{
  "snapshot_id": "epoch-frontiermath-tier4-v2-20260912",
  "retrieved_at": "2026-09-12T14:17:53.565263+00:00",
  "benchmark": "FrontierMath Tier 4 (v2)",
  "license": "CC-BY-4.0",
  "attribution": "Epoch AI, Capabilities & Benchmarking. Retrieved 12 September 2026.",
  "source_url": "https://epoch.ai/data/benchmark_data.zip",
  "methodology_url": "https://epoch.ai/benchmarks/about",
  "benchmark_url": "https://epoch.ai/benchmarks/frontiermath-tier-4-v2",
  "raw_csv_sha256": "0c5a60359ff520a571a00b1c9d0001893d950607b5f7085ca96add349c6e0ca9",
  "selection_rule": "All rows of frontiermath_tier_4_v2.csv, no best-of-run selection; ambiguous duplicate configuration fails validation. Exact model/effort match only.",
  "results": [
    {
      "model_version": "qwen3.8-max-0902_xhigh",
      "organization": "Alibaba",
      "accuracy": 0.34146341463414637,
      "standard_error": 0.07497768853141171,
      "run_id": "kZUdrqdGtCEJa3ip6L9YnH",
      "started_at": "2026-09-02T18:04:58.000Z",
      "release_date": "2026-09-01"
    },
    {
      "model_version": "gemini-3.8-flash_high",
      "organization": "Google DeepMind",
      "accuracy": 0.21951219512195122,
      "standard_error": 0.06544589202438408,
      "run_id": "Lgy52ZT4YpFrUqqKVu2qUs",
      "started_at": "2026-09-02T17:15:36.000Z",
      "release_date": "2026-09-02"
    },
    {
      "model_version": "claude-fable-5-1_max",
      "organization": "Anthropic",
      "accuracy": 0.8780487804878049,
      "standard_error": 0.051739520574625435,
      "run_id": "T56N3X4KLA3xz7tf5aCMGb",
      "started_at": "2026-09-01T18:38:24.000Z",
      "release_date": "2026-09-01"
    },
    {
      "model_version": "gpt-6-astra_none",
      "organization": "OpenAI",
      "accuracy": 0.8292682926829268,
      "standard_error": 0.05949419959829497,
      "run_id": "4BCZTu7HQKmhdwRR2W4bzT",
      "started_at": "2026-08-30T02:22:12.000Z",
      "release_date": "2026-09-03"
    },
    {
      "model_version": "gpt-6-astra_low",
      "organization": "OpenAI",
      "accuracy": 0.8780487804878049,
      "standard_error": 0.051739520574625435,
      "run_id": "kRPSnHnDdheajNdqmKsMjs",
      "started_at": "2026-08-30T02:22:12.000Z",
      "release_date": "2026-09-03"
    },
    {
      "model_version": "gpt-6-astra_medium",
      "organization": "OpenAI",
      "accuracy": 0.975609756097561,
      "standard_error": 0.024390243902439022,
      "run_id": "5sNRWemuVwb4adBaSwrLFf",
      "started_at": "2026-08-30T02:22:12.000Z",
      "release_date": "2026-09-03"
    },
    {
      "model_version": "gpt-6-astra_high",
      "organization": "OpenAI",
      "accuracy": 0.976,
      "standard_error": 0.024,
      "run_id": "f46NQY9jSxyKqo9qzrmk6F",
      "started_at": "2026-08-30T02:22:12.000Z",
      "release_date": "2026-09-03"
    },
    {
      "model_version": "gpt-6-astra_xhigh",
      "organization": "OpenAI",
      "accuracy": 0.976,
      "standard_error": 0.024,
      "run_id": "AfKJ2DGMhkjZmeFZxGoY3w",
      "started_at": "2026-08-30T02:22:12.000Z",
      "release_date": "2026-09-03"
    },
    {
      "model_version": "gpt-6-astra_max",
      "organization": "OpenAI",
      "accuracy": 0.976,
      "standard_error": 0.024,
      "run_id": "f7BCT27LzcAtaLDP2tcrWF",
      "started_at": "2026-08-30T02:22:12.000Z",
      "release_date": "2026-09-03"
    },
    {
      "model_version": "glm-5.3-flash_max",
      "organization": "Z.ai (Zhipu AI)",
      "accuracy": 0.17073170731707318,
      "standard_error": 0.05949419959829495,
      "run_id": "iCysWvFJevNyw2dbYosU6J",
      "started_at": "2026-08-27T13:29:52.000Z",
      "release_date": "2026-08-20"
    },
    {
      "model_version": "glm-5.3_max",
      "organization": "Z.ai (Zhipu AI)",
      "accuracy": 0.2926829268292683,
      "standard_error": 0.07194088392074452,
      "run_id": "4S3ZrJDkwh4gmGrmJQdvzR",
      "started_at": "2026-08-25T14:20:20.000Z",
      "release_date": "2026-08-14"
    },
    {
      "model_version": "deepseek-v4-pro-0813_max",
      "organization": "DeepSeek",
      "accuracy": 0.2682926829268293,
      "standard_error": 0.07005564203095158,
      "run_id": "a4uV8HMMBAcHXLVSmZfBgX",
      "started_at": "2026-08-19T13:35:48.000Z",
      "release_date": "2026-08-13"
    },
    {
      "model_version": "gemini-3.7-flash_high",
      "organization": "Google DeepMind",
      "accuracy": 0.36585365853658536,
      "standard_error": 0.07615851217559022,
      "run_id": "KptKXK5nE2Ns2WbiaSSnbc",
      "started_at": "2026-08-14T19:59:58.000Z",
      "release_date": "2026-08-13"
    },
    {
      "model_version": "inkling-small_xhigh",
      "organization": "Thinking Machines",
      "accuracy": 0.17073170731707318,
      "standard_error": 0.05949419959829495,
      "run_id": "DVD8G8DHNyzkowBzK7JvMt",
      "started_at": "2026-08-14T19:59:09.000Z",
      "release_date": "2026-07-15"
    },
    {
      "model_version": "grok-4.6_xhigh",
      "organization": "xAI",
      "accuracy": 0.3170731707317073,
      "standard_error": 0.07357611282438223,
      "run_id": "bg3JRdd33iWfvRoGSvyXXE",
      "started_at": "2026-08-14T15:46:50.000Z",
      "release_date": "2026-08-12"
    },
    {
      "model_version": "Inkling_xhigh",
      "organization": "Thinking Machines",
      "accuracy": 0.04878048780487805,
      "standard_error": 0.03405912205797303,
      "run_id": "jvJPoRYkPH9Sb5MX252oHp",
      "started_at": "2026-08-06T20:14:32.000Z",
      "release_date": "2026-07-15"
    },
    {
      "model_version": "qwen3.8-max_xhigh",
      "organization": "Alibaba",
      "accuracy": 0.4634146341463415,
      "standard_error": 0.07884502378168105,
      "run_id": "PBRdMZyz4L4CqPpXyLD2Lk",
      "started_at": "2026-08-04T19:32:51.000Z",
      "release_date": "2026-08-02"
    },
    {
      "model_version": "gpt-5.5-instant",
      "organization": "OpenAI",
      "accuracy": 0.024390243902439025,
      "standard_error": 0.024390243902439022,
      "run_id": "2UmuBfhBvYbbnTg5BnQqov",
      "started_at": "2026-08-02T03:23:36.000Z",
      "release_date": "2026-05-05"
    },
    {
      "model_version": "deepseek-v4-flash-0731_max",
      "organization": "DeepSeek",
      "accuracy": 0.24390243902439024,
      "standard_error": 0.06789956540036612,
      "run_id": "WYs83xB6NwrkMNyQJtuMcQ",
      "started_at": "2026-08-02T02:15:37.000Z",
      "release_date": "2026-07-31"
    },
    {
      "model_version": "gemini-3.5-flash-lite_high",
      "organization": "Google DeepMind",
      "accuracy": 0.0,
      "standard_error": 0.0,
      "run_id": "juNVfKBcYTwaKnk9GRA2zL",
      "started_at": "2026-08-02T01:20:17.000Z",
      "release_date": "2026-07-21"
    },
    {
      "model_version": "gemini-3.6-flash_high",
      "organization": "Google DeepMind",
      "accuracy": 0.21951219512195122,
      "standard_error": 0.06544589202438408,
      "run_id": "JCMjXU6FRiiLV3xvs9R2Xv",
      "started_at": "2026-08-02T00:36:45.000Z",
      "release_date": "2026-07-21"
    },
    {
      "model_version": "claude-opus-5_max",
      "organization": "Anthropic",
      "accuracy": 0.7317073170731707,
      "standard_error": 0.07005564203095158,
      "run_id": "ji3kMn49Jcr9t4whxcpjj4",
      "started_at": "2026-07-24T19:54:20.000Z",
      "release_date": "2026-07-24"
    },
    {
      "model_version": "kimi-k3_max",
      "organization": "Moonshot",
      "accuracy": 0.3902439024390244,
      "standard_error": 0.07712872341874095,
      "run_id": "gZERSAQwxBAevQzBsWfUie",
      "started_at": "2026-07-17T01:48:42.000Z",
      "release_date": "2026-07-16"
    },
    {
      "model_version": "grok-4.20-0309-reasoning",
      "organization": "xAI",
      "accuracy": 0.17073170731707318,
      "standard_error": 0.05949419959829495,
      "run_id": "SX5FCWMk3HF83gGsP3WFA6",
      "started_at": "2026-07-13T15:51:35.000Z",
      "release_date": "2026-02-17"
    },
    {
      "model_version": "gpt-5.6-sol_promax",
      "organization": "OpenAI",
      "accuracy": 0.8048780487804879,
      "standard_error": 0.06265967111543964,
      "run_id": "cdWwJSH3jVfrKyhUHmT54T",
      "started_at": "2026-07-09T20:34:41.000Z",
      "release_date": "2026-07-09"
    },
    {
      "model_version": "gpt-5.6-terra_max",
      "organization": "OpenAI",
      "accuracy": 0.7073170731707317,
      "standard_error": 0.07194088392074452,
      "run_id": "ZUygpkXqg3NjchXXfedAWV",
      "started_at": "2026-07-09T15:58:30.000Z",
      "release_date": "2026-07-09"
    },
    {
      "model_version": "gpt-5.6-luna_max",
      "organization": "OpenAI",
      "accuracy": 0.6097560975609756,
      "standard_error": 0.07712872341874097,
      "run_id": "NLiiJJSyA9wmg2FHENFpAq",
      "started_at": "2026-07-09T14:18:35.000Z",
      "release_date": "2026-07-09"
    },
    {
      "model_version": "gpt-5.6-sol_max",
      "organization": "OpenAI",
      "accuracy": 0.8292682926829268,
      "standard_error": 0.05949419959829497,
      "run_id": "QJ5rJMVypB4PMxPcBGftc8",
      "started_at": "2026-07-09T02:45:09.000Z",
      "release_date": "2026-07-09"
    },
    {
      "model_version": "grok-4.5_high",
      "organization": "xAI",
      "accuracy": 0.24390243902439024,
      "standard_error": 0.06789956540036612,
      "run_id": "7vThbrtbczScg2UcsAYXtF",
      "started_at": "2026-07-09T01:53:42.000Z",
      "release_date": "2026-07-08"
    },
    {
      "model_version": "claude-sonnet-5_max",
      "organization": "Anthropic",
      "accuracy": 0.2926829268292683,
      "standard_error": 0.07194088392074452,
      "run_id": "bqiVinHXFcgyFGPstJwXJN",
      "started_at": "2026-06-30T21:54:08.000Z",
      "release_date": "2026-06-30"
    },
    {
      "model_version": "glm-5.2_max",
      "organization": "Z.ai (Zhipu AI)",
      "accuracy": 0.2926829268292683,
      "standard_error": 0.07194088392074452,
      "run_id": "JebA8dYLVov6H49ZQSLXh5",
      "started_at": "2026-06-19T16:13:05.000Z",
      "release_date": "2026-06-16"
    },
    {
      "model_version": "grok-4.3_high",
      "organization": "xAI",
      "accuracy": 0.14634146341463414,
      "standard_error": 0.055885069450680974,
      "run_id": "3iRQkf5v7Wu7YuZhxa76Gg",
      "started_at": "2026-06-17T17:10:48.000Z",
      "release_date": "2026-04-17"
    },
    {
      "model_version": "deepseek-v4-pro_max",
      "organization": "DeepSeek",
      "accuracy": 0.024390243902439025,
      "standard_error": 0.024390243902439025,
      "run_id": "RAZ6VMfGtQSE8EvUZRRz4Q",
      "started_at": "2026-06-17T02:30:14.000Z",
      "release_date": "2026-04-24"
    },
    {
      "model_version": "qwen3.7-max",
      "organization": "Alibaba",
      "accuracy": 0.34146341463414637,
      "standard_error": 0.07497768853141171,
      "run_id": "FHZCKbtfvpKGXU99vZFnXN",
      "started_at": "2026-06-13T03:08:22.000Z",
      "release_date": "2026-05-19"
    },
    {
      "model_version": "gpt-5.4-pro-2026-03-05_xhigh",
      "organization": "OpenAI",
      "accuracy": 0.5853658536585366,
      "standard_error": 0.07789619230571372,
      "run_id": "W6b3C8AeNC7CAFs5DQma8N",
      "started_at": "2026-06-13T02:38:02.000Z",
      "release_date": "2026-03-05"
    },
    {
      "model_version": "gpt-5.2-pro-2025-12-11_xhigh",
      "organization": "OpenAI",
      "accuracy": 0.46,
      "standard_error": 0.07789619230571372,
      "run_id": "oALBfwNc7dcNSDJSSBJXYe",
      "started_at": "2026-06-13T02:37:55.000Z",
      "release_date": "2025-12-11"
    },
    {
      "model_version": "kimi-k2.7-code",
      "organization": "Moonshot",
      "accuracy": 0.12195121951219512,
      "standard_error": 0.051739520574625435,
      "run_id": "UTVSgirxQpTgmTHbgE9EUk",
      "started_at": "2026-06-13T01:39:44.000Z",
      "release_date": "2026-06-12"
    },
    {
      "model_version": "gpt-5.5-pro_xhigh",
      "organization": "OpenAI",
      "accuracy": 0.7804878048780488,
      "standard_error": 0.06544589202438408,
      "run_id": "SFKXCPUu73mwxMoUotiZM2",
      "started_at": "2026-06-12T21:33:12.000Z",
      "release_date": "2026-04-23"
    },
    {
      "model_version": "gpt-5-pro-2025-10-06_high",
      "organization": "OpenAI",
      "accuracy": 0.1951219512195122,
      "standard_error": 0.06265967111543964,
      "run_id": "HfDahF4ZXTN6bjfDFwcHTU",
      "started_at": "2026-06-12T20:32:51.000Z",
      "release_date": "2025-10-07"
    },
    {
      "model_version": "gpt-5-mini-2025-08-07_high",
      "organization": "OpenAI",
      "accuracy": 0.12195121951219512,
      "standard_error": 0.051739520574625435,
      "run_id": "2WxRbjq5ZWTgiXZPqYUG4R",
      "started_at": "2026-06-12T16:48:15.000Z",
      "release_date": "2025-08-07"
    },
    {
      "model_version": "gpt-5.4-nano-2026-03-17_high",
      "organization": "OpenAI",
      "accuracy": 0.12195121951219512,
      "standard_error": 0.051739520574625435,
      "run_id": "SF6FYnnKztPCDWQc4v4EJa",
      "started_at": "2026-06-12T16:48:06.000Z",
      "release_date": "2026-03-17"
    },
    {
      "model_version": "gpt-5-nano-2025-08-07_high",
      "organization": "OpenAI",
      "accuracy": 0.024390243902439025,
      "standard_error": 0.024390243902439025,
      "run_id": "P5v3FUnVVTbaqaCr43deKi",
      "started_at": "2026-06-12T16:47:59.000Z",
      "release_date": "2025-08-07"
    },
    {
      "model_version": "gpt-5.4-mini-2026-03-17_xhigh",
      "organization": "OpenAI",
      "accuracy": 0.0975609756097561,
      "standard_error": 0.04691557088212523,
      "run_id": "jtQ6rDiNxjNKkVMEPBE3Ub",
      "started_at": "2026-06-12T16:40:39.000Z",
      "release_date": "2026-03-17"
    },
    {
      "model_version": "gdm-ai-co-mathematician",
      "organization": "Google DeepMind",
      "accuracy": 0.756,
      "standard_error": 0.067,
      "run_id": "gdm-ai-co-mathematician-manual-run-v2",
      "started_at": "2026-06-12T14:01:58.023Z",
      "release_date": "2026-05-08"
    },
    {
      "model_version": "gemini-2.5-pro",
      "organization": "Google DeepMind",
      "accuracy": 0.0,
      "standard_error": 0.0,
      "run_id": "2z7kL9DpfvY44T9Wo99FYx",
      "started_at": "2026-06-11T23:56:12.000Z",
      "release_date": "2025-06-17"
    },
    {
      "model_version": "gemini-3-flash-preview",
      "organization": "Google DeepMind",
      "accuracy": 0.17073170731707318,
      "standard_error": 0.05949419959829497,
      "run_id": "ZczEhu69FQ9Cnns8NT6ew4",
      "started_at": "2026-06-11T23:44:30.000Z",
      "release_date": "2025-12-17"
    },
    {
      "model_version": "gemini-3.1-pro-preview",
      "organization": "Google DeepMind",
      "accuracy": 0.2682926829268293,
      "standard_error": 0.07005564203095158,
      "run_id": "eXVWr3RBeSsWy5DAU8ed46",
      "started_at": "2026-06-11T23:35:57.000Z",
      "release_date": "2026-02-19"
    },
    {
      "model_version": "claude-opus-4-1-20250805_32K",
      "organization": "Anthropic",
      "accuracy": 0.024390243902439025,
      "standard_error": 0.024390243902439022,
      "run_id": "9zT72xRR6wZhiwraJTNPyz",
      "started_at": "2026-06-11T21:15:07.000Z",
      "release_date": "2025-08-05"
    },
    {
      "model_version": "claude-sonnet-4-5-20250929_32K",
      "organization": "Anthropic",
      "accuracy": 0.024390243902439025,
      "standard_error": 0.024390243902439022,
      "run_id": "d3m9Xip8cof4s77z6LKkEh",
      "started_at": "2026-06-11T21:10:25.000Z",
      "release_date": "2025-09-29"
    },
    {
      "model_version": "gpt-5.4-2026-03-05_xhigh",
      "organization": "OpenAI",
      "accuracy": 0.49,
      "standard_error": 0.08,
      "run_id": "doiQDghqnLzbE8CFijLNQJ",
      "started_at": "2026-06-11T17:55:45.000Z",
      "release_date": "2026-03-05"
    },
    {
      "model_version": "gpt-5.5_xhigh",
      "organization": "OpenAI",
      "accuracy": 0.725,
      "standard_error": 0.07149950690165274,
      "run_id": "PKSMspJiJdNThB6qNvaBSM",
      "started_at": "2026-06-11T17:47:36.000Z",
      "release_date": "2026-04-23"
    },
    {
      "model_version": "gpt-5.2-2025-12-11_xhigh",
      "organization": "OpenAI",
      "accuracy": 0.317,
      "standard_error": 0.073,
      "run_id": "J3NUxsK2GXq9Whkjn7mEVm",
      "started_at": "2026-06-11T14:48:04.000Z",
      "release_date": "2025-12-11"
    },
    {
      "model_version": "o3-mini-2025-01-31_high",
      "organization": "OpenAI",
      "accuracy": 0.0,
      "standard_error": 0.0,
      "run_id": "iR3Ut7aYEtz8dTYQFzq7rG",
      "started_at": "2026-06-11T13:04:17.103Z",
      "release_date": "2025-01-31"
    },
    {
      "model_version": "o4-mini-2025-04-16_high",
      "organization": "OpenAI",
      "accuracy": 0.04878048780487805,
      "standard_error": 0.03405912205797303,
      "run_id": "TomkQucNryhFGkaUnUGeLR",
      "started_at": "2026-06-11T13:02:32.385Z",
      "release_date": "2025-04-16"
    },
    {
      "model_version": "claude-opus-4-5-20251101_32K",
      "organization": "Anthropic",
      "accuracy": 0.04878048780487805,
      "standard_error": 0.03405912205797303,
      "run_id": "azk4theXxAsizmoi2HDD49",
      "started_at": "2026-06-11T01:55:25.000Z",
      "release_date": "2025-11-24"
    },
    {
      "model_version": "gpt-5-2025-08-07_high",
      "organization": "OpenAI",
      "accuracy": 0.21951219512195122,
      "standard_error": 0.06544589202438408,
      "run_id": "mSS7C84hJyYi4VnJtgzYvW",
      "started_at": "2026-06-11T01:44:04.000Z",
      "release_date": "2025-08-07"
    },
    {
      "model_version": "claude-opus-4-6_max",
      "organization": "Anthropic",
      "accuracy": 0.2682926829268293,
      "standard_error": 0.07005564203095158,
      "run_id": "2hbbp8z6PhB7c2T33os9q5",
      "started_at": "2026-06-11T00:19:44.000Z",
      "release_date": "2026-02-05"
    },
    {
      "model_version": "claude-opus-4-7_max",
      "organization": "Anthropic",
      "accuracy": 0.3170731707317073,
      "standard_error": 0.07357611282438223,
      "run_id": "P4LXQ9PyDDziEANMB5jVRd",
      "started_at": "2026-06-10T21:41:54.000Z",
      "release_date": "2026-04-16"
    },
    {
      "model_version": "kimi-k2.6",
      "organization": "Moonshot",
      "accuracy": 0.2564102564102564,
      "standard_error": 0.07083413480167725,
      "run_id": "eVKA2eavdRAbSN6AnzyBxA",
      "started_at": "2026-06-10T19:47:36.000Z",
      "release_date": "2026-04-20"
    },
    {
      "model_version": "gemini-3.5-flash_high",
      "organization": "Google DeepMind",
      "accuracy": 0.2682926829268293,
      "standard_error": 0.07005564203095158,
      "run_id": "cXn5zzns6VFwbTgotDwZjV",
      "started_at": "2026-06-10T17:32:38.000Z",
      "release_date": "2026-05-19"
    },
    {
      "model_version": "claude-opus-4-8_max",
      "organization": "Anthropic",
      "accuracy": 0.5609756097560976,
      "standard_error": 0.07846686801046543,
      "run_id": "JiqgqBTfkGx797uowZgJbk",
      "started_at": "2026-06-10T15:25:15.000Z",
      "release_date": "2026-05-28"
    },
    {
      "model_version": "claude-fable-5_max",
      "organization": "Anthropic",
      "accuracy": 0.902,
      "standard_error": 0.046,
      "run_id": "ReDAtJGGfRtrEGGWf4Zbsd",
      "started_at": "2026-06-09T18:12:35.000Z",
      "release_date": "2026-06-09"
    }
  ]
}
