration_per_task_ms": 374447.82616940583,
      "avg_output_tokens_per_task": 27636.46776232617,
      "last_run_timestamp": "2026-10-31T16:12:07.129Z"
    },
    {
      "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": 0.6478676666666666,
      "accuracy_std_dev": 0.008199376656524262,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0038777457912457913,
      "avg_duration_per_task_ms": 63026.43939393939,
      "avg_output_tokens_per_task": 2350.077441077441,
      "last_run_timestamp": "2026-08-29T16:10:45.295Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "auto-routing",
      "accuracy": 0.6477275,
      "accuracy_std_dev": 0.005503645223849386,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.0027929646464646468,
      "avg_duration_per_task_ms": 157689.33585858587,
      "avg_output_tokens_per_task": 9994.479797979799,
      "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": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "AtlasCloud",
      "accuracy": 0.6467880041128085,
      "accuracy_std_dev": 0.20710201321517321,
      "run_count": 9,
      "total_questions": 1702,
      "avg_cost_per_task_usd": 0.011494767920094007,
      "avg_duration_per_task_ms": 300785.6809635723,
      "avg_output_tokens_per_task": 13387.064629847238,
      "last_run_timestamp": "2026-10-17T08:17:57.529Z"
    },
    {
      "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": 0.6453423333333334,
      "accuracy_std_dev": 0.010677151752951496,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0023666683501683498,
      "avg_duration_per_task_ms": 54700.6936026936,
      "avg_output_tokens_per_task": 2768.026936026936,
      "last_run_timestamp": "2026-08-29T16:11:23.388Z"
    },
    {
      "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": 0.6436586666666667,
      "accuracy_std_dev": 0.0015872190215034077,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.002485070707070707,
      "avg_duration_per_task_ms": 52022.1632996633,
      "avg_output_tokens_per_task": 2681.813131313131,
      "last_run_timestamp": "2026-08-29T17:03:33.297Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "Ambient",
      "accuracy": 0.641414,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.00284089898989899,
      "avg_duration_per_task_ms": 209352.64646464647,
      "avg_output_tokens_per_task": 9266.040404040405,
      "last_run_timestamp": "2026-10-09T15:00:06.492Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Darkbloom",
      "accuracy": 0.64025675,
      "accuracy_std_dev": 0.01324879906585875,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.001088048611111111,
      "avg_duration_per_task_ms": 331146.75505050505,
      "avg_output_tokens_per_task": 11572.473484848484,
      "last_run_timestamp": "2026-10-31T16:31:13.438Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.640151625,
      "accuracy_std_dev": 0.013825012268868864,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.007202421717171718,
      "avg_duration_per_task_ms": 241781.60606060605,
      "avg_output_tokens_per_task": 11537.880681818182,
      "last_run_timestamp": "2026-10-03T10:20:18.247Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-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": "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": 0.63846825,
      "accuracy_std_dev": 0.015215861565731332,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.0046192222222222225,
      "avg_duration_per_task_ms": 77763.38257575757,
      "avg_output_tokens_per_task": 8686.127525252525,
      "last_run_timestamp": "2026-10-03T16:51:25.541Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4.1-2025-04-14",
      "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Phala",
      "accuracy": 0.6361454238683127,
      "accuracy_std_dev": 0.218773277941707,
      "run_count": 8,
      "total_questions": 1458,
      "avg_cost_per_task_usd": 0.00909512757201646,
      "avg_duration_per_task_ms": 157677.57133058985,
      "avg_output_tokens_per_task": 14981.154320987655,
      "last_run_timestamp": "2026-10-31T16:54:17.849Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "Darkbloom",
      "accuracy": 0.6351663037907099,
      "accuracy_std_dev": 0.019472921033065377,
      "run_count": 10,
      "total_questions": 1873,
      "avg_cost_per_task_usd": 0.005367699412706888,
      "avg_duration_per_task_ms": 807188.147357181,
      "avg_output_tokens_per_task": 35119.06513614522,
      "last_run_timestamp": "2026-10-31T08:20:09.538Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.6350172000000001,
      "accuracy_std_dev": 0.013898819394466565,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.0016177010101010101,
      "avg_duration_per_task_ms": 77808.62626262626,
      "avg_output_tokens_per_task": 8177.061616161616,
      "last_run_timestamp": "2026-10-03T16:47:32.559Z"
    },
    {
      "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": 0.63320725,
      "accuracy_std_dev": 0.035387886365781995,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.006817837121212121,
      "avg_duration_per_task_ms": 104450.63320707071,
      "avg_output_tokens_per_task": 9507.679924242424,
      "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": 0.6328369143389199,
      "accuracy_std_dev": 0.012371127066233214,
      "run_count": 3,
      "total_questions": 537,
      "avg_cost_per_task_usd": 0.00797138175046555,
      "avg_duration_per_task_ms": 28551.49720670391,
      "avg_output_tokens_per_task": 13203.731843575419,
      "last_run_timestamp": "2026-08-29T16:23:45.123Z"
    },
    {
      "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": "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": 0.6314185,
      "accuracy_std_dev": 0.01109086049411856,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.019078623737373738,
      "avg_duration_per_task_ms": 506361.1357323232,
      "avg_output_tokens_per_task": 95407.51325757576,
      "last_run_timestamp": "2026-10-03T17:08:02.072Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "DekaLLM",
      "accuracy": 0.6306148156028369,
      "accuracy_std_dev": 0.05281625698168657,
      "run_count": 3,
      "total_questions": 282,
      "avg_cost_per_task_usd": 0.0037584255319148938,
      "avg_duration_per_task_ms": 208452.59574468085,
      "avg_output_tokens_per_task": 12308.468085106382,
      "last_run_timestamp": "2026-10-23T09:29:14.171Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "SiliconFlow",
      "accuracy": 0.6305678657378088,
      "accuracy_std_dev": 0.007960405916746635,
      "run_count": 8,
      "total_questions": 1579,
      "avg_cost_per_task_usd": 0.0018545345155161492,
      "avg_duration_per_task_ms": 230896.0607979734,
      "avg_output_tokens_per_task": 10115.67764407853,
      "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": "SiliconFlow",
      "accuracy": 0.62962975,
      "accuracy_std_dev": 0.007985647073813095,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.0027725871212121215,
      "avg_duration_per_task_ms": 281279.59217171714,
      "avg_output_tokens_per_task": 9027.892676767677,
      "last_run_timestamp": "2026-10-03T17:03:50.406Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "Darkbloom",
      "accuracy": 0.6295420719476744,
      "accuracy_std_dev": 0.1792338661514529,
      "run_count": 14,
      "total_questions": 2752,
      "avg_cost_per_task_usd": 0.02676043095930232,
      "avg_duration_per_task_ms": 725458.0534156977,
      "avg_output_tokens_per_task": 37903.851744186046,
      "last_run_timestamp": "2026-10-31T16:08:33.702Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "Venice",
      "accuracy": 0.6292411856287424,
      "accuracy_std_dev": 0.0117815,
      "run_count": 2,
      "total_questions": 334,
      "avg_cost_per_task_usd": 0.0071094161676646716,
      "avg_duration_per_task_ms": 112258.81736526947,
      "avg_output_tokens_per_task": 28232.077844311378,
      "last_run_timestamp": "2026-08-15T08:02:44.624Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.629208875,
      "accuracy_std_dev": 0.022106924353228682,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.017882390151515153,
      "avg_duration_per_task_ms": 634871.8630050505,
      "avg_output_tokens_per_task": 87147.22348484848,
      "last_run_timestamp": "2026-10-03T17:40:51.448Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3-nano-30b-a3b",
      "model_slug": "nvidia/nemotron-3-nano-30b-a3b",
      "model_name": "NVIDIA: Nemotron 3 Nano 30B A3B",
      "provider_name": "Nebius",
      "accuracy": 0.627104375,
      "accuracy_std_dev": 0.008333089267155095,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.024497946969696968,
      "avg_duration_per_task_ms": 1060409.1414141415,
      "avg_output_tokens_per_task": 101874.34027777778,
      "last_run_timestamp": "2026-10-23T09:03:34.804Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "CoreWeave",
      "accuracy": 0.6240183333333333,
      "accuracy_std_dev": 0.009654676391377534,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0036675572390572394,
      "avg_duration_per_task_ms": 146319.07912457912,
      "avg_output_tokens_per_task": 12576.434343434343,
      "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": 0.6236870803858521,
      "accuracy_std_dev": 0.010127786036444495,
      "run_count": 8,
      "total_questions": 1555,
      "avg_cost_per_task_usd": 0.016994219935691317,
      "avg_duration_per_task_ms": 65807.6733118971,
      "avg_output_tokens_per_task": 56415.3536977492,
      "last_run_timestamp": "2026-10-31T16:52:47.514Z"
    },
    {
      "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": 0.6232324,
      "accuracy_std_dev": 0.01591442443319895,
      "run_count": 5,
      "total_questions": 990,
      "avg_cost_per_task_usd": 0.004449116161616162,
      "avg_duration_per_task_ms": 258518.11111111112,
      "avg_output_tokens_per_task": 14559.569696969696,
      "last_run_timestamp": "2026-10-23T08:08:11.470Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-chat-v3",
      "model_slug": "deepseek/deepseek-chat",
      "model_name": "DeepSeek: DeepSeek V3",
      "provider_name": "auto-routing",
      "accuracy": 0.6212125000000001,
      "accuracy_std_dev": 0.011784500000000031,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.0023548611111111114,
      "avg_duration_per_task_ms": 64938.90151515151,
      "avg_output_tokens_per_task": 2165.318181818182,
      "last_run_timestamp": "2026-10-01T19:21:21.945Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.621212,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.00552190909090909,
      "avg_duration_per_task_ms": 174055.64141414143,
      "avg_output_tokens_per_task": 10415.944444444445,
      "last_run_timestamp": "2026-10-04T03:09:40.101Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Fireworks",
      "accuracy": 0.6182949425287356,
      "accuracy_std_dev": 0.0013475000000000015,
      "run_count": 2,
      "total_questions": 348,
      "avg_cost_per_task_usd": 0.011105244252873564,
      "avg_duration_per_task_ms": 345775.4454022989,
      "avg_output_tokens_per_task": 38606.68103448276,
      "last_run_timestamp": "2026-08-22T17:46:40.285Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-32b-04-28",
      "model_slug": "qwen/qwen3-32b",
      "model_name": "Qwen: Qwen3 32B",
      "provider_name": "Nebius",
      "accuracy": 0.6150393333333333,
      "accuracy_std_dev": 0.00649574344389378,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.005692112794612795,
      "avg_duration_per_task_ms": 660625.7861952862,
      "avg_output_tokens_per_task": 18940.488215488214,
      "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-vl-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 VL 30B A3B Instruct",
      "provider_name": "SiliconFlow",
      "accuracy": 0.6134706293436293,
      "accuracy_std_dev": 0.013891280772484577,
      "run_count": 4,
      "total_questions": 777,
      "avg_cost_per_task_usd": 0.01592537323037323,
      "avg_duration_per_task_ms": 820331.4504504504,
      "avg_output_tokens_per_task": 16891.465894465895,
      "last_run_timestamp": "2026-10-03T11:06:38.068Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-14b-04-28",
      "model_slug": "qwen/qwen3-14b",
      "model_name": "Qwen: Qwen3 14B",
      "provider_name": "DeepInfra",
      "accuracy": 0.6123375670628184,
      "accuracy_std_dev": 0.008273645159312946,
      "run_count": 3,
      "total_questions": 589,
      "avg_cost_per_task_usd": 0.003321100169779287,
      "avg_duration_per_task_ms": 426798.44821731746,
      "avg_output_tokens_per_task": 13430.818336162989,
      "last_run_timestamp": "2026-10-08T22:16:42.056Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "Google",
      "accuracy": 0.6121632499999999,
      "accuracy_std_dev": 0.004096520863794059,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.004063032828282829,
      "avg_duration_per_task_ms": 22356.58080808081,
      "avg_output_tokens_per_task": 2158.7285353535353,
      "last_run_timestamp": "2026-10-03T17:03:50.598Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-30b-a3b-04-28",
      "model_slug": "qwen/qwen3-30b-a3b",
      "model_name": "Qwen: Qwen3 30B A3B",
      "provider_name": "DeepInfra",
      "accuracy": 0.609428,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.00778269696969697,
      "avg_duration_per_task_ms": 302947.1666666667,
      "avg_output_tokens_per_task": 15369.68686868687,
      "last_run_timestamp": "2026-10-08T22:57:53.933Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "Novita",
      "accuracy": 0.6077442500000001,
      "accuracy_std_dev": 0.01208155948905189,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.003638679292929293,
      "avg_duration_per_task_ms": 32468.945707070707,
      "avg_output_tokens_per_task": 2148.3560606060605,
      "last_run_timestamp": "2026-10-03T17:03:29.436Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "auto-routing",
      "accuracy": 0.6069025,
      "accuracy_std_dev": 0.006127792118699825,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.0035186742424242424,
      "avg_duration_per_task_ms": 45117.36994949495,
      "avg_output_tokens_per_task": 2127.872474747475,
      "last_run_timestamp": "2026-10-03T16:57:23.266Z"
    },
    {
      "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/qwen3-32b-04-28",
      "model_slug": "qwen/qwen3-32b",
      "model_name": "Qwen: Qwen3 32B",
      "provider_name": "auto-routing",
      "accuracy": 0.6043773333333333,
      "accuracy_std_dev": 0.010910481270574418,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.008336993265993266,
      "avg_duration_per_task_ms": 317962.5740740741,
      "avg_output_tokens_per_task": 17357.656565656565,
      "last_run_timestamp": "2026-08-29T19:52:22.642Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 VL 30B A3B Instruct",
      "provider_name": "DeepInfra",
      "accuracy": 0.6037849337670017,
      "accuracy_std_dev": 0.011765991415941101,
      "run_count": 9,
      "total_questions": 1691,
      "avg_cost_per_task_usd": 0.008547784742755766,
      "avg_duration_per_task_ms": 560925.4151389711,
      "avg_output_tokens_per_task": 14044.128326434062,
      "last_run_timestamp": "2026-10-08T22:50:36.237Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "DeepInfra",
      "accuracy": 0.6026385750873108,
      "accuracy_std_dev": 0.021655756090240754,
      "run_count": 5,
      "total_questions": 859,
      "avg_cost_per_task_usd": 0.0021882596041909196,
      "avg_duration_per_task_ms": 53677.2398137369,
      "avg_output_tokens_per_task": 2050.151338766007,
      "last_run_timestamp": "2026-10-08T23:15:07.689Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "Alibaba",
      "accuracy": 0.6010102499999999,
      "accuracy_std_dev": 0.0035714785016152262,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.011003395202020202,
      "avg_duration_per_task_ms": 31387.02398989899,
      "avg_output_tokens_per_task": 2097.1805555555557,
      "last_run_timestamp": "2026-10-03T17:43:44.972Z"
    },
    {
      "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": 0.6010099999999999,
      "accuracy_std_dev": 0.006733999999999962,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.04278589646464647,
      "avg_duration_per_task_ms": 337432.404040404,
      "avg_output_tokens_per_task": 21288.575757575756,
      "last_run_timestamp": "2026-08-22T10:29:25.897Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Io Net",
      "accuracy": 0.600457,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 146,
      "avg_cost_per_task_usd": 0.06895466438356164,
      "avg_duration_per_task_ms": 858101.3082191781,
      "avg_output_tokens_per_task": 24023.73287671233,
      "last_run_timestamp": "2026-08-15T09:29:41.209Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Nebius",
      "accuracy": 0.599327,
      "accuracy_std_dev": 0.0033669999999999534,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.014991068181818182,
      "avg_duration_per_task_ms": 284433.01515151514,
      "avg_output_tokens_per_task": 12302.075757575758,
      "last_run_timestamp": "2026-08-15T15:26:03.892Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.599327,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.012440297979797978,
      "avg_duration_per_task_ms": 230792.08080808082,
      "avg_output_tokens_per_task": 41242.71717171717,
      "last_run_timestamp": "2026-10-04T03:37:16.537Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-14b-04-28",
      "model_slug": "qwen/qwen3-14b",
      "model_name": "Qwen: Qwen3 14B",
      "provider_name": "NextBit",
      "accuracy": 0.5978837142857143,
      "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": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "Darkbloom",
      "accuracy": 0.5972300452232899,
      "accuracy_std_dev": 0.06739364692023929,
      "run_count": 9,
      "total_questions": 1769,
      "avg_cost_per_task_usd": 0.012534433578292821,
      "avg_duration_per_task_ms": 1310890.6302996043,
      "avg_output_tokens_per_task": 57642.5850763143,
      "last_run_timestamp": "2026-10-31T16:05:14.631Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "Venice",
      "accuracy": 0.5972225000000001,
      "accuracy_std_dev": 0.00323265575804168,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.0032439330808080808,
      "avg_duration_per_task_ms": 69237.93308080808,
      "avg_output_tokens_per_task": 2070.3876262626263,
      "last_run_timestamp": "2026-10-03T16:59:44.383Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-2.5-flash-lite",
      "model_slug": "google/gemini-2.5-flash-lite",
      "model_name": "Google: Gemini 2.5 Flash Lite",
      "provider_name": "Google AI Studio",
      "accuracy": 0.5963805,
      "accuracy_std_dev": 0.0012624999999999997,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.01735105808080808,
      "avg_duration_per_task_ms": 112232.29292929293,
      "avg_output_tokens_per_task": 43198.833333333336,
      "last_run_timestamp": "2026-10-01T19:14:48.696Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": 0.5955725338441891,
      "accuracy_std_dev": 0.008950333597134817,
      "run_count": 8,
      "total_questions": 1566,
      "avg_cost_per_task_usd": 0.009035680076628352,
      "avg_duration_per_task_ms": 471811.50638569606,
      "avg_output_tokens_per_task": 44984.78607918263,
      "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": 0.5892258750000001,
      "accuracy_std_dev": 0.008204338493100791,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.01031350505050505,
      "avg_duration_per_task_ms": 10608.848484848484,
      "avg_output_tokens_per_task": 13296.869318181818,
      "last_run_timestamp": "2026-10-31T16:21:16.340Z"
    },
    {
      "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": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Novita",
      "accuracy": 0.58585875,
      "accuracy_std_dev": 0.007241239789393828,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.006404845959595959,
      "avg_duration_per_task_ms": 643938.6085858586,
      "avg_output_tokens_per_task": 42440.19065656565,
      "last_run_timestamp": "2026-10-03T16:12:07.492Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "Mancer 2",
      "accuracy": 0.5851989731343283,
      "accuracy_std_dev": 0.25257942345483236,
      "run_count": 7,
      "total_questions": 1340,
      "avg_cost_per_task_usd": 0.0922404328358209,
      "avg_duration_per_task_ms": 1890268.2298507462,
      "avg_output_tokens_per_task": 40999.71268656717,
      "last_run_timestamp": "2026-10-31T16:02:18.348Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "Fireworks",
      "accuracy": 0.5810130708860759,
      "accuracy_std_dev": 0.012019500000000016,
      "run_count": 2,
      "total_questions": 395,
      "avg_cost_per_task_usd": 0.01962629113924051,
      "avg_duration_per_task_ms": 376708.9949367089,
      "avg_output_tokens_per_task": 97953.7417721519,
      "last_run_timestamp": "2026-10-31T08:09:25.783Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.7-flash-20260119",
      "model_slug": "z-ai/glm-4.7-flash",
      "model_name": "Z.ai: GLM 4.7 Flash",
      "provider_name": "Cloudflare",
      "accuracy": 0.5803875000000001,
      "accuracy_std_dev": 0.04095041650887082,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.02901417803030303,
      "avg_duration_per_task_ms": 1378907.4116161617,
      "avg_output_tokens_per_task": 72425.33712121213,
      "last_run_timestamp": "2026-10-13T21:28:41.390Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.7-flash-20260119",
      "model_slug": "z-ai/glm-4.7-flash",
      "model_name": "Z.ai: GLM 4.7 Flash",
      "provider_name": "Novita",
      "accuracy": 0.579686,
      "accuracy_std_dev": 0.008054024625407281,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.02485565488215488,
      "avg_duration_per_task_ms": 1088356.1313131314,
      "avg_output_tokens_per_task": 66893.9074074074,
      "last_run_timestamp": "2026-08-30T00:32:50.853Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Google",
      "accuracy": 0.5774414999999999,
      "accuracy_std_dev": 0.005050500000000013,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.010289136363636365,
      "avg_duration_per_task_ms": 331432.74242424243,
      "avg_output_tokens_per_task": 40883.856060606064,
      "last_run_timestamp": "2026-10-17T16:06:05.579Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-chat-v3-0324",
      "model_slug": "deepseek/deepseek-chat-v3-0324",
      "model_name": "DeepSeek: DeepSeek V3 0324",
      "provider_name": "DeepInfra",
      "accuracy": 0.5764006434634974,
      "accuracy_std_dev": 0.0627820184359255,
      "run_count": 3,
      "total_questions": 589,
      "avg_cost_per_task_usd": 0.003614684210526316,
      "avg_duration_per_task_ms": 138004.24957555177,
      "avg_output_tokens_per_task": 3859.125636672326,
      "last_run_timestamp": "2026-08-29T20:34:10.117Z"
    },
    {
      "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": 0.5702345561377246,
      "accuracy_std_dev": 0.07436004550317579,
      "run_count": 8,
      "total_questions": 1336,
      "avg_cost_per_task_usd": 0.06098996482035929,
      "avg_duration_per_task_ms": 369346.24326347304,
      "avg_output_tokens_per_task": 23120.943113772457,
      "last_run_timestamp": "2026-10-31T16:21:16.340Z"
    },
    {
      "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": 0.5628510000000001,
      "accuracy_std_dev": 0.07456152319170169,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.005432700336700336,
      "avg_duration_per_task_ms": 118723.56734006734,
      "avg_output_tokens_per_task": 4911.058922558923,
      "last_run_timestamp": "2026-08-29T20:57:58.443Z"
    },
    {
      "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": 0.55681825,
      "accuracy_std_dev": 0.011441432698202594,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.004685652777777778,
      "avg_duration_per_task_ms": 548781.2613636364,
      "avg_output_tokens_per_task": 16501.491161616163,
      "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": 0.5514771309523809,
      "accuracy_std_dev": 0.02267678126927133,
      "run_count": 8,
      "total_questions": 1512,
      "avg_cost_per_task_usd": 0.005655453042328042,
      "avg_duration_per_task_ms": 802376.1481481482,
      "avg_output_tokens_per_task": 14488.589285714286,
      "last_run_timestamp": "2026-11-04T18:37:19.993Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "DeepInfra",
      "accuracy": 0.5420490229357798,
      "accuracy_std_dev": 0.06975384672944968,
      "run_count": 3,
      "total_questions": 218,
      "avg_cost_per_task_usd": 0.03485861009174312,
      "avg_duration_per_task_ms": 397279.0229357798,
      "avg_output_tokens_per_task": 16660.94495412844,
      "last_run_timestamp": "2026-08-29T19:07:49.385Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "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": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Coder 30B A3B Instruct",
      "provider_name": "DigitalOcean",
      "accuracy": 0.53367,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.004812772727272727,
      "avg_duration_per_task_ms": 168981.19191919192,
      "avg_output_tokens_per_task": 2773.419191919192,
      "last_run_timestamp": "2026-08-15T09:11:29.276Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4o-2024-08-06",
      "model_slug": "openai/gpt-4o-2024-08-06",
      "model_name": "OpenAI: GPT-4o (2024-08-06)",
      "provider_name": "OpenAI",
      "accuracy": 0.5280586666666666,
      "accuracy_std_dev": 0.004827144451490511,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.01799902861952862,
      "avg_duration_per_task_ms": 14939.786195286195,
      "avg_output_tokens_per_task": 1607.080808080808,
      "last_run_timestamp": "2026-10-02T23:16:52.382Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4o",
      "model_slug": "openai/gpt-4o",
      "model_name": "OpenAI: GPT-4o",
      "provider_name": "OpenAI",
      "accuracy": 0.5266555,
      "accuracy_std_dev": 0.023895888382383546,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.01547567760942761,
      "avg_duration_per_task_ms": 13716.89898989899,
      "avg_output_tokens_per_task": 1354.1801346801346,
      "last_run_timestamp": "2026-10-02T23:17:37.980Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.7-flash-20260119",
      "model_slug": "z-ai/glm-4.7-flash",
      "model_name": "Z.ai: GLM 4.7 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.522278736040609,
      "accuracy_std_dev": 0.00921274336027126,
      "run_count": 3,
      "total_questions": 591,
      "avg_cost_per_task_usd": 0.016910294416243657,
      "avg_duration_per_task_ms": 664622.7783417936,
      "avg_output_tokens_per_task": 43236.38747884941,
      "last_run_timestamp": "2026-08-30T00:11:29.373Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Coder 30B A3B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.521886,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.004345590909090909,
      "avg_duration_per_task_ms": 107214.82828282828,
      "avg_output_tokens_per_task": 2763.3333333333335,
      "last_run_timestamp": "2026-08-15T08:46:42.897Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "DigitalOcean",
      "accuracy": 0.5216982222222223,
      "accuracy_std_dev": 0.1325981414523495,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.003293574635241302,
      "avg_duration_per_task_ms": 195367.39393939395,
      "avg_output_tokens_per_task": 7908.62962962963,
      "last_run_timestamp": "2026-10-31T16:10:36.257Z"
    },
    {
      "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": "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": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Coder 30B A3B Instruct",
      "provider_name": "Alibaba",
      "accuracy": 0.506734,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.004105035353535354,
      "avg_duration_per_task_ms": 27074.26767676768,
      "avg_output_tokens_per_task": 2644.373737373737,
      "last_run_timestamp": "2026-08-15T09:25:42.974Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Coder 30B A3B Instruct",
      "provider_name": "SiliconFlow",
      "accuracy": 0.506734,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0008276868686868686,
      "avg_duration_per_task_ms": 96676.70202020202,
      "avg_output_tokens_per_task": 2752.9141414141413,
      "last_run_timestamp": "2026-08-15T08:49:03.156Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4.1-nano-2025-04-14",
      "model_slug": "openai/gpt-4.1-nano",
      "model_name": "OpenAI: GPT-4.1 Nano",
      "provider_name": "OpenAI",
      "accuracy": 0.5050510000000001,
      "accuracy_std_dev": 0.002749143987983661,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0008273518518518518,
      "avg_duration_per_task_ms": 18715.93265993266,
      "avg_output_tokens_per_task": 1873.8602693602693,
      "last_run_timestamp": "2026-10-02T23:16:50.042Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "inclusionai/ling-3.0-flash-fin-20260827",
      "model_slug": "inclusionai/ling-3.0-flash-fin",
      "model_name": "inclusionAI: Ling 3.0 Flash Fin",
      "provider_name": "auto-routing",
      "accuracy": 0.505051,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.009755954545454545,
      "avg_duration_per_task_ms": 700759.0555555555,
      "avg_output_tokens_per_task": 64762.606060606064,
      "last_run_timestamp": "2026-10-01T08:24:17.043Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "openai/gpt-4.1-nano-2025-04-14",
      "model_slug": "openai/gpt-4.1-nano",
      "model_name": "OpenAI: GPT-4.1 Nano",
      "provider_name": "Azure",
      "accuracy": 0.4995792500000001,
      "accuracy_std_dev": 0.02025051336849268,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.0008581388888888889,
      "avg_duration_per_task_ms": 12992.723484848484,
      "avg_output_tokens_per_task": 1853.0088383838383,
      "last_run_timestamp": "2026-10-01T19:12:29.219Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "Groq",
      "accuracy": 0.49831649999999994,
      "accuracy_std_dev": 0.012284988196982519,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.0023916161616161615,
      "avg_duration_per_task_ms": 9167.109217171717,
      "avg_output_tokens_per_task": 2380.501262626263,
      "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": "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": 0.490313482051282,
      "accuracy_std_dev": 0.025671764346586447,
      "run_count": 3,
      "total_questions": 585,
      "avg_cost_per_task_usd": 0.010895215384615384,
      "avg_duration_per_task_ms": 415746.58461538464,
      "avg_output_tokens_per_task": 27278.131623931626,
      "last_run_timestamp": "2026-08-30T00:23:23.003Z"
    },
    {
      "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": 0.486532,
      "accuracy_std_dev": 0.021886000000000017,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.025879300505050506,
      "avg_duration_per_task_ms": 4030738.6136363638,
      "avg_output_tokens_per_task": 21539.305555555555,
      "last_run_timestamp": "2026-08-22T17:37:26.215Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "AkashML",
      "accuracy": 0.4851015826366559,
      "accuracy_std_dev": 0.011963524563621507,
      "run_count": 16,
      "total_questions": 3110,
      "avg_cost_per_task_usd": 0.0013306218649517683,
      "avg_duration_per_task_ms": 59520.47717041801,
      "avg_output_tokens_per_task": 2348.433118971061,
      "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": "CoreWeave",
      "accuracy": 0.483796,
      "accuracy_std_dev": 0.011661703413309726,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.0025953238636363635,
      "avg_duration_per_task_ms": 48652.757575757576,
      "avg_output_tokens_per_task": 2847.4810606060605,
      "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": "SambaNova",
      "accuracy": 0.48190212499999996,
      "accuracy_std_dev": 0.010369324127414242,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.002355522095959596,
      "avg_duration_per_task_ms": 15986.89898989899,
      "avg_output_tokens_per_task": 2214.277777777778,
      "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": 0.47966760309278356,
      "accuracy_std_dev": 0.006495603179587451,
      "run_count": 3,
      "total_questions": 582,
      "avg_cost_per_task_usd": 0.01110826116838488,
      "avg_duration_per_task_ms": 372116.1821305842,
      "avg_output_tokens_per_task": 27698.08591065292,
      "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": 0.47706212499999995,
      "accuracy_std_dev": 0.013881069307131027,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.001330816919191919,
      "avg_duration_per_task_ms": 45550.52525252525,
      "avg_output_tokens_per_task": 2388.7910353535353,
      "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": 0.47579937499999997,
      "accuracy_std_dev": 0.009595087765850562,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.0013503068181818184,
      "avg_duration_per_task_ms": 81673.49116161616,
      "avg_output_tokens_per_task": 2351.566919191919,
      "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": 0.475589,
      "accuracy_std_dev": 0.006204488912607278,
      "run_count": 6,
      "total_questions": 1188,
      "avg_cost_per_task_usd": 0.003132998316498317,
      "avg_duration_per_task_ms": 67179.39141414141,
      "avg_output_tokens_per_task": 2206.4680134680134,
      "last_run_timestamp": "2026-10-17T16:05:55.476Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "Cerebras",
      "accuracy": 0.474747,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.04817519191919192,
      "avg_duration_per_task_ms": 40570.833333333336,
      "avg_output_tokens_per_task": 16881.383838383837,
      "last_run_timestamp": "2026-08-15T14:46:10.940Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "Nebius",
      "accuracy": 0.47306366666666666,
      "accuracy_std_dev": 0.011903906959015132,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0010633148148148147,
      "avg_duration_per_task_ms": 225625.12962962964,
      "avg_output_tokens_per_task": 2395.355218855219,
      "last_run_timestamp": "2026-08-29T20:09:29.346Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "Novita",
      "accuracy": 0.472853375,
      "accuracy_std_dev": 0.009170414043780961,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.0010971363636363637,
      "avg_duration_per_task_ms": 63777.9494949495,
      "avg_output_tokens_per_task": 2470.3926767676767,
      "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": "Crusoe",
      "accuracy": 0.4662730708263069,
      "accuracy_std_dev": 0.0036637559355878987,
      "run_count": 3,
      "total_questions": 593,
      "avg_cost_per_task_usd": 0.0023418077571669476,
      "avg_duration_per_task_ms": 38653.266441821244,
      "avg_output_tokens_per_task": 2860.9241146711634,
      "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": 0.4654881111111111,
      "accuracy_std_dev": 0.011153042389554961,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.0008001296296296297,
      "avg_duration_per_task_ms": 150622.7519640853,
      "avg_output_tokens_per_task": 2249.726150392817,
      "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": "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": 0.44871055972696244,
      "accuracy_std_dev": 0.08315656909813463,
      "run_count": 9,
      "total_questions": 1758,
      "avg_cost_per_task_usd": 0.004654043230944255,
      "avg_duration_per_task_ms": 44940.08703071672,
      "avg_output_tokens_per_task": 1962.6114903299203,
      "last_run_timestamp": "2026-10-31T16:05:36.136Z"
    },
    {
      "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-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "Venice",
      "accuracy": 0.34701175,
      "accuracy_std_dev": 0.09152431512547636,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.008584518939393939,
      "avg_duration_per_task_ms": 487705.029040404,
      "avg_output_tokens_per_task": 21225.517045454544,
      "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": 0.29037048571428564,
      "accuracy_std_dev": 0.015956919278462715,
      "run_count": 8,
      "total_questions": 1575,
      "avg_cost_per_task_usd": 0.002851500317460318,
      "avg_duration_per_task_ms": 40947.175873015876,
      "avg_output_tokens_per_task": 35610.56507936508,
      "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": 0.27925087499999995,
      "accuracy_std_dev": 0.021335446769622016,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.001127453914141414,
      "avg_duration_per_task_ms": 148962.23926767678,
      "avg_output_tokens_per_task": 16512.824494949495,
      "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": "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": 0.2711902753623188,
      "accuracy_std_dev": 0.014229495825133793,
      "run_count": 8,
      "total_questions": 1518,
      "avg_cost_per_task_usd": 0.010431668642951253,
      "avg_duration_per_task_ms": 429362.8353096179,
      "avg_output_tokens_per_task": 46706.179183135704,
      "last_run_timestamp": "2026-10-31T16:04:29.124Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemma-3-4b-it",
      "model_slug": "google/gemma-3-4b-it",
      "model_name": "Google: Gemma 3 4B",
      "provider_name": "Parasail",
      "accuracy": 0.271044,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.00011306565656565657,
      "avg_duration_per_task_ms": 10846.11616161616,
      "avg_output_tokens_per_task": 1804.4646464646464,
      "last_run_timestamp": "2026-10-15T03:00:03.049Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Coder 30B A3B Instruct",
      "provider_name": "Novita",
      "accuracy": 0.265993,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.0007544898989898989,
      "avg_duration_per_task_ms": 47786.38888888889,
      "avg_output_tokens_per_task": 2583.777777777778,
      "last_run_timestamp": "2026-08-15T09:28:07.208Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.1-8b-instruct",
      "model_slug": "meta-llama/llama-3.1-8b-instruct",
      "model_name": "Meta: Llama 3.1 8B Instruct",
      "provider_name": "DeepInfra",
      "accuracy": 0.2644967777777778,
      "accuracy_std_dev": 0.010114842370142964,
      "run_count": 9,
      "total_questions": 1782,
      "avg_cost_per_task_usd": 0.0003960353535353535,
      "avg_duration_per_task_ms": 363717.83613916946,
      "avg_output_tokens_per_task": 9497.86251402918,
      "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": 0.256734125,
      "accuracy_std_dev": 0.014383764471075542,
      "run_count": 8,
      "total_questions": 1584,
      "avg_cost_per_task_usd": 0.00042851388888888887,
      "avg_duration_per_task_ms": 73900.38257575757,
      "avg_output_tokens_per_task": 8246.635732323231,
      "last_run_timestamp": "2026-10-31T16:55:07.715Z"
    },
    {
      "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": "deepseek/deepseek-chat-v3-0324",
      "model_slug": "deepseek/deepseek-chat-v3-0324",
      "model_name": "DeepSeek: DeepSeek V3 0324",
      "provider_name": "Novita",
      "accuracy": 0.157127,
      "accuracy_std_dev": 0.0633250504987297,
      "run_count": 3,
      "total_questions": 594,
      "avg_cost_per_task_usd": 0.0016609797979797978,
      "avg_duration_per_task_ms": 58054.12962962963,
      "avg_output_tokens_per_task": 1445.550505050505,
      "last_run_timestamp": "2026-08-29T18:43:09.114Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "meta-llama/llama-3.1-8b-instruct",
      "model_slug": "meta-llama/llama-3.1-8b-instruct",
      "model_name": "Meta: Llama 3.1 8B Instruct",
      "provider_name": "Cloudflare",
      "accuracy": 0.1409935,
      "accuracy_std_dev": 0.13931077227641084,
      "run_count": 4,
      "total_questions": 792,
      "avg_cost_per_task_usd": 0.000907405303030303,
      "avg_duration_per_task_ms": 385614.86616161617,
      "avg_output_tokens_per_task": 2945.2651515151515,
      "last_run_timestamp": "2026-10-24T16:03:19.773Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.1-flash-lite-image-20260630",
      "model_slug": "google/gemini-3.1-flash-lite-image",
      "model_name": "Google: Nano Banana 2 Lite (Gemini 3.1 Flash Lite Image)",
      "provider_name": "Google AI Studio",
      "accuracy": 0.1161615,
      "accuracy_std_dev": 0.0016835000000000044,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.017850737373737373,
      "avg_duration_per_task_ms": 46853.982323232325,
      "avg_output_tokens_per_task": 11776.444444444445,
      "last_run_timestamp": "2026-10-01T16:00:58.389Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.1-flash-lite-image-20260630",
      "model_slug": "google/gemini-3.1-flash-lite-image",
      "model_name": "Google: Nano Banana 2 Lite (Gemini 3.1 Flash Lite Image)",
      "provider_name": "Google",
      "accuracy": 0.112795,
      "accuracy_std_dev": 0.006733999999999997,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.018038613636363634,
      "avg_duration_per_task_ms": 21125.79292929293,
      "avg_output_tokens_per_task": 11795.222222222223,
      "last_run_timestamp": "2026-10-01T16:00:58.364Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "google/gemini-3.1-flash-lite-image-20260630",
      "model_slug": "google/gemini-3.1-flash-lite-image",
      "model_name": "Google: Nano Banana 2 Lite (Gemini 3.1 Flash Lite Image)",
      "provider_name": "auto-routing",
      "accuracy": 0.102694,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.017984406565656568,
      "avg_duration_per_task_ms": 21554.590909090908,
      "avg_output_tokens_per_task": 11812.823232323231,
      "last_run_timestamp": "2026-10-01T16:02:13.023Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "AkashML",
      "accuracy": 0.013468000000000003,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.006156141414141415,
      "avg_duration_per_task_ms": 168896.5202020202,
      "avg_output_tokens_per_task": 14359.661616161617,
      "last_run_timestamp": "2026-08-15T15:24:35.389Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-v3.1-terminus",
      "model_slug": "deepseek/deepseek-v3.1-terminus",
      "model_name": "DeepSeek: DeepSeek V3.1 Terminus",
      "provider_name": "SiliconFlow",
      "accuracy": 0.0025255000000000004,
      "accuracy_std_dev": 0.0008415000000000001,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.0034195530303030306,
      "avg_duration_per_task_ms": 159765.32575757575,
      "avg_output_tokens_per_task": 3219.0782828282827,
      "last_run_timestamp": "2026-08-22T17:29:06.480Z"
    },
    {
      "benchmark": "gpqa_diamond",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/gpqa-diamond",
      "model_permaslug": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "SiliconFlow",
      "accuracy": 0.001684,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.003405252525252525,
      "avg_duration_per_task_ms": 146900.18181818182,
      "avg_output_tokens_per_task": 3204.777777777778,
      "last_run_timestamp": "2026-10-31T16:04:43.943Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-5-fable-20260609",
      "model_slug": "anthropic/claude-fable-5",
      "model_name": "Anthropic: Claude Fable 5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.82,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.8143706199999999,
      "avg_duration_per_task_ms": 141768.16,
      "avg_output_tokens_per_task": 5650.52,
      "last_run_timestamp": "2026-08-15T08:09:14.229Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "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": "google/gemini-3.7-flash-20260813",
      "model_slug": "google/gemini-3.7-flash",
      "model_name": "Google: Gemini 3.7 Flash",
      "provider_name": "Google AI Studio",
      "accuracy": 0.806667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.20567228,
      "avg_duration_per_task_ms": 69455.26,
      "avg_output_tokens_per_task": 15737.28,
      "last_run_timestamp": "2026-08-15T00:15:11.753Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "DigitalOcean",
      "accuracy": 0.80208325,
      "accuracy_std_dev": 0.012126500481899135,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.1338215925,
      "avg_duration_per_task_ms": 2777506.835,
      "avg_output_tokens_per_task": 27899.5725,
      "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": "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": "anthropic/claude-5-fable-20260609",
      "model_slug": "anthropic/claude-fable-5",
      "model_name": "Anthropic: Claude Fable 5",
      "provider_name": "Azure",
      "accuracy": 0.7981818181818182,
      "accuracy_std_dev": 0.021851955343414428,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.8809048418181818,
      "avg_duration_per_task_ms": 125745.05454545455,
      "avg_output_tokens_per_task": 5732.8836363636365,
      "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": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "BaseTen",
      "accuracy": 0.7981221830985916,
      "accuracy_std_dev": 0.05639978873679898,
      "run_count": 3,
      "total_questions": 142,
      "avg_cost_per_task_usd": 0.007624964788732395,
      "avg_duration_per_task_ms": 30939.802816901407,
      "avg_output_tokens_per_task": 8761.732394366198,
      "last_run_timestamp": "2026-10-31T11:56:20.980Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Sail Research",
      "accuracy": 0.7977776666666667,
      "accuracy_std_dev": 0.01911642045177097,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.008363246666666668,
      "avg_duration_per_task_ms": 272879.14,
      "avg_output_tokens_per_task": 12625.026666666667,
      "last_run_timestamp": "2026-08-22T16:20:38.832Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "Phala",
      "accuracy": 0.7956450603015076,
      "accuracy_std_dev": 0.0288553624912251,
      "run_count": 8,
      "total_questions": 398,
      "avg_cost_per_task_usd": 0.2041434145728643,
      "avg_duration_per_task_ms": 814395.7738693467,
      "avg_output_tokens_per_task": 35286.695979899494,
      "last_run_timestamp": "2026-10-31T19:01:09.612Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-5-fable-20260609",
      "model_slug": "anthropic/claude-fable-5",
      "model_name": "Anthropic: Claude Fable 5",
      "provider_name": "auto-routing",
      "accuracy": 0.7939393636363636,
      "accuracy_std_dev": 0.02791815599951441,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.8490899109090908,
      "avg_duration_per_task_ms": 118340.14,
      "avg_output_tokens_per_task": 5735.276363636363,
      "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": "anthropic/claude-5-fable-20260609",
      "model_slug": "anthropic/claude-fable-5",
      "model_name": "Anthropic: Claude Fable 5",
      "provider_name": "Anthropic",
      "accuracy": 0.7933333636363636,
      "accuracy_std_dev": 0.01705605730990246,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.8212979,
      "avg_duration_per_task_ms": 123271.04,
      "avg_output_tokens_per_task": 5720.44,
      "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": "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-fable-5.1-20260831",
      "model_slug": "anthropic/claude-fable-5.1",
      "model_name": "Anthropic: Claude Fable 5.1",
      "provider_name": "auto-routing",
      "accuracy": 0.79166675,
      "accuracy_std_dev": 0.012801703448662584,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.71616581,
      "avg_duration_per_task_ms": 111830.815,
      "avg_output_tokens_per_task": 6900.955,
      "last_run_timestamp": "2026-10-01T08:03:01.529Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-fable-5.1-20260831",
      "model_slug": "anthropic/claude-fable-5.1",
      "model_name": "Anthropic: Claude Fable 5.1",
      "provider_name": "Anthropic",
      "accuracy": 0.7906666,
      "accuracy_std_dev": 0.009977824303925176,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.814870324,
      "avg_duration_per_task_ms": 182126.652,
      "avg_output_tokens_per_task": 7046.908,
      "last_run_timestamp": "2026-10-01T08:05:22.427Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Chutes",
      "accuracy": 0.7901233981481481,
      "accuracy_std_dev": 0.021051257864696603,
      "run_count": 11,
      "total_questions": 540,
      "avg_cost_per_task_usd": 0.07926520925925926,
      "avg_duration_per_task_ms": 1152985.5148148148,
      "avg_output_tokens_per_task": 25849.22962962963,
      "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": 0.7896970909090909,
      "accuracy_std_dev": 0.02153015652712829,
      "run_count": 22,
      "total_questions": 1100,
      "avg_cost_per_task_usd": 0.8954311818181818,
      "avg_duration_per_task_ms": 92327.7890909091,
      "avg_output_tokens_per_task": 5722.323636363636,
      "last_run_timestamp": "2026-10-31T12:30:34.973Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "Alibaba",
      "accuracy": 0.789473611528822,
      "accuracy_std_dev": 0.025156094260387483,
      "run_count": 8,
      "total_questions": 399,
      "avg_cost_per_task_usd": 0.18347024812030074,
      "avg_duration_per_task_ms": 432487.7944862155,
      "avg_output_tokens_per_task": 30439.06015037594,
      "last_run_timestamp": "2026-10-31T19:12:41.306Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "Xiaomi",
      "accuracy": 0.788889,
      "accuracy_std_dev": 0.015234869040023505,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.057788776666666666,
      "avg_duration_per_task_ms": 893038.1233333333,
      "avg_output_tokens_per_task": 40447.416666666664,
      "last_run_timestamp": "2026-10-31T13:13:33.031Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "z-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": 0.788194,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 48,
      "avg_cost_per_task_usd": 0.0031695833333333333,
      "avg_duration_per_task_ms": 66197.89583333333,
      "avg_output_tokens_per_task": 2922.6041666666665,
      "last_run_timestamp": "2026-11-04T18:22:51.105Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-sonnet-5.5-20260928",
      "model_slug": "anthropic/claude-sonnet-5.5",
      "model_name": "Anthropic: Claude Sonnet 5.5",
      "provider_name": "auto-routing",
      "accuracy": 0.788,
      "accuracy_std_dev": 0.02246951515275753,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.15925035199999998,
      "avg_duration_per_task_ms": 71120.884,
      "avg_output_tokens_per_task": 6245.168,
      "last_run_timestamp": "2026-10-31T11:10:36.361Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "Alibaba",
      "accuracy": 0.7874073333333333,
      "accuracy_std_dev": 0.02836162518850661,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.08114153333333332,
      "avg_duration_per_task_ms": 209436.06444444443,
      "avg_output_tokens_per_task": 25962.27111111111,
      "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": "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": "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": "anthropic/claude-opus-5-20260723",
      "model_slug": "anthropic/claude-opus-5",
      "model_name": "Anthropic: Claude Opus 5",
      "provider_name": "auto-routing",
      "accuracy": 0.7866666363636363,
      "accuracy_std_dev": 0.020695005011191054,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.4729785890909092,
      "avg_duration_per_task_ms": 117016.24545454545,
      "avg_output_tokens_per_task": 6994.309090909091,
      "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": "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": "auto-routing",
      "accuracy": 0.785,
      "accuracy_std_dev": 0.023979192370678373,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.20099723,
      "avg_duration_per_task_ms": 767265.06,
      "avg_output_tokens_per_task": 32060.4625,
      "last_run_timestamp": "2026-10-31T19:47:27.263Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Together",
      "accuracy": 0.7844443333333334,
      "accuracy_std_dev": 0.01632984089872824,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.0785912688888889,
      "avg_duration_per_task_ms": 47362.52222222222,
      "avg_output_tokens_per_task": 4720.888888888889,
      "last_run_timestamp": "2026-10-31T11:58:26.584Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-sonnet-5.5-20260928",
      "model_slug": "anthropic/claude-sonnet-5.5",
      "model_name": "Anthropic: Claude Sonnet 5.5",
      "provider_name": "Azure",
      "accuracy": 0.7833333333333333,
      "accuracy_std_dev": 0.011706202852429232,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.1729425,
      "avg_duration_per_task_ms": 66541.35,
      "avg_output_tokens_per_task": 6203.445,
      "last_run_timestamp": "2026-10-31T12:38:55.501Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.5-flash-lite-20260721",
      "model_slug": "google/gemini-3.5-flash-lite",
      "model_name": "Google: Gemini 3.5 Flash Lite",
      "provider_name": "auto-routing",
      "accuracy": 0.7833330000000001,
      "accuracy_std_dev": 0.009999999999999953,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.07443635,
      "avg_duration_per_task_ms": 98397.15,
      "avg_output_tokens_per_task": 21518.67,
      "last_run_timestamp": "2026-10-01T19:04:42.463Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.7-flash-20260813",
      "model_slug": "google/gemini-3.7-flash",
      "model_name": "Google: Gemini 3.7 Flash",
      "provider_name": "Google",
      "accuracy": 0.7831052054794521,
      "accuracy_std_dev": 0.017155537965592597,
      "run_count": 3,
      "total_questions": 146,
      "avg_cost_per_task_usd": 0.14992966438356164,
      "avg_duration_per_task_ms": 183129.3904109589,
      "avg_output_tokens_per_task": 16313.164383561643,
      "last_run_timestamp": "2026-10-15T21:26:28.195Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Novita",
      "accuracy": 0.7825,
      "accuracy_std_dev": 0.023790007240436065,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.08953825,
      "avg_duration_per_task_ms": 274829.5175,
      "avg_output_tokens_per_task": 19992.6725,
      "last_run_timestamp": "2026-10-31T12:57:09.932Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Novita",
      "accuracy": 0.7822223333333334,
      "accuracy_std_dev": 0.01571327975171178,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.011317603333333334,
      "avg_duration_per_task_ms": 599501.3033333333,
      "avg_output_tokens_per_task": 21464.33,
      "last_run_timestamp": "2026-10-31T14:08:11.994Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "Novita",
      "accuracy": 0.781333,
      "accuracy_std_dev": 0.027129319932501044,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.0587014,
      "avg_duration_per_task_ms": 871009.496,
      "avg_output_tokens_per_task": 40629.384,
      "last_run_timestamp": "2026-10-31T13:41:45.309Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Fireworks",
      "accuracy": 0.7811976407035177,
      "accuracy_std_dev": 0.02110393286871771,
      "run_count": 16,
      "total_questions": 796,
      "avg_cost_per_task_usd": 0.09273568467336683,
      "avg_duration_per_task_ms": 66515.5364321608,
      "avg_output_tokens_per_task": 4679.810301507538,
      "last_run_timestamp": "2026-10-31T13:12:01.731Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Tenstorrent",
      "accuracy": 0.7809525714285713,
      "accuracy_std_dev": 0.03331975778019121,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.04230577142857143,
      "avg_duration_per_task_ms": 122423.61428571428,
      "avg_output_tokens_per_task": 4845.648571428572,
      "last_run_timestamp": "2026-11-04T18:25:27.180Z"
    },
    {
      "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": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.78,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.0042799399999999994,
      "avg_duration_per_task_ms": 627315.22,
      "avg_output_tokens_per_task": 12675.1,
      "last_run_timestamp": "2026-10-31T21:10:18.835Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "Parasail",
      "accuracy": 0.7790587291066282,
      "accuracy_std_dev": 0.03375729268813859,
      "run_count": 7,
      "total_questions": 347,
      "avg_cost_per_task_usd": 0.19038813832853027,
      "avg_duration_per_task_ms": 636053.7665706052,
      "avg_output_tokens_per_task": 30091.458213256483,
      "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": "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": "anthropic/claude-4.7-opus-20260416",
      "model_slug": "anthropic/claude-opus-4.7",
      "model_name": "Anthropic: Claude Opus 4.7",
      "provider_name": "Anthropic",
      "accuracy": 0.7777776666666667,
      "accuracy_std_dev": 0.01791624892071129,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.4406913983333334,
      "avg_duration_per_task_ms": 87633.54833333334,
      "avg_output_tokens_per_task": 5698.321666666667,
      "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": "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": "qwen/qwen3.5-122b-a10b-20260224",
      "model_slug": "qwen/qwen3.5-122b-a10b",
      "model_name": "Qwen: Qwen3.5-122B-A10B",
      "provider_name": "Novita",
      "accuracy": 0.7773700055045871,
      "accuracy_std_dev": 0.023259459839311162,
      "run_count": 11,
      "total_questions": 545,
      "avg_cost_per_task_usd": 0.19957610458715597,
      "avg_duration_per_task_ms": 504017.81834862387,
      "avg_output_tokens_per_task": 27873.480733944954,
      "last_run_timestamp": "2026-10-31T18:53:29.091Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-122b-a10b-20260224",
      "model_slug": "qwen/qwen3.5-122b-a10b",
      "model_name": "Qwen: Qwen3.5-122B-A10B",
      "provider_name": "auto-routing",
      "accuracy": 0.7769697272727273,
      "accuracy_std_dev": 0.023503914842545274,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.17000764545454541,
      "avg_duration_per_task_ms": 562646.2527272727,
      "avg_output_tokens_per_task": 35727.95818181818,
      "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": "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": "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": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7758335,
      "accuracy_std_dev": 0.016308646311082982,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.15398643500000003,
      "avg_duration_per_task_ms": 330861.565,
      "avg_output_tokens_per_task": 19948.0325,
      "last_run_timestamp": "2026-10-31T13:41:05.689Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Cloudflare",
      "accuracy": 0.7757695461847389,
      "accuracy_std_dev": 0.016169521055368324,
      "run_count": 10,
      "total_questions": 498,
      "avg_cost_per_task_usd": 0.08735203012048193,
      "avg_duration_per_task_ms": 139136.04417670684,
      "avg_output_tokens_per_task": 4443.712851405622,
      "last_run_timestamp": "2026-10-31T12:29:23.419Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "StreamLake",
      "accuracy": 0.7757574545454545,
      "accuracy_std_dev": 0.02074823590031709,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.08103076363636365,
      "avg_duration_per_task_ms": 336159.44,
      "avg_output_tokens_per_task": 20208.643636363635,
      "last_run_timestamp": "2026-10-31T12:24:53.587Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "DeepInfra",
      "accuracy": 0.7753625384615385,
      "accuracy_std_dev": 0.017323958686038136,
      "run_count": 6,
      "total_questions": 299,
      "avg_cost_per_task_usd": 0.06710363879598663,
      "avg_duration_per_task_ms": 1125092.9698996656,
      "avg_output_tokens_per_task": 36573.67558528428,
      "last_run_timestamp": "2026-10-24T13:26:35.887Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-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": "anthropic/claude-opus-5-20260723",
      "model_slug": "anthropic/claude-opus-5",
      "model_name": "Anthropic: Claude Opus 5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7752688387096777,
      "accuracy_std_dev": 0.0207523765107747,
      "run_count": 31,
      "total_questions": 1550,
      "avg_cost_per_task_usd": 0.4981299012903227,
      "avg_duration_per_task_ms": 112970.55806451612,
      "avg_output_tokens_per_task": 7038.44064516129,
      "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": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "auto-routing",
      "accuracy": 0.7751514545454545,
      "accuracy_std_dev": 0.026375843918816165,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.10104016181818182,
      "avg_duration_per_task_ms": 294074.4890909091,
      "avg_output_tokens_per_task": 21520.51818181818,
      "last_run_timestamp": "2026-10-31T11:29:11.154Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-opus-5-20260723",
      "model_slug": "anthropic/claude-opus-5",
      "model_name": "Anthropic: Claude Opus 5",
      "provider_name": "Azure",
      "accuracy": 0.7750724782608696,
      "accuracy_std_dev": 0.020450702198786246,
      "run_count": 23,
      "total_questions": 1150,
      "avg_cost_per_task_usd": 0.47903272521739126,
      "avg_duration_per_task_ms": 101622.89130434782,
      "avg_output_tokens_per_task": 7123.294782608696,
      "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": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "AtlasCloud",
      "accuracy": 0.775,
      "accuracy_std_dev": 0.01787294665828779,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.09322250500000001,
      "avg_duration_per_task_ms": 167383.47,
      "avg_output_tokens_per_task": 5123.47,
      "last_run_timestamp": "2026-10-31T12:32:51.992Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-fable-5.1-20260831",
      "model_slug": "anthropic/claude-fable-5.1",
      "model_name": "Anthropic: Claude Fable 5.1",
      "provider_name": "Google",
      "accuracy": 0.7746668,
      "accuracy_std_dev": 0.012927338819726197,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.758212416,
      "avg_duration_per_task_ms": 112131.876,
      "avg_output_tokens_per_task": 6866.944,
      "last_run_timestamp": "2026-10-01T08:05:27.141Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.8-opus-20260528",
      "model_slug": "anthropic/claude-opus-4.8",
      "model_name": "Anthropic: Claude Opus 4.8",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.7745454545454545,
      "accuracy_std_dev": 0.02425759601131023,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.48932630545454553,
      "avg_duration_per_task_ms": 124123.69272727273,
      "avg_output_tokens_per_task": 7977.827272727272,
      "last_run_timestamp": "2026-10-31T11:58:26.584Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-opus-5-20260723",
      "model_slug": "anthropic/claude-opus-5",
      "model_name": "Anthropic: Claude Opus 5",
      "provider_name": "Google",
      "accuracy": 0.7743749687500002,
      "accuracy_std_dev": 0.022737010850269506,
      "run_count": 32,
      "total_questions": 1600,
      "avg_cost_per_task_usd": 0.49075661062500003,
      "avg_duration_per_task_ms": 106422.6925,
      "avg_output_tokens_per_task": 7062.208125,
      "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": "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": "anthropic/claude-4.8-opus-20260528",
      "model_slug": "anthropic/claude-opus-4.8",
      "model_name": "Anthropic: Claude Opus 4.8",
      "provider_name": "Anthropic",
      "accuracy": 0.7733334166666667,
      "accuracy_std_dev": 0.024343346495015628,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.506005475,
      "avg_duration_per_task_ms": 120787.225,
      "avg_output_tokens_per_task": 7991.263333333333,
      "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": "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-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Ionstream",
      "accuracy": 0.7733333333333333,
      "accuracy_std_dev": 0.04109609335312651,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.08230668000000001,
      "avg_duration_per_task_ms": 584422.56,
      "avg_output_tokens_per_task": 10999.886666666667,
      "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": "Baidu",
      "accuracy": 0.7733332857142857,
      "accuracy_std_dev": 0.022537305934283946,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.08668051142857143,
      "avg_duration_per_task_ms": 364495.4428571428,
      "avg_output_tokens_per_task": 20483.94,
      "last_run_timestamp": "2026-10-24T12:45:56.375Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "ModelRun",
      "accuracy": 0.77333325,
      "accuracy_std_dev": 0.029814193116492343,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.06547900499999999,
      "avg_duration_per_task_ms": 79239.345,
      "avg_output_tokens_per_task": 12411.3875,
      "last_run_timestamp": "2026-10-31T11:09:23.381Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Io Net",
      "accuracy": 0.7733331666666667,
      "accuracy_std_dev": 0.021081851068448003,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.05726335,
      "avg_duration_per_task_ms": 158508.54333333333,
      "avg_output_tokens_per_task": 4588.7266666666665,
      "last_run_timestamp": "2026-10-31T13:11:06.716Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Reka",
      "accuracy": 0.773333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.012345539999999999,
      "avg_duration_per_task_ms": 657478.84,
      "avg_output_tokens_per_task": 14196.68,
      "last_run_timestamp": "2026-10-03T00:00:06.435Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "DeepInfra",
      "accuracy": 0.773333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.06913468,
      "avg_duration_per_task_ms": 170689.28,
      "avg_output_tokens_per_task": 8815.54,
      "last_run_timestamp": "2026-08-15T08:17:17.868Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "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": "meta/muse-glimmer-30b-20260810",
      "model_slug": "meta/muse-glimmer-30b",
      "model_name": "Meta: Muse Glimmer 30B",
      "provider_name": "DeepInfra",
      "accuracy": 0.7722222222222223,
      "accuracy_std_dev": 0.03095202181433487,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.03593436666666667,
      "avg_duration_per_task_ms": 446735.1666666667,
      "avg_output_tokens_per_task": 13815.786666666667,
      "last_run_timestamp": "2026-10-08T22:42:47.098Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.8-opus-20260528",
      "model_slug": "anthropic/claude-opus-4.8",
      "model_name": "Anthropic: Claude Opus 4.8",
      "provider_name": "Azure",
      "accuracy": 0.7721739130434783,
      "accuracy_std_dev": 0.03024227286994588,
      "run_count": 23,
      "total_questions": 1150,
      "avg_cost_per_task_usd": 0.5151446756521739,
      "avg_duration_per_task_ms": 124707.92,
      "avg_output_tokens_per_task": 7934.98347826087,
      "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": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "auto-routing",
      "accuracy": 0.7721212727272727,
      "accuracy_std_dev": 0.023236674253863547,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.14458932363636365,
      "avg_duration_per_task_ms": 821555.8181818182,
      "avg_output_tokens_per_task": 31749.51818181818,
      "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": "Relace",
      "accuracy": 0.7720001999999999,
      "accuracy_std_dev": 0.024367767418456705,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.14564374000000002,
      "avg_duration_per_task_ms": 314489.068,
      "avg_output_tokens_per_task": 20798.24,
      "last_run_timestamp": "2026-10-31T13:16:57.928Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-fable-5.1-20260831",
      "model_slug": "anthropic/claude-fable-5.1",
      "model_name": "Anthropic: Claude Fable 5.1",
      "provider_name": "Azure",
      "accuracy": 0.772,
      "accuracy_std_dev": 0.02679938639596063,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.808827636,
      "avg_duration_per_task_ms": 183544.5,
      "avg_output_tokens_per_task": 6940.7,
      "last_run_timestamp": "2026-10-01T08:03:00.737Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Parasail",
      "accuracy": 0.7718182727272727,
      "accuracy_std_dev": 0.031151820244762325,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.06332707454545455,
      "avg_duration_per_task_ms": 243359.18363636362,
      "avg_output_tokens_per_task": 11661.423636363636,
      "last_run_timestamp": "2026-10-31T12:57:09.934Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "auto-routing",
      "accuracy": 0.771666625,
      "accuracy_std_dev": 0.02127857229666442,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.058498104999999995,
      "avg_duration_per_task_ms": 220433.295,
      "avg_output_tokens_per_task": 12747.45,
      "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": "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": "anthropic/claude-opus-5-20260723",
      "model_slug": "anthropic/claude-opus-5",
      "model_name": "Anthropic: Claude Opus 5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.7715151818181818,
      "accuracy_std_dev": 0.01635249901846076,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.4578688290909091,
      "avg_duration_per_task_ms": 118180.02,
      "avg_output_tokens_per_task": 7063.84,
      "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": "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": "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-4.7-opus-20260416",
      "model_slug": "anthropic/claude-opus-4.7",
      "model_name": "Anthropic: Claude Opus 4.7",
      "provider_name": "Azure",
      "accuracy": 0.7709090909090909,
      "accuracy_std_dev": 0.015380107124905002,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.43817922727272723,
      "avg_duration_per_task_ms": 92556,
      "avg_output_tokens_per_task": 5682.427272727273,
      "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": "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": "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": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "DeepInfra",
      "accuracy": 0.7702915342789598,
      "accuracy_std_dev": 0.030639008330912567,
      "run_count": 17,
      "total_questions": 846,
      "avg_cost_per_task_usd": 0.034585855791962175,
      "avg_duration_per_task_ms": 444223.6111111111,
      "avg_output_tokens_per_task": 12313.30023640662,
      "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": "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": "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": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "OpenInference",
      "accuracy": 0.77,
      "accuracy_std_dev": 0.019051314722786625,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.01876148,
      "avg_duration_per_task_ms": 512061.9,
      "avg_output_tokens_per_task": 12105.946666666667,
      "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": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Together",
      "accuracy": 0.77,
      "accuracy_std_dev": 0.010000000000000009,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.07660454,
      "avg_duration_per_task_ms": 184215.64,
      "avg_output_tokens_per_task": 8506.78,
      "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": "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": "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": 0.7696968636363637,
      "accuracy_std_dev": 0.030466247715218248,
      "run_count": 22,
      "total_questions": 1100,
      "avg_cost_per_task_usd": 0.4521496381818182,
      "avg_duration_per_task_ms": 78848.03,
      "avg_output_tokens_per_task": 5649.991818181818,
      "last_run_timestamp": "2026-10-31T12:41:58.029Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "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-5.3-20260816",
      "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": "anthropic/claude-opus-5-20260723",
      "model_slug": "anthropic/claude-opus-5",
      "model_name": "Anthropic: Claude Opus 5",
      "provider_name": "Anthropic",
      "accuracy": 0.769090909090909,
      "accuracy_std_dev": 0.02562719156909333,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.4768810909090909,
      "avg_duration_per_task_ms": 108340.30181818182,
      "avg_output_tokens_per_task": 7142.923636363636,
      "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": "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": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7688886666666668,
      "accuracy_std_dev": 0.03456950772316872,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.07998294666666667,
      "avg_duration_per_task_ms": 191648.96,
      "avg_output_tokens_per_task": 9368.36,
      "last_run_timestamp": "2026-08-29T16:11:58.076Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "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": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Baidu",
      "accuracy": 0.768333375,
      "accuracy_std_dev": 0.026193847349795237,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0433664925,
      "avg_duration_per_task_ms": 202126.965,
      "avg_output_tokens_per_task": 13303.9375,
      "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": "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": "Google",
      "accuracy": 0.7681249687499998,
      "accuracy_std_dev": 0.028172817762078275,
      "run_count": 32,
      "total_questions": 1600,
      "avg_cost_per_task_usd": 0.5299744849999999,
      "avg_duration_per_task_ms": 112471.520625,
      "avg_output_tokens_per_task": 8064.054375,
      "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-4.8-opus-20260528",
      "model_slug": "anthropic/claude-opus-4.8",
      "model_name": "Anthropic: Claude Opus 4.8",
      "provider_name": "auto-routing",
      "accuracy": 0.767879,
      "accuracy_std_dev": 0.0204628104530234,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.4928555472727273,
      "avg_duration_per_task_ms": 123502.80727272727,
      "avg_output_tokens_per_task": 8045.849090909091,
      "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": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Venice",
      "accuracy": 0.767500125,
      "accuracy_std_dev": 0.028369943260594926,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.141204825,
      "avg_duration_per_task_ms": 285064.7125,
      "avg_output_tokens_per_task": 12155.7,
      "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": "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": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "GMICloud",
      "accuracy": 0.7674074444444444,
      "accuracy_std_dev": 0.02972210199726465,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.08416672666666668,
      "avg_duration_per_task_ms": 371249.3911111111,
      "avg_output_tokens_per_task": 18973.266666666666,
      "last_run_timestamp": "2026-10-31T20:27:36.016Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Mistral",
      "accuracy": 0.7670707272727272,
      "accuracy_std_dev": 0.02101397565441647,
      "run_count": 33,
      "total_questions": 1650,
      "avg_cost_per_task_usd": 0.06510748121212123,
      "avg_duration_per_task_ms": 67556.39090909091,
      "avg_output_tokens_per_task": 4633.900606060606,
      "last_run_timestamp": "2026-11-02T21:36:47.807Z"
    },
    {
      "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": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Morph",
      "accuracy": 0.7666669999999999,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.0213333,
      "avg_duration_per_task_ms": 951098.34,
      "avg_output_tokens_per_task": 12473.42,
      "last_run_timestamp": "2026-08-15T08:15:52.453Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-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": "google/gemini-3.6-flash-20260721",
      "model_slug": "google/gemini-3.6-flash",
      "model_name": "Google: Gemini 3.6 Flash",
      "provider_name": "Google",
      "accuracy": 0.7666668,
      "accuracy_std_dev": 0.007302967435227963,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.220517392,
      "avg_duration_per_task_ms": 158612.068,
      "avg_output_tokens_per_task": 23228.836,
      "last_run_timestamp": "2026-10-01T19:04:58.090Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "NextBit",
      "accuracy": 0.76666675,
      "accuracy_std_dev": 0.024267078752653787,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.036940722499999995,
      "avg_duration_per_task_ms": 209470.17,
      "avg_output_tokens_per_task": 17358.2575,
      "last_run_timestamp": "2026-10-03T08:39:48.671Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Morph",
      "accuracy": 0.7666666363636364,
      "accuracy_std_dev": 0.038454932300439754,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.08742998,
      "avg_duration_per_task_ms": 299403.17272727273,
      "avg_output_tokens_per_task": 15964.536363636364,
      "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": "google/gemini-3.8-flash-20260902",
      "model_slug": "google/gemini-3.8-flash",
      "model_name": "Google: Gemini 3.8 Flash",
      "provider_name": "Google",
      "accuracy": 0.7666665,
      "accuracy_std_dev": 0.013333499999999998,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.27949224,
      "avg_duration_per_task_ms": 363191.33,
      "avg_output_tokens_per_task": 35460.86,
      "last_run_timestamp": "2026-10-15T21:31:19.497Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "DekaLLM",
      "accuracy": 0.7666665,
      "accuracy_std_dev": 0.026666499999999982,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01352679,
      "avg_duration_per_task_ms": 90848.6,
      "avg_output_tokens_per_task": 4507.38,
      "last_run_timestamp": "2026-10-10T11:18:50.119Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "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": 0.7658823529411766,
      "accuracy_std_dev": 0.020244567150490694,
      "run_count": 34,
      "total_questions": 1700,
      "avg_cost_per_task_usd": 0.5334031917647061,
      "avg_duration_per_task_ms": 118335.29588235295,
      "avg_output_tokens_per_task": 8074.314705882353,
      "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": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Novita",
      "accuracy": 0.765833375,
      "accuracy_std_dev": 0.019843192579178773,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.1200979175,
      "avg_duration_per_task_ms": 261332.9925,
      "avg_output_tokens_per_task": 20111.6625,
      "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": "anthropic/claude-4.5-opus-20251124",
      "model_slug": "anthropic/claude-opus-4.5",
      "model_name": "Anthropic: Claude Opus 4.5",
      "provider_name": "Anthropic",
      "accuracy": 0.7655556666666666,
      "accuracy_std_dev": 0.02746477429767488,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.5335597466666667,
      "avg_duration_per_task_ms": 180574.04166666666,
      "avg_output_tokens_per_task": 10505.758333333333,
      "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": "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": "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": "anthropic/claude-4.7-opus-20260416",
      "model_slug": "anthropic/claude-opus-4.7",
      "model_name": "Anthropic: Claude Opus 4.7",
      "provider_name": "Google",
      "accuracy": 0.7648485454545456,
      "accuracy_std_dev": 0.020842423656674335,
      "run_count": 33,
      "total_questions": 1650,
      "avg_cost_per_task_usd": 0.5573711296969697,
      "avg_duration_per_task_ms": 70870.49272727272,
      "avg_output_tokens_per_task": 5721.953939393939,
      "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": "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": "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": 0.7642425454545454,
      "accuracy_std_dev": 0.021324467855935902,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.42425751818181817,
      "avg_duration_per_task_ms": 89884.58,
      "avg_output_tokens_per_task": 5563.18,
      "last_run_timestamp": "2026-10-31T11:48:47.453Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7642422727272727,
      "accuracy_std_dev": 0.025309747612300426,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.02170633272727273,
      "avg_duration_per_task_ms": 276388.2545454545,
      "avg_output_tokens_per_task": 15679.149090909092,
      "last_run_timestamp": "2026-10-31T13:19:25.796Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "Venice",
      "accuracy": 0.7641668749999999,
      "accuracy_std_dev": 0.026233564961883768,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.3343550775,
      "avg_duration_per_task_ms": 660538.1825,
      "avg_output_tokens_per_task": 28260.275,
      "last_run_timestamp": "2026-10-31T19:49:31.051Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7641665000000001,
      "accuracy_std_dev": 0.021586383312171588,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0185151775,
      "avg_duration_per_task_ms": 270096.6575,
      "avg_output_tokens_per_task": 14323.5575,
      "last_run_timestamp": "2026-10-31T12:30:14.640Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "Novita",
      "accuracy": 0.764,
      "accuracy_std_dev": 0.020483142122242875,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.080228768,
      "avg_duration_per_task_ms": 494377.212,
      "avg_output_tokens_per_task": 14669.972,
      "last_run_timestamp": "2026-10-31T12:27:34.738Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Alibaba",
      "accuracy": 0.7639395454545455,
      "accuracy_std_dev": 0.035811554789337924,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.19443994909090911,
      "avg_duration_per_task_ms": 474708.6345454545,
      "avg_output_tokens_per_task": 27695.465454545454,
      "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": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7637037777777778,
      "accuracy_std_dev": 0.021798347556015337,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.06269429111111112,
      "avg_duration_per_task_ms": 113855.65555555555,
      "avg_output_tokens_per_task": 4583.311111111111,
      "last_run_timestamp": "2026-10-31T12:40:08.732Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Friendli",
      "accuracy": 0.7637036666666667,
      "accuracy_std_dev": 0.01835498758739252,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.029433213333333333,
      "avg_duration_per_task_ms": 122522.84,
      "avg_output_tokens_per_task": 12030.735555555555,
      "last_run_timestamp": "2026-10-31T19:08:17.847Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7633335,
      "accuracy_std_dev": 0.06984115707983368,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.251395065,
      "avg_duration_per_task_ms": 485523.98,
      "avg_output_tokens_per_task": 29105.6325,
      "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": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "Darkbloom",
      "accuracy": 0.763333375,
      "accuracy_std_dev": 0.024267095921934618,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0630558525,
      "avg_duration_per_task_ms": 1180279.39,
      "avg_output_tokens_per_task": 60139.9825,
      "last_run_timestamp": "2026-10-31T19:16:07.249Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Together",
      "accuracy": 0.7633333333333333,
      "accuracy_std_dev": 0.03134832186896135,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.014250108888888892,
      "avg_duration_per_task_ms": 231681.52,
      "avg_output_tokens_per_task": 15957.917777777779,
      "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": "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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "AkashML",
      "accuracy": 0.763333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.05078266,
      "avg_duration_per_task_ms": 205280.74,
      "avg_output_tokens_per_task": 7724.94,
      "last_run_timestamp": "2026-08-15T08:10:05.152Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.7630303636363637,
      "accuracy_std_dev": 0.027171604447132024,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.5365658963636364,
      "avg_duration_per_task_ms": 173544.41818181818,
      "avg_output_tokens_per_task": 10675.270909090908,
      "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": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Cloudflare",
      "accuracy": 0.7630302727272728,
      "accuracy_std_dev": 0.03170123269496259,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.15556449636363637,
      "avg_duration_per_task_ms": 593095.6363636364,
      "avg_output_tokens_per_task": 12420.629090909091,
      "last_run_timestamp": "2026-10-31T13:02:37.599Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "GMICloud",
      "accuracy": 0.7629628888888889,
      "accuracy_std_dev": 0.03837586761797183,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.00882314,
      "avg_duration_per_task_ms": 182772.50444444444,
      "avg_output_tokens_per_task": 15534.135555555555,
      "last_run_timestamp": "2026-10-31T19:07:06.878Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Together",
      "accuracy": 0.7629628888888889,
      "accuracy_std_dev": 0.03191221842089971,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.02681685777777778,
      "avg_duration_per_task_ms": 156543.7,
      "avg_output_tokens_per_task": 9000.691111111111,
      "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-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "NextBit",
      "accuracy": 0.7628571428571429,
      "accuracy_std_dev": 0.023328310442945685,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.11986269428571429,
      "avg_duration_per_task_ms": 415507.14,
      "avg_output_tokens_per_task": 18962.734285714287,
      "last_run_timestamp": "2026-10-03T08:33:31.578Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-pro-20260921",
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "model_name": "Xiaomi: MiMo-V2.6-Pro",
      "provider_name": "GMICloud",
      "accuracy": 0.7627778333333333,
      "accuracy_std_dev": 0.03313374040368652,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.07174823666666666,
      "avg_duration_per_task_ms": 893696.3066666666,
      "avg_output_tokens_per_task": 51834.78666666667,
      "last_run_timestamp": "2026-10-27T03:58:39.326Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "GMICloud",
      "accuracy": 0.7626668,
      "accuracy_std_dev": 0.014966487007978884,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.009859876,
      "avg_duration_per_task_ms": 578544.56,
      "avg_output_tokens_per_task": 20173.776,
      "last_run_timestamp": "2026-10-31T14:48:08.689Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Z.AI",
      "accuracy": 0.7626666,
      "accuracy_std_dev": 0.02954472165785287,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.009234846,
      "avg_duration_per_task_ms": 195074.734,
      "avg_output_tokens_per_task": 4475.264,
      "last_run_timestamp": "2026-10-31T12:25:34.861Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.7625418160535117,
      "accuracy_std_dev": 0.02147753020972668,
      "run_count": 12,
      "total_questions": 598,
      "avg_cost_per_task_usd": 0.09006360702341139,
      "avg_duration_per_task_ms": 48936.09866220736,
      "avg_output_tokens_per_task": 4348.224080267559,
      "last_run_timestamp": "2026-10-31T12:39:14.326Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Crusoe",
      "accuracy": 0.7625417625418061,
      "accuracy_std_dev": 0.01910679234955175,
      "run_count": 6,
      "total_questions": 299,
      "avg_cost_per_task_usd": 0.03341546153846154,
      "avg_duration_per_task_ms": 144903.0602006689,
      "avg_output_tokens_per_task": 21777.678929765887,
      "last_run_timestamp": "2026-10-31T12:18:15.192Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Parasail",
      "accuracy": 0.7625,
      "accuracy_std_dev": 0.028467730270430754,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0217337575,
      "avg_duration_per_task_ms": 233098.835,
      "avg_output_tokens_per_task": 12085.595,
      "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-k3-20260715",
      "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-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "DeepInfra",
      "accuracy": 0.7622220000000001,
      "accuracy_std_dev": 0.008314764498569207,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.007347476666666667,
      "avg_duration_per_task_ms": 300690.79,
      "avg_output_tokens_per_task": 12348.97,
      "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": "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": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Venice",
      "accuracy": 0.761904902255639,
      "accuracy_std_dev": 0.022085869925347657,
      "run_count": 8,
      "total_questions": 399,
      "avg_cost_per_task_usd": 0.010246486215538846,
      "avg_duration_per_task_ms": 307475.68922305765,
      "avg_output_tokens_per_task": 12562.348370927319,
      "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": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7619047142857143,
      "accuracy_std_dev": 0.04042854562846226,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.11324955142857143,
      "avg_duration_per_task_ms": 271951.16285714286,
      "avg_output_tokens_per_task": 11844.885714285714,
      "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": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Baidu",
      "accuracy": 0.7617847878787878,
      "accuracy_std_dev": 0.03876959831491683,
      "run_count": 4,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.02413183333333333,
      "avg_duration_per_task_ms": 113963.81313131313,
      "avg_output_tokens_per_task": 15997.81313131313,
      "last_run_timestamp": "2026-10-24T12:54:33.303Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "DekaLLM",
      "accuracy": 0.761784404040404,
      "accuracy_std_dev": 0.046884570628299496,
      "run_count": 4,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.03727159090909091,
      "avg_duration_per_task_ms": 138498.65151515152,
      "avg_output_tokens_per_task": 4381.035353535353,
      "last_run_timestamp": "2026-10-31T12:40:46.071Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "GMICloud",
      "accuracy": 0.76166675,
      "accuracy_std_dev": 0.025404142635027726,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.02248819,
      "avg_duration_per_task_ms": 170461.53666666665,
      "avg_output_tokens_per_task": 16469.198333333334,
      "last_run_timestamp": "2026-10-31T12:11:52.708Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Ionstream",
      "accuracy": 0.7614249072580644,
      "accuracy_std_dev": 0.03191360961423825,
      "run_count": 10,
      "total_questions": 496,
      "avg_cost_per_task_usd": 0.04908138104838709,
      "avg_duration_per_task_ms": 385516.27419354836,
      "avg_output_tokens_per_task": 12910.963709677419,
      "last_run_timestamp": "2026-10-31T13:30:45.600Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.5-opus-20251124",
      "model_slug": "anthropic/claude-opus-4.5",
      "model_name": "Anthropic: Claude Opus 4.5",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.7612120909090909,
      "accuracy_std_dev": 0.02288640210595304,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.54388782,
      "avg_duration_per_task_ms": 180374.2618181818,
      "avg_output_tokens_per_task": 10717.156363636364,
      "last_run_timestamp": "2026-10-31T12:58:32.935Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Cloudflare",
      "accuracy": 0.7612120909090909,
      "accuracy_std_dev": 0.01903078652391983,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.03308226727272728,
      "avg_duration_per_task_ms": 220782.66727272727,
      "avg_output_tokens_per_task": 15395.681818181818,
      "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": "qwen/qwen3.5-122b-a10b-20260224",
      "model_slug": "qwen/qwen3.5-122b-a10b",
      "model_name": "Qwen: Qwen3.5-122B-A10B",
      "provider_name": "DeepInfra",
      "accuracy": 0.7611108333333334,
      "accuracy_std_dev": 0.02586761405320473,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.2186762766666667,
      "avg_duration_per_task_ms": 724933.3133333334,
      "avg_output_tokens_per_task": 57844.28333333333,
      "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": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Reka",
      "accuracy": 0.7609786762860727,
      "accuracy_std_dev": 0.02511202768812936,
      "run_count": 16,
      "total_questions": 797,
      "avg_cost_per_task_usd": 0.054708953575909665,
      "avg_duration_per_task_ms": 274645.58720200753,
      "avg_output_tokens_per_task": 12041.819322459221,
      "last_run_timestamp": "2026-10-31T12:07:50.786Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/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": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Relace",
      "accuracy": 0.7607406666666667,
      "accuracy_std_dev": 0.034633054297881376,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.006436806666666667,
      "avg_duration_per_task_ms": 125547.04444444444,
      "avg_output_tokens_per_task": 4262.8533333333335,
      "last_run_timestamp": "2026-10-24T14:25:23.734Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Novita",
      "accuracy": 0.7606060909090909,
      "accuracy_std_dev": 0.02369062226226981,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.03764717818181818,
      "avg_duration_per_task_ms": 169270.9309090909,
      "avg_output_tokens_per_task": 16441.54909090909,
      "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": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "AkashML",
      "accuracy": 0.7604166354166666,
      "accuracy_std_dev": 0.010476531320050537,
      "run_count": 4,
      "total_questions": 192,
      "avg_cost_per_task_usd": 0.29691326041666666,
      "avg_duration_per_task_ms": 310718.984375,
      "avg_output_tokens_per_task": 7030.182291666667,
      "last_run_timestamp": "2026-10-31T12:13:14.359Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Mancer 2",
      "accuracy": 0.7603703333333334,
      "accuracy_std_dev": 0.033126907327897526,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.08524436222222224,
      "avg_duration_per_task_ms": 566210.18,
      "avg_output_tokens_per_task": 12520.904444444444,
      "last_run_timestamp": "2026-10-31T12:26:16.049Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Darkbloom",
      "accuracy": 0.7601351621621621,
      "accuracy_std_dev": 0.012280818789432917,
      "run_count": 6,
      "total_questions": 296,
      "avg_cost_per_task_usd": 0.032259290540540546,
      "avg_duration_per_task_ms": 475460.4324324324,
      "avg_output_tokens_per_task": 12381.702702702703,
      "last_run_timestamp": "2026-10-31T12:56:10.594Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "GMICloud",
      "accuracy": 0.7600002499999999,
      "accuracy_std_dev": 0.008164965810553043,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.015787874999999996,
      "avg_duration_per_task_ms": 119015.24,
      "avg_output_tokens_per_task": 12260.915,
      "last_run_timestamp": "2026-10-31T13:06:19.395Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Alibaba",
      "accuracy": 0.7600002499999999,
      "accuracy_std_dev": 0.027080189561513434,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.02989183,
      "avg_duration_per_task_ms": 125270.795,
      "avg_output_tokens_per_task": 12269.195,
      "last_run_timestamp": "2026-10-31T12:01:23.978Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "auto-routing",
      "accuracy": 0.7600001999999999,
      "accuracy_std_dev": 0.03126929626582598,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.114411384,
      "avg_duration_per_task_ms": 874872.748,
      "avg_output_tokens_per_task": 24203.156,
      "last_run_timestamp": "2026-10-31T12:44:40.958Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "SiliconFlow",
      "accuracy": 0.76,
      "accuracy_std_dev": 0.01333300000000004,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.054461335,
      "avg_duration_per_task_ms": 91417,
      "avg_output_tokens_per_task": 15101.33,
      "last_run_timestamp": "2026-10-03T08:13:12.597Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "StreamLake",
      "accuracy": 0.76,
      "accuracy_std_dev": 0.03311047725083071,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.01778674,
      "avg_duration_per_task_ms": 293727.4111111111,
      "avg_output_tokens_per_task": 11448.377777777778,
      "last_run_timestamp": "2026-10-31T20:10:47.431Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-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": "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": "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-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "Cerebras",
      "accuracy": 0.76,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.40321123999999997,
      "avg_duration_per_task_ms": 34129.92,
      "avg_output_tokens_per_task": 9245.3,
      "last_run_timestamp": "2026-08-15T14:16:11.865Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Makora",
      "accuracy": 0.76,
      "accuracy_std_dev": 0.0066669999999999785,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.07038507,
      "avg_duration_per_task_ms": 256302.49,
      "avg_output_tokens_per_task": 10045.27,
      "last_run_timestamp": "2026-08-26T00:00:05.638Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Novita",
      "accuracy": 0.7599999000000001,
      "accuracy_std_dev": 0.030840204640858015,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.008153706000000002,
      "avg_duration_per_task_ms": 200971.108,
      "avg_output_tokens_per_task": 4219.412,
      "last_run_timestamp": "2026-10-31T12:49:39.925Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "AtlasCloud",
      "accuracy": 0.759999875,
      "accuracy_std_dev": 0.030368121078844755,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.10768833249999998,
      "avg_duration_per_task_ms": 237288.72,
      "avg_output_tokens_per_task": 11168.8675,
      "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": "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": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "CoreWeave",
      "accuracy": 0.7594444166666666,
      "accuracy_std_dev": 0.02168086358388558,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.021428943333333332,
      "avg_duration_per_task_ms": 120315.54166666667,
      "avg_output_tokens_per_task": 17324.703333333335,
      "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": "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": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Sail Research",
      "accuracy": 0.7593939090909091,
      "accuracy_std_dev": 0.01916536035974823,
      "run_count": 22,
      "total_questions": 1100,
      "avg_cost_per_task_usd": 0.01099011090909091,
      "avg_duration_per_task_ms": 321334.9527272727,
      "avg_output_tokens_per_task": 14815.935454545455,
      "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": "z-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": "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": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7585184444444445,
      "accuracy_std_dev": 0.023680696011313677,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.010843522222222221,
      "avg_duration_per_task_ms": 1028208.5311111112,
      "avg_output_tokens_per_task": 13132.244444444445,
      "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": "meta/muse-glimmer-30b-20260810",
      "model_slug": "meta/muse-glimmer-30b",
      "model_name": "Meta: Muse Glimmer 30B",
      "provider_name": "Phala",
      "accuracy": 0.758333375,
      "accuracy_std_dev": 0.012360224066916245,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.034803532500000005,
      "avg_duration_per_task_ms": 275650.13,
      "avg_output_tokens_per_task": 13717.48,
      "last_run_timestamp": "2026-10-03T08:20:09.018Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.6-opus-20260205",
      "model_slug": "anthropic/claude-opus-4.6",
      "model_name": "Anthropic: Claude Opus 4.6",
      "provider_name": "Anthropic",
      "accuracy": 0.7583333333333333,
      "accuracy_std_dev": 0.027772194995874723,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.5207035750000001,
      "avg_duration_per_task_ms": 214690.15666666668,
      "avg_output_tokens_per_task": 9895.341666666667,
      "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-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Relace",
      "accuracy": 0.7583333333333333,
      "accuracy_std_dev": 0.03125451671181126,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.06620967333333333,
      "avg_duration_per_task_ms": 161431.49,
      "avg_output_tokens_per_task": 10316.483333333334,
      "last_run_timestamp": "2026-10-31T19:43:05.932Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.1-20251113",
      "model_slug": "openai/gpt-5.1",
      "model_name": "OpenAI: GPT-5.1",
      "provider_name": "Azure",
      "accuracy": 0.7583332500000001,
      "accuracy_std_dev": 0.014433785597946933,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.42515429,
      "avg_duration_per_task_ms": 460450.54,
      "avg_output_tokens_per_task": 34342.89,
      "last_run_timestamp": "2026-10-01T16:01:13.647Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Baidu",
      "accuracy": 0.7580951428571429,
      "accuracy_std_dev": 0.016605241912709113,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.048255082857142854,
      "avg_duration_per_task_ms": 69808.27142857143,
      "avg_output_tokens_per_task": 4328.277142857143,
      "last_run_timestamp": "2026-10-31T12:44:24.094Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.7578788636363636,
      "accuracy_std_dev": 0.03116509593832564,
      "run_count": 22,
      "total_questions": 1100,
      "avg_cost_per_task_usd": 0.5560903545454545,
      "avg_duration_per_task_ms": 181840.00545454546,
      "avg_output_tokens_per_task": 10618.228181818182,
      "last_run_timestamp": "2026-10-31T12:37:18.963Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Azure",
      "accuracy": 0.7577777777777778,
      "accuracy_std_dev": 0.026480824650543643,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.025315553333333334,
      "avg_duration_per_task_ms": 238703.38,
      "avg_output_tokens_per_task": 12990.084444444445,
      "last_run_timestamp": "2026-10-31T19:02:15.155Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Azure",
      "accuracy": 0.7577777777777778,
      "accuracy_std_dev": 0.01859267080077988,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.1710706822222222,
      "avg_duration_per_task_ms": 189011.57111111112,
      "avg_output_tokens_per_task": 11157.495555555555,
      "last_run_timestamp": "2026-10-31T19:18:33.522Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "CoreWeave",
      "accuracy": 0.7577776666666667,
      "accuracy_std_dev": 0.040123113009281265,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.15586289333333334,
      "avg_duration_per_task_ms": 442954.5133333333,
      "avg_output_tokens_per_task": 26206.566666666666,
      "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": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Crusoe",
      "accuracy": 0.7577776666666667,
      "accuracy_std_dev": 0.003142853940113798,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.09565631333333334,
      "avg_duration_per_task_ms": 177852.71333333335,
      "avg_output_tokens_per_task": 9687.686666666666,
      "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": "Cloudflare",
      "accuracy": 0.7575,
      "accuracy_std_dev": 0.02184208476199102,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0299009825,
      "avg_duration_per_task_ms": 207569.0425,
      "avg_output_tokens_per_task": 13839.6975,
      "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": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Cerebras",
      "accuracy": 0.7575,
      "accuracy_std_dev": 0.05171382256708548,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.26214604,
      "avg_duration_per_task_ms": 26048.875,
      "avg_output_tokens_per_task": 12109.95,
      "last_run_timestamp": "2026-10-31T11:18:07.571Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7575,
      "accuracy_std_dev": 0.023078766095699303,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.012481125,
      "avg_duration_per_task_ms": 215467.87,
      "avg_output_tokens_per_task": 4890.4975,
      "last_run_timestamp": "2026-10-31T11:56:22.568Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "NextBit",
      "accuracy": 0.7573333999999999,
      "accuracy_std_dev": 0.03991088294989224,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.023026752,
      "avg_duration_per_task_ms": 161149.652,
      "avg_output_tokens_per_task": 18473.992,
      "last_run_timestamp": "2026-10-03T08:55:31.322Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "Alibaba",
      "accuracy": 0.7571111333333334,
      "accuracy_std_dev": 0.02820195562927429,
      "run_count": 15,
      "total_questions": 750,
      "avg_cost_per_task_usd": 0.09072873866666667,
      "avg_duration_per_task_ms": 188985.284,
      "avg_output_tokens_per_task": 11040.473333333333,
      "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": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "StreamLake",
      "accuracy": 0.757037111111111,
      "accuracy_std_dev": 0.029162999653078087,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.04487014222222222,
      "avg_duration_per_task_ms": 185453.6288888889,
      "avg_output_tokens_per_task": 9571.857777777777,
      "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": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "Phala",
      "accuracy": 0.7569698181818182,
      "accuracy_std_dev": 0.01591999626437932,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.07314548,
      "avg_duration_per_task_ms": 513687.81272727274,
      "avg_output_tokens_per_task": 32116.561818181817,
      "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": "z-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": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "DeepInfra",
      "accuracy": 0.7568491678082192,
      "accuracy_std_dev": 0.014345100754845415,
      "run_count": 6,
      "total_questions": 292,
      "avg_cost_per_task_usd": 0.03330520547945206,
      "avg_duration_per_task_ms": 2320790.7157534244,
      "avg_output_tokens_per_task": 51570.45205479452,
      "last_run_timestamp": "2026-10-09T04:05:56.641Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Cloudflare",
      "accuracy": 0.756708080495356,
      "accuracy_std_dev": 0.027120750325406145,
      "run_count": 13,
      "total_questions": 646,
      "avg_cost_per_task_usd": 0.19143365634674925,
      "avg_duration_per_task_ms": 498299.07585139317,
      "avg_output_tokens_per_task": 20833.128482972137,
      "last_run_timestamp": "2026-10-31T14:12:32.839Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta/muse-spark-open-20261015",
      "model_slug": "meta/muse-spark-open",
      "model_name": "Meta: Muse Spark Open",
      "provider_name": "auto-routing",
      "accuracy": 0.7566669999999999,
      "accuracy_std_dev": 0.010000000000000009,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.08839812,
      "avg_duration_per_task_ms": 73380.11,
      "avg_output_tokens_per_task": 8719.83,
      "last_run_timestamp": "2026-10-31T12:44:24.100Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7566666666666667,
      "accuracy_std_dev": 0.022026835486641215,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.06927284666666667,
      "avg_duration_per_task_ms": 313911.26,
      "avg_output_tokens_per_task": 14319.526666666667,
      "last_run_timestamp": "2026-10-31T12:43:48.271Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Reka",
      "accuracy": 0.7564444666666666,
      "accuracy_std_dev": 0.023958871344776582,
      "run_count": 15,
      "total_questions": 750,
      "avg_cost_per_task_usd": 0.04415017200000001,
      "avg_duration_per_task_ms": 138405.18933333334,
      "avg_output_tokens_per_task": 4510.322666666667,
      "last_run_timestamp": "2026-10-31T12:43:47.323Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Venice",
      "accuracy": 0.7563636363636363,
      "accuracy_std_dev": 0.033917706275022164,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.019490098181818183,
      "avg_duration_per_task_ms": 208148.6509090909,
      "avg_output_tokens_per_task": 15196.68909090909,
      "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": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Alibaba",
      "accuracy": 0.7562961111111112,
      "accuracy_std_dev": 0.02966657881426859,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.03134207111111111,
      "avg_duration_per_task_ms": 175492.92222222223,
      "avg_output_tokens_per_task": 14182.893333333333,
      "last_run_timestamp": "2026-10-31T11:54:37.503Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Alibaba",
      "accuracy": 0.7561521722595078,
      "accuracy_std_dev": 0.03542861834034983,
      "run_count": 9,
      "total_questions": 447,
      "avg_cost_per_task_usd": 0.10875903355704698,
      "avg_duration_per_task_ms": 155583.68456375838,
      "avg_output_tokens_per_task": 9207.604026845638,
      "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-pro-20260423",
      "model_slug": "deepseek/deepseek-v4-pro",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0423",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7559789775784755,
      "accuracy_std_dev": 0.038608976715106755,
      "run_count": 9,
      "total_questions": 446,
      "avg_cost_per_task_usd": 0.090114966367713,
      "avg_duration_per_task_ms": 617883.6928251121,
      "avg_output_tokens_per_task": 10068.533632286995,
      "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": "z-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": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "AtlasCloud",
      "accuracy": 0.75583325,
      "accuracy_std_dev": 0.028855413030963544,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.012369372500000001,
      "avg_duration_per_task_ms": 223619.505,
      "avg_output_tokens_per_task": 13737.5625,
      "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-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "GMICloud",
      "accuracy": 0.7557575454545454,
      "accuracy_std_dev": 0.021134116315326572,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.0626866909090909,
      "avg_duration_per_task_ms": 166564.66363636364,
      "avg_output_tokens_per_task": 4635.490909090909,
      "last_run_timestamp": "2026-10-31T13:36:23.199Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "StreamLake",
      "accuracy": 0.7555555,
      "accuracy_std_dev": 0.015947487129638963,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.0058668516666666665,
      "avg_duration_per_task_ms": 265133.9683333333,
      "avg_output_tokens_per_task": 15836.96,
      "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": "z-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": "Mancer 2",
      "accuracy": 0.7553333,
      "accuracy_std_dev": 0.03334003466119973,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.04678315799999999,
      "avg_duration_per_task_ms": 241838.904,
      "avg_output_tokens_per_task": 14405.774,
      "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": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Makora",
      "accuracy": 0.7552378571428572,
      "accuracy_std_dev": 0.014126057041900291,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.009880265714285713,
      "avg_duration_per_task_ms": 245506.26571428572,
      "avg_output_tokens_per_task": 16074.99142857143,
      "last_run_timestamp": "2026-10-31T12:27:44.073Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "Darkbloom",
      "accuracy": 0.7552085625,
      "accuracy_std_dev": 0.029782500000000045,
      "run_count": 2,
      "total_questions": 96,
      "avg_cost_per_task_usd": 0.02997340625,
      "avg_duration_per_task_ms": 1679708.6145833333,
      "avg_output_tokens_per_task": 72130.22916666667,
      "last_run_timestamp": "2026-10-31T20:01:41.151Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta/muse-spark-open-20261015",
      "model_slug": "meta/muse-spark-open",
      "model_name": "Meta: Muse Spark Open",
      "provider_name": "Modal",
      "accuracy": 0.7551024285714285,
      "accuracy_std_dev": 0.011805499999999969,
      "run_count": 2,
      "total_questions": 98,
      "avg_cost_per_task_usd": 0.27501580612244897,
      "avg_duration_per_task_ms": 65997.72448979592,
      "avg_output_tokens_per_task": 8842.591836734693,
      "last_run_timestamp": "2026-10-31T12:36:21.331Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-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": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Modal",
      "accuracy": 0.7549998,
      "accuracy_std_dev": 0.029636688974310207,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.014858368,
      "avg_duration_per_task_ms": 71120.474,
      "avg_output_tokens_per_task": 4319.272,
      "last_run_timestamp": "2026-10-31T11:35:00.891Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Venice",
      "accuracy": 0.7547041935483871,
      "accuracy_std_dev": 0.023888446777469652,
      "run_count": 10,
      "total_questions": 496,
      "avg_cost_per_task_usd": 0.01950188306451613,
      "avg_duration_per_task_ms": 216554.71370967742,
      "avg_output_tokens_per_task": 4241.487903225807,
      "last_run_timestamp": "2026-10-31T11:51:07.775Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "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-sonnet-5-20260630",
      "model_slug": "anthropic/claude-sonnet-5",
      "model_name": "Anthropic: Claude Sonnet 5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7541666875,
      "accuracy_std_dev": 0.02875898591575064,
      "run_count": 32,
      "total_questions": 1600,
      "avg_cost_per_task_usd": 0.20787477937499993,
      "avg_duration_per_task_ms": 114719.823125,
      "avg_output_tokens_per_task": 7707.7675,
      "last_run_timestamp": "2026-10-31T12:48:39.187Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "Sail Research",
      "accuracy": 0.7540880930817611,
      "accuracy_std_dev": 0.02434701736834258,
      "run_count": 16,
      "total_questions": 795,
      "avg_cost_per_task_usd": 0.08681218238993711,
      "avg_duration_per_task_ms": 616702.9220125786,
      "avg_output_tokens_per_task": 21133.411320754716,
      "last_run_timestamp": "2026-11-06T22:40:08.138Z"
    },
    {
      "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": 0.7539393636363636,
      "accuracy_std_dev": 0.030411866372049668,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.5373833490909091,
      "avg_duration_per_task_ms": 189228.24181818182,
      "avg_output_tokens_per_task": 10661.903636363637,
      "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": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "DeepInfra",
      "accuracy": 0.7538888333333333,
      "accuracy_std_dev": 0.023007058767667125,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.007775635,
      "avg_duration_per_task_ms": 226184.26333333334,
      "avg_output_tokens_per_task": 14761.815,
      "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": "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": "tencent/hy4-preview-20260827",
      "model_slug": "tencent/hy4-preview",
      "model_name": "Tencent: Hy4 preview",
      "provider_name": "GMICloud",
      "accuracy": 0.7533335,
      "accuracy_std_dev": 0.00666650000000002,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.07113415,
      "avg_duration_per_task_ms": 677434.95,
      "avg_output_tokens_per_task": 13100.77,
      "last_run_timestamp": "2026-10-31T13:06:57.362Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Ambient",
      "accuracy": 0.7533333333333333,
      "accuracy_std_dev": 0.019625871060640607,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.06505746,
      "avg_duration_per_task_ms": 250536.72666666665,
      "avg_output_tokens_per_task": 8651.606666666667,
      "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": "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": "AtlasCloud",
      "accuracy": 0.7533332727272728,
      "accuracy_std_dev": 0.037172733573000175,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.044545774545454546,
      "avg_duration_per_task_ms": 178434.63818181818,
      "avg_output_tokens_per_task": 19162.814545454545,
      "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": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "auto-routing",
      "accuracy": 0.7533332500000001,
      "accuracy_std_dev": 0.030731887168501403,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0298026275,
      "avg_duration_per_task_ms": 427042.9575,
      "avg_output_tokens_per_task": 11770.6425,
      "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-v4.1-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": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "CoreWeave",
      "accuracy": 0.7528342925170068,
      "accuracy_std_dev": 0.024943599437308363,
      "run_count": 3,
      "total_questions": 147,
      "avg_cost_per_task_usd": 0.1277965918367347,
      "avg_duration_per_task_ms": 243630.26530612246,
      "avg_output_tokens_per_task": 45142.87074829932,
      "last_run_timestamp": "2026-08-29T17:18:52.408Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-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": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Venice",
      "accuracy": 0.7525,
      "accuracy_std_dev": 0.022715180925979885,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.02207197,
      "avg_duration_per_task_ms": 477011.815,
      "avg_output_tokens_per_task": 13991.755,
      "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": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.752499875,
      "accuracy_std_dev": 0.028173595055643437,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.18478309000000004,
      "avg_duration_per_task_ms": 179248.5325,
      "avg_output_tokens_per_task": 9658.775,
      "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": "anthropic/claude-opus-5.5-20260921",
      "model_slug": "anthropic/claude-opus-5.5",
      "model_name": "Anthropic: Claude Opus 5.5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7523809047619048,
      "accuracy_std_dev": 0.017612644913472624,
      "run_count": 21,
      "total_questions": 1050,
      "avg_cost_per_task_usd": 0.36889707904761904,
      "avg_duration_per_task_ms": 141180.9876190476,
      "avg_output_tokens_per_task": 7127.063809523809,
      "last_run_timestamp": "2026-10-31T12:26:17.125Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "GMICloud",
      "accuracy": 0.7523076923076923,
      "accuracy_std_dev": 0.0332661532107905,
      "run_count": 13,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.10102499846153845,
      "avg_duration_per_task_ms": 302464.04,
      "avg_output_tokens_per_task": 18714.755384615386,
      "last_run_timestamp": "2026-10-31T13:11:23.394Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "Google",
      "accuracy": 0.7522221666666666,
      "accuracy_std_dev": 0.016960131174971006,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.12217692666666667,
      "avg_duration_per_task_ms": 121673.44,
      "avg_output_tokens_per_task": 10724.803333333333,
      "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/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "Darkbloom",
      "accuracy": 0.7519260539291217,
      "accuracy_std_dev": 0.019651146282719818,
      "run_count": 13,
      "total_questions": 649,
      "avg_cost_per_task_usd": 0.027055047765793528,
      "avg_duration_per_task_ms": 539313.875192604,
      "avg_output_tokens_per_task": 22031.878274268103,
      "last_run_timestamp": "2026-10-31T19:23:15.539Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "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": "anthropic/claude-sonnet-5-20260630",
      "model_slug": "anthropic/claude-sonnet-5",
      "model_name": "Anthropic: Claude Sonnet 5",
      "provider_name": "Azure",
      "accuracy": 0.7513043478260869,
      "accuracy_std_dev": 0.029647460490422974,
      "run_count": 23,
      "total_questions": 1150,
      "avg_cost_per_task_usd": 0.20092619999999997,
      "avg_duration_per_task_ms": 96197.9252173913,
      "avg_output_tokens_per_task": 7585.031304347826,
      "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": "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": "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": 0.751111,
      "accuracy_std_dev": 0.022662199996175744,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.05099817333333333,
      "avg_duration_per_task_ms": 373986.0133333333,
      "avg_output_tokens_per_task": 8579.546666666667,
      "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": "anthropic/claude-sonnet-5-20260630",
      "model_slug": "anthropic/claude-sonnet-5",
      "model_name": "Anthropic: Claude Sonnet 5",
      "provider_name": "auto-routing",
      "accuracy": 0.7509091818181818,
      "accuracy_std_dev": 0.030784008944486355,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.20063707090909091,
      "avg_duration_per_task_ms": 117571.03454545455,
      "avg_output_tokens_per_task": 7764.232727272727,
      "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": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "Parasail",
      "accuracy": 0.750909,
      "accuracy_std_dev": 0.015640538557339927,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.03659924545454545,
      "avg_duration_per_task_ms": 266725.77636363637,
      "avg_output_tokens_per_task": 21362.425454545453,
      "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": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "CoreWeave",
      "accuracy": 0.750833375,
      "accuracy_std_dev": 0.029237916546573147,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.025213935,
      "avg_duration_per_task_ms": 300217.2775,
      "avg_output_tokens_per_task": 10899.815,
      "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": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Decart",
      "accuracy": 0.7506668,
      "accuracy_std_dev": 0.025854254755455656,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.264866352,
      "avg_duration_per_task_ms": 264537.464,
      "avg_output_tokens_per_task": 10440.212,
      "last_run_timestamp": "2026-10-24T11:49:28.605Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Near AI",
      "accuracy": 0.7506666,
      "accuracy_std_dev": 0.03785357285435553,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.007070539999999999,
      "avg_duration_per_task_ms": 292989.264,
      "avg_output_tokens_per_task": 3954.392,
      "last_run_timestamp": "2026-10-31T13:04:37.434Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7504761428571428,
      "accuracy_std_dev": 0.04779025223509368,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.020752462857142853,
      "avg_duration_per_task_ms": 333009.8742857143,
      "avg_output_tokens_per_task": 11171.528571428571,
      "last_run_timestamp": "2026-10-31T13:00:18.168Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "ModelRun",
      "accuracy": 0.7502563846153846,
      "accuracy_std_dev": 0.02447608447990521,
      "run_count": 13,
      "total_questions": 650,
      "avg_cost_per_task_usd": 0.09372697076923077,
      "avg_duration_per_task_ms": 135890.0169230769,
      "avg_output_tokens_per_task": 12287.435384615384,
      "last_run_timestamp": "2026-10-31T18:56:42.115Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "SiliconFlow",
      "accuracy": 0.750000125,
      "accuracy_std_dev": 0.019720301160463437,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.013962239999999999,
      "avg_duration_per_task_ms": 241952.985,
      "avg_output_tokens_per_task": 14693.71,
      "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": "deepseek/deepseek-v4-pro-20260813",
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "model_name": "DeepSeek: DeepSeek V4 Pro 0813",
      "provider_name": "BaseTen",
      "accuracy": 0.75,
      "accuracy_std_dev": 0.0033330000000000304,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.28014248,
      "avg_duration_per_task_ms": 223020.56,
      "avg_output_tokens_per_task": 18715.38,
      "last_run_timestamp": "2026-09-12T08:12:13.731Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.4-20260305",
      "model_slug": "openai/gpt-5.4",
      "model_name": "OpenAI: GPT-5.4",
      "provider_name": "auto-routing",
      "accuracy": 0.75,
      "accuracy_std_dev": 0.0033330000000000304,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.29130333999999997,
      "avg_duration_per_task_ms": 176962.37,
      "avg_output_tokens_per_task": 10386.26,
      "last_run_timestamp": "2026-10-01T19:04:36.169Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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-sonnet-5-20260630",
      "model_slug": "anthropic/claude-sonnet-5",
      "model_name": "Anthropic: Claude Sonnet 5",
      "provider_name": "Google",
      "accuracy": 0.7498989696969697,
      "accuracy_std_dev": 0.03411693737496916,
      "run_count": 33,
      "total_questions": 1650,
      "avg_cost_per_task_usd": 0.20618370909090916,
      "avg_duration_per_task_ms": 91953.36,
      "avg_output_tokens_per_task": 7458.127878787879,
      "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/qwen3.5-122b-a10b-20260224",
      "model_slug": "qwen/qwen3.5-122b-a10b",
      "model_name": "Qwen: Qwen3.5-122B-A10B",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7496970909090909,
      "accuracy_std_dev": 0.10488520113867744,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.1517481509090909,
      "avg_duration_per_task_ms": 559680.3290909091,
      "avg_output_tokens_per_task": 28098.66,
      "last_run_timestamp": "2026-10-31T19:37:30.834Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-opus-5.5-20260921",
      "model_slug": "anthropic/claude-opus-5.5",
      "model_name": "Anthropic: Claude Opus 5.5",
      "provider_name": "Azure",
      "accuracy": 0.749524,
      "accuracy_std_dev": 0.021633867377662402,
      "run_count": 14,
      "total_questions": 700,
      "avg_cost_per_task_usd": 0.35241304,
      "avg_duration_per_task_ms": 87048.52428571429,
      "avg_output_tokens_per_task": 6993.714285714285,
      "last_run_timestamp": "2026-10-31T11:29:30.760Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "CoreWeave",
      "accuracy": 0.7495238571428572,
      "accuracy_std_dev": 0.021338220105372136,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.016605834285714285,
      "avg_duration_per_task_ms": 72972.84285714285,
      "avg_output_tokens_per_task": 4253.202857142857,
      "last_run_timestamp": "2026-10-31T12:03:51.953Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.5-sonnet-20250929",
      "model_slug": "anthropic/claude-sonnet-4.5",
      "model_name": "Anthropic: Claude Sonnet 4.5",
      "provider_name": "Anthropic",
      "accuracy": 0.7494446666666666,
      "accuracy_std_dev": 0.029496932253861397,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.34673691166666676,
      "avg_duration_per_task_ms": 225396.725,
      "avg_output_tokens_per_task": 10875.54,
      "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": "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": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "CoreWeave",
      "accuracy": 0.748889,
      "accuracy_std_dev": 0.01911599421078245,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.019773013333333336,
      "avg_duration_per_task_ms": 225130.51333333334,
      "avg_output_tokens_per_task": 12421.726666666667,
      "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": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "DeepInfra",
      "accuracy": 0.748889,
      "accuracy_std_dev": 0.015713326891527487,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.03620722,
      "avg_duration_per_task_ms": 114735.25333333333,
      "avg_output_tokens_per_task": 14087.753333333334,
      "last_run_timestamp": "2026-10-31T11:35:01.591Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Baidu",
      "accuracy": 0.7488888888888889,
      "accuracy_std_dev": 0.029814247982557492,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.012581893333333335,
      "avg_duration_per_task_ms": 105161.7,
      "avg_output_tokens_per_task": 15935.586666666666,
      "last_run_timestamp": "2026-10-24T11:51:53.770Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.7488236470588235,
      "accuracy_std_dev": 0.024613068112981307,
      "run_count": 34,
      "total_questions": 1700,
      "avg_cost_per_task_usd": 0.3698944605882353,
      "avg_duration_per_task_ms": 152128.24411764706,
      "avg_output_tokens_per_task": 11726.421176470589,
      "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": "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": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Moonshot AI",
      "accuracy": 0.7485714285714286,
      "accuracy_std_dev": 0.04934029368103327,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.11107009428571428,
      "avg_duration_per_task_ms": 240868.85714285713,
      "avg_output_tokens_per_task": 11857.671428571428,
      "last_run_timestamp": "2026-10-24T19:31:21.574Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-sonnet-5-20260630",
      "model_slug": "anthropic/claude-sonnet-5",
      "model_name": "Anthropic: Claude Sonnet 5",
      "provider_name": "Anthropic",
      "accuracy": 0.748484909090909,
      "accuracy_std_dev": 0.035629023673135175,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.19773472545454546,
      "avg_duration_per_task_ms": 96148.09090909091,
      "avg_output_tokens_per_task": 7605.130909090909,
      "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": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Wafer",
      "accuracy": 0.748333375,
      "accuracy_std_dev": 0.024664456546503833,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0092696725,
      "avg_duration_per_task_ms": 91481.9675,
      "avg_output_tokens_per_task": 13410.6275,
      "last_run_timestamp": "2026-10-31T13:03:50.748Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta/muse-glimmer-30b-20260810",
      "model_slug": "meta/muse-glimmer-30b",
      "model_name": "Meta: Muse Glimmer 30B",
      "provider_name": "Together",
      "accuracy": 0.748333375,
      "accuracy_std_dev": 0.018782375754530515,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0418515775,
      "avg_duration_per_task_ms": 191146.0675,
      "avg_output_tokens_per_task": 12827.9725,
      "last_run_timestamp": "2026-10-03T08:33:09.729Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Modal",
      "accuracy": 0.748322067114094,
      "accuracy_std_dev": 0.02148614665313445,
      "run_count": 3,
      "total_questions": 149,
      "avg_cost_per_task_usd": 0.029814261744966444,
      "avg_duration_per_task_ms": 257681.3489932886,
      "avg_output_tokens_per_task": 10900.818791946309,
      "last_run_timestamp": "2026-10-31T12:26:05.030Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "auto-routing",
      "accuracy": 0.7481482222222222,
      "accuracy_std_dev": 0.020068464436510978,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.19089935333333333,
      "avg_duration_per_task_ms": 286228.22222222225,
      "avg_output_tokens_per_task": 11572.273333333333,
      "last_run_timestamp": "2026-10-31T12:37:16.632Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "auto-routing",
      "accuracy": 0.7480002,
      "accuracy_std_dev": 0.020396130352593862,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.030171100000000003,
      "avg_duration_per_task_ms": 197857.312,
      "avg_output_tokens_per_task": 18663.832,
      "last_run_timestamp": "2026-10-03T08:18:33.163Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "BaseTen",
      "accuracy": 0.74791675,
      "accuracy_std_dev": 0.031222055284486006,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.01487202,
      "avg_duration_per_task_ms": 51526.7675,
      "avg_output_tokens_per_task": 4120.2525,
      "last_run_timestamp": "2026-10-31T13:13:13.022Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Inceptron",
      "accuracy": 0.7478510515759312,
      "accuracy_std_dev": 0.031000808296901307,
      "run_count": 7,
      "total_questions": 349,
      "avg_cost_per_task_usd": 0.018540833810888252,
      "avg_duration_per_task_ms": 126193.2005730659,
      "avg_output_tokens_per_task": 4035.12893982808,
      "last_run_timestamp": "2026-10-31T11:27:58.748Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Crusoe",
      "accuracy": 0.7477776666666667,
      "accuracy_std_dev": 0.028393017748891882,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.020487799999999997,
      "avg_duration_per_task_ms": 112394.16,
      "avg_output_tokens_per_task": 6879.653333333334,
      "last_run_timestamp": "2026-10-31T12:42:03.915Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-122b-a10b-20260224",
      "model_slug": "qwen/qwen3.5-122b-a10b",
      "model_name": "Qwen: Qwen3.5-122B-A10B",
      "provider_name": "Alibaba",
      "accuracy": 0.7476435861182519,
      "accuracy_std_dev": 0.0440966339389981,
      "run_count": 8,
      "total_questions": 389,
      "avg_cost_per_task_usd": 0.10830866066838046,
      "avg_duration_per_task_ms": 421756.3444730077,
      "avg_output_tokens_per_task": 20888.023136246786,
      "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": "x-ai/grok-4.6-20260810",
      "model_slug": "x-ai/grok-4.6",
      "model_name": "SpaceXAI: Grok 4.6",
      "provider_name": "xAI",
      "accuracy": 0.747619,
      "accuracy_std_dev": 0.025305279849978234,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.33171734571428574,
      "avg_duration_per_task_ms": 235739.5942857143,
      "avg_output_tokens_per_task": 12164.34,
      "last_run_timestamp": "2026-10-01T16:01:04.865Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Alibaba",
      "accuracy": 0.7474576237288135,
      "accuracy_std_dev": 0.020378547413068143,
      "run_count": 6,
      "total_questions": 295,
      "avg_cost_per_task_usd": 0.010754061016949153,
      "avg_duration_per_task_ms": 127184.37966101695,
      "avg_output_tokens_per_task": 12239.813559322034,
      "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": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Wafer",
      "accuracy": 0.7473684210526316,
      "accuracy_std_dev": 0.02465141176979054,
      "run_count": 19,
      "total_questions": 950,
      "avg_cost_per_task_usd": 0.02123823368421052,
      "avg_duration_per_task_ms": 106810.42210526316,
      "avg_output_tokens_per_task": 14307.62947368421,
      "last_run_timestamp": "2026-10-31T13:23:12.329Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.5-opus-20251124",
      "model_slug": "anthropic/claude-opus-4.5",
      "model_name": "Anthropic: Claude Opus 4.5",
      "provider_name": "Azure",
      "accuracy": 0.7472727272727273,
      "accuracy_std_dev": 0.024197018843767105,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.5306860090909091,
      "avg_duration_per_task_ms": 186964.4309090909,
      "avg_output_tokens_per_task": 10402.76909090909,
      "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": "Parasail",
      "accuracy": 0.7472727272727273,
      "accuracy_std_dev": 0.02643158433008411,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.018947145454545452,
      "avg_duration_per_task_ms": 214213.9818181818,
      "avg_output_tokens_per_task": 15639.545454545454,
      "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": "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-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "StreamLake",
      "accuracy": 0.746667,
      "accuracy_std_dev": 0.016329931618554536,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.020258,
      "avg_duration_per_task_ms": 489464.5733333333,
      "avg_output_tokens_per_task": 14818,
      "last_run_timestamp": "2026-08-29T16:31:42.782Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "Relace",
      "accuracy": 0.746667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.00930136,
      "avg_duration_per_task_ms": 89811.58,
      "avg_output_tokens_per_task": 12020.36,
      "last_run_timestamp": "2026-10-24T15:18:33.444Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "AtlasCloud",
      "accuracy": 0.746667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.07114796,
      "avg_duration_per_task_ms": 129363.94,
      "avg_output_tokens_per_task": 8164.92,
      "last_run_timestamp": "2026-08-15T08:13:33.738Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Io Net",
      "accuracy": 0.746667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.473986,
      "avg_duration_per_task_ms": 617970.5,
      "avg_output_tokens_per_task": 17752,
      "last_run_timestamp": "2026-11-04T17:25:46.000Z"
    },
    {
      "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": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "Baidu",
      "accuracy": 0.7466667499999999,
      "accuracy_std_dev": 0.022607717464120523,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.06205405,
      "avg_duration_per_task_ms": 201734.45,
      "avg_output_tokens_per_task": 11582.5475,
      "last_run_timestamp": "2026-10-31T19:58:43.811Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.6-sonnet-20260217",
      "model_slug": "anthropic/claude-sonnet-4.6",
      "model_name": "Anthropic: Claude Sonnet 4.6",
      "provider_name": "Azure",
      "accuracy": 0.7466667272727272,
      "accuracy_std_dev": 0.01969470010430084,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.3479023018181817,
      "avg_duration_per_task_ms": 215177.51454545456,
      "avg_output_tokens_per_task": 11904.012727272728,
      "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": "meta/muse-glimmer-30b-20260810",
      "model_slug": "meta/muse-glimmer-30b",
      "model_name": "Meta: Muse Glimmer 30B",
      "provider_name": "auto-routing",
      "accuracy": 0.7466665714285714,
      "accuracy_std_dev": 0.03380611384628165,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.03743281428571429,
      "avg_duration_per_task_ms": 197093.65714285715,
      "avg_output_tokens_per_task": 13601.3,
      "last_run_timestamp": "2026-10-03T08:15:57.213Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Reka",
      "accuracy": 0.7466665625000001,
      "accuracy_std_dev": 0.022360772946078904,
      "run_count": 16,
      "total_questions": 800,
      "avg_cost_per_task_usd": 0.013514558749999999,
      "avg_duration_per_task_ms": 139486.7775,
      "avg_output_tokens_per_task": 15486.175,
      "last_run_timestamp": "2026-10-31T13:03:01.037Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "DeepInfra",
      "accuracy": 0.7466665,
      "accuracy_std_dev": 0.026666500000000037,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.00520861,
      "avg_duration_per_task_ms": 148091.91,
      "avg_output_tokens_per_task": 12193.49,
      "last_run_timestamp": "2026-10-31T11:47:33.345Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta/muse-glimmer-30b-20260810",
      "model_slug": "meta/muse-glimmer-30b",
      "model_name": "Meta: Muse Glimmer 30B",
      "provider_name": "Fireworks",
      "accuracy": 0.7466665,
      "accuracy_std_dev": 0.011547101609639251,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.04422623666666667,
      "avg_duration_per_task_ms": 266653.04,
      "avg_output_tokens_per_task": 13360.296666666667,
      "last_run_timestamp": "2026-09-19T08:01:36.427Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.5-20260423",
      "model_slug": "openai/gpt-5.5",
      "model_name": "OpenAI: GPT-5.5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7466665,
      "accuracy_std_dev": 0.00666650000000002,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.58250779,
      "avg_duration_per_task_ms": 102747.13,
      "avg_output_tokens_per_task": 7546.85,
      "last_run_timestamp": "2026-10-01T19:04:42.459Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Morph",
      "accuracy": 0.7466663333333334,
      "accuracy_std_dev": 0.033993463423951875,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.01425808,
      "avg_duration_per_task_ms": 139346.08,
      "avg_output_tokens_per_task": 12850.746666666666,
      "last_run_timestamp": "2026-10-31T12:18:46.817Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Wafer",
      "accuracy": 0.7463769999999998,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 46,
      "avg_cost_per_task_usd": 0.05924082608695652,
      "avg_duration_per_task_ms": 289784.32608695654,
      "avg_output_tokens_per_task": 9531.695652173914,
      "last_run_timestamp": "2026-08-15T14:23:29.125Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Phala",
      "accuracy": 0.7460605454545455,
      "accuracy_std_dev": 0.03392861625494335,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.03215642727272727,
      "avg_duration_per_task_ms": 233795.59636363637,
      "avg_output_tokens_per_task": 16719.21090909091,
      "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": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Venice",
      "accuracy": 0.745833375,
      "accuracy_std_dev": 0.03307177485007986,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0937693475,
      "avg_duration_per_task_ms": 255498.785,
      "avg_output_tokens_per_task": 6773.36,
      "last_run_timestamp": "2026-10-31T19:33:49.408Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-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": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "DeepInfra",
      "accuracy": 0.7457911447811448,
      "accuracy_std_dev": 0.035718498385523795,
      "run_count": 6,
      "total_questions": 297,
      "avg_cost_per_task_usd": 0.0683820404040404,
      "avg_duration_per_task_ms": 479113.58585858584,
      "avg_output_tokens_per_task": 10595.451178451178,
      "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": "openai/gpt-5.5-20260423",
      "model_slug": "openai/gpt-5.5",
      "model_name": "OpenAI: GPT-5.5",
      "provider_name": "Azure",
      "accuracy": 0.7457142857142857,
      "accuracy_std_dev": 0.029692331634924993,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.5093009542857143,
      "avg_duration_per_task_ms": 175938.91142857142,
      "avg_output_tokens_per_task": 7450.982857142857,
      "last_run_timestamp": "2026-10-01T19:05:47.886Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "auto-routing",
      "accuracy": 0.7454546363636363,
      "accuracy_std_dev": 0.027239077558379347,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.02595640363636364,
      "avg_duration_per_task_ms": 149834.41636363635,
      "avg_output_tokens_per_task": 9437.201818181818,
      "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": 0.7454545454545455,
      "accuracy_std_dev": 0.02753421112779068,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.5068107745454545,
      "avg_duration_per_task_ms": 192436.82727272727,
      "avg_output_tokens_per_task": 10076.376363636364,
      "last_run_timestamp": "2026-10-31T12:10:56.773Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "Together",
      "accuracy": 0.7453394630872483,
      "accuracy_std_dev": 0.030816218886236597,
      "run_count": 9,
      "total_questions": 447,
      "avg_cost_per_task_usd": 0.06666300223713646,
      "avg_duration_per_task_ms": 1014852.2460850112,
      "avg_output_tokens_per_task": 75497.7091722595,
      "last_run_timestamp": "2026-10-24T19:38:23.433Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-2.4t-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": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Baidu",
      "accuracy": 0.745,
      "accuracy_std_dev": 0.024664481886915845,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0069834499999999996,
      "avg_duration_per_task_ms": 152934.4,
      "avg_output_tokens_per_task": 12714.7475,
      "last_run_timestamp": "2026-10-31T19:25:36.518Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Novita",
      "accuracy": 0.745,
      "accuracy_std_dev": 0.03840572873934308,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.014579895,
      "avg_duration_per_task_ms": 156494.315,
      "avg_output_tokens_per_task": 14989.845,
      "last_run_timestamp": "2026-10-03T16:15:22.012Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Mistral",
      "accuracy": 0.745,
      "accuracy_std_dev": 0.008660254037844393,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.033786329999999996,
      "avg_duration_per_task_ms": 75426.93,
      "avg_output_tokens_per_task": 13154.035,
      "last_run_timestamp": "2026-11-02T21:36:49.825Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "SambaNova",
      "accuracy": 0.745,
      "accuracy_std_dev": 0.026822730826670128,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.15284815,
      "avg_duration_per_task_ms": 494527.49,
      "avg_output_tokens_per_task": 28246.935,
      "last_run_timestamp": "2026-10-17T19:47:21.934Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Phala",
      "accuracy": 0.744999875,
      "accuracy_std_dev": 0.030867001888900313,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0114373625,
      "avg_duration_per_task_ms": 404697.66,
      "avg_output_tokens_per_task": 5754.185,
      "last_run_timestamp": "2026-10-31T12:24:48.023Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Morph",
      "accuracy": 0.7448979591836735,
      "accuracy_std_dev": 0.02585427538338681,
      "run_count": 5,
      "total_questions": 245,
      "avg_cost_per_task_usd": 0.01656998775510204,
      "avg_duration_per_task_ms": 900521.2571428572,
      "avg_output_tokens_per_task": 22085.71020408163,
      "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": "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": "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": "anthropic/claude-4.6-opus-20260205",
      "model_slug": "anthropic/claude-opus-4.6",
      "model_name": "Anthropic: Claude Opus 4.6",
      "provider_name": "Azure",
      "accuracy": 0.7442423636363636,
      "accuracy_std_dev": 0.025784247154460715,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.5240637272727273,
      "avg_duration_per_task_ms": 222945.58545454545,
      "avg_output_tokens_per_task": 9999.456363636364,
      "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": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "AtlasCloud",
      "accuracy": 0.74416675,
      "accuracy_std_dev": 0.04241809591657196,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.11267837500000001,
      "avg_duration_per_task_ms": 196024.8775,
      "avg_output_tokens_per_task": 9614.775,
      "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": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "auto-routing",
      "accuracy": 0.744166625,
      "accuracy_std_dev": 0.03381937006708397,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0447137225,
      "avg_duration_per_task_ms": 180624.91,
      "avg_output_tokens_per_task": 9433.0525,
      "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": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "AkashML",
      "accuracy": 0.744012664310954,
      "accuracy_std_dev": 0.013345937402400357,
      "run_count": 17,
      "total_questions": 849,
      "avg_cost_per_task_usd": 0.0449638492343934,
      "avg_duration_per_task_ms": 327500.66313309775,
      "avg_output_tokens_per_task": 26020.465253239105,
      "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": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Morph",
      "accuracy": 0.74400005,
      "accuracy_std_dev": 0.022050473415496093,
      "run_count": 20,
      "total_questions": 1000,
      "avg_cost_per_task_usd": 0.43035259,
      "avg_duration_per_task_ms": 608815.123,
      "avg_output_tokens_per_task": 10549.796,
      "last_run_timestamp": "2026-10-31T11:24:57.351Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Phala",
      "accuracy": 0.744,
      "accuracy_std_dev": 0.04013309523821955,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.12688716400000002,
      "avg_duration_per_task_ms": 700323.218,
      "avg_output_tokens_per_task": 36239.088,
      "last_run_timestamp": "2026-10-31T12:47:12.433Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Io Net",
      "accuracy": 0.7438595789473685,
      "accuracy_std_dev": 0.004074000000000022,
      "run_count": 2,
      "total_questions": 95,
      "avg_cost_per_task_usd": 0.022313326315789476,
      "avg_duration_per_task_ms": 210423.8947368421,
      "avg_output_tokens_per_task": 9630.8,
      "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": "z-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": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Venice",
      "accuracy": 0.7436363636363637,
      "accuracy_std_dev": 0.025324294459427084,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.14332595090909092,
      "avg_duration_per_task_ms": 579280.3727272727,
      "avg_output_tokens_per_task": 26391.907272727272,
      "last_run_timestamp": "2026-10-31T19:26:38.406Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "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": "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": "anthropic/claude-4.6-opus-20260205",
      "model_slug": "anthropic/claude-opus-4.6",
      "model_name": "Anthropic: Claude Opus 4.6",
      "provider_name": "Google",
      "accuracy": 0.7430302272727273,
      "accuracy_std_dev": 0.0266459682076121,
      "run_count": 22,
      "total_questions": 1100,
      "avg_cost_per_task_usd": 0.5452918245454546,
      "avg_duration_per_task_ms": 200617.91363636364,
      "avg_output_tokens_per_task": 9759.658181818182,
      "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": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "ModelRun",
      "accuracy": 0.7429168125,
      "accuracy_std_dev": 0.02570869361037903,
      "run_count": 16,
      "total_questions": 800,
      "avg_cost_per_task_usd": 0.05779907625,
      "avg_duration_per_task_ms": 136331.88625,
      "avg_output_tokens_per_task": 9998.7625,
      "last_run_timestamp": "2026-10-31T19:13:33.926Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "GMICloud",
      "accuracy": 0.7426667,
      "accuracy_std_dev": 0.028472143112347543,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.020416034,
      "avg_duration_per_task_ms": 153083.812,
      "avg_output_tokens_per_task": 16185.95,
      "last_run_timestamp": "2026-10-31T12:48:17.235Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-2.4t-a95b-20260812",
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "model_name": "Qwen: Qwen3.8 2.4T A95B",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7426666000000001,
      "accuracy_std_dev": 0.04434197309367279,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.16675142799999998,
      "avg_duration_per_task_ms": 302047.356,
      "avg_output_tokens_per_task": 9682.606,
      "last_run_timestamp": "2026-10-31T11:21:43.861Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Sail Research",
      "accuracy": 0.7425926666666667,
      "accuracy_std_dev": 0.03165311687654157,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.013586744444444443,
      "avg_duration_per_task_ms": 236886.98,
      "avg_output_tokens_per_task": 22693.746666666666,
      "last_run_timestamp": "2026-10-24T12:29:34.092Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Mara",
      "accuracy": 0.7425405959079282,
      "accuracy_std_dev": 0.026283120985422872,
      "run_count": 8,
      "total_questions": 391,
      "avg_cost_per_task_usd": 0.12785004603580563,
      "avg_duration_per_task_ms": 203562.63938618926,
      "avg_output_tokens_per_task": 10880.268542199488,
      "last_run_timestamp": "2026-10-31T19:27:07.786Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "auto-routing",
      "accuracy": 0.742500125,
      "accuracy_std_dev": 0.027271738271686614,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.041572885,
      "avg_duration_per_task_ms": 596372.7425,
      "avg_output_tokens_per_task": 15967.995,
      "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": "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": 0.7424243636363637,
      "accuracy_std_dev": 0.03232057912192878,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.5451988545454546,
      "avg_duration_per_task_ms": 180223.92,
      "avg_output_tokens_per_task": 10245.276363636363,
      "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": "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-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7422223333333333,
      "accuracy_std_dev": 0.03456997912768952,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.06651346666666666,
      "avg_duration_per_task_ms": 1404270.7533333334,
      "avg_output_tokens_per_task": 9023.02,
      "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": "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": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Modal",
      "accuracy": 0.7418518888888889,
      "accuracy_std_dev": 0.020793479247347656,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.3703597533333333,
      "avg_duration_per_task_ms": 185758.07111111112,
      "avg_output_tokens_per_task": 10963.044444444444,
      "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": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Phala",
      "accuracy": 0.741666625,
      "accuracy_std_dev": 0.025331164930464133,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.021842495,
      "avg_duration_per_task_ms": 141327.99,
      "avg_output_tokens_per_task": 16250.775,
      "last_run_timestamp": "2026-10-31T12:24:54.205Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "CoreWeave",
      "accuracy": 0.741666625,
      "accuracy_std_dev": 0.030866893899198455,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.021115824999999998,
      "avg_duration_per_task_ms": 150544.0025,
      "avg_output_tokens_per_task": 8219.325,
      "last_run_timestamp": "2026-10-31T18:57:59.813Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Relace",
      "accuracy": 0.7416481202672606,
      "accuracy_std_dev": 0.022563192218990436,
      "run_count": 9,
      "total_questions": 449,
      "avg_cost_per_task_usd": 0.21138012694877506,
      "avg_duration_per_task_ms": 402109.59020044544,
      "avg_output_tokens_per_task": 9801.538975501113,
      "last_run_timestamp": "2026-10-31T11:26:03.627Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "Venice",
      "accuracy": 0.7416426103151863,
      "accuracy_std_dev": 0.022067987204666104,
      "run_count": 7,
      "total_questions": 349,
      "avg_cost_per_task_usd": 0.051895842406876794,
      "avg_duration_per_task_ms": 765180.4412607449,
      "avg_output_tokens_per_task": 14064.899713467048,
      "last_run_timestamp": "2026-10-24T19:40:41.765Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Baidu",
      "accuracy": 0.7414965918367348,
      "accuracy_std_dev": 0.01208300000000001,
      "run_count": 2,
      "total_questions": 98,
      "avg_cost_per_task_usd": 0.014484438775510205,
      "avg_duration_per_task_ms": 90891.24489795919,
      "avg_output_tokens_per_task": 9684.755102040815,
      "last_run_timestamp": "2026-10-31T13:18:30.658Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "Novita",
      "accuracy": 0.7414814444444444,
      "accuracy_std_dev": 0.026810305535625297,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.02054646222222222,
      "avg_duration_per_task_ms": 418578.0088888889,
      "avg_output_tokens_per_task": 11914.037777777778,
      "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": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "DeepInfra",
      "accuracy": 0.74136442746114,
      "accuracy_std_dev": 0.023598813561278888,
      "run_count": 8,
      "total_questions": 386,
      "avg_cost_per_task_usd": 0.1499042176165803,
      "avg_duration_per_task_ms": 675715.4041450777,
      "avg_output_tokens_per_task": 29694.012953367877,
      "last_run_timestamp": "2026-10-31T19:55:58.516Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta/muse-spark-open-20261015",
      "model_slug": "meta/muse-spark-open",
      "model_name": "Meta: Muse Spark Open",
      "provider_name": "Fireworks",
      "accuracy": 0.7413334,
      "accuracy_std_dev": 0.02436751571744648,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.097602524,
      "avg_duration_per_task_ms": 217804.16,
      "avg_output_tokens_per_task": 9920.02,
      "last_run_timestamp": "2026-10-31T14:40:25.043Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "DeepInfra",
      "accuracy": 0.7413333,
      "accuracy_std_dev": 0.03095523075685272,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.40647999399999996,
      "avg_duration_per_task_ms": 588538.78,
      "avg_output_tokens_per_task": 9896.416,
      "last_run_timestamp": "2026-10-31T12:27:47.241Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.6-sonnet-20260217",
      "model_slug": "anthropic/claude-sonnet-4.6",
      "model_name": "Anthropic: Claude Sonnet 4.6",
      "provider_name": "Claude Platform on AWS",
      "accuracy": 0.7412121818181818,
      "accuracy_std_dev": 0.0275340250329006,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.3449351727272727,
      "avg_duration_per_task_ms": 200135.41454545455,
      "avg_output_tokens_per_task": 11800.205454545454,
      "last_run_timestamp": "2026-10-31T13:12:27.229Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Nebius",
      "accuracy": 0.7410254102564102,
      "accuracy_std_dev": 0.015687179899124656,
      "run_count": 4,
      "total_questions": 195,
      "avg_cost_per_task_usd": 0.010222671794871796,
      "avg_duration_per_task_ms": 114819.93846153846,
      "avg_output_tokens_per_task": 3734.005128205128,
      "last_run_timestamp": "2026-10-31T12:06:01.335Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7409522857142857,
      "accuracy_std_dev": 0.024018230038953373,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.05641216857142857,
      "avg_duration_per_task_ms": 852526.1028571428,
      "avg_output_tokens_per_task": 15826,
      "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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7408025986622073,
      "accuracy_std_dev": 0.026595914006744867,
      "run_count": 12,
      "total_questions": 598,
      "avg_cost_per_task_usd": 0.04194225585284281,
      "avg_duration_per_task_ms": 163070.16889632106,
      "avg_output_tokens_per_task": 8777.19397993311,
      "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": "z-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": "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": 0.7406060909090909,
      "accuracy_std_dev": 0.026123868248489987,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.3297550236363636,
      "avg_duration_per_task_ms": 195835.06181818183,
      "avg_output_tokens_per_task": 11555.365454545454,
      "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-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Inceptron",
      "accuracy": 0.7403698151001541,
      "accuracy_std_dev": 0.028547774310492945,
      "run_count": 13,
      "total_questions": 649,
      "avg_cost_per_task_usd": 0.012115494607087829,
      "avg_duration_per_task_ms": 425113.8289676425,
      "avg_output_tokens_per_task": 14181.97688751926,
      "last_run_timestamp": "2026-10-31T11:58:45.245Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "CoreWeave",
      "accuracy": 0.7400002499999999,
      "accuracy_std_dev": 0.029999944444740217,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.08284509749999999,
      "avg_duration_per_task_ms": 74658.595,
      "avg_output_tokens_per_task": 11791.0375,
      "last_run_timestamp": "2026-10-31T19:56:16.658Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "GMICloud",
      "accuracy": 0.7400001666666666,
      "accuracy_std_dev": 0.01154690915954955,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.034900296666666664,
      "avg_duration_per_task_ms": 129527.63333333333,
      "avg_output_tokens_per_task": 16132.923333333334,
      "last_run_timestamp": "2026-10-27T00:42:49.766Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Cloudflare",
      "accuracy": 0.7400001111111111,
      "accuracy_std_dev": 0.024545206467724207,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.07987433111111111,
      "avg_duration_per_task_ms": 314745.96222222224,
      "avg_output_tokens_per_task": 11321.113333333333,
      "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": "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": "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": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Morph",
      "accuracy": 0.74,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.1350024,
      "avg_duration_per_task_ms": 461018.86,
      "avg_output_tokens_per_task": 26035.68,
      "last_run_timestamp": "2026-08-15T08:15:35.382Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "auto-routing",
      "accuracy": 0.74,
      "accuracy_std_dev": 0.03471220655619578,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.02934289777777778,
      "avg_duration_per_task_ms": 461721.2688888889,
      "avg_output_tokens_per_task": 15007.28,
      "last_run_timestamp": "2026-10-17T13:55:17.194Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-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": "Together",
      "accuracy": 0.7399999090909091,
      "accuracy_std_dev": 0.02785246847377528,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.3035846727272727,
      "avg_duration_per_task_ms": 295316.7327272727,
      "avg_output_tokens_per_task": 9413.458181818181,
      "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": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Makora",
      "accuracy": 0.7399329261744967,
      "accuracy_std_dev": 0.017150616176167584,
      "run_count": 6,
      "total_questions": 298,
      "avg_cost_per_task_usd": 0.24555212751677855,
      "avg_duration_per_task_ms": 357614.23825503356,
      "avg_output_tokens_per_task": 12408.355704697986,
      "last_run_timestamp": "2026-10-31T12:24:39.056Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7395555333333333,
      "accuracy_std_dev": 0.016947170784004716,
      "run_count": 15,
      "total_questions": 750,
      "avg_cost_per_task_usd": 0.013297554666666666,
      "avg_duration_per_task_ms": 100947.12933333333,
      "avg_output_tokens_per_task": 10746.964,
      "last_run_timestamp": "2026-10-31T12:45:41.031Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.6-sonnet-20260217",
      "model_slug": "anthropic/claude-sonnet-4.6",
      "model_name": "Anthropic: Claude Sonnet 4.6",
      "provider_name": "Anthropic",
      "accuracy": 0.7394444166666667,
      "accuracy_std_dev": 0.02168094045645598,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.3449865733333333,
      "avg_duration_per_task_ms": 207685.65666666668,
      "avg_output_tokens_per_task": 11866.296666666667,
      "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": "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": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Phala",
      "accuracy": 0.73916675,
      "accuracy_std_dev": 0.032220958069826236,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.1383197475,
      "avg_duration_per_task_ms": 366177.2025,
      "avg_output_tokens_per_task": 11823.9525,
      "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": "xiaomi/mimo-v2.6-flash-20260921",
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "model_name": "Xiaomi: MiMo-V2.6-Flash",
      "provider_name": "Relace",
      "accuracy": 0.738889,
      "accuracy_std_dev": 0.024695368688615833,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.01095846,
      "avg_duration_per_task_ms": 321436.35333333333,
      "avg_output_tokens_per_task": 22570.04,
      "last_run_timestamp": "2026-10-31T12:52:06.048Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Decart",
      "accuracy": 0.7386668,
      "accuracy_std_dev": 0.03166503013357166,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.009092168000000001,
      "avg_duration_per_task_ms": 101935.708,
      "avg_output_tokens_per_task": 5027.116,
      "last_run_timestamp": "2026-10-17T11:17:02.460Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "auto-routing",
      "accuracy": 0.7386667,
      "accuracy_std_dev": 0.04987086078673599,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.009059300000000001,
      "avg_duration_per_task_ms": 243713.362,
      "avg_output_tokens_per_task": 15953.852,
      "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": "anthropic/claude-4.6-sonnet-20260217",
      "model_slug": "anthropic/claude-sonnet-4.6",
      "model_name": "Anthropic: Claude Sonnet 4.6",
      "provider_name": "Google",
      "accuracy": 0.7385858181818183,
      "accuracy_std_dev": 0.02930192686750755,
      "run_count": 33,
      "total_questions": 1650,
      "avg_cost_per_task_usd": 0.3660526951515151,
      "avg_duration_per_task_ms": 198575.79333333333,
      "avg_output_tokens_per_task": 11916.818181818182,
      "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": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "Xiaomi",
      "accuracy": 0.7385184444444445,
      "accuracy_std_dev": 0.025293917789377812,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.008203653333333333,
      "avg_duration_per_task_ms": 425108.4222222222,
      "avg_output_tokens_per_task": 14620.724444444444,
      "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": "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": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "DeepInfra",
      "accuracy": 0.7383335,
      "accuracy_std_dev": 0.0016665000000000152,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.08170183999999998,
      "avg_duration_per_task_ms": 327224.56,
      "avg_output_tokens_per_task": 8775.96,
      "last_run_timestamp": "2026-08-22T16:12:20.461Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "SiliconFlow",
      "accuracy": 0.738333375,
      "accuracy_std_dev": 0.03244649655948039,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.049700385,
      "avg_duration_per_task_ms": 377861.96,
      "avg_output_tokens_per_task": 11236.51,
      "last_run_timestamp": "2026-10-31T21:59:10.427Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "SiliconFlow",
      "accuracy": 0.73833325,
      "accuracy_std_dev": 0.03460099147694328,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0732567425,
      "avg_duration_per_task_ms": 220285.0375,
      "avg_output_tokens_per_task": 9565.1675,
      "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": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "StreamLake",
      "accuracy": 0.7382021797752809,
      "accuracy_std_dev": 0.021880590311850887,
      "run_count": 9,
      "total_questions": 445,
      "avg_cost_per_task_usd": 0.020498339325842697,
      "avg_duration_per_task_ms": 307841.11235955055,
      "avg_output_tokens_per_task": 10994.101123595505,
      "last_run_timestamp": "2026-10-17T11:09:59.070Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "auto-routing",
      "accuracy": 0.7381817272727274,
      "accuracy_std_dev": 0.026375810495400075,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.0789221490909091,
      "avg_duration_per_task_ms": 120883.72,
      "avg_output_tokens_per_task": 8767.690909090908,
      "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-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7380954285714285,
      "accuracy_std_dev": 0.030801812534131074,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.0719204,
      "avg_duration_per_task_ms": 327417.77714285714,
      "avg_output_tokens_per_task": 11436.125714285714,
      "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-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "StreamLake",
      "accuracy": 0.7380953482142857,
      "accuracy_std_dev": 0.031652037557723764,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.004402678571428571,
      "avg_duration_per_task_ms": 256685.421875,
      "avg_output_tokens_per_task": 12156.511160714286,
      "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": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Parasail",
      "accuracy": 0.7380952719449226,
      "accuracy_std_dev": 0.03197285191855047,
      "run_count": 12,
      "total_questions": 581,
      "avg_cost_per_task_usd": 0.024856838209982787,
      "avg_duration_per_task_ms": 93869.10499139415,
      "avg_output_tokens_per_task": 8326.619621342514,
      "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-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": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Moonshot AI",
      "accuracy": 0.737882,
      "accuracy_std_dev": 0.023142126657293677,
      "run_count": 9,
      "total_questions": 447,
      "avg_cost_per_task_usd": 0.3130514742729307,
      "avg_duration_per_task_ms": 411436.096196868,
      "avg_output_tokens_per_task": 9782.814317673377,
      "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-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": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "Novita",
      "accuracy": 0.7377779999999999,
      "accuracy_std_dev": 0.044333155109917456,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.010008253333333335,
      "avg_duration_per_task_ms": 438344.2577777778,
      "avg_output_tokens_per_task": 14883.606666666667,
      "last_run_timestamp": "2026-10-17T11:26:35.790Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "StreamLake",
      "accuracy": 0.7377777777777778,
      "accuracy_std_dev": 0.03189484265094273,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.06719376444444446,
      "avg_duration_per_task_ms": 260527.6288888889,
      "avg_output_tokens_per_task": 9620.40888888889,
      "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": "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": "z-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": "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": "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": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "Xiaomi",
      "accuracy": 0.7370371111111111,
      "accuracy_std_dev": 0.03175710369680895,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.024656559999999994,
      "avg_duration_per_task_ms": 423744.3888888889,
      "avg_output_tokens_per_task": 15866.275555555556,
      "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": "DigitalOcean",
      "accuracy": 0.7370184522613066,
      "accuracy_std_dev": 0.037544250024970534,
      "run_count": 8,
      "total_questions": 398,
      "avg_cost_per_task_usd": 0.07583668341708542,
      "avg_duration_per_task_ms": 861965.7211055276,
      "avg_output_tokens_per_task": 14681.494974874371,
      "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": "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": "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-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "auto-routing",
      "accuracy": 0.7366665,
      "accuracy_std_dev": 0.02285212122539177,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0637652175,
      "avg_duration_per_task_ms": 239799.505,
      "avg_output_tokens_per_task": 12644.0625,
      "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": "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": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Minimax",
      "accuracy": 0.7363637272727273,
      "accuracy_std_dev": 0.029727967744168908,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.029319285454545447,
      "avg_duration_per_task_ms": 101893.45090909091,
      "avg_output_tokens_per_task": 10418.469090909091,
      "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-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Inceptron",
      "accuracy": 0.7362762003395587,
      "accuracy_std_dev": 0.03269045093786308,
      "run_count": 12,
      "total_questions": 589,
      "avg_cost_per_task_usd": 0.06555029541595926,
      "avg_duration_per_task_ms": 138699.1324278438,
      "avg_output_tokens_per_task": 8434.507640067912,
      "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-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.736,
      "accuracy_std_dev": 0.011623577573191504,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.02217724,
      "avg_duration_per_task_ms": 74896.184,
      "avg_output_tokens_per_task": 4355.024,
      "last_run_timestamp": "2026-10-31T12:05:37.260Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "Google",
      "accuracy": 0.7359999333333334,
      "accuracy_std_dev": 0.0231964569463145,
      "run_count": 15,
      "total_questions": 750,
      "avg_cost_per_task_usd": 0.013030309333333335,
      "avg_duration_per_task_ms": 73246.75466666666,
      "avg_output_tokens_per_task": 10779.536,
      "last_run_timestamp": "2026-10-31T11:42:11.484Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "auto-routing",
      "accuracy": 0.7359999,
      "accuracy_std_dev": 0.02831560386588993,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.260319256,
      "avg_duration_per_task_ms": 297852.642,
      "avg_output_tokens_per_task": 10753.23,
      "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": "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": "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": 0.735833375,
      "accuracy_std_dev": 0.017618592217722016,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.11823662750000001,
      "avg_duration_per_task_ms": 341338.0975,
      "avg_output_tokens_per_task": 9727.67,
      "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": "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": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "GMICloud",
      "accuracy": 0.7355555555555555,
      "accuracy_std_dev": 0.03220320511692074,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.04873943777777778,
      "avg_duration_per_task_ms": 232537.14444444445,
      "avg_output_tokens_per_task": 10149.82,
      "last_run_timestamp": "2026-10-31T19:47:34.571Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Relace",
      "accuracy": 0.7355555,
      "accuracy_std_dev": 0.037646835160253966,
      "run_count": 18,
      "total_questions": 900,
      "avg_cost_per_task_usd": 0.009423405555555554,
      "avg_duration_per_task_ms": 235235.8688888889,
      "avg_output_tokens_per_task": 15545.602222222222,
      "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": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Ambient",
      "accuracy": 0.7355553333333333,
      "accuracy_std_dev": 0.02997928502964824,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.008365386666666667,
      "avg_duration_per_task_ms": 195637.39333333334,
      "avg_output_tokens_per_task": 12883.673333333334,
      "last_run_timestamp": "2026-08-29T09:31:13.940Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7355553333333333,
      "accuracy_std_dev": 0.016629618082071397,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.05735756666666667,
      "avg_duration_per_task_ms": 219541.36,
      "avg_output_tokens_per_task": 9149.013333333334,
      "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": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "auto-routing",
      "accuracy": 0.7352380000000001,
      "accuracy_std_dev": 0.02932336720092016,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.011415382857142857,
      "avg_duration_per_task_ms": 224754.26285714286,
      "avg_output_tokens_per_task": 14326.351428571428,
      "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": "openai/gpt-5.4-20260305",
      "model_slug": "openai/gpt-5.4",
      "model_name": "OpenAI: GPT-5.4",
      "provider_name": "Azure",
      "accuracy": 0.7352380000000001,
      "accuracy_std_dev": 0.01698343330593503,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.3052330542857143,
      "avg_duration_per_task_ms": 191492.84285714285,
      "avg_output_tokens_per_task": 10672.025714285714,
      "last_run_timestamp": "2026-10-01T19:05:18.005Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "GMICloud",
      "accuracy": 0.735000125,
      "accuracy_std_dev": 0.04186617734650937,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0273002175,
      "avg_duration_per_task_ms": 144987.4475,
      "avg_output_tokens_per_task": 12792.885,
      "last_run_timestamp": "2026-10-31T18:53:34.804Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Fireworks",
      "accuracy": 0.7349398099062917,
      "accuracy_std_dev": 0.02455646627618342,
      "run_count": 15,
      "total_questions": 747,
      "avg_cost_per_task_usd": 0.05549164792503348,
      "avg_duration_per_task_ms": 157974.1485943775,
      "avg_output_tokens_per_task": 30263.555555555555,
      "last_run_timestamp": "2026-10-31T12:41:59.422Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Phala",
      "accuracy": 0.7349397911646587,
      "accuracy_std_dev": 0.047428285455095254,
      "run_count": 10,
      "total_questions": 498,
      "avg_cost_per_task_usd": 0.2969604819277108,
      "avg_duration_per_task_ms": 386565.9397590361,
      "avg_output_tokens_per_task": 11351.530120481928,
      "last_run_timestamp": "2026-10-31T12:42:56.001Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-3.8-flash-20260902",
      "model_slug": "google/gemini-3.8-flash",
      "model_name": "Google: Gemini 3.8 Flash",
      "provider_name": "Google AI Studio",
      "accuracy": 0.734694,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 49,
      "avg_cost_per_task_usd": 0.25339910204081634,
      "avg_duration_per_task_ms": 146531.16326530612,
      "avg_output_tokens_per_task": 29831.102040816328,
      "last_run_timestamp": "2026-10-01T08:03:06.861Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Venice",
      "accuracy": 0.7344445,
      "accuracy_std_dev": 0.0325272326558839,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.07984409666666667,
      "avg_duration_per_task_ms": 507527.12666666665,
      "avg_output_tokens_per_task": 12020.823333333334,
      "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": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Novita",
      "accuracy": 0.734166625,
      "accuracy_std_dev": 0.031169207785639575,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0986136275,
      "avg_duration_per_task_ms": 266155.0625,
      "avg_output_tokens_per_task": 12567.385,
      "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": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "Groq",
      "accuracy": 0.7341136153846154,
      "accuracy_std_dev": 0.026553848166927203,
      "run_count": 6,
      "total_questions": 299,
      "avg_cost_per_task_usd": 0.12951078260869564,
      "avg_duration_per_task_ms": 62609.57525083612,
      "avg_output_tokens_per_task": 9507.267558528429,
      "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": "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": "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": 0.7337547565858799,
      "accuracy_std_dev": 0.027618715667845153,
      "run_count": 19,
      "total_questions": 949,
      "avg_cost_per_task_usd": 0.12936480400421493,
      "avg_duration_per_task_ms": 156235.47629083245,
      "avg_output_tokens_per_task": 9321.471022128557,
      "last_run_timestamp": "2026-10-24T19:48:18.775Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Inceptron",
      "accuracy": 0.7335003934837092,
      "accuracy_std_dev": 0.03251829565197682,
      "run_count": 8,
      "total_questions": 399,
      "avg_cost_per_task_usd": 0.06496012280701755,
      "avg_duration_per_task_ms": 149415.91228070174,
      "avg_output_tokens_per_task": 11151.812030075187,
      "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": "google/gemini-3-flash-preview-20251217",
      "model_slug": "google/gemini-3-flash-preview",
      "model_name": "Google: Gemini 3 Flash Preview",
      "provider_name": "Google",
      "accuracy": 0.7333335,
      "accuracy_std_dev": 0.00666650000000002,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.12301775,
      "avg_duration_per_task_ms": 155307.55,
      "avg_output_tokens_per_task": 18812.75,
      "last_run_timestamp": "2026-10-01T19:05:47.885Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta/muse-spark-open-20261015",
      "model_slug": "meta/muse-spark-open",
      "model_name": "Meta: Muse Spark Open",
      "provider_name": "Together",
      "accuracy": 0.7333335,
      "accuracy_std_dev": 0.00666650000000002,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.08901925000000001,
      "avg_duration_per_task_ms": 93612.32,
      "avg_output_tokens_per_task": 8634.62,
      "last_run_timestamp": "2026-10-31T11:29:29.503Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Phala",
      "accuracy": 0.7333335,
      "accuracy_std_dev": 0.013333499999999998,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.08394800999999999,
      "avg_duration_per_task_ms": 134148.76,
      "avg_output_tokens_per_task": 8723.94,
      "last_run_timestamp": "2026-10-17T19:55:39.408Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-sol-20260709",
      "model_slug": "openai/gpt-5.6-sol",
      "model_name": "OpenAI: GPT-5.6 Sol",
      "provider_name": "auto-routing",
      "accuracy": 0.7333335,
      "accuracy_std_dev": 0.026666499999999982,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.12687428,
      "avg_duration_per_task_ms": 110480.49,
      "avg_output_tokens_per_task": 5152.54,
      "last_run_timestamp": "2026-10-01T19:04:55.696Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7333334,
      "accuracy_std_dev": 0.03451234883690184,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.30572972400000004,
      "avg_duration_per_task_ms": 282521.796,
      "avg_output_tokens_per_task": 10813.856,
      "last_run_timestamp": "2026-10-31T12:51:11.835Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "Novita",
      "accuracy": 0.733333375,
      "accuracy_std_dev": 0.027688655921412557,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.051658964999999994,
      "avg_duration_per_task_ms": 354319.435,
      "avg_output_tokens_per_task": 10072.31,
      "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-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Decart",
      "accuracy": 0.7333333333333333,
      "accuracy_std_dev": 0.005443174461857933,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.008480786666666665,
      "avg_duration_per_task_ms": 397846.0266666667,
      "avg_output_tokens_per_task": 19127.326666666668,
      "last_run_timestamp": "2026-08-22T08:38:00.859Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Mancer 2",
      "accuracy": 0.7333333333333333,
      "accuracy_std_dev": 0.014401903168987377,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.06451945333333334,
      "avg_duration_per_task_ms": 876085.0533333333,
      "avg_output_tokens_per_task": 15144.64,
      "last_run_timestamp": "2026-10-31T12:31:12.958Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Crusoe",
      "accuracy": 0.7333333333333333,
      "accuracy_std_dev": 0.03569390445004426,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.021479246666666667,
      "avg_duration_per_task_ms": 120781.28,
      "avg_output_tokens_per_task": 11581.666666666666,
      "last_run_timestamp": "2026-10-31T11:22:01.481Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "InferenceNet",
      "accuracy": 0.7333333333333333,
      "accuracy_std_dev": 0.05760637824832321,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.04898614,
      "avg_duration_per_task_ms": 106829.63333333333,
      "avg_output_tokens_per_task": 9897.326666666666,
      "last_run_timestamp": "2026-10-10T11:51:48.927Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-20260816",
      "model_slug": "z-ai/glm-5.3",
      "model_name": "Z.ai: GLM 5.3",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7333333333333333,
      "accuracy_std_dev": 0.024944382578492966,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.09260214,
      "avg_duration_per_task_ms": 326558.9733333333,
      "avg_output_tokens_per_task": 4729.153333333334,
      "last_run_timestamp": "2026-10-31T12:28:59.426Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "Cohere",
      "accuracy": 0.7333333333333333,
      "accuracy_std_dev": 0.009428090415820642,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.024898453333333334,
      "avg_duration_per_task_ms": 37126.9,
      "avg_output_tokens_per_task": 4153.473333333333,
      "last_run_timestamp": "2026-10-31T12:07:50.460Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-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": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Phala",
      "accuracy": 0.7333332,
      "accuracy_std_dev": 0.03211094655970144,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.18171748399999998,
      "avg_duration_per_task_ms": 476804.58,
      "avg_output_tokens_per_task": 9681.516,
      "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-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "NextBit",
      "accuracy": 0.7333330000000001,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01186947,
      "avg_duration_per_task_ms": 176514.74,
      "avg_output_tokens_per_task": 14201.61,
      "last_run_timestamp": "2026-08-29T16:50:07.898Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7333330000000001,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.058005,
      "avg_duration_per_task_ms": 191243.62,
      "avg_output_tokens_per_task": 30812.32,
      "last_run_timestamp": "2026-09-12T08:09:12.993Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "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": "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": "z-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": 0.7327778333333334,
      "accuracy_std_dev": 0.03057555204416685,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.01934296,
      "avg_duration_per_task_ms": 97612.35666666667,
      "avg_output_tokens_per_task": 5244.636666666666,
      "last_run_timestamp": "2026-11-02T21:39:06.869Z"
    },
    {
      "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": 0.7326191428571428,
      "accuracy_std_dev": 0.015745039894052536,
      "run_count": 14,
      "total_questions": 700,
      "avg_cost_per_task_usd": 0.2114459285714286,
      "avg_duration_per_task_ms": 238357.1,
      "avg_output_tokens_per_task": 9801.7,
      "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": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "OpenInference",
      "accuracy": 0.7325,
      "accuracy_std_dev": 0.04596937990662913,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.012367795,
      "avg_duration_per_task_ms": 953069.875,
      "avg_output_tokens_per_task": 22737,
      "last_run_timestamp": "2026-10-17T10:32:07.729Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7324560809716599,
      "accuracy_std_dev": 0.04384289047314286,
      "run_count": 10,
      "total_questions": 494,
      "avg_cost_per_task_usd": 0.02773825708502024,
      "avg_duration_per_task_ms": 111349.57489878543,
      "avg_output_tokens_per_task": 10355.305668016194,
      "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": "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": "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": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "Azure",
      "accuracy": 0.7313332,
      "accuracy_std_dev": 0.020231941744676914,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.012560144,
      "avg_duration_per_task_ms": 74716.332,
      "avg_output_tokens_per_task": 10525.228,
      "last_run_timestamp": "2026-10-31T11:19:23.393Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7311827782258065,
      "accuracy_std_dev": 0.07837008853695393,
      "run_count": 10,
      "total_questions": 496,
      "avg_cost_per_task_usd": 0.009883709677419354,
      "avg_duration_per_task_ms": 170348.1431451613,
      "avg_output_tokens_per_task": 4275.5161290322585,
      "last_run_timestamp": "2026-10-31T12:31:51.998Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Wafer",
      "accuracy": 0.7311637573964497,
      "accuracy_std_dev": 0.02813377440238986,
      "run_count": 17,
      "total_questions": 845,
      "avg_cost_per_task_usd": 0.024811487573964494,
      "avg_duration_per_task_ms": 209572.125443787,
      "avg_output_tokens_per_task": 21468.289940828403,
      "last_run_timestamp": "2026-10-31T11:27:16.882Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "Z.AI",
      "accuracy": 0.7311111666666666,
      "accuracy_std_dev": 0.03022542058387644,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.05052855666666667,
      "avg_duration_per_task_ms": 436577.48333333334,
      "avg_output_tokens_per_task": 9693.223333333333,
      "last_run_timestamp": "2026-10-24T20:04:11.026Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "Novita",
      "accuracy": 0.7311110000000001,
      "accuracy_std_dev": 0.013698950130089036,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.06301382666666666,
      "avg_duration_per_task_ms": 511947.82666666666,
      "avg_output_tokens_per_task": 14587.22,
      "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-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "OpenInference",
      "accuracy": 0.7311110000000001,
      "accuracy_std_dev": 0.020608136952831695,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.028709900000000003,
      "avg_duration_per_task_ms": 638312.6933333334,
      "avg_output_tokens_per_task": 61738.18,
      "last_run_timestamp": "2026-10-31T12:28:49.803Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-sol-20260709",
      "model_slug": "openai/gpt-5.6-sol",
      "model_name": "OpenAI: GPT-5.6 Sol",
      "provider_name": "OpenAI",
      "accuracy": 0.7311110000000001,
      "accuracy_std_dev": 0.013698950130089036,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.12928292,
      "avg_duration_per_task_ms": 109953.54,
      "avg_output_tokens_per_task": 5131.753333333333,
      "last_run_timestamp": "2026-10-02T23:08:37.533Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Novita",
      "accuracy": 0.730909090909091,
      "accuracy_std_dev": 0.04442884014497901,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.02853827818181818,
      "avg_duration_per_task_ms": 115537.2709090909,
      "avg_output_tokens_per_task": 9990.294545454546,
      "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": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Crusoe",
      "accuracy": 0.73083325,
      "accuracy_std_dev": 0.029988453797878633,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.097229255,
      "avg_duration_per_task_ms": 318293.035,
      "avg_output_tokens_per_task": 11028.215,
      "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": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7306667,
      "accuracy_std_dev": 0.06533322244930216,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.15851144199999997,
      "avg_duration_per_task_ms": 1396906.662,
      "avg_output_tokens_per_task": 28393.436,
      "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": "anthropic/claude-4.5-sonnet-20250929",
      "model_slug": "anthropic/claude-sonnet-4.5",
      "model_name": "Anthropic: Claude Sonnet 4.5",
      "provider_name": "Google",
      "accuracy": 0.730303,
      "accuracy_std_dev": 0.02426511211221128,
      "run_count": 22,
      "total_questions": 1100,
      "avg_cost_per_task_usd": 0.364338510909091,
      "avg_duration_per_task_ms": 226168.5118181818,
      "avg_output_tokens_per_task": 10866.631818181819,
      "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-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-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7299998750000001,
      "accuracy_std_dev": 0.03299833750144659,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.15967390250000002,
      "avg_duration_per_task_ms": 222567.4375,
      "avg_output_tokens_per_task": 9532.0625,
      "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": "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": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "DeepInfra",
      "accuracy": 0.7294444999999999,
      "accuracy_std_dev": 0.04038785494737249,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.06920636333333333,
      "avg_duration_per_task_ms": 306647.6433333333,
      "avg_output_tokens_per_task": 12851.353333333333,
      "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": "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": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "auto-routing",
      "accuracy": 0.728889,
      "accuracy_std_dev": 0.03987630256458365,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.01001504888888889,
      "avg_duration_per_task_ms": 391679.0466666667,
      "avg_output_tokens_per_task": 14314.308888888889,
      "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": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "DeepInfra",
      "accuracy": 0.7288886666666667,
      "accuracy_std_dev": 0.0318947884596981,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.035202753333333336,
      "avg_duration_per_task_ms": 333081.32666666666,
      "avg_output_tokens_per_task": 8294.333333333334,
      "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": "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": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "DigitalOcean",
      "accuracy": 0.728735503448276,
      "accuracy_std_dev": 0.0506044752324886,
      "run_count": 3,
      "total_questions": 145,
      "avg_cost_per_task_usd": 0.07558873793103447,
      "avg_duration_per_task_ms": 486950.46206896554,
      "avg_output_tokens_per_task": 7447.337931034483,
      "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": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "auto-routing",
      "accuracy": 0.7286996367713005,
      "accuracy_std_dev": 0.04475223154653922,
      "run_count": 9,
      "total_questions": 446,
      "avg_cost_per_task_usd": 0.03329973766816144,
      "avg_duration_per_task_ms": 1152472.0291479821,
      "avg_output_tokens_per_task": 52238.542600896864,
      "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": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Parasail",
      "accuracy": 0.7285714285714285,
      "accuracy_std_dev": 0.035717562390258945,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.07543138571428572,
      "avg_duration_per_task_ms": 221880.41142857142,
      "avg_output_tokens_per_task": 11178.914285714285,
      "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": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.7283335,
      "accuracy_std_dev": 0.024626886091898288,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.04604961333333333,
      "avg_duration_per_task_ms": 118630.39,
      "avg_output_tokens_per_task": 26239.556666666667,
      "last_run_timestamp": "2026-10-31T12:19:28.999Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-haiku-5.5-20261007",
      "model_slug": "anthropic/claude-haiku-5.5",
      "model_name": "Anthropic: Claude Haiku 5.5",
      "provider_name": "auto-routing",
      "accuracy": 0.72833325,
      "accuracy_std_dev": 0.041197373838480274,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.011627464999999998,
      "avg_duration_per_task_ms": 86463.365,
      "avg_output_tokens_per_task": 10669.235,
      "last_run_timestamp": "2026-10-31T12:31:32.731Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.728,
      "accuracy_std_dev": 0.02362657637492154,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.027497304,
      "avg_duration_per_task_ms": 93734.552,
      "avg_output_tokens_per_task": 12939.792,
      "last_run_timestamp": "2026-10-31T12:24:43.560Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-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": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "Friendli",
      "accuracy": 0.7277778333333333,
      "accuracy_std_dev": 0.01651796253998121,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.10709389666666666,
      "avg_duration_per_task_ms": 343686.82666666666,
      "avg_output_tokens_per_task": 14435.816666666668,
      "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-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "Alibaba",
      "accuracy": 0.7274997500000001,
      "accuracy_std_dev": 0.030034653596928695,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.04475519,
      "avg_duration_per_task_ms": 324824.455,
      "avg_output_tokens_per_task": 13203.02,
      "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": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "DeepInfra",
      "accuracy": 0.7273503282051282,
      "accuracy_std_dev": 0.04018278252249339,
      "run_count": 4,
      "total_questions": 195,
      "avg_cost_per_task_usd": 0.017554456410256408,
      "avg_duration_per_task_ms": 289835.1384615385,
      "avg_output_tokens_per_task": 6267.041025641026,
      "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": "openai/gpt-5.6-sol-20260709",
      "model_slug": "openai/gpt-5.6-sol",
      "model_name": "OpenAI: GPT-5.6 Sol",
      "provider_name": "Azure",
      "accuracy": 0.7271428571428571,
      "accuracy_std_dev": 0.01855313018202104,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.31224118,
      "avg_duration_per_task_ms": 126481.89142857143,
      "avg_output_tokens_per_task": 5238.86,
      "last_run_timestamp": "2026-10-01T19:05:16.562Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-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-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": "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": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Baidu",
      "accuracy": 0.72666675,
      "accuracy_std_dev": 0.02748728999351337,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.068011135,
      "avg_duration_per_task_ms": 209802.9225,
      "avg_output_tokens_per_task": 13044.05,
      "last_run_timestamp": "2026-10-31T19:39:42.479Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.1-terminus",
      "model_slug": "deepseek/deepseek-v3.1-terminus",
      "model_name": "DeepSeek: DeepSeek V3.1 Terminus",
      "provider_name": "Novita",
      "accuracy": 0.7266666666666667,
      "accuracy_std_dev": 0.023726809342082957,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.06235516666666667,
      "avg_duration_per_task_ms": 501815.64666666667,
      "avg_output_tokens_per_task": 14648.713333333333,
      "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": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "StreamLake",
      "accuracy": 0.7266666666666667,
      "accuracy_std_dev": 0.039752443827043155,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.0691682688888889,
      "avg_duration_per_task_ms": 289187.1177777778,
      "avg_output_tokens_per_task": 12696.368888888888,
      "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": "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": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Cloudflare",
      "accuracy": 0.7266665833333333,
      "accuracy_std_dev": 0.02538595412315748,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.07330062000000001,
      "avg_duration_per_task_ms": 226287.41166666665,
      "avg_output_tokens_per_task": 8532.536666666667,
      "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": "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": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Alibaba",
      "accuracy": 0.726330531092437,
      "accuracy_std_dev": 0.031188754641318872,
      "run_count": 12,
      "total_questions": 595,
      "avg_cost_per_task_usd": 0.07654581848739497,
      "avg_duration_per_task_ms": 189475.03361344538,
      "avg_output_tokens_per_task": 8239.426890756302,
      "last_run_timestamp": "2026-10-31T19:01:31.914Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Venice",
      "accuracy": 0.7260605454545455,
      "accuracy_std_dev": 0.023519428697311804,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.09003248909090909,
      "avg_duration_per_task_ms": 647311.3927272727,
      "avg_output_tokens_per_task": 9449.287272727273,
      "last_run_timestamp": "2026-10-31T19:40:42.793Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-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": "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": 0.7255072173913043,
      "accuracy_std_dev": 0.03099958588076344,
      "run_count": 23,
      "total_questions": 1150,
      "avg_cost_per_task_usd": 0.35199391739130437,
      "avg_duration_per_task_ms": 204615.2852173913,
      "avg_output_tokens_per_task": 10253.284347826087,
      "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": "anthropic/claude-4.5-sonnet-20250929",
      "model_slug": "anthropic/claude-sonnet-4.5",
      "model_name": "Anthropic: Claude Sonnet 4.5",
      "provider_name": "Azure",
      "accuracy": 0.7254545454545455,
      "accuracy_std_dev": 0.030460198481307724,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.34895750909090917,
      "avg_duration_per_task_ms": 232559.38545454547,
      "avg_output_tokens_per_task": 10874.458181818181,
      "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": "z-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": "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": 0.7248484545454545,
      "accuracy_std_dev": 0.03023008675404398,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.32869586,
      "avg_duration_per_task_ms": 207276.92,
      "avg_output_tokens_per_task": 10330.558181818182,
      "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": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "Venice",
      "accuracy": 0.7244444444444444,
      "accuracy_std_dev": 0.02948105899692626,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.027190664444444446,
      "avg_duration_per_task_ms": 241956.15333333332,
      "avg_output_tokens_per_task": 9273.511111111111,
      "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": "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": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "Novita",
      "accuracy": 0.7241666250000001,
      "accuracy_std_dev": 0.024252670244415866,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.09570978000000002,
      "avg_duration_per_task_ms": 206305.7525,
      "avg_output_tokens_per_task": 12767.7075,
      "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": "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": "meta/muse-spark-open-20261015",
      "model_slug": "meta/muse-spark-open",
      "model_name": "Meta: Muse Spark Open",
      "provider_name": "Makora",
      "accuracy": 0.723360643442623,
      "accuracy_std_dev": 0.019132270021092652,
      "run_count": 5,
      "total_questions": 244,
      "avg_cost_per_task_usd": 0.0756424631147541,
      "avg_duration_per_task_ms": 76631.30327868853,
      "avg_output_tokens_per_task": 8337.151639344262,
      "last_run_timestamp": "2026-10-31T11:46:18.075Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Phala",
      "accuracy": 0.7233335,
      "accuracy_std_dev": 0.023333500000000007,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.02063872,
      "avg_duration_per_task_ms": 638594.61,
      "avg_output_tokens_per_task": 13867.61,
      "last_run_timestamp": "2026-08-29T16:33:36.025Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-terra-20260709",
      "model_slug": "openai/gpt-5.6-terra",
      "model_name": "OpenAI: GPT-5.6 Terra",
      "provider_name": "auto-routing",
      "accuracy": 0.7233335,
      "accuracy_std_dev": 0.0033334999999999892,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.11131322,
      "avg_duration_per_task_ms": 83286.18,
      "avg_output_tokens_per_task": 3736.91,
      "last_run_timestamp": "2026-10-01T19:04:40.360Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "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": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "Venice",
      "accuracy": 0.723333375,
      "accuracy_std_dev": 0.029814290943847292,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.041186885,
      "avg_duration_per_task_ms": 602276.8525,
      "avg_output_tokens_per_task": 9143.275,
      "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": 0.723333375,
      "accuracy_std_dev": 0.04203188274077637,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.10373765500000001,
      "avg_duration_per_task_ms": 200657.97,
      "avg_output_tokens_per_task": 11980.6125,
      "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": "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": 0.7232361605839416,
      "accuracy_std_dev": 0.08188802109119138,
      "run_count": 11,
      "total_questions": 548,
      "avg_cost_per_task_usd": 0.04147013868613139,
      "avg_duration_per_task_ms": 272167.5200729927,
      "avg_output_tokens_per_task": 24002.558394160584,
      "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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7230302727272727,
      "accuracy_std_dev": 0.036221982175797626,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.0770231490909091,
      "avg_duration_per_task_ms": 156976.8818181818,
      "avg_output_tokens_per_task": 8259.927272727273,
      "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": "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-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "AkashML",
      "accuracy": 0.7222223333333333,
      "accuracy_std_dev": 0.03456997912768952,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.011571553333333331,
      "avg_duration_per_task_ms": 541667.7866666666,
      "avg_output_tokens_per_task": 18235.293333333335,
      "last_run_timestamp": "2026-08-29T08:43:07.210Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Io Net",
      "accuracy": 0.7222223333333333,
      "accuracy_std_dev": 0.011330936570096713,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.018792733333333332,
      "avg_duration_per_task_ms": 301192.17333333334,
      "avg_output_tokens_per_task": 19692.3,
      "last_run_timestamp": "2026-10-01T16:00:06.404Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "SambaNova",
      "accuracy": 0.7222223333333333,
      "accuracy_std_dev": 0.011330936570096713,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.03683,
      "avg_duration_per_task_ms": 122927.25333333333,
      "avg_output_tokens_per_task": 9303.18,
      "last_run_timestamp": "2026-10-31T12:49:19.749Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Baidu",
      "accuracy": 0.722083375,
      "accuracy_std_dev": 0.03692844197206775,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.089901815,
      "avg_duration_per_task_ms": 164446.19,
      "avg_output_tokens_per_task": 10597.3975,
      "last_run_timestamp": "2026-10-31T19:56:54.162Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "ModelRun",
      "accuracy": 0.722,
      "accuracy_std_dev": 0.019843879550128302,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.06701492399999999,
      "avg_duration_per_task_ms": 78660.404,
      "avg_output_tokens_per_task": 8431.312,
      "last_run_timestamp": "2026-09-26T08:08:19.642Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7218181818181818,
      "accuracy_std_dev": 0.026680618940681487,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.08085123454545454,
      "avg_duration_per_task_ms": 229489.58727272728,
      "avg_output_tokens_per_task": 8760.714545454546,
      "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": "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": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "Darkbloom",
      "accuracy": 0.7214610958904109,
      "accuracy_std_dev": 0.02944405326868039,
      "run_count": 3,
      "total_questions": 146,
      "avg_cost_per_task_usd": 0.01699455479452055,
      "avg_duration_per_task_ms": 840567.9794520548,
      "avg_output_tokens_per_task": 10799.780821917808,
      "last_run_timestamp": "2026-10-31T13:36:58.095Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Reka",
      "accuracy": 0.7213334,
      "accuracy_std_dev": 0.01146975927559073,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.012899843999999999,
      "avg_duration_per_task_ms": 152142.784,
      "avg_output_tokens_per_task": 13063.108,
      "last_run_timestamp": "2026-10-31T13:10:04.012Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "DeepInfra",
      "accuracy": 0.7213332,
      "accuracy_std_dev": 0.030811331249395918,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.033439392,
      "avg_duration_per_task_ms": 410344.808,
      "avg_output_tokens_per_task": 8978.336,
      "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": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "Phala",
      "accuracy": 0.721250125,
      "accuracy_std_dev": 0.025052106116240477,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.21457064250000002,
      "avg_duration_per_task_ms": 1269183.795,
      "avg_output_tokens_per_task": 14006.3875,
      "last_run_timestamp": "2026-10-31T19:47:23.162Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Makora",
      "accuracy": 0.7210882448979592,
      "accuracy_std_dev": 0.0076780598851532815,
      "run_count": 5,
      "total_questions": 245,
      "avg_cost_per_task_usd": 0.009489710204081632,
      "avg_duration_per_task_ms": 145499.9755102041,
      "avg_output_tokens_per_task": 16260.81224489796,
      "last_run_timestamp": "2026-09-26T08:06:43.933Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "auto-routing",
      "accuracy": 0.720833375,
      "accuracy_std_dev": 0.014312870571076042,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.05266061,
      "avg_duration_per_task_ms": 391408.44,
      "avg_output_tokens_per_task": 9353.0625,
      "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-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Fireworks",
      "accuracy": 0.7201178737373739,
      "accuracy_std_dev": 0.03208603854872086,
      "run_count": 8,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.0926496590909091,
      "avg_duration_per_task_ms": 92170.53535353535,
      "avg_output_tokens_per_task": 7432.570707070707,
      "last_run_timestamp": "2026-09-26T08:26:47.303Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-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": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "Novita",
      "accuracy": 0.72,
      "accuracy_std_dev": 0.035694340251642157,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.05599238666666667,
      "avg_duration_per_task_ms": 288072.32666666666,
      "avg_output_tokens_per_task": 8411.226666666667,
      "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": "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": "z-ai/glm-5-20260211",
      "model_slug": "z-ai/glm-5",
      "model_name": "Z.ai: GLM 5",
      "provider_name": "Venice",
      "accuracy": 0.72,
      "accuracy_std_dev": 0.03375490985700519,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.08901049818181819,
      "avg_duration_per_task_ms": 307122.0472727273,
      "avg_output_tokens_per_task": 9253.218181818182,
      "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": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "Moonshot AI",
      "accuracy": 0.7199996666666667,
      "accuracy_std_dev": 0.024944382578492966,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.046407413333333335,
      "avg_duration_per_task_ms": 163709.00666666665,
      "avg_output_tokens_per_task": 8100.34,
      "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": "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": "moonshotai/kimi-k3-20260715",
      "model_slug": "moonshotai/kimi-k3",
      "model_name": "MoonshotAI: Kimi K3",
      "provider_name": "DigitalOcean",
      "accuracy": 0.71944425,
      "accuracy_std_dev": 0.06232871962576722,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.27328554166666674,
      "avg_duration_per_task_ms": 424932.33166666667,
      "avg_output_tokens_per_task": 10565.703333333333,
      "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": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Novita",
      "accuracy": 0.7193939090909091,
      "accuracy_std_dev": 0.021547202760681436,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.08196447454545454,
      "avg_duration_per_task_ms": 229606.52181818182,
      "avg_output_tokens_per_task": 9247.236363636364,
      "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": 0.7193939090909091,
      "accuracy_std_dev": 0.026278368554779535,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.04867809818181818,
      "avg_duration_per_task_ms": 168623.93818181817,
      "avg_output_tokens_per_task": 9688.069090909092,
      "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": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "Novita",
      "accuracy": 0.7191666250000001,
      "accuracy_std_dev": 0.02145742393052752,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0557140475,
      "avg_duration_per_task_ms": 330536.0525,
      "avg_output_tokens_per_task": 9856.7875,
      "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": "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": 0.7191665,
      "accuracy_std_dev": 0.035968585831945094,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.09730926000000001,
      "avg_duration_per_task_ms": 222643.045,
      "avg_output_tokens_per_task": 10115.1975,
      "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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Crusoe",
      "accuracy": 0.7183335,
      "accuracy_std_dev": 0.031402344215838394,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.114731285,
      "avg_duration_per_task_ms": 88444.29,
      "avg_output_tokens_per_task": 8961.775,
      "last_run_timestamp": "2026-09-12T08:12:28.966Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "auto-routing",
      "accuracy": 0.71833325,
      "accuracy_std_dev": 0.023033749394909635,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.059744359999999996,
      "avg_duration_per_task_ms": 328747.2175,
      "avg_output_tokens_per_task": 9865.755,
      "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": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Novita",
      "accuracy": 0.71833325,
      "accuracy_std_dev": 0.028037588857416033,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.1325301525,
      "avg_duration_per_task_ms": 255309.835,
      "avg_output_tokens_per_task": 10427.27,
      "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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Baidu",
      "accuracy": 0.7182456315789474,
      "accuracy_std_dev": 0.026827730127829193,
      "run_count": 19,
      "total_questions": 950,
      "avg_cost_per_task_usd": 0.06829176842105264,
      "avg_duration_per_task_ms": 115296.11368421053,
      "avg_output_tokens_per_task": 8771.72947368421,
      "last_run_timestamp": "2026-10-31T19:54:14.505Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3-20260531",
      "model_slug": "minimax/minimax-m3",
      "model_name": "MiniMax: MiniMax M3",
      "provider_name": "StreamLake",
      "accuracy": 0.717778,
      "accuracy_std_dev": 0.046930323977857506,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.045013793333333336,
      "avg_duration_per_task_ms": 196458.52,
      "avg_output_tokens_per_task": 18662.633333333335,
      "last_run_timestamp": "2026-10-31T19:03:35.054Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Ionstream",
      "accuracy": 0.7177779999999999,
      "accuracy_std_dev": 0.03499564699597175,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.039637780000000004,
      "avg_duration_per_task_ms": 223451.97333333333,
      "avg_output_tokens_per_task": 45456.886666666665,
      "last_run_timestamp": "2026-10-31T11:34:19.759Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Chutes",
      "accuracy": 0.717499875,
      "accuracy_std_dev": 0.033654675052054425,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.051500805000000004,
      "avg_duration_per_task_ms": 406565.7275,
      "avg_output_tokens_per_task": 11732.8625,
      "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": "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": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "Novita",
      "accuracy": 0.71666675,
      "accuracy_std_dev": 0.02624654474759486,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.026279967499999998,
      "avg_duration_per_task_ms": 219262.365,
      "avg_output_tokens_per_task": 10080.3425,
      "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": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Venice",
      "accuracy": 0.7166665000000001,
      "accuracy_std_dev": 0.0033334999999999892,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.022047789999999998,
      "avg_duration_per_task_ms": 63248.07,
      "avg_output_tokens_per_task": 8992.32,
      "last_run_timestamp": "2026-10-31T12:43:48.267Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "StreamLake",
      "accuracy": 0.7166665000000001,
      "accuracy_std_dev": 0.0033334999999999892,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.015074639999999999,
      "avg_duration_per_task_ms": 124785.23,
      "avg_output_tokens_per_task": 11839.88,
      "last_run_timestamp": "2026-10-31T12:28:39.529Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-sol-pro-20260709",
      "model_slug": "openai/gpt-5.6-sol-pro",
      "model_name": "OpenAI: GPT-5.6 Sol Pro",
      "provider_name": "OpenAI",
      "accuracy": 0.7166665000000001,
      "accuracy_std_dev": 0.03666649999999999,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.71669186,
      "avg_duration_per_task_ms": 277432.68,
      "avg_output_tokens_per_task": 22919.22,
      "last_run_timestamp": "2026-10-02T23:08:29.026Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Relace",
      "accuracy": 0.7166665000000001,
      "accuracy_std_dev": 0.02603405888927041,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.13610847,
      "avg_duration_per_task_ms": 230555.42,
      "avg_output_tokens_per_task": 11694.86,
      "last_run_timestamp": "2026-10-31T19:46:31.005Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Venice",
      "accuracy": 0.7159998000000001,
      "accuracy_std_dev": 0.03295793548085197,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.24450749400000002,
      "avg_duration_per_task_ms": 314057.314,
      "avg_output_tokens_per_task": 21575.564,
      "last_run_timestamp": "2026-10-31T12:46:35.433Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "Venice",
      "accuracy": 0.71583325,
      "accuracy_std_dev": 0.03770315157831637,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0668666975,
      "avg_duration_per_task_ms": 253791.08,
      "avg_output_tokens_per_task": 27006.42,
      "last_run_timestamp": "2026-10-31T19:22:37.818Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Alibaba",
      "accuracy": 0.7157778333333332,
      "accuracy_std_dev": 0.032271291757105434,
      "run_count": 30,
      "total_questions": 1500,
      "avg_cost_per_task_usd": 0.116951148,
      "avg_duration_per_task_ms": 125547.258,
      "avg_output_tokens_per_task": 8555.182666666668,
      "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-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": "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": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "Parasail",
      "accuracy": 0.7148900152284264,
      "accuracy_std_dev": 0.09117406530359592,
      "run_count": 8,
      "total_questions": 394,
      "avg_cost_per_task_usd": 0.14045050253807106,
      "avg_duration_per_task_ms": 1841531.8324873096,
      "avg_output_tokens_per_task": 104496.4695431472,
      "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": "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": 0.7145454545454546,
      "accuracy_std_dev": 0.032388915902494195,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.046899645454545454,
      "avg_duration_per_task_ms": 102311.42727272728,
      "avg_output_tokens_per_task": 8424.598181818183,
      "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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "SiliconFlow",
      "accuracy": 0.7145453636363637,
      "accuracy_std_dev": 0.028258254765877216,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.0781004418181818,
      "avg_duration_per_task_ms": 129072.70181818181,
      "avg_output_tokens_per_task": 9377.12,
      "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": "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": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "Minimax",
      "accuracy": 0.7125000625,
      "accuracy_std_dev": 0.023790114515983167,
      "run_count": 16,
      "total_questions": 800,
      "avg_cost_per_task_usd": 0.0387893925,
      "avg_duration_per_task_ms": 225024.16875,
      "avg_output_tokens_per_task": 9749.72875,
      "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": "openai/gpt-5.6-terra-20260709",
      "model_slug": "openai/gpt-5.6-terra",
      "model_name": "OpenAI: GPT-5.6 Terra",
      "provider_name": "Azure",
      "accuracy": 0.712381,
      "accuracy_std_dev": 0.016106167044245584,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.11192505714285715,
      "avg_duration_per_task_ms": 96202.06,
      "avg_output_tokens_per_task": 3694.017142857143,
      "last_run_timestamp": "2026-10-01T19:04:52.651Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "CoreWeave",
      "accuracy": 0.7122223333333333,
      "accuracy_std_dev": 0.04016943897071779,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.07876202666666668,
      "avg_duration_per_task_ms": 110695.52666666667,
      "avg_output_tokens_per_task": 8630.058333333332,
      "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": "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": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7111111666666666,
      "accuracy_std_dev": 0.06618325102425,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.10397005333333333,
      "avg_duration_per_task_ms": 211488.60333333333,
      "avg_output_tokens_per_task": 11044.173333333334,
      "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": "Wafer",
      "accuracy": 0.7109090454545455,
      "accuracy_std_dev": 0.02554809684866649,
      "run_count": 22,
      "total_questions": 1100,
      "avg_cost_per_task_usd": 0.0988174109090909,
      "avg_duration_per_task_ms": 108917.78818181818,
      "avg_output_tokens_per_task": 9669.932727272728,
      "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": "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": "GMICloud",
      "accuracy": 0.7096295555555556,
      "accuracy_std_dev": 0.018083757273747114,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.022423851111111108,
      "avg_duration_per_task_ms": 274816.1711111111,
      "avg_output_tokens_per_task": 10081.8,
      "last_run_timestamp": "2026-10-31T19:56:09.033Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash-20260826",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "Z.ai: GLM 5.3 Flash",
      "provider_name": "GMICloud",
      "accuracy": 0.7093333,
      "accuracy_std_dev": 0.07889230465267193,
      "run_count": 10,
      "total_questions": 500,
      "avg_cost_per_task_usd": 0.010360992,
      "avg_duration_per_task_ms": 240638.89,
      "avg_output_tokens_per_task": 9031.174,
      "last_run_timestamp": "2026-10-31T12:26:55.059Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "auto-routing",
      "accuracy": 0.709166625,
      "accuracy_std_dev": 0.011989726507905632,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0270094325,
      "avg_duration_per_task_ms": 227764.9775,
      "avg_output_tokens_per_task": 10406.5575,
      "last_run_timestamp": "2026-10-31T19:47:12.389Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Ambient",
      "accuracy": 0.709166625,
      "accuracy_std_dev": 0.04593912865667321,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0539720225,
      "avg_duration_per_task_ms": 239606.0525,
      "avg_output_tokens_per_task": 8198.2525,
      "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": "qwen/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "AtlasCloud",
      "accuracy": 0.7090909090909091,
      "accuracy_std_dev": 0.15487416012037558,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.062087180000000006,
      "avg_duration_per_task_ms": 147570.8818181818,
      "avg_output_tokens_per_task": 16917.33090909091,
      "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": "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": "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": "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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Phala",
      "accuracy": 0.7074073333333334,
      "accuracy_std_dev": 0.018972248885616532,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.09800756444444444,
      "avg_duration_per_task_ms": 168166.58444444445,
      "avg_output_tokens_per_task": 11095.48,
      "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/gemini-3.5-flash-20260519",
      "model_slug": "google/gemini-3.5-flash",
      "model_name": "Google: Gemini 3.5 Flash",
      "provider_name": "Google AI Studio",
      "accuracy": 0.706667,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.44632813,
      "avg_duration_per_task_ms": 132811.5,
      "avg_output_tokens_per_task": 20668.24,
      "last_run_timestamp": "2026-10-01T19:04:52.658Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-sol-pro-20260709",
      "model_slug": "openai/gpt-5.6-sol-pro",
      "model_name": "OpenAI: GPT-5.6 Sol Pro",
      "provider_name": "auto-routing",
      "accuracy": 0.706667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.71176192,
      "avg_duration_per_task_ms": 285827.28,
      "avg_output_tokens_per_task": 22895.76,
      "last_run_timestamp": "2026-10-01T16:00:34.660Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "ModelRun",
      "accuracy": 0.706667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.1838298,
      "avg_duration_per_task_ms": 329811.9,
      "avg_output_tokens_per_task": 9540.84,
      "last_run_timestamp": "2026-08-15T00:16:18.405Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-astra-pro-20260903",
      "model_slug": "openai/gpt-6-astra-pro",
      "model_name": "OpenAI: GPT-6 Astra Pro",
      "provider_name": "OpenAI",
      "accuracy": 0.7066669999999999,
      "accuracy_std_dev": 0.019999999999999962,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 3.04321209,
      "avg_duration_per_task_ms": 317897.29,
      "avg_output_tokens_per_task": 13272.53,
      "last_run_timestamp": "2026-10-02T23:13:16.494Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6.1-sol-20260929",
      "model_slug": "openai/gpt-6.1-sol",
      "model_name": "OpenAI: GPT-6.1 Sol",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.7066669999999999,
      "accuracy_std_dev": 0.019999999999999962,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.1042244,
      "avg_duration_per_task_ms": 226024.67,
      "avg_output_tokens_per_task": 3543.23,
      "last_run_timestamp": "2026-10-30T00:17:12.390Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "AtlasCloud",
      "accuracy": 0.70666675,
      "accuracy_std_dev": 0.05477228618049002,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.031395105,
      "avg_duration_per_task_ms": 328094.585,
      "avg_output_tokens_per_task": 10267.135,
      "last_run_timestamp": "2026-10-31T20:10:49.937Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "GMICloud",
      "accuracy": 0.70666675,
      "accuracy_std_dev": 0.027080004924067123,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.017216295,
      "avg_duration_per_task_ms": 228198.02,
      "avg_output_tokens_per_task": 15657.435,
      "last_run_timestamp": "2026-10-31T13:09:38.196Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "DigitalOcean",
      "accuracy": 0.7066666666666667,
      "accuracy_std_dev": 0.014401440259764123,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.015417799999999999,
      "avg_duration_per_task_ms": 75266.33333333333,
      "avg_output_tokens_per_task": 12018.02,
      "last_run_timestamp": "2026-10-31T11:20:39.476Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5-2025-08-07",
      "model_slug": "openai/gpt-5",
      "model_name": "OpenAI: GPT-5",
      "provider_name": "OpenAI",
      "accuracy": 0.7066666666666667,
      "accuracy_std_dev": 0.005443174461857933,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.5994524266666668,
      "avg_duration_per_task_ms": 636541.5466666666,
      "avg_output_tokens_per_task": 54354.86,
      "last_run_timestamp": "2026-10-02T23:22:41.519Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-luna-pro-20260709",
      "model_slug": "openai/gpt-5.6-luna-pro",
      "model_name": "OpenAI: GPT-5.6 Luna Pro",
      "provider_name": "OpenAI",
      "accuracy": 0.7066666666666667,
      "accuracy_std_dev": 0.01962587106064057,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.08575882,
      "avg_duration_per_task_ms": 316120.8933333333,
      "avg_output_tokens_per_task": 32977.22666666667,
      "last_run_timestamp": "2026-10-02T23:12:53.530Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-31b-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": "openai/gpt-5.6-luna-20260709",
      "model_slug": "openai/gpt-5.6-luna",
      "model_name": "OpenAI: GPT-5.6 Luna",
      "provider_name": "Azure",
      "accuracy": 0.7057145714285713,
      "accuracy_std_dev": 0.030327119046712443,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.012995502857142857,
      "avg_duration_per_task_ms": 102610.46857142857,
      "avg_output_tokens_per_task": 4898.9228571428575,
      "last_run_timestamp": "2026-10-01T19:05:16.550Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "DeepInfra",
      "accuracy": 0.7055555833333333,
      "accuracy_std_dev": 0.03440877813014756,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.041761648333333325,
      "avg_duration_per_task_ms": 145144.42333333334,
      "avg_output_tokens_per_task": 8526.811666666666,
      "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-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "StreamLake",
      "accuracy": 0.7055555,
      "accuracy_std_dev": 0.02979349502408202,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.03607245166666667,
      "avg_duration_per_task_ms": 168553.58333333334,
      "avg_output_tokens_per_task": 9088.338333333333,
      "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": "openai/gpt-5-mini-2025-08-07",
      "model_slug": "openai/gpt-5-mini",
      "model_name": "OpenAI: GPT-5 Mini",
      "provider_name": "Azure",
      "accuracy": 0.70499975,
      "accuracy_std_dev": 0.011902485735656231,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.13260128,
      "avg_duration_per_task_ms": 477326.39,
      "avg_output_tokens_per_task": 55745.89,
      "last_run_timestamp": "2026-10-01T19:05:16.547Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Chutes",
      "accuracy": 0.7047618571428571,
      "accuracy_std_dev": 0.03161536123210892,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.08984225142857143,
      "avg_duration_per_task_ms": 859687.0971428572,
      "avg_output_tokens_per_task": 9438.171428571428,
      "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": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "DeepInfra",
      "accuracy": 0.7044443333333332,
      "accuracy_std_dev": 0.022662526857249517,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.02059840666666667,
      "avg_duration_per_task_ms": 207682.6,
      "avg_output_tokens_per_task": 9037.553333333333,
      "last_run_timestamp": "2026-08-29T16:45:30.973Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Z.AI",
      "accuracy": 0.7042425454545455,
      "accuracy_std_dev": 0.024662964506480832,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.12288133272727272,
      "avg_duration_per_task_ms": 163283.9090909091,
      "avg_output_tokens_per_task": 11693.267272727273,
      "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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Together",
      "accuracy": 0.7042424545454545,
      "accuracy_std_dev": 0.0349627848858641,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.10044568909090908,
      "avg_duration_per_task_ms": 91693.66545454545,
      "avg_output_tokens_per_task": 8696.44,
      "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": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "Venice",
      "accuracy": 0.70416675,
      "accuracy_std_dev": 0.03264911731100092,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.07095855000000001,
      "avg_duration_per_task_ms": 325685.87,
      "avg_output_tokens_per_task": 10970.505,
      "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": "deepseek/deepseek-v4.1-flash-20261017",
      "model_slug": "deepseek/deepseek-v4.1-flash-1017",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash 1017",
      "provider_name": "Makora",
      "accuracy": 0.7039998000000001,
      "accuracy_std_dev": 0.025508095447524113,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.021704752,
      "avg_duration_per_task_ms": 119683.016,
      "avg_output_tokens_per_task": 11904.78,
      "last_run_timestamp": "2026-10-31T12:25:42.323Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.6-luna-pro-20260709",
      "model_slug": "openai/gpt-5.6-luna-pro",
      "model_name": "OpenAI: GPT-5.6 Luna Pro",
      "provider_name": "auto-routing",
      "accuracy": 0.7033335,
      "accuracy_std_dev": 0.0033335000000000448,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.08534018,
      "avg_duration_per_task_ms": 347658.09,
      "avg_output_tokens_per_task": 32908.85,
      "last_run_timestamp": "2026-10-01T16:00:47.532Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.7-plus-20260602",
      "model_slug": "qwen/qwen3.7-plus",
      "model_name": "Qwen: Qwen3.7 Plus",
      "provider_name": "auto-routing",
      "accuracy": 0.703333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.05425128,
      "avg_duration_per_task_ms": 396980.46,
      "avg_output_tokens_per_task": 20673.9,
      "last_run_timestamp": "2026-10-01T08:05:20.277Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Venice",
      "accuracy": 0.7030302727272727,
      "accuracy_std_dev": 0.04093488390134421,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.11610903454545456,
      "avg_duration_per_task_ms": 126684.9490909091,
      "avg_output_tokens_per_task": 9547.561818181819,
      "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-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Nebius",
      "accuracy": 0.702963,
      "accuracy_std_dev": 0.02847753091278963,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.2806454911111111,
      "avg_duration_per_task_ms": 494767.4533333333,
      "avg_output_tokens_per_task": 9478.277777777777,
      "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": "openai/gpt-5.6-terra-20260709",
      "model_slug": "openai/gpt-5.6-terra",
      "model_name": "OpenAI: GPT-5.6 Terra",
      "provider_name": "OpenAI",
      "accuracy": 0.7022223333333333,
      "accuracy_std_dev": 0.021999169459070246,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.10818771333333332,
      "avg_duration_per_task_ms": 79804.25333333333,
      "avg_output_tokens_per_task": 3542.24,
      "last_run_timestamp": "2026-10-02T23:10:36.500Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-sol-20260922",
      "model_slug": "openai/gpt-6-sol",
      "model_name": "OpenAI: GPT-6 Sol",
      "provider_name": "OpenAI",
      "accuracy": 0.7022223333333333,
      "accuracy_std_dev": 0.011330936570096727,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.13357544,
      "avg_duration_per_task_ms": 105586.74666666667,
      "avg_output_tokens_per_task": 5260.826666666667,
      "last_run_timestamp": "2026-10-02T23:07:17.250Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "StreamLake",
      "accuracy": 0.7022223333333333,
      "accuracy_std_dev": 0.046294969462734856,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.13726572666666667,
      "avg_duration_per_task_ms": 333190.6066666667,
      "avg_output_tokens_per_task": 10328.673333333334,
      "last_run_timestamp": "2026-08-30T00:01:09.490Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Parasail",
      "accuracy": 0.7012120909090909,
      "accuracy_std_dev": 0.026102771387012606,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.11443949090909093,
      "avg_duration_per_task_ms": 89100.04545454546,
      "avg_output_tokens_per_task": 9820.098181818183,
      "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-4.7-20251222",
      "model_slug": "z-ai/glm-4.7",
      "model_name": "Z.ai: GLM 4.7",
      "provider_name": "Z.AI",
      "accuracy": 0.7008332500000001,
      "accuracy_std_dev": 0.04182460234344735,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.06761195,
      "avg_duration_per_task_ms": 346627.2725,
      "avg_output_tokens_per_task": 10080.1925,
      "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": "anthropic/claude-4-sonnet-20250522",
      "model_slug": "anthropic/claude-sonnet-4",
      "model_name": "Anthropic: Claude Sonnet 4",
      "provider_name": "auto-routing",
      "accuracy": 0.7,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.29769207999999997,
      "avg_duration_per_task_ms": 128375.6,
      "avg_output_tokens_per_task": 8500.86,
      "last_run_timestamp": "2026-09-01T12:41:24.616Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Google",
      "accuracy": 0.7,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.0355865,
      "avg_duration_per_task_ms": 154491.8,
      "avg_output_tokens_per_task": 12574.4,
      "last_run_timestamp": "2026-10-31T11:15:31.400Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.3-flash",
      "model_slug": "z-ai/glm-5.3-flash",
      "model_name": "GLM 5.3 Flash",
      "provider_name": "digitalocean",
      "accuracy": 0.7,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.00357748,
      "avg_duration_per_task_ms": 48793.08,
      "avg_output_tokens_per_task": 1567.76,
      "last_run_timestamp": "2026-10-01T23:50:45.022Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "inclusionai/ling-3.0-flash-vl-20260910",
      "model_slug": "inclusionai/ling-3.0-flash-vl",
      "model_name": "inclusionAI: Ling 3.0 Flash VL",
      "provider_name": "auto-routing",
      "accuracy": 0.7,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.0030046400000000003,
      "avg_duration_per_task_ms": 98848.12,
      "avg_output_tokens_per_task": 17113.74,
      "last_run_timestamp": "2026-10-01T08:00:19.901Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.7,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.12726106,
      "avg_duration_per_task_ms": 181226.44,
      "avg_output_tokens_per_task": 9838.56,
      "last_run_timestamp": "2026-10-04T03:00:16.405Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m3.1-flash-20261019",
      "model_slug": "minimax/minimax-m3.1-flash",
      "model_name": "MiniMax: MiniMax M3.1 Flash",
      "provider_name": "StreamLake",
      "accuracy": 0.7,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.0161093,
      "avg_duration_per_task_ms": 215051.72,
      "avg_output_tokens_per_task": 15584.08,
      "last_run_timestamp": "2026-10-31T12:24:55.227Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "Phala",
      "accuracy": 0.7,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.07376152,
      "avg_duration_per_task_ms": 750952.84,
      "avg_output_tokens_per_task": 11282.28,
      "last_run_timestamp": "2026-08-29T16:44:16.336Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5-2025-08-07",
      "model_slug": "openai/gpt-5",
      "model_name": "OpenAI: GPT-5",
      "provider_name": "Azure",
      "accuracy": 0.7,
      "accuracy_std_dev": 0.004714280910170736,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.634261665,
      "avg_duration_per_task_ms": 673734.145,
      "avg_output_tokens_per_task": 52923.67,
      "last_run_timestamp": "2026-10-01T16:01:02.669Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6.1-sol-20260929",
      "model_slug": "openai/gpt-6.1-sol",
      "model_name": "OpenAI: GPT-6.1 Sol",
      "provider_name": "auto-routing",
      "accuracy": 0.7,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.1012818,
      "avg_duration_per_task_ms": 144103.72,
      "avg_output_tokens_per_task": 3638.14,
      "last_run_timestamp": "2026-10-01T08:03:00.822Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6.1-terra-20261026",
      "model_slug": "openai/gpt-6.1-terra",
      "model_name": "OpenAI: GPT-6.1 Terra",
      "provider_name": "OpenAI",
      "accuracy": 0.7,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.05778106,
      "avg_duration_per_task_ms": 103456.94,
      "avg_output_tokens_per_task": 4693.4,
      "last_run_timestamp": "2026-10-28T00:06:02.481Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Decart",
      "accuracy": 0.6992855714285715,
      "accuracy_std_dev": 0.022473227207102254,
      "run_count": 14,
      "total_questions": 700,
      "avg_cost_per_task_usd": 0.1477709457142857,
      "avg_duration_per_task_ms": 78438.86571428571,
      "avg_output_tokens_per_task": 10556.58,
      "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": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "DeepInfra",
      "accuracy": 0.6987950361445784,
      "accuracy_std_dev": 0.0178749221984321,
      "run_count": 5,
      "total_questions": 249,
      "avg_cost_per_task_usd": 0.04930324899598394,
      "avg_duration_per_task_ms": 404321.7309236948,
      "avg_output_tokens_per_task": 8360.08032128514,
      "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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Inceptron",
      "accuracy": 0.6983862420701168,
      "accuracy_std_dev": 0.027878199943260105,
      "run_count": 12,
      "total_questions": 599,
      "avg_cost_per_task_usd": 0.07633183305509182,
      "avg_duration_per_task_ms": 154743.70951585978,
      "avg_output_tokens_per_task": 9542.741235392321,
      "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": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "auto-routing",
      "accuracy": 0.698333375,
      "accuracy_std_dev": 0.030867217869195378,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.05358634250000001,
      "avg_duration_per_task_ms": 348281.08,
      "avg_output_tokens_per_task": 9167.5225,
      "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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "CoreWeave",
      "accuracy": 0.698333375,
      "accuracy_std_dev": 0.043301188398638384,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.06877143999999999,
      "avg_duration_per_task_ms": 73336.0625,
      "avg_output_tokens_per_task": 9928.085,
      "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": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "Parasail",
      "accuracy": 0.698333,
      "accuracy_std_dev": 0.0050000000000000044,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.05153802,
      "avg_duration_per_task_ms": 227792.72,
      "avg_output_tokens_per_task": 6678.6,
      "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": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "DigitalOcean",
      "accuracy": 0.6977779999999999,
      "accuracy_std_dev": 0.03095182238038119,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.051447179999999995,
      "avg_duration_per_task_ms": 808105.6666666666,
      "avg_output_tokens_per_task": 8854.186666666666,
      "last_run_timestamp": "2026-08-29T16:47:05.512Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-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-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "Venice",
      "accuracy": 0.6975001249999999,
      "accuracy_std_dev": 0.025806460486850466,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.045653685,
      "avg_duration_per_task_ms": 547622.5375,
      "avg_output_tokens_per_task": 9024.185,
      "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": "BaseTen",
      "accuracy": 0.6969061874251495,
      "accuracy_std_dev": 0.026180217814685852,
      "run_count": 34,
      "total_questions": 1670,
      "avg_cost_per_task_usd": 0.10252084191616766,
      "avg_duration_per_task_ms": 73055.79640718563,
      "avg_output_tokens_per_task": 8098.897005988024,
      "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": "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-4.7-flash-20260119",
      "model_slug": "z-ai/glm-4.7-flash",
      "model_name": "Z.ai: GLM 4.7 Flash",
      "provider_name": "Venice",
      "accuracy": 0.6966665000000001,
      "accuracy_std_dev": 0.0033334999999999892,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.0061224999999999995,
      "avg_duration_per_task_ms": 182716.13,
      "avg_output_tokens_per_task": 9606.88,
      "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": "nvidia/nemotron-3.5-super-vl-20261015",
      "model_slug": "nvidia/nemotron-3.5-super-vl",
      "model_name": "NVIDIA: Nemotron 3.5 Super VL",
      "provider_name": "Io Net",
      "accuracy": 0.6963472739726028,
      "accuracy_std_dev": 0.05315590618172168,
      "run_count": 3,
      "total_questions": 146,
      "avg_cost_per_task_usd": 0.01628368493150685,
      "avg_duration_per_task_ms": 275676.23287671234,
      "avg_output_tokens_per_task": 12531.828767123288,
      "last_run_timestamp": "2026-10-31T13:29:36.200Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Reka",
      "accuracy": 0.6958335,
      "accuracy_std_dev": 0.03781341986980283,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.132227115,
      "avg_duration_per_task_ms": 205681.995,
      "avg_output_tokens_per_task": 9407.695,
      "last_run_timestamp": "2026-10-01T22:50:26.018Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "moonshotai/kimi-k2.6-20260420",
      "model_slug": "moonshotai/kimi-k2.6",
      "model_name": "MoonshotAI: Kimi K2.6",
      "provider_name": "Decart",
      "accuracy": 0.6955782448979592,
      "accuracy_std_dev": 0.13306624277967724,
      "run_count": 4,
      "total_questions": 196,
      "avg_cost_per_task_usd": 0.07090773979591837,
      "avg_duration_per_task_ms": 158871.75,
      "avg_output_tokens_per_task": 16373.045918367347,
      "last_run_timestamp": "2026-10-03T16:16:30.145Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "StreamLake",
      "accuracy": 0.6955556666666667,
      "accuracy_std_dev": 0.021998731726069,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.04300826666666667,
      "avg_duration_per_task_ms": 173663.04666666666,
      "avg_output_tokens_per_task": 8296.866666666667,
      "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": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Crusoe",
      "accuracy": 0.6948719487179488,
      "accuracy_std_dev": 0.16500680809647672,
      "run_count": 4,
      "total_questions": 195,
      "avg_cost_per_task_usd": 0.024616958974358973,
      "avg_duration_per_task_ms": 297869.91282051284,
      "avg_output_tokens_per_task": 9525.128205128205,
      "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": "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": 0.694073888888889,
      "accuracy_std_dev": 0.014209898212509368,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.0217813,
      "avg_duration_per_task_ms": 480864.6,
      "avg_output_tokens_per_task": 18593.075555555555,
      "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": "Cloudflare",
      "accuracy": 0.6940476428571428,
      "accuracy_std_dev": 0.033242348446365685,
      "run_count": 14,
      "total_questions": 700,
      "avg_cost_per_task_usd": 0.12397131285714287,
      "avg_duration_per_task_ms": 129105.92571428571,
      "avg_output_tokens_per_task": 8940.347142857143,
      "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": "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": "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": "qwen/qwen3.6-27b-20260422",
      "model_slug": "qwen/qwen3.6-27b",
      "model_name": "Qwen: Qwen3.6 27B",
      "provider_name": "Io Net",
      "accuracy": 0.693333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.13946158,
      "avg_duration_per_task_ms": 819620.14,
      "avg_output_tokens_per_task": 28549.28,
      "last_run_timestamp": "2026-08-15T08:15:24.293Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "reflection/beam-1.0-501b-a23b-20261016",
      "model_slug": "reflection/beam-1.0-501b-a23b",
      "model_name": "Reflection: Beam 1.0 501B A23B",
      "provider_name": "auto-routing",
      "accuracy": 0.693333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.02754302,
      "avg_duration_per_task_ms": 209408.8,
      "avg_output_tokens_per_task": 14914.54,
      "last_run_timestamp": "2026-10-24T12:03:24.771Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260423",
      "model_slug": "deepseek/deepseek-v4-flash",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0423",
      "provider_name": "Mancer 2",
      "accuracy": 0.6928572857142857,
      "accuracy_std_dev": 0.07897025990969059,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.047876402857142855,
      "avg_duration_per_task_ms": 481015.8685714286,
      "avg_output_tokens_per_task": 15710.031428571428,
      "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": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Venice",
      "accuracy": 0.6923937315436242,
      "accuracy_std_dev": 0.032633747756987186,
      "run_count": 3,
      "total_questions": 149,
      "avg_cost_per_task_usd": 0.01647934899328859,
      "avg_duration_per_task_ms": 211946.26845637584,
      "avg_output_tokens_per_task": 8844.79865771812,
      "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": "qwen/qwen3.8-27b-20260814",
      "model_slug": "qwen/qwen3.8-27b",
      "model_name": "Qwen: Qwen3.8 27B",
      "provider_name": "Tenstorrent",
      "accuracy": 0.6919048571428571,
      "accuracy_std_dev": 0.1628112211782438,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.01966778285714286,
      "avg_duration_per_task_ms": 333360.0657142857,
      "avg_output_tokens_per_task": 11239.56,
      "last_run_timestamp": "2026-11-04T18:04:38.320Z"
    },
    {
      "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": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "AtlasCloud",
      "accuracy": 0.69166675,
      "accuracy_std_dev": 0.09607597832542482,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.09531616,
      "avg_duration_per_task_ms": 242770.45,
      "avg_output_tokens_per_task": 22128.1325,
      "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": "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": 0.6873332000000001,
      "accuracy_std_dev": 0.026686126789526168,
      "run_count": 30,
      "total_questions": 1500,
      "avg_cost_per_task_usd": 0.10512817333333335,
      "avg_duration_per_task_ms": 76744.592,
      "avg_output_tokens_per_task": 9600.512666666667,
      "last_run_timestamp": "2026-11-02T21:38:02.811Z"
    },
    {
      "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-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Minimax",
      "accuracy": 0.6866668333333333,
      "accuracy_std_dev": 0.029313263337134965,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.03567404666666667,
      "avg_duration_per_task_ms": 155529.32333333333,
      "avg_output_tokens_per_task": 9866.113333333333,
      "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": "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": 0.6866665000000001,
      "accuracy_std_dev": 0.04496909432599238,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0149933775,
      "avg_duration_per_task_ms": 333684.6375,
      "avg_output_tokens_per_task": 19970.01,
      "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": "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": "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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "GMICloud",
      "accuracy": 0.6838888333333334,
      "accuracy_std_dev": 0.04670285039094391,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.09925613333333333,
      "avg_duration_per_task_ms": 184779.46,
      "avg_output_tokens_per_task": 7335.005,
      "last_run_timestamp": "2026-10-31T19:10:43.801Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "Nebius",
      "accuracy": 0.6835019898989899,
      "accuracy_std_dev": 0.023980399500164742,
      "run_count": 4,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.03288487878787879,
      "avg_duration_per_task_ms": 455208.9646464646,
      "avg_output_tokens_per_task": 20918.757575757576,
      "last_run_timestamp": "2026-11-05T15:52:10.043Z"
    },
    {
      "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": "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": "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": 0.6830808080808081,
      "accuracy_std_dev": 0.024660860760271505,
      "run_count": 8,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.020343022727272727,
      "avg_duration_per_task_ms": 329622.595959596,
      "avg_output_tokens_per_task": 16902.575757575756,
      "last_run_timestamp": "2026-11-04T19:25:46.206Z"
    },
    {
      "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": 0.682995527027027,
      "accuracy_std_dev": 0.017120226436067428,
      "run_count": 6,
      "total_questions": 296,
      "avg_cost_per_task_usd": 0.03385346621621622,
      "avg_duration_per_task_ms": 512642.8918918919,
      "avg_output_tokens_per_task": 19026.391891891893,
      "last_run_timestamp": "2026-10-17T19:21:01.980Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-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": "z-ai/glm-5.2-20260616",
      "model_slug": "z-ai/glm-5.2",
      "model_name": "Z.ai: GLM 5.2",
      "provider_name": "Friendli",
      "accuracy": 0.68222225,
      "accuracy_std_dev": 0.06499753651629805,
      "run_count": 12,
      "total_questions": 600,
      "avg_cost_per_task_usd": 0.108739995,
      "avg_duration_per_task_ms": 80556.50333333333,
      "avg_output_tokens_per_task": 8974.771666666667,
      "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": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "Google",
      "accuracy": 0.6816577724867724,
      "accuracy_std_dev": 0.02279793656232948,
      "run_count": 4,
      "total_questions": 189,
      "avg_cost_per_task_usd": 0.13611332804232804,
      "avg_duration_per_task_ms": 393574.9259259259,
      "avg_output_tokens_per_task": 10439.26455026455,
      "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": "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": "z-ai/glm-5.1-20260406",
      "model_slug": "z-ai/glm-5.1",
      "model_name": "Z.ai: GLM 5.1",
      "provider_name": "Friendli",
      "accuracy": 0.6800001111111111,
      "accuracy_std_dev": 0.0689962496749633,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.09411114000000001,
      "avg_duration_per_task_ms": 131955.26888888888,
      "avg_output_tokens_per_task": 8809.797777777778,
      "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": "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": 0.6787879090909091,
      "accuracy_std_dev": 0.033492690704511906,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.016563767272727275,
      "avg_duration_per_task_ms": 290847.2818181818,
      "avg_output_tokens_per_task": 19763.932727272728,
      "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": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "Darkbloom",
      "accuracy": 0.678333,
      "accuracy_std_dev": 0.0050000000000000044,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.0069528,
      "avg_duration_per_task_ms": 471037.61,
      "avg_output_tokens_per_task": 19230.47,
      "last_run_timestamp": "2026-10-31T13:27:50.961Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4-flash-vision-exp-20260821",
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "model_name": "DeepSeek: DeepSeek V4 Flash Vision Exp",
      "provider_name": "Fireworks",
      "accuracy": 0.6777776666666667,
      "accuracy_std_dev": 0.013698571612479227,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.0340762,
      "avg_duration_per_task_ms": 244093.44,
      "avg_output_tokens_per_task": 27086.806666666667,
      "last_run_timestamp": "2026-09-19T08:01:53.393Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "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": "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": 0.6763826771300447,
      "accuracy_std_dev": 0.027382172972416883,
      "run_count": 9,
      "total_questions": 446,
      "avg_cost_per_task_usd": 0.021115630044843046,
      "avg_duration_per_task_ms": 298545.9484304933,
      "avg_output_tokens_per_task": 16751.952914798207,
      "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": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "GMICloud",
      "accuracy": 0.67583325,
      "accuracy_std_dev": 0.10893112439145894,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0162695775,
      "avg_duration_per_task_ms": 256380.655,
      "avg_output_tokens_per_task": 10959.865,
      "last_run_timestamp": "2026-10-31T19:58:02.198Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "qwen/qwen3.5-397b-a17b-20260216",
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "model_name": "Qwen: Qwen3.5 397B A17B",
      "provider_name": "Chutes",
      "accuracy": 0.6736109999999998,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 48,
      "avg_cost_per_task_usd": 0.11439916666666666,
      "avg_duration_per_task_ms": 982669.25,
      "avg_output_tokens_per_task": 30745.75,
      "last_run_timestamp": "2026-08-15T14:15:18.888Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.7-flash-20260119",
      "model_slug": "z-ai/glm-4.7-flash",
      "model_name": "Z.ai: GLM 4.7 Flash",
      "provider_name": "DeepInfra",
      "accuracy": 0.6733780939597316,
      "accuracy_std_dev": 0.01094232519470459,
      "run_count": 3,
      "total_questions": 149,
      "avg_cost_per_task_usd": 0.006236456375838926,
      "avg_duration_per_task_ms": 189635.3489932886,
      "avg_output_tokens_per_task": 9352.58389261745,
      "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": "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": 0.6733333333333333,
      "accuracy_std_dev": 0.014401903168987377,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.007971766666666666,
      "avg_duration_per_task_ms": 205309.70666666667,
      "avg_output_tokens_per_task": 9557.78,
      "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-5.6-sol-pro-20260709",
      "model_slug": "openai/gpt-5.6-sol-pro",
      "model_name": "OpenAI: GPT-5.6 Sol Pro",
      "provider_name": "Azure",
      "accuracy": 0.6733330000000001,
      "accuracy_std_dev": 0.019999999999999962,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 1.5819717300000002,
      "avg_duration_per_task_ms": 314318.37,
      "avg_output_tokens_per_task": 23725.55,
      "last_run_timestamp": "2026-10-01T16:01:13.647Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-6-luna-20260922",
      "model_slug": "openai/gpt-6-luna",
      "model_name": "OpenAI: GPT-6 Luna",
      "provider_name": "auto-routing",
      "accuracy": 0.673333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.00928508,
      "avg_duration_per_task_ms": 121838.98,
      "avg_output_tokens_per_task": 8898.44,
      "last_run_timestamp": "2026-10-01T08:00:22.469Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "thinkingmachines/inkling-20260715",
      "model_slug": "thinkingmachines/inkling",
      "model_name": "Thinking Machines: Inkling",
      "provider_name": "auto-routing",
      "accuracy": 0.672,
      "accuracy_std_dev": 0.028409779224766934,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.07962264,
      "avg_duration_per_task_ms": 80557.024,
      "avg_output_tokens_per_task": 3531.236,
      "last_run_timestamp": "2026-10-01T16:00:45.064Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.7-flash-20260119",
      "model_slug": "z-ai/glm-4.7-flash",
      "model_name": "Z.ai: GLM 4.7 Flash",
      "provider_name": "Cloudflare",
      "accuracy": 0.6716665,
      "accuracy_std_dev": 0.01443369899748499,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.014968054999999997,
      "avg_duration_per_task_ms": 194795.94,
      "avg_output_tokens_per_task": 10086.53,
      "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": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Inceptron",
      "accuracy": 0.67,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.022004160000000002,
      "avg_duration_per_task_ms": 232974.22,
      "avg_output_tokens_per_task": 9887.82,
      "last_run_timestamp": "2026-08-15T14:15:19.430Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "GMICloud",
      "accuracy": 0.6690909090909091,
      "accuracy_std_dev": 0.14703195632270777,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.12102286727272728,
      "avg_duration_per_task_ms": 359186.4090909091,
      "avg_output_tokens_per_task": 8810.841818181818,
      "last_run_timestamp": "2026-10-31T19:48:58.813Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "DigitalOcean",
      "accuracy": 0.6688886666666667,
      "accuracy_std_dev": 0.006285236475706792,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.028048506666666667,
      "avg_duration_per_task_ms": 240999.82,
      "avg_output_tokens_per_task": 10014.44,
      "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-m2.7-20260318",
      "model_slug": "minimax/minimax-m2.7",
      "model_name": "MiniMax: MiniMax M2.7",
      "provider_name": "SambaNova",
      "accuracy": 0.666667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.0755104,
      "avg_duration_per_task_ms": 77993.72,
      "avg_output_tokens_per_task": 14223.7,
      "last_run_timestamp": "2026-08-15T14:12:46.929Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-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-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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Near AI",
      "accuracy": 0.66583325,
      "accuracy_std_dev": 0.03639120582211448,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.006095729999999999,
      "avg_duration_per_task_ms": 325369.79,
      "avg_output_tokens_per_task": 18643.27,
      "last_run_timestamp": "2026-10-31T19:46:03.854Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "auto-routing",
      "accuracy": 0.6644446666666667,
      "accuracy_std_dev": 0.02997950740459595,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.022058193333333333,
      "avg_duration_per_task_ms": 220747.36,
      "avg_output_tokens_per_task": 10146.033333333333,
      "last_run_timestamp": "2026-08-29T16:11:34.887Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4.5-haiku-20251001",
      "model_slug": "anthropic/claude-haiku-4.5",
      "model_name": "Anthropic: Claude Haiku 4.5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.6643136764705884,
      "accuracy_std_dev": 0.02945620942699677,
      "run_count": 34,
      "total_questions": 1700,
      "avg_cost_per_task_usd": 0.12474326823529412,
      "avg_duration_per_task_ms": 118894.24411764706,
      "avg_output_tokens_per_task": 12635.786470588235,
      "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": "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": 0.6622223333333334,
      "accuracy_std_dev": 0.016629618082071397,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.008384560000000001,
      "avg_duration_per_task_ms": 353358.7133333333,
      "avg_output_tokens_per_task": 13464.853333333333,
      "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": "anthropic/claude-4.5-haiku-20251001",
      "model_slug": "anthropic/claude-haiku-4.5",
      "model_name": "Anthropic: Claude Haiku 4.5",
      "provider_name": "auto-routing",
      "accuracy": 0.6618180909090909,
      "accuracy_std_dev": 0.03062851381683575,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.11598973272727274,
      "avg_duration_per_task_ms": 123395.72363636363,
      "avg_output_tokens_per_task": 12658.365454545454,
      "last_run_timestamp": "2026-10-31T12:42:18.189Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.4-mini-20260317",
      "model_slug": "openai/gpt-5.4-mini",
      "model_name": "OpenAI: GPT-5.4 Mini",
      "provider_name": "Azure",
      "accuracy": 0.6613334000000001,
      "accuracy_std_dev": 0.01758814752724117,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.10186603200000001,
      "avg_duration_per_task_ms": 156255.32,
      "avg_output_tokens_per_task": 13208.728,
      "last_run_timestamp": "2026-10-01T19:04:42.436Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "DekaLLM",
      "accuracy": 0.6604760000000001,
      "accuracy_std_dev": 0.025350253033394797,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.008717225714285714,
      "avg_duration_per_task_ms": 574345.4257142857,
      "avg_output_tokens_per_task": 17634.762857142858,
      "last_run_timestamp": "2026-10-31T20:07:05.983Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Parasail",
      "accuracy": 0.6603702222222223,
      "accuracy_std_dev": 0.014353851715192269,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.02473104222222222,
      "avg_duration_per_task_ms": 151337.52666666667,
      "avg_output_tokens_per_task": 16990.524444444443,
      "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": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Friendli",
      "accuracy": 0.66,
      "accuracy_std_dev": 0.005443582705045132,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.029202206666666668,
      "avg_duration_per_task_ms": 237856.5,
      "avg_output_tokens_per_task": 10714.64,
      "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": "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": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "SiliconFlow",
      "accuracy": 0.6577835126903554,
      "accuracy_std_dev": 0.125359473887098,
      "run_count": 8,
      "total_questions": 394,
      "avg_cost_per_task_usd": 0.02656375888324873,
      "avg_duration_per_task_ms": 1104661.6015228427,
      "avg_output_tokens_per_task": 34353.058375634515,
      "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": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "SiliconFlow",
      "accuracy": 0.6577776666666667,
      "accuracy_std_dev": 0.016629618082071397,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.022382520000000003,
      "avg_duration_per_task_ms": 235251.81333333332,
      "avg_output_tokens_per_task": 9436.286666666667,
      "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": "anthropic/claude-4.5-haiku-20251001",
      "model_slug": "anthropic/claude-haiku-4.5",
      "model_name": "Anthropic: Claude Haiku 4.5",
      "provider_name": "Anthropic",
      "accuracy": 0.6575758181818182,
      "accuracy_std_dev": 0.02640359841460382,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.11578603454545454,
      "avg_duration_per_task_ms": 129057.64181818182,
      "avg_output_tokens_per_task": 12484.370909090909,
      "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": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "DigitalOcean",
      "accuracy": 0.6575,
      "accuracy_std_dev": 0.0675721425182005,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.056107382500000004,
      "avg_duration_per_task_ms": 767450.61,
      "avg_output_tokens_per_task": 18037.37,
      "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": "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": "Google",
      "accuracy": 0.656161606060606,
      "accuracy_std_dev": 0.026312321205253775,
      "run_count": 33,
      "total_questions": 1650,
      "avg_cost_per_task_usd": 0.12037706606060608,
      "avg_duration_per_task_ms": 121292.18242424242,
      "avg_output_tokens_per_task": 12576.998787878787,
      "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": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "AtlasCloud",
      "accuracy": 0.65583325,
      "accuracy_std_dev": 0.09059597044674504,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.04999349,
      "avg_duration_per_task_ms": 312377.32,
      "avg_output_tokens_per_task": 7667.325,
      "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": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "Phala",
      "accuracy": 0.6557823102040816,
      "accuracy_std_dev": 0.009318408855593308,
      "run_count": 5,
      "total_questions": 245,
      "avg_cost_per_task_usd": 0.014797057142857142,
      "avg_duration_per_task_ms": 277907.2775510204,
      "avg_output_tokens_per_task": 20245.579591836733,
      "last_run_timestamp": "2026-10-10T11:18:54.565Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "nvidia/nemotron-3.5-lightning-20260807",
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "model_name": "NVIDIA: Nemotron 3.5 Lightning",
      "provider_name": "DeepInfra",
      "accuracy": 0.6555556666666666,
      "accuracy_std_dev": 0.03807065880409451,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.021039175555555557,
      "avg_duration_per_task_ms": 248791.5977777778,
      "avg_output_tokens_per_task": 26269.908888888887,
      "last_run_timestamp": "2026-10-31T12:25:21.635Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.4-nano-20260317",
      "model_slug": "openai/gpt-5.4-nano",
      "model_name": "OpenAI: GPT-5.4 Nano",
      "provider_name": "OpenAI",
      "accuracy": 0.6555556666666666,
      "accuracy_std_dev": 0.012570472951413636,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.030086579999999998,
      "avg_duration_per_task_ms": 125366.48,
      "avg_output_tokens_per_task": 16133.046666666667,
      "last_run_timestamp": "2026-10-02T23:09:54.584Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "AkashML",
      "accuracy": 0.65541675,
      "accuracy_std_dev": 0.02332950117474654,
      "run_count": 16,
      "total_questions": 800,
      "avg_cost_per_task_usd": 0.0099308575,
      "avg_duration_per_task_ms": 308960.4,
      "avg_output_tokens_per_task": 17384.11375,
      "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": "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "auto-routing",
      "accuracy": 0.65416675,
      "accuracy_std_dev": 0.03373725891262509,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.010956202499999998,
      "avg_duration_per_task_ms": 539774.1775,
      "avg_output_tokens_per_task": 18259.87,
      "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": "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": "anthropic/claude-4.5-haiku-20251001",
      "model_slug": "anthropic/claude-haiku-4.5",
      "model_name": "Anthropic: Claude Haiku 4.5",
      "provider_name": "Azure",
      "accuracy": 0.6533334545454546,
      "accuracy_std_dev": 0.023268429213552766,
      "run_count": 11,
      "total_questions": 550,
      "avg_cost_per_task_usd": 0.1143488090909091,
      "avg_duration_per_task_ms": 136176.96727272728,
      "avg_output_tokens_per_task": 12485.16,
      "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": "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": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "Darkbloom",
      "accuracy": 0.6518188195991091,
      "accuracy_std_dev": 0.02538535107955173,
      "run_count": 9,
      "total_questions": 449,
      "avg_cost_per_task_usd": 0.012973,
      "avg_duration_per_task_ms": 774691.5612472161,
      "avg_output_tokens_per_task": 26354.719376391982,
      "last_run_timestamp": "2026-10-31T19:22:53.448Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-26b-a4b-it-20260403",
      "model_slug": "google/gemma-4-26b-a4b-it",
      "model_name": "Google: Gemma 4 26B A4B ",
      "provider_name": "AkashML",
      "accuracy": 0.65166675,
      "accuracy_std_dev": 0.028819327042585528,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.01676374,
      "avg_duration_per_task_ms": 406825.29,
      "avg_output_tokens_per_task": 27732.185,
      "last_run_timestamp": "2026-10-31T19:20:36.815Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "thinkingmachines/inkling-small-20260730",
      "model_slug": "thinkingmachines/inkling-small",
      "model_name": "Thinking Machines: Inkling Small",
      "provider_name": "Together",
      "accuracy": 0.65166675,
      "accuracy_std_dev": 0.04506177549972369,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.04474220500000001,
      "avg_duration_per_task_ms": 212693.005,
      "avg_output_tokens_per_task": 3236.785,
      "last_run_timestamp": "2026-09-12T08:12:16.067Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-9b-20260310",
      "model_slug": "qwen/qwen3.5-9b",
      "model_name": "Qwen: Qwen3.5-9B",
      "provider_name": "Venice",
      "accuracy": 0.65029765625,
      "accuracy_std_dev": 0.07423887380727727,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.025986263392857145,
      "avg_duration_per_task_ms": 280869.27901785716,
      "avg_output_tokens_per_task": 24958.395089285714,
      "last_run_timestamp": "2026-10-17T19:25:29.604Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.4-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": "deepseek/deepseek-v4-flash-20260731",
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "model_name": "DeepSeek: DeepSeek V4 Flash 0731",
      "provider_name": "DigitalOcean",
      "accuracy": 0.6480655267857143,
      "accuracy_std_dev": 0.12510981874337443,
      "run_count": 9,
      "total_questions": 448,
      "avg_cost_per_task_usd": 0.04457617410714286,
      "avg_duration_per_task_ms": 1953706.560267857,
      "avg_output_tokens_per_task": 78316.31919642857,
      "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Groq",
      "accuracy": 0.6479167499999999,
      "accuracy_std_dev": 0.03374491834243344,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0332404675,
      "avg_duration_per_task_ms": 56284.1875,
      "avg_output_tokens_per_task": 14744.0125,
      "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-nano-20260317",
      "model_slug": "openai/gpt-5.4-nano",
      "model_name": "OpenAI: GPT-5.4 Nano",
      "provider_name": "auto-routing",
      "accuracy": 0.6466665,
      "accuracy_std_dev": 0.026666499999999982,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.0295618,
      "avg_duration_per_task_ms": 121397.75,
      "avg_output_tokens_per_task": 15953.93,
      "last_run_timestamp": "2026-10-01T19:04:28.503Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v4.1-flash-20260910",
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "model_name": "DeepSeek: DeepSeek V4.1 Flash",
      "provider_name": "Krea",
      "accuracy": 0.6455556666666666,
      "accuracy_std_dev": 0.09964144484545016,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.11177372666666668,
      "avg_duration_per_task_ms": 710817.72,
      "avg_output_tokens_per_task": 87845.5,
      "last_run_timestamp": "2026-09-29T19:45:03.655Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "DeepInfra",
      "accuracy": 0.6421009059720457,
      "accuracy_std_dev": 0.04038730565548258,
      "run_count": 16,
      "total_questions": 787,
      "avg_cost_per_task_usd": 0.0215750292249047,
      "avg_duration_per_task_ms": 328009.2465057179,
      "avg_output_tokens_per_task": 17335.555273189326,
      "last_run_timestamp": "2026-10-31T20:19:17.569Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5.4-nano-20260317",
      "model_slug": "openai/gpt-5.4-nano",
      "model_name": "OpenAI: GPT-5.4 Nano",
      "provider_name": "Azure",
      "accuracy": 0.6413336000000001,
      "accuracy_std_dev": 0.014847377621654282,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.03101342,
      "avg_duration_per_task_ms": 232783.988,
      "avg_output_tokens_per_task": 15818.512,
      "last_run_timestamp": "2026-10-01T19:05:16.543Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "anthropic/claude-4-sonnet-20250522",
      "model_slug": "anthropic/claude-sonnet-4",
      "model_name": "Anthropic: Claude Sonnet 4",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.64,
      "accuracy_std_dev": 0.020000000000000018,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.29234208,
      "avg_duration_per_task_ms": 227202.07,
      "avg_output_tokens_per_task": 8702.01,
      "last_run_timestamp": "2026-09-01T13:00:51.629Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "ibm-granite/granite-4.2-8b-20260831",
      "model_slug": "ibm-granite/granite-4.2-8b",
      "model_name": "IBM: Granite 4.2 8B",
      "provider_name": "DeepInfra",
      "accuracy": 0.64,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.01684376,
      "avg_duration_per_task_ms": 1178898.4,
      "avg_output_tokens_per_task": 43776.78,
      "last_run_timestamp": "2026-10-09T00:39:20.903Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "inclusionai/ling-3.0-flash-fin-20260827",
      "model_slug": "inclusionai/ling-3.0-flash-fin",
      "model_name": "inclusionAI: Ling 3.0 Flash Fin",
      "provider_name": "DeepInfra",
      "accuracy": 0.638298,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 47,
      "avg_cost_per_task_usd": 0.020538510638297872,
      "avg_duration_per_task_ms": 686659.4680851063,
      "avg_output_tokens_per_task": 47094.31914893617,
      "last_run_timestamp": "2026-10-09T00:23:16.829Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "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": 0.634583375,
      "accuracy_std_dev": 0.061732736285008256,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.024563202499999996,
      "avg_duration_per_task_ms": 408597.23,
      "avg_output_tokens_per_task": 15674.2925,
      "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": 0.6333333333333333,
      "accuracy_std_dev": 0.019625871060640607,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.02406874,
      "avg_duration_per_task_ms": 135941.56666666668,
      "avg_output_tokens_per_task": 10001.226666666667,
      "last_run_timestamp": "2026-08-29T16:49:08.574Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-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/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "SiliconFlow",
      "accuracy": 0.6322223333333333,
      "accuracy_std_dev": 0.13740533890848478,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.05561755666666665,
      "avg_duration_per_task_ms": 502939.72,
      "avg_output_tokens_per_task": 9552.823333333334,
      "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": "deepseek/deepseek-v3.2-20251201",
      "model_slug": "deepseek/deepseek-v3.2",
      "model_name": "DeepSeek: DeepSeek V3.2",
      "provider_name": "AtlasCloud",
      "accuracy": 0.63083325,
      "accuracy_std_dev": 0.12670763646654254,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.051196177499999995,
      "avg_duration_per_task_ms": 879758.5,
      "avg_output_tokens_per_task": 27053.105,
      "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "SambaNova",
      "accuracy": 0.6282699518987342,
      "accuracy_std_dev": 0.030128786590858582,
      "run_count": 8,
      "total_questions": 395,
      "avg_cost_per_task_usd": 0.040355772151898735,
      "avg_duration_per_task_ms": 83816.96455696202,
      "avg_output_tokens_per_task": 19846.754430379748,
      "last_run_timestamp": "2026-10-31T18:59:19.181Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "StreamLake",
      "accuracy": 0.6266666666666667,
      "accuracy_std_dev": 0.0356942779945968,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.021180666666666667,
      "avg_duration_per_task_ms": 241944.27333333335,
      "avg_output_tokens_per_task": 9711.62,
      "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": "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Cerebras",
      "accuracy": 0.6225,
      "accuracy_std_dev": 0.024706510614613316,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.068399865,
      "avg_duration_per_task_ms": 32162.2675,
      "avg_output_tokens_per_task": 21719.0175,
      "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": "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": 0.02177351432359967,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.028176926666666664,
      "avg_duration_per_task_ms": 213805.89333333334,
      "avg_output_tokens_per_task": 9686.44,
      "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": "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": 0.6166665,
      "accuracy_std_dev": 0.01666650000000003,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.21128705,
      "avg_duration_per_task_ms": 934328.99,
      "avg_output_tokens_per_task": 27359.88,
      "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": "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": 0.6166665,
      "accuracy_std_dev": 0.0033334999999999892,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.12254571,
      "avg_duration_per_task_ms": 607769.66,
      "avg_output_tokens_per_task": 27082.3,
      "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/qwen3-235b-a22b-thinking-2507",
      "model_slug": "qwen/qwen3-235b-a22b-thinking-2507",
      "model_name": "Qwen: Qwen3 235B A22B Thinking 2507",
      "provider_name": "auto-routing",
      "accuracy": 0.6133335,
      "accuracy_std_dev": 0.013333499999999998,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.12268729,
      "avg_duration_per_task_ms": 618681.49,
      "avg_output_tokens_per_task": 27773.81,
      "last_run_timestamp": "2026-08-22T16:01:40.197Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "thinkingmachines/inkling-small-20260730",
      "model_slug": "thinkingmachines/inkling-small",
      "model_name": "Thinking Machines: Inkling Small",
      "provider_name": "DeepInfra",
      "accuracy": 0.6076188571428572,
      "accuracy_std_dev": 0.03274303997331158,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.046068719999999994,
      "avg_duration_per_task_ms": 191059.44,
      "avg_output_tokens_per_task": 3405.154285714286,
      "last_run_timestamp": "2026-10-08T22:20:40.622Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-next-2025-02-03",
      "model_slug": "qwen/qwen3-coder-next",
      "model_name": "Qwen: Qwen3 Coder Next",
      "provider_name": "StreamLake",
      "accuracy": 0.6073332,
      "accuracy_std_dev": 0.020912477934477324,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.0484729,
      "avg_duration_per_task_ms": 95144.364,
      "avg_output_tokens_per_task": 4713.688,
      "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": "qwen/qwen3-coder-next-2025-02-03",
      "model_slug": "qwen/qwen3-coder-next",
      "model_name": "Qwen: Qwen3 Coder Next",
      "provider_name": "auto-routing",
      "accuracy": 0.605,
      "accuracy_std_dev": 0.02598076211353318,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.047004014999999996,
      "avg_duration_per_task_ms": 99148.785,
      "avg_output_tokens_per_task": 4762.08,
      "last_run_timestamp": "2026-10-03T16:07:50.329Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-2.5-pro",
      "model_slug": "google/gemini-2.5-pro",
      "model_name": "Google: Gemini 2.5 Pro",
      "provider_name": "Google",
      "accuracy": 0.6000001666666667,
      "accuracy_std_dev": 0.03771239112642186,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.19008634000000002,
      "avg_duration_per_task_ms": 186975.22,
      "avg_output_tokens_per_task": 11452.75,
      "last_run_timestamp": "2026-10-01T19:04:55.725Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Mancer 2",
      "accuracy": 0.59966491959799,
      "accuracy_std_dev": 0.08352485507909829,
      "run_count": 8,
      "total_questions": 398,
      "avg_cost_per_task_usd": 0.028319288944723613,
      "avg_duration_per_task_ms": 1740003.0175879397,
      "avg_output_tokens_per_task": 55612.85427135678,
      "last_run_timestamp": "2026-10-31T19:40:44.417Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "moonshotai/kimi-k2.5-0127",
      "model_slug": "moonshotai/kimi-k2.5",
      "model_name": "MoonshotAI: Kimi K2.5",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.5975,
      "accuracy_std_dev": 0.036959023857510084,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.09500824249999999,
      "avg_duration_per_task_ms": 57134.1525,
      "avg_output_tokens_per_task": 3617.775,
      "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": "moonshotai/kimi-k2.7-code-20260612",
      "model_slug": "moonshotai/kimi-k2.7-code",
      "model_name": "MoonshotAI: Kimi K2.7 Code",
      "provider_name": "BaseTen",
      "accuracy": 0.593333,
      "accuracy_std_dev": 0.020000000000000018,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.04552234,
      "avg_duration_per_task_ms": 32685.39,
      "avg_output_tokens_per_task": 3125.98,
      "last_run_timestamp": "2026-09-26T08:22:32.737Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.6-20251208",
      "model_slug": "z-ai/glm-4.6v",
      "model_name": "Z.ai: GLM 4.6V",
      "provider_name": "Z.AI",
      "accuracy": 0.593333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.01896666,
      "avg_duration_per_task_ms": 306972.98,
      "avg_output_tokens_per_task": 7436.64,
      "last_run_timestamp": "2026-10-01T08:02:59.889Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-2.5-flash",
      "model_slug": "google/gemini-2.5-flash",
      "model_name": "Google: Gemini 2.5 Flash",
      "provider_name": "Google",
      "accuracy": 0.591111,
      "accuracy_std_dev": 0.006285295405415193,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.029060479999999996,
      "avg_duration_per_task_ms": 77755.87,
      "avg_output_tokens_per_task": 7763.98,
      "last_run_timestamp": "2026-10-01T19:05:46.198Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-next-2025-02-03",
      "model_slug": "qwen/qwen3-coder-next",
      "model_name": "Qwen: Qwen3 Coder Next",
      "provider_name": "Parasail",
      "accuracy": 0.58916675,
      "accuracy_std_dev": 0.017058058540979977,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.020928054999999997,
      "avg_duration_per_task_ms": 95089.825,
      "avg_output_tokens_per_task": 4516.935,
      "last_run_timestamp": "2026-10-03T16:14:27.398Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-235b-a22b-thinking-2507",
      "model_slug": "qwen/qwen3-235b-a22b-thinking-2507",
      "model_name": "Qwen: Qwen3 235B A22B Thinking 2507",
      "provider_name": "Novita",
      "accuracy": 0.5833335,
      "accuracy_std_dev": 0.0033335000000000448,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.14855325,
      "avg_duration_per_task_ms": 501107.83,
      "avg_output_tokens_per_task": 23998.36,
      "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": "ibm-granite/granite-4.2-8b-20260831",
      "model_slug": "ibm-granite/granite-4.2-8b",
      "model_name": "IBM: Granite 4.2 8B",
      "provider_name": "auto-routing",
      "accuracy": 0.583333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.015297799999999999,
      "avg_duration_per_task_ms": 833284.24,
      "avg_output_tokens_per_task": 38596.74,
      "last_run_timestamp": "2026-10-01T08:04:13.086Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-2.5-pro",
      "model_slug": "google/gemini-2.5-pro",
      "model_name": "Google: Gemini 2.5 Pro",
      "provider_name": "auto-routing",
      "accuracy": 0.58,
      "accuracy_std_dev": 0.026666999999999996,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.20638665,
      "avg_duration_per_task_ms": 179102.02,
      "avg_output_tokens_per_task": 12109.41,
      "last_run_timestamp": "2026-10-01T19:04:42.461Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-r1-0528",
      "model_slug": "deepseek/deepseek-r1-0528",
      "model_name": "DeepSeek: R1 0528",
      "provider_name": "Novita",
      "accuracy": 0.5799547837837838,
      "accuracy_std_dev": 0.055903572770937965,
      "run_count": 6,
      "total_questions": 296,
      "avg_cost_per_task_usd": 0.19823583783783785,
      "avg_duration_per_task_ms": 2181766.8243243243,
      "avg_output_tokens_per_task": 43743.00675675676,
      "last_run_timestamp": "2026-10-17T19:42:10.129Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-next-2025-02-03",
      "model_slug": "qwen/qwen3-coder-next",
      "model_name": "Qwen: Qwen3 Coder Next",
      "provider_name": "Alibaba",
      "accuracy": 0.575129585492228,
      "accuracy_std_dev": 0.04407725366331256,
      "run_count": 4,
      "total_questions": 193,
      "avg_cost_per_task_usd": 0.06909092746113989,
      "avg_duration_per_task_ms": 82524.34196891192,
      "avg_output_tokens_per_task": 4092.2538860103627,
      "last_run_timestamp": "2026-10-03T16:05:18.093Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-2.5-flash",
      "model_slug": "google/gemini-2.5-flash",
      "model_name": "Google: Gemini 2.5 Flash",
      "provider_name": "auto-routing",
      "accuracy": 0.5666665000000001,
      "accuracy_std_dev": 0.006666499999999964,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.028862480000000003,
      "avg_duration_per_task_ms": 75034.33,
      "avg_output_tokens_per_task": 7701.06,
      "last_run_timestamp": "2026-10-01T19:04:40.340Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Cerebras",
      "accuracy": 0.5648772953020134,
      "accuracy_std_dev": 0.0424873867714486,
      "run_count": 3,
      "total_questions": 149,
      "avg_cost_per_task_usd": 0.1486826308724832,
      "avg_duration_per_task_ms": 18002.88590604027,
      "avg_output_tokens_per_task": 2546.3288590604025,
      "last_run_timestamp": "2026-08-29T17:07:32.456Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "xiaomi/mimo-v2.5-20260422",
      "model_slug": "xiaomi/mimo-v2.5",
      "model_name": "Xiaomi: MiMo-V2.5",
      "provider_name": "DeepInfra",
      "accuracy": 0.545139,
      "accuracy_std_dev": 0.11788546679722656,
      "run_count": 3,
      "total_questions": 144,
      "avg_cost_per_task_usd": 0.05024318055555556,
      "avg_duration_per_task_ms": 408710.0763888889,
      "avg_output_tokens_per_task": 26837.208333333332,
      "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": "qwen/qwen3-32b-04-28",
      "model_slug": "qwen/qwen3-32b",
      "model_name": "Qwen: Qwen3 32B",
      "provider_name": "SiliconFlow",
      "accuracy": 0.543478,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 46,
      "avg_cost_per_task_usd": 0.053712021739130436,
      "avg_duration_per_task_ms": 3699508.3260869565,
      "avg_output_tokens_per_task": 39104.608695652176,
      "last_run_timestamp": "2026-08-15T14:12:30.435Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/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": 0.5390477142857143,
      "accuracy_std_dev": 0.06584534894240396,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.03078722857142857,
      "avg_duration_per_task_ms": 382475.1971428571,
      "avg_output_tokens_per_task": 16102.351428571428,
      "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": "xiaomi/mimo-v2.5-pro-20260422",
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "model_name": "Xiaomi: MiMo-V2.5-Pro",
      "provider_name": "AtlasCloud",
      "accuracy": 0.5384294010025062,
      "accuracy_std_dev": 0.2628453105716364,
      "run_count": 8,
      "total_questions": 399,
      "avg_cost_per_task_usd": 0.014483468671679198,
      "avg_duration_per_task_ms": 268751.037593985,
      "avg_output_tokens_per_task": 7227.022556390977,
      "last_run_timestamp": "2026-10-17T11:22:12.284Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/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-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Novita",
      "accuracy": 0.5362941747851002,
      "accuracy_std_dev": 0.05033110710574323,
      "run_count": 7,
      "total_questions": 349,
      "avg_cost_per_task_usd": 0.010147799426934096,
      "avg_duration_per_task_ms": 180290.85100286533,
      "avg_output_tokens_per_task": 16030.240687679083,
      "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": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "auto-routing",
      "accuracy": 0.5346869695431471,
      "accuracy_std_dev": 0.0181709902798389,
      "run_count": 8,
      "total_questions": 394,
      "avg_cost_per_task_usd": 0.01636143654822335,
      "avg_duration_per_task_ms": 1046919.6649746193,
      "avg_output_tokens_per_task": 68140.03553299492,
      "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": "minimax/minimax-m2.5-20260211",
      "model_slug": "minimax/minimax-m2.5",
      "model_name": "MiniMax: MiniMax M2.5",
      "provider_name": "Nebius",
      "accuracy": 0.5333335,
      "accuracy_std_dev": 0.026666500000000037,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.10219375,
      "avg_duration_per_task_ms": 375443.12,
      "avg_output_tokens_per_task": 17495.06,
      "last_run_timestamp": "2026-08-15T14:16:11.776Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "SiliconFlow",
      "accuracy": 0.5300001249999999,
      "accuracy_std_dev": 0.02027592305196919,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.066882445,
      "avg_duration_per_task_ms": 284395.4725,
      "avg_output_tokens_per_task": 4047.5225,
      "last_run_timestamp": "2026-10-31T19:28:47.220Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Darkbloom",
      "accuracy": 0.5285472267002518,
      "accuracy_std_dev": 0.03054949850959095,
      "run_count": 8,
      "total_questions": 397,
      "avg_cost_per_task_usd": 0.0026572493702770776,
      "avg_duration_per_task_ms": 371704.879093199,
      "avg_output_tokens_per_task": 6936.692695214106,
      "last_run_timestamp": "2026-10-31T19:13:20.193Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4.1-2025-04-14",
      "model_slug": "openai/gpt-4.1",
      "model_name": "OpenAI: GPT-4.1",
      "provider_name": "Azure",
      "accuracy": 0.522831205479452,
      "accuracy_std_dev": 0.027543964303555765,
      "run_count": 3,
      "total_questions": 146,
      "avg_cost_per_task_usd": 0.11947870547945207,
      "avg_duration_per_task_ms": 69921.19178082192,
      "avg_output_tokens_per_task": 2362.404109589041,
      "last_run_timestamp": "2026-10-01T19:05:18.003Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.5200001428571428,
      "accuracy_std_dev": 0.0256965935620634,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.028399222857142855,
      "avg_duration_per_task_ms": 421712.7342857143,
      "avg_output_tokens_per_task": 58758.36571428571,
      "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": "deepseek/deepseek-r1-0528",
      "model_slug": "deepseek/deepseek-r1-0528",
      "model_name": "DeepSeek: R1 0528",
      "provider_name": "auto-routing",
      "accuracy": 0.5016778724832214,
      "accuracy_std_dev": 0.03760315973167088,
      "run_count": 6,
      "total_questions": 298,
      "avg_cost_per_task_usd": 0.16471912416107382,
      "avg_duration_per_task_ms": 1996208.6208053692,
      "avg_output_tokens_per_task": 38529.68791946309,
      "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": 0.49888883333333334,
      "accuracy_std_dev": 0.028130992093043726,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.026234426666666665,
      "avg_duration_per_task_ms": 626626.7566666667,
      "avg_output_tokens_per_task": 52920.51,
      "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": 0.4931975918367347,
      "accuracy_std_dev": 0.018578984814838502,
      "run_count": 4,
      "total_questions": 196,
      "avg_cost_per_task_usd": 0.022155443877551022,
      "avg_duration_per_task_ms": 1297132.918367347,
      "avg_output_tokens_per_task": 110956.9693877551,
      "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": "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": 0.48047614285714285,
      "accuracy_std_dev": 0.02256761169355368,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.02984426,
      "avg_duration_per_task_ms": 354083.21428571426,
      "avg_output_tokens_per_task": 63173.58857142857,
      "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": "google/gemma-3-12b-it",
      "model_slug": "google/gemma-3-12b-it",
      "model_name": "Google: Gemma 3 12B",
      "provider_name": "DeepInfra",
      "accuracy": 0.48,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.011001320000000002,
      "avg_duration_per_task_ms": 80205.64,
      "avg_output_tokens_per_task": 3762.88,
      "last_run_timestamp": "2026-10-08T22:08:03.352Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "auto-routing",
      "accuracy": 0.48,
      "accuracy_std_dev": 0.03496018370232054,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.052433775,
      "avg_duration_per_task_ms": 89289.75,
      "avg_output_tokens_per_task": 2759.585,
      "last_run_timestamp": "2026-10-03T16:07:50.332Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_slug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_name": "Qwen: Qwen3 VL 235B A22B Instruct",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.48,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.07899766,
      "avg_duration_per_task_ms": 132704.5,
      "avg_output_tokens_per_task": 3161.38,
      "last_run_timestamp": "2026-10-04T03:00:16.401Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-r1-0528",
      "model_slug": "deepseek/deepseek-r1-0528",
      "model_name": "DeepSeek: R1 0528",
      "provider_name": "SiliconFlow",
      "accuracy": 0.47983850806451617,
      "accuracy_std_dev": 0.010328084689815433,
      "run_count": 5,
      "total_questions": 248,
      "avg_cost_per_task_usd": 0.16880521774193546,
      "avg_duration_per_task_ms": 2361254.620967742,
      "avg_output_tokens_per_task": 44884.471774193546,
      "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/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Crusoe",
      "accuracy": 0.47916700000000007,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 48,
      "avg_cost_per_task_usd": 0.040637479166666664,
      "avg_duration_per_task_ms": 59566.541666666664,
      "avg_output_tokens_per_task": 2608.1666666666665,
      "last_run_timestamp": "2026-08-15T14:08:14.681Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-v3.1-terminus",
      "model_slug": "deepseek/deepseek-v3.1-terminus",
      "model_name": "DeepSeek: DeepSeek V3.1 Terminus",
      "provider_name": "SiliconFlow",
      "accuracy": 0.4777779999999999,
      "accuracy_std_dev": 0.01571332689152746,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.06735880666666666,
      "avg_duration_per_task_ms": 286720.9,
      "avg_output_tokens_per_task": 4080.2733333333335,
      "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-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": "qwen/qwen3-32b-04-28",
      "model_slug": "qwen/qwen3-32b",
      "model_name": "Qwen: Qwen3 32B",
      "provider_name": "Nebius",
      "accuracy": 0.47555566666666665,
      "accuracy_std_dev": 0.008314853589945066,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.023880926666666667,
      "avg_duration_per_task_ms": 868904.6933333334,
      "avg_output_tokens_per_task": 23095.28,
      "last_run_timestamp": "2026-08-29T16:13:54.382Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemini-2.5-flash-lite",
      "model_slug": "google/gemini-2.5-flash-lite",
      "model_name": "Google: Gemini 2.5 Flash Lite",
      "provider_name": "Google",
      "accuracy": 0.4754382631578947,
      "accuracy_std_dev": 0.054999999999999966,
      "run_count": 2,
      "total_questions": 95,
      "avg_cost_per_task_usd": 0.01757382105263158,
      "avg_duration_per_task_ms": 167109.45263157896,
      "avg_output_tokens_per_task": 28512.663157894738,
      "last_run_timestamp": "2026-10-01T19:06:10.500Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-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": 0.475,
      "accuracy_std_dev": 0.025980762113533153,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.034838315,
      "avg_duration_per_task_ms": 40251.235,
      "avg_output_tokens_per_task": 3653.015,
      "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/qwen3-32b-04-28",
      "model_slug": "qwen/qwen3-32b",
      "model_name": "Qwen: Qwen3 32B",
      "provider_name": "auto-routing",
      "accuracy": 0.471111,
      "accuracy_std_dev": 0.008314764498569181,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.021167693333333334,
      "avg_duration_per_task_ms": 745067.6466666666,
      "avg_output_tokens_per_task": 19698.2,
      "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-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "Google",
      "accuracy": 0.47101399999999993,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 46,
      "avg_cost_per_task_usd": 0.08111417391304347,
      "avg_duration_per_task_ms": 51984.04347826087,
      "avg_output_tokens_per_task": 3590.7391304347825,
      "last_run_timestamp": "2026-08-15T14:15:58.674Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-30b-a3b-04-28",
      "model_slug": "qwen/qwen3-30b-a3b",
      "model_name": "Qwen: Qwen3 30B A3B",
      "provider_name": "DeepInfra",
      "accuracy": 0.47,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.03551402,
      "avg_duration_per_task_ms": 437816.58,
      "avg_output_tokens_per_task": 20922.04,
      "last_run_timestamp": "2026-10-08T23:09:09.727Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Parasail",
      "accuracy": 0.46820784971098267,
      "accuracy_std_dev": 0.04097451831844482,
      "run_count": 7,
      "total_questions": 346,
      "avg_cost_per_task_usd": 0.027368598265895955,
      "avg_duration_per_task_ms": 2298741.832369942,
      "avg_output_tokens_per_task": 136175.1647398844,
      "last_run_timestamp": "2026-10-24T19:05:04.624Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3.5-35b-a3b-20260224",
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "model_name": "Qwen: Qwen3.5-35B-A3B",
      "provider_name": "AkashML",
      "accuracy": 0.466667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.03567932,
      "avg_duration_per_task_ms": 645558.66,
      "avg_output_tokens_per_task": 31749.44,
      "last_run_timestamp": "2026-08-15T14:08:35.086Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.46666650000000004,
      "accuracy_std_dev": 0.02108179836375446,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.023839585000000003,
      "avg_duration_per_task_ms": 125170.4,
      "avg_output_tokens_per_task": 3749.6,
      "last_run_timestamp": "2026-10-03T16:13:56.619Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-32b-04-28",
      "model_slug": "qwen/qwen3-32b",
      "model_name": "Qwen: Qwen3 32B",
      "provider_name": "DeepInfra",
      "accuracy": 0.4655553333333334,
      "accuracy_std_dev": 0.012861993011280265,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.020687186666666666,
      "avg_duration_per_task_ms": 683546.4733333334,
      "avg_output_tokens_per_task": 19239.053333333333,
      "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": "openai/gpt-5-nano-2025-08-07",
      "model_slug": "openai/gpt-5-nano",
      "model_name": "OpenAI: GPT-5 Nano",
      "provider_name": "OpenAI",
      "accuracy": 0.4644446666666666,
      "accuracy_std_dev": 0.029979285029648195,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.04325750666666667,
      "avg_duration_per_task_ms": 738509.7133333334,
      "avg_output_tokens_per_task": 100007.46,
      "last_run_timestamp": "2026-10-02T23:22:06.591Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-next-80b-a3b-instruct-2509",
      "model_slug": "qwen/qwen3-next-80b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Next 80B A3B Instruct",
      "provider_name": "Alibaba",
      "accuracy": 0.46333349999999995,
      "accuracy_std_dev": 0.046308230610227374,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.025355005,
      "avg_duration_per_task_ms": 58190.01,
      "avg_output_tokens_per_task": 3726.025,
      "last_run_timestamp": "2026-10-03T16:10:49.378Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-5-nano-2025-08-07",
      "model_slug": "openai/gpt-5-nano",
      "model_name": "OpenAI: GPT-5 Nano",
      "provider_name": "Azure",
      "accuracy": 0.46166674999999996,
      "accuracy_std_dev": 0.04010416645047619,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.04448468,
      "avg_duration_per_task_ms": 738816.635,
      "avg_output_tokens_per_task": 95390.315,
      "last_run_timestamp": "2026-10-01T19:06:01.717Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "mistralai/mistral-small-2603",
      "model_slug": "mistralai/mistral-small-2603",
      "model_name": "Mistral: Mistral Small 4",
      "provider_name": "auto-routing",
      "accuracy": 0.46,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.011152899999999999,
      "avg_duration_per_task_ms": 87357.46,
      "avg_output_tokens_per_task": 9087.9,
      "last_run_timestamp": "2026-09-01T13:02:23.335Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4.1-2025-04-14",
      "model_slug": "openai/gpt-4.1",
      "model_name": "OpenAI: GPT-4.1",
      "provider_name": "FakeProvider",
      "accuracy": 0.46,
      "accuracy_std_dev": 0,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": null,
      "avg_duration_per_task_ms": 158124.84,
      "avg_output_tokens_per_task": 4515,
      "last_run_timestamp": "2026-10-01T19:06:51.244Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Coder 30B A3B Instruct",
      "provider_name": "Novita",
      "accuracy": 0.46,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.00632216,
      "avg_duration_per_task_ms": 38771.86,
      "avg_output_tokens_per_task": 1211.54,
      "last_run_timestamp": "2026-08-15T08:16:13.926Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-chat-v3.1",
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "model_name": "DeepSeek: DeepSeek V3.1",
      "provider_name": "auto-routing",
      "accuracy": 0.4597700689655172,
      "accuracy_std_dev": 0.17656167724922991,
      "run_count": 7,
      "total_questions": 348,
      "avg_cost_per_task_usd": 0.07236286206896553,
      "avg_duration_per_task_ms": 580100.433908046,
      "avg_output_tokens_per_task": 10024.175287356322,
      "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Together",
      "accuracy": 0.45875,
      "accuracy_std_dev": 0.14443306119704724,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.0279354525,
      "avg_duration_per_task_ms": 219700.04,
      "avg_output_tokens_per_task": 18638.5075,
      "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": "Parasail",
      "accuracy": 0.45833349999999995,
      "accuracy_std_dev": 0.036552937258857866,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.019015,
      "avg_duration_per_task_ms": 100068.645,
      "avg_output_tokens_per_task": 3610.935,
      "last_run_timestamp": "2026-10-03T16:12:20.062Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Coder 30B A3B Instruct",
      "provider_name": "DigitalOcean",
      "accuracy": 0.453333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.03731064,
      "avg_duration_per_task_ms": 268154,
      "avg_output_tokens_per_task": 3892.22,
      "last_run_timestamp": "2026-08-15T08:15:42.180Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Groq",
      "accuracy": 0.45114957241379305,
      "accuracy_std_dev": 0.023020145226518452,
      "run_count": 6,
      "total_questions": 290,
      "avg_cost_per_task_usd": 0.0516089,
      "avg_duration_per_task_ms": 110224.24482758621,
      "avg_output_tokens_per_task": 64863.19655172414,
      "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": "openai/gpt-4.1-mini-2025-04-14",
      "model_slug": "openai/gpt-4.1-mini",
      "model_name": "OpenAI: GPT-4.1 Mini",
      "provider_name": "OpenAI",
      "accuracy": 0.45111100000000004,
      "accuracy_std_dev": 0.01571332689152746,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.023389446666666668,
      "avg_duration_per_task_ms": 54689.44666666666,
      "avg_output_tokens_per_task": 2398.8933333333334,
      "last_run_timestamp": "2026-10-02T23:07:02.872Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-32b-04-28",
      "model_slug": "qwen/qwen3-32b",
      "model_name": "Qwen: Qwen3 32B",
      "provider_name": "Tenstorrent",
      "accuracy": 0.4497488994974874,
      "accuracy_std_dev": 0.02675352548730728,
      "run_count": 4,
      "total_questions": 199,
      "avg_cost_per_task_usd": 0.023178929648241205,
      "avg_duration_per_task_ms": 1975750.9296482413,
      "avg_output_tokens_per_task": 14402.733668341709,
      "last_run_timestamp": "2026-11-04T19:34:19.992Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Parasail",
      "accuracy": 0.4497487512562814,
      "accuracy_std_dev": 0.043480269304018805,
      "run_count": 8,
      "total_questions": 398,
      "avg_cost_per_task_usd": 0.015482552763819097,
      "avg_duration_per_task_ms": 129064.97989949748,
      "avg_output_tokens_per_task": 3318.743718592965,
      "last_run_timestamp": "2026-10-31T18:58:15.016Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Coder 30B A3B Instruct",
      "provider_name": "Alibaba",
      "accuracy": 0.446667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.087889,
      "avg_duration_per_task_ms": 78557.44,
      "avg_output_tokens_per_task": 4155.32,
      "last_run_timestamp": "2026-08-15T08:11:06.981Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-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": 0.4466668,
      "accuracy_std_dev": 0.030768057029328312,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.021730787999999997,
      "avg_duration_per_task_ms": 222860.848,
      "avg_output_tokens_per_task": 3530.168,
      "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": "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/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Google",
      "accuracy": 0.442176887755102,
      "accuracy_std_dev": 0.05508412293199829,
      "run_count": 4,
      "total_questions": 196,
      "avg_cost_per_task_usd": 0.049728214285714285,
      "avg_duration_per_task_ms": 74838.58163265306,
      "avg_output_tokens_per_task": 3054.173469387755,
      "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": "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": 0.4414284285714286,
      "accuracy_std_dev": 0.029160501195267273,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.039515631428571424,
      "avg_duration_per_task_ms": 280456.21142857143,
      "avg_output_tokens_per_task": 3528.7685714285712,
      "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": "qwen/qwen3-14b-04-28",
      "model_slug": "qwen/qwen3-14b",
      "model_name": "Qwen: Qwen3 14B",
      "provider_name": "auto-routing",
      "accuracy": 0.44,
      "accuracy_std_dev": 0.0066669999999999785,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.03130282,
      "avg_duration_per_task_ms": 680376.23,
      "avg_output_tokens_per_task": 16688.33,
      "last_run_timestamp": "2026-10-01T16:00:49.878Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "Google",
      "accuracy": 0.43999974999999997,
      "accuracy_std_dev": 0.047081383440671115,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.050863820000000004,
      "avg_duration_per_task_ms": 37018.795,
      "avg_output_tokens_per_task": 2487.295,
      "last_run_timestamp": "2026-10-03T16:12:29.903Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-chat-v3-0324",
      "model_slug": "deepseek/deepseek-chat-v3-0324",
      "model_name": "DeepSeek: DeepSeek V3 0324",
      "provider_name": "SiliconFlow",
      "accuracy": 0.4377776666666667,
      "accuracy_std_dev": 0.01911642045177101,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.10492657333333334,
      "avg_duration_per_task_ms": 892034.1333333333,
      "avg_output_tokens_per_task": 22149.68666666667,
      "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/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Nebius",
      "accuracy": 0.4372221666666667,
      "accuracy_std_dev": 0.029465364687016663,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.053356796666666664,
      "avg_duration_per_task_ms": 150036.94,
      "avg_output_tokens_per_task": 3684.5133333333333,
      "last_run_timestamp": "2026-10-17T19:11:29.194Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "Novita",
      "accuracy": 0.43666675,
      "accuracy_std_dev": 0.08698667522493028,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.22938595999999997,
      "avg_duration_per_task_ms": 120240.42,
      "avg_output_tokens_per_task": 3735.89,
      "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": "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": 0.43378998630136983,
      "accuracy_std_dev": 0.04334328866413141,
      "run_count": 3,
      "total_questions": 146,
      "avg_cost_per_task_usd": 0.05113669178082191,
      "avg_duration_per_task_ms": 70393.37671232877,
      "avg_output_tokens_per_task": 2340.2671232876714,
      "last_run_timestamp": "2026-08-29T16:52:55.844Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-chat-v3-0324",
      "model_slug": "deepseek/deepseek-chat-v3-0324",
      "model_name": "DeepSeek: DeepSeek V3 0324",
      "provider_name": "auto-routing",
      "accuracy": 0.43333333333333335,
      "accuracy_std_dev": 0.02372680934208295,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.07764511333333333,
      "avg_duration_per_task_ms": 794132.34,
      "avg_output_tokens_per_task": 20856.233333333334,
      "last_run_timestamp": "2026-08-29T16:11:46.632Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4o",
      "model_slug": "openai/gpt-4o",
      "model_name": "OpenAI: GPT-4o",
      "provider_name": "OpenAI",
      "accuracy": 0.4322221666666667,
      "accuracy_std_dev": 0.01651776073823432,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.21270556333333332,
      "avg_duration_per_task_ms": 46305.06,
      "avg_output_tokens_per_task": 2266.36,
      "last_run_timestamp": "2026-10-02T23:10:05.271Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "google/gemma-4-31b-it-20260402",
      "model_slug": "google/gemma-4-31b-it",
      "model_name": "Google: Gemma 4 31B",
      "provider_name": "Parasail",
      "accuracy": 0.43142857142857144,
      "accuracy_std_dev": 0.030750162879276607,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.007827288571428571,
      "avg_duration_per_task_ms": 194648.64,
      "avg_output_tokens_per_task": 4575.702857142857,
      "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": "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": "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": 0.43027193877551023,
      "accuracy_std_dev": 0.04094127400008455,
      "run_count": 4,
      "total_questions": 196,
      "avg_cost_per_task_usd": 0.06303705612244899,
      "avg_duration_per_task_ms": 258525.35714285713,
      "avg_output_tokens_per_task": 2431.734693877551,
      "last_run_timestamp": "2026-10-05T20:09:54.588Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-chat-v3",
      "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": "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": 0.42888899999999996,
      "accuracy_std_dev": 0.01571332689152746,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.05722675333333334,
      "avg_duration_per_task_ms": 196077.84666666668,
      "avg_output_tokens_per_task": 2331.213333333333,
      "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": "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": 0.4283335,
      "accuracy_std_dev": 0.013016997436044923,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.025223935000000003,
      "avg_duration_per_task_ms": 228221.25,
      "avg_output_tokens_per_task": 3524.31,
      "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-coder-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Coder 30B A3B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.42666699999999996,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.02109498,
      "avg_duration_per_task_ms": 190969.76,
      "avg_output_tokens_per_task": 3644.16,
      "last_run_timestamp": "2026-08-15T08:02:48.904Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": "StreamLake",
      "accuracy": 0.4253334,
      "accuracy_std_dev": 0.03222148627608603,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.052599823999999996,
      "avg_duration_per_task_ms": 127600.244,
      "avg_output_tokens_per_task": 3365.64,
      "last_run_timestamp": "2026-10-03T16:19:54.632Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-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": 0.42499975000000006,
      "accuracy_std_dev": 0.011902485735656231,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.03875372,
      "avg_duration_per_task_ms": 152646.77,
      "avg_output_tokens_per_task": 3373.325,
      "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": 0.4242807969543147,
      "accuracy_std_dev": 0.029334918915006395,
      "run_count": 8,
      "total_questions": 394,
      "avg_cost_per_task_usd": 0.005077210659898477,
      "avg_duration_per_task_ms": 361279.1116751269,
      "avg_output_tokens_per_task": 4832.657360406091,
      "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": "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": 0.42333344444444443,
      "accuracy_std_dev": 0.030307058217558316,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.024007242222222225,
      "avg_duration_per_task_ms": 421871.2622222222,
      "avg_output_tokens_per_task": 3556.5755555555556,
      "last_run_timestamp": "2026-10-31T19:22:37.207Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-480b-a35b-07-25",
      "model_slug": "qwen/qwen3-coder",
      "model_name": "Qwen: Qwen3 Coder 480B A35B",
      "provider_name": "Alibaba",
      "accuracy": 0.42333325,
      "accuracy_std_dev": 0.05763875528832576,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.563229095,
      "avg_duration_per_task_ms": 116619.195,
      "avg_output_tokens_per_task": 3538.205,
      "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": "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": "deepseek/deepseek-chat-v3-0324",
      "model_slug": "deepseek/deepseek-chat-v3-0324",
      "model_name": "DeepSeek: DeepSeek V3 0324",
      "provider_name": "DeepInfra",
      "accuracy": 0.42176859183673465,
      "accuracy_std_dev": 0.029027999999999998,
      "run_count": 2,
      "total_questions": 98,
      "avg_cost_per_task_usd": 0.033904112244897956,
      "avg_duration_per_task_ms": 521650.17346938775,
      "avg_output_tokens_per_task": 5726.459183673469,
      "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": "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": "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/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "StreamLake",
      "accuracy": 0.4149376141078838,
      "accuracy_std_dev": 0.02248141178485016,
      "run_count": 5,
      "total_questions": 241,
      "avg_cost_per_task_usd": 0.01079711203319502,
      "avg_duration_per_task_ms": 83125.57676348547,
      "avg_output_tokens_per_task": 3009.3568464730292,
      "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/qwen3.6-35b-a3b-20260415",
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "model_name": "Qwen: Qwen3.6 35B A3B",
      "provider_name": "Io Net",
      "accuracy": 0.41363020242914983,
      "accuracy_std_dev": 0.021563261057641545,
      "run_count": 5,
      "total_questions": 247,
      "avg_cost_per_task_usd": 0.0543252995951417,
      "avg_duration_per_task_ms": 608248.056680162,
      "avg_output_tokens_per_task": 23841.582995951416,
      "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-vl-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 VL 30B A3B Instruct",
      "provider_name": "Novita",
      "accuracy": 0.4132554619883041,
      "accuracy_std_dev": 0.026064865319476507,
      "run_count": 7,
      "total_questions": 342,
      "avg_cost_per_task_usd": 0.06276042690058481,
      "avg_duration_per_task_ms": 235574.4707602339,
      "avg_output_tokens_per_task": 5011.222222222223,
      "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-vl-235b-a22b-instruct",
      "model_slug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_name": "Qwen: Qwen3 VL 235B A22B Instruct",
      "provider_name": "Venice",
      "accuracy": 0.41,
      "accuracy_std_dev": 0.0317979384394649,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.07720068,
      "avg_duration_per_task_ms": 347548.415,
      "avg_output_tokens_per_task": 16051.36,
      "last_run_timestamp": "2026-10-03T16:12:29.905Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_slug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_name": "Qwen: Qwen3 VL 235B A22B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.40833325000000004,
      "accuracy_std_dev": 0.01907573356879102,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.075146255,
      "avg_duration_per_task_ms": 435207.485,
      "avg_output_tokens_per_task": 11776.71,
      "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/qwen3-235b-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/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "AtlasCloud",
      "accuracy": 0.406667,
      "accuracy_std_dev": 0.0565685424949238,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.055826546666666664,
      "avg_duration_per_task_ms": 125468.13333333333,
      "avg_output_tokens_per_task": 3336.5933333333332,
      "last_run_timestamp": "2026-08-29T16:12:10.467Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_slug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_name": "Qwen: Qwen3 VL 235B A22B Instruct",
      "provider_name": "Novita",
      "accuracy": 0.40499975000000005,
      "accuracy_std_dev": 0.028431130239360863,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.07697062,
      "avg_duration_per_task_ms": 115787.03,
      "avg_output_tokens_per_task": 3079.1,
      "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": "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": "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": 0.40166650000000004,
      "accuracy_std_dev": 0.00866035026716588,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.034973950000000004,
      "avg_duration_per_task_ms": 178771.545,
      "avg_output_tokens_per_task": 3509.995,
      "last_run_timestamp": "2026-10-03T16:08:50.061Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_slug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_name": "Qwen: Qwen3 VL 235B A22B Instruct",
      "provider_name": "Alibaba",
      "accuracy": 0.40000025,
      "accuracy_std_dev": 0.018257509870940786,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.06178068,
      "avg_duration_per_task_ms": 95175.335,
      "avg_output_tokens_per_task": 2906.275,
      "last_run_timestamp": "2026-10-03T16:12:33.500Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "Novita",
      "accuracy": 0.39926732967032963,
      "accuracy_std_dev": 0.017592999999999998,
      "run_count": 2,
      "total_questions": 91,
      "avg_cost_per_task_usd": 0.009130032967032966,
      "avg_duration_per_task_ms": 125085.65934065935,
      "avg_output_tokens_per_task": 2027.8351648351647,
      "last_run_timestamp": "2026-10-03T16:15:51.302Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "deepseek/deepseek-chat-v3",
      "model_slug": "deepseek/deepseek-chat",
      "model_name": "DeepSeek: DeepSeek V3",
      "provider_name": "auto-routing",
      "accuracy": 0.39833349999999995,
      "accuracy_std_dev": 0.018333500000000003,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.06903712000000001,
      "avg_duration_per_task_ms": 315593.59,
      "avg_output_tokens_per_task": 7308.28,
      "last_run_timestamp": "2026-10-01T19:04:42.463Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "SiliconFlow",
      "accuracy": 0.39666674999999996,
      "accuracy_std_dev": 0.02768865592056609,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.027719889999999997,
      "avg_duration_per_task_ms": 277229.35,
      "avg_output_tokens_per_task": 4205.25,
      "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": "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": 0.3933333333333333,
      "accuracy_std_dev": 0.009428090415820642,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.015131493333333334,
      "avg_duration_per_task_ms": 100578.96,
      "avg_output_tokens_per_task": 613.0666666666667,
      "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/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "auto-routing",
      "accuracy": 0.39166674999999995,
      "accuracy_std_dev": 0.127355609962567,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.020578967500000003,
      "avg_duration_per_task_ms": 226485.375,
      "avg_output_tokens_per_task": 3325.03,
      "last_run_timestamp": "2026-10-31T19:18:14.606Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-30b-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": 0.3866665,
      "accuracy_std_dev": 0.00816496581438035,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.025150575,
      "avg_duration_per_task_ms": 193350.84,
      "avg_output_tokens_per_task": 4372.65,
      "last_run_timestamp": "2026-10-03T16:03:16.197Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_slug": "qwen/qwen3-vl-235b-a22b-instruct",
      "model_name": "Qwen: Qwen3 VL 235B A22B Instruct",
      "provider_name": "DeepInfra",
      "accuracy": 0.38279137398373986,
      "accuracy_std_dev": 0.019122202306219867,
      "run_count": 5,
      "total_questions": 246,
      "avg_cost_per_task_usd": 0.05721597154471545,
      "avg_duration_per_task_ms": 352471.3292682927,
      "avg_output_tokens_per_task": 3819.8211382113823,
      "last_run_timestamp": "2026-10-08T23:21:09.489Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "model_name": "Qwen: Qwen3 30B A3B Instruct 2507",
      "provider_name": "CoreWeave",
      "accuracy": 0.38,
      "accuracy_std_dev": 0.005443582705045132,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.028440120000000006,
      "avg_duration_per_task_ms": 82079.82666666666,
      "avg_output_tokens_per_task": 4234.44,
      "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-coder-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-coder-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 Coder 30B A3B Instruct",
      "provider_name": "SiliconFlow",
      "accuracy": 0.38,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.020695199999999997,
      "avg_duration_per_task_ms": 192718,
      "avg_output_tokens_per_task": 3853.66,
      "last_run_timestamp": "2026-08-15T08:02:44.458Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "AkashML",
      "accuracy": 0.3738005794743429,
      "accuracy_std_dev": 0.022831389807948622,
      "run_count": 16,
      "total_questions": 799,
      "avg_cost_per_task_usd": 0.013106496871088861,
      "avg_duration_per_task_ms": 40746.92365456821,
      "avg_output_tokens_per_task": 767.6082603254067,
      "last_run_timestamp": "2026-10-31T19:10:53.562Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-235b-a22b-07-25",
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "model_name": "Qwen: Qwen3 235B A22B Instruct 2507",
      "provider_name": "Alibaba",
      "accuracy": 0.37,
      "accuracy_std_dev": 0.0481895677973978,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.05701098999999999,
      "avg_duration_per_task_ms": 93841.05,
      "avg_output_tokens_per_task": 3849.31,
      "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-v3.1",
      "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": 0.36799980000000004,
      "accuracy_std_dev": 0.019504976005112133,
      "run_count": 5,
      "total_questions": 250,
      "avg_cost_per_task_usd": 0.032627564,
      "avg_duration_per_task_ms": 131122.06,
      "avg_output_tokens_per_task": 4369.504,
      "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": "Alibaba",
      "accuracy": 0.36666650000000006,
      "accuracy_std_dev": 0.02943920288917484,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.05628207,
      "avg_duration_per_task_ms": 69776.165,
      "avg_output_tokens_per_task": 4432.06,
      "last_run_timestamp": "2026-10-03T16:05:34.166Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "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": 0.35678391959798994,
      "accuracy_std_dev": 0.031780867542438175,
      "run_count": 8,
      "total_questions": 398,
      "avg_cost_per_task_usd": 0.07306190452261307,
      "avg_duration_per_task_ms": 404962.46482412063,
      "avg_output_tokens_per_task": 8082.469849246231,
      "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": 0.3534484195402299,
      "accuracy_std_dev": 0.02330663243957542,
      "run_count": 7,
      "total_questions": 348,
      "avg_cost_per_task_usd": 0.07739280172413791,
      "avg_duration_per_task_ms": 206617.80747126436,
      "avg_output_tokens_per_task": 7421.793103448276,
      "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": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "CoreWeave",
      "accuracy": 0.35333325000000004,
      "accuracy_std_dev": 0.034480082818744796,
      "run_count": 8,
      "total_questions": 400,
      "avg_cost_per_task_usd": 0.09413336750000001,
      "avg_duration_per_task_ms": 27568.8675,
      "avg_output_tokens_per_task": 1082.295,
      "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": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "Amazon Bedrock",
      "accuracy": 0.3345927808564232,
      "accuracy_std_dev": 0.13502996743634121,
      "run_count": 16,
      "total_questions": 794,
      "avg_cost_per_task_usd": 0.03743722166246851,
      "avg_duration_per_task_ms": 517162.338790932,
      "avg_output_tokens_per_task": 129436.19017632242,
      "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-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.333333,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.051354699999999996,
      "avg_duration_per_task_ms": 358714.54,
      "avg_output_tokens_per_task": 69285.14,
      "last_run_timestamp": "2026-10-04T03:00:10.240Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "auto-routing",
      "accuracy": 0.32996641919191916,
      "accuracy_std_dev": 0.04648910770394905,
      "run_count": 8,
      "total_questions": 396,
      "avg_cost_per_task_usd": 0.05162643181818182,
      "avg_duration_per_task_ms": 79675.6186868687,
      "avg_output_tokens_per_task": 1323.4722222222222,
      "last_run_timestamp": "2026-10-31T19:07:35.559Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-vl-8b-instruct",
      "model_slug": "qwen/qwen3-vl-8b-instruct",
      "model_name": "Qwen: Qwen3 VL 8B Instruct",
      "provider_name": "PrimeIntellect",
      "accuracy": 0.326667,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.02872976,
      "avg_duration_per_task_ms": 133695.86,
      "avg_output_tokens_per_task": 3408.56,
      "last_run_timestamp": "2026-10-04T03:00:16.402Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_slug": "qwen/qwen3-vl-30b-a3b-instruct",
      "model_name": "Qwen: Qwen3 VL 30B A3B Instruct",
      "provider_name": "Alibaba",
      "accuracy": 0.32666671428571425,
      "accuracy_std_dev": 0.02666675000131537,
      "run_count": 7,
      "total_questions": 350,
      "avg_cost_per_task_usd": 0.07288134285714286,
      "avg_duration_per_task_ms": 131296.76,
      "avg_output_tokens_per_task": 7380.06,
      "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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "CoreWeave",
      "accuracy": 0.3249161313131313,
      "accuracy_std_dev": 0.3215350090739537,
      "run_count": 4,
      "total_questions": 198,
      "avg_cost_per_task_usd": 0.004530868686868687,
      "avg_duration_per_task_ms": 258173.5101010101,
      "avg_output_tokens_per_task": 8291.585858585859,
      "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": "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": 0.3150683150684931,
      "accuracy_std_dev": 0.04608944973514969,
      "run_count": 3,
      "total_questions": 146,
      "avg_cost_per_task_usd": 0.01928564383561644,
      "avg_duration_per_task_ms": 173538.2602739726,
      "avg_output_tokens_per_task": 1715.5753424657535,
      "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": "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-4o-mini",
      "model_slug": "openai/gpt-4o-mini",
      "model_name": "OpenAI: GPT-4o-mini",
      "provider_name": "auto-routing",
      "accuracy": 0.30666649999999995,
      "accuracy_std_dev": 0.006666499999999992,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.01779966,
      "avg_duration_per_task_ms": 55498.41,
      "avg_output_tokens_per_task": 3406.81,
      "last_run_timestamp": "2026-10-01T19:04:58.075Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "meta-llama/llama-3.3-70b-instruct",
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "model_name": "Meta: Llama 3.3 70B Instruct",
      "provider_name": "Together",
      "accuracy": 0.29,
      "accuracy_std_dev": 0.02821877494387971,
      "run_count": 6,
      "total_questions": 300,
      "avg_cost_per_task_usd": 0.17210771,
      "avg_duration_per_task_ms": 134359.03333333333,
      "avg_output_tokens_per_task": 1879.0433333333333,
      "last_run_timestamp": "2026-10-17T19:19:08.724Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4o-mini",
      "model_slug": "openai/gpt-4o-mini",
      "model_name": "OpenAI: GPT-4o-mini",
      "provider_name": "OpenAI",
      "accuracy": 0.2844446666666667,
      "accuracy_std_dev": 0.02997950740459592,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.025635820000000004,
      "avg_duration_per_task_ms": 67806.75333333333,
      "avg_output_tokens_per_task": 4107.166666666667,
      "last_run_timestamp": "2026-10-02T23:13:47.090Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen-2.5-72b-instruct",
      "model_slug": "qwen/qwen-2.5-72b-instruct",
      "model_name": "Qwen2.5 72B Instruct",
      "provider_name": "Novita",
      "accuracy": 0.2733335,
      "accuracy_std_dev": 0.013333499999999998,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.07945186,
      "avg_duration_per_task_ms": 632916.72,
      "avg_output_tokens_per_task": 5086.04,
      "last_run_timestamp": "2026-10-01T16:00:45.063Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen3-coder-next-2025-02-03",
      "model_slug": "qwen/qwen3-coder-next",
      "model_name": "Qwen: Qwen3 Coder Next",
      "provider_name": "Novita",
      "accuracy": 0.25166675,
      "accuracy_std_dev": 0.06998009439253064,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.014117339999999999,
      "avg_duration_per_task_ms": 22717.98,
      "avg_output_tokens_per_task": 1149.03,
      "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": "DeepInfra",
      "accuracy": 0.25,
      "accuracy_std_dev": null,
      "run_count": 1,
      "total_questions": 50,
      "avg_cost_per_task_usd": 0.0713485,
      "avg_duration_per_task_ms": 660516.14,
      "avg_output_tokens_per_task": 6662.1,
      "last_run_timestamp": "2026-09-01T12:59:14.290Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "z-ai/glm-4.6",
      "model_slug": "z-ai/glm-4.6",
      "model_name": "Z.ai: GLM 4.6",
      "provider_name": "AtlasCloud",
      "accuracy": 0.24222200000000002,
      "accuracy_std_dev": 0.10100740224689805,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.14696888000000002,
      "avg_duration_per_task_ms": 457890.76,
      "avg_output_tokens_per_task": 10647.186666666666,
      "last_run_timestamp": "2026-08-29T16:22:37.844Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/qwen-2.5-72b-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": "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": 0.23504264102564099,
      "accuracy_std_dev": 0.060676152257291324,
      "run_count": 4,
      "total_questions": 195,
      "avg_cost_per_task_usd": 0.004586492307692308,
      "avg_duration_per_task_ms": 16018.45641025641,
      "avg_output_tokens_per_task": 1693.246153846154,
      "last_run_timestamp": "2026-10-24T19:30:38.210Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "BaseTen",
      "accuracy": 0.22578841216216217,
      "accuracy_std_dev": 0.3171451865228027,
      "run_count": 6,
      "total_questions": 296,
      "avg_cost_per_task_usd": 0.007187790540540541,
      "avg_duration_per_task_ms": 27424.14864864865,
      "avg_output_tokens_per_task": 7417.02027027027,
      "last_run_timestamp": "2026-10-31T19:36:09.009Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/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": 0.19574940268456376,
      "accuracy_std_dev": 0.18279813689209806,
      "run_count": 6,
      "total_questions": 298,
      "avg_cost_per_task_usd": 0.2264336979865772,
      "avg_duration_per_task_ms": 1382638.3389261744,
      "avg_output_tokens_per_task": 8657.936241610738,
      "last_run_timestamp": "2026-10-03T08:09:23.857Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "qwen/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": "openai/gpt-oss-120b",
      "model_slug": "openai/gpt-oss-120b",
      "model_name": "OpenAI: gpt-oss-120b",
      "provider_name": "Nebius",
      "accuracy": 0.17185177777777777,
      "accuracy_std_dev": 0.22268352465714294,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.05974023333333334,
      "avg_duration_per_task_ms": 93439.47333333333,
      "avg_output_tokens_per_task": 15042.673333333334,
      "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": "deepseek/deepseek-v3.1-terminus",
      "model_slug": "deepseek/deepseek-v3.1-terminus",
      "model_name": "DeepSeek: DeepSeek V3.1 Terminus",
      "provider_name": "AtlasCloud",
      "accuracy": 0.162222,
      "accuracy_std_dev": 0.10740478872936718,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.010201566666666667,
      "avg_duration_per_task_ms": 48420.386666666665,
      "avg_output_tokens_per_task": 838.2466666666667,
      "last_run_timestamp": "2026-08-29T16:17:08.780Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-oss-20b",
      "model_slug": "openai/gpt-oss-20b",
      "model_name": "OpenAI: gpt-oss-20b",
      "provider_name": "CoreWeave",
      "accuracy": 0.15221598843930634,
      "accuracy_std_dev": 0.22421725648430124,
      "run_count": 7,
      "total_questions": 346,
      "avg_cost_per_task_usd": 0.006188393063583815,
      "avg_duration_per_task_ms": 494644.65606936417,
      "avg_output_tokens_per_task": 41928.95953757226,
      "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": "openai/gpt-4.1-nano-2025-04-14",
      "model_slug": "openai/gpt-4.1-nano",
      "model_name": "OpenAI: GPT-4.1 Nano",
      "provider_name": "Azure",
      "accuracy": 0.14,
      "accuracy_std_dev": 0.014142135623730954,
      "run_count": 4,
      "total_questions": 200,
      "avg_cost_per_task_usd": 0.0072234,
      "avg_duration_per_task_ms": 76455.195,
      "avg_output_tokens_per_task": 2322.375,
      "last_run_timestamp": "2026-10-01T19:04:58.093Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4.1-nano-2025-04-14",
      "model_slug": "openai/gpt-4.1-nano",
      "model_name": "OpenAI: GPT-4.1 Nano",
      "provider_name": "OpenAI",
      "accuracy": 0.111111,
      "accuracy_std_dev": 0.01571332689152746,
      "run_count": 3,
      "total_questions": 150,
      "avg_cost_per_task_usd": 0.007342933333333333,
      "avg_duration_per_task_ms": 57269.24,
      "avg_output_tokens_per_task": 2382.0866666666666,
      "last_run_timestamp": "2026-10-02T23:07:59.112Z"
    },
    {
      "benchmark": "tau_bench_verified_airline",
      "benchmark_page_url": "https://openrouter.ai/benchmarks/tau2-bench-airline",
      "model_permaslug": "openai/gpt-4.1-nano-2025-04-14",
      "model_slug": "openai/gpt-4.1-nano",
      "model_name": "OpenAI: GPT-4.1 Nano",
      "provider_name": "auto-routing",
      "accuracy": 0.11,
      "accuracy_std_dev": 0.009999999999999995,
      "run_count": 2,
      "total_questions": 100,
      "avg_cost_per_task_usd": 0.006762440000000001,
      "avg_duration_per_task_ms": 64306.02,
      "avg_output_tokens_per_task": 2312.51,
      "last_run_timestamp": "2026-10-01T19:04:36.194Z"
    },
    {
      "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": 0.10666677777777778,
      "accuracy_std_dev": 0.18531230895963338,
      "run_count": 9,
      "total_questions": 450,
      "avg_cost_per_task_usd": 0.0021768155555555555,
      "avg_duration_per_task_ms": 5484.648888888889,
      "avg_output_tokens_per_task": 710.5644444444445,
      "last_run_timestamp": "2026-10-17T19:48:24.252Z"
    }
  ]
}
