{
  "collectedAt": "2026-10-01",
  "revises": "data/research/scores-terminal-bench4-round1-2026-09-10.json",
  "revisionNote": "Full snapshot of the tbench.ai official board (27 rows vs round1's 18): adds Fable 5.1 xhigh/high/medium/low and Opus 5 xhigh/high/medium/low (entries updated 2026-09-17) and Grok Build · Grok 4.7 xhigh (2026-09-21).",
  "extraction": {
    "method": "curl download of https://www.tbench.ai/leaderboard on 2026-10-01; entries parsed from the SSR RSC payload.",
    "tempDir": "_build/boards-round5/",
    "latestEntryUpdatedAt": "2026-09-21T21:28:10.802412+00:00",
    "entries": 27
  },
  "boards": [
    {
      "boardId": "terminal_bench_4",
      "name": "Terminal-Bench 4.0",
      "url": "https://www.tbench.ai/leaderboard",
      "metric": "Resolution Rate %",
      "snapshotDate": "2026-09-21",
      "note": "Official board (tbench.ai), latest entry updated_at 2026-09-21T21:28:10.802412+00:00; checked 2026-10-01. v4.0 = 66 tasks, 330 trials per row. AA-harness rows live in aa_terminal_bench_4."
    }
  ],
  "scores": [
    {
      "boardId": "terminal_bench_4",
      "model": "gpt-6-astra",
      "variantLabel": "Codex - GPT-6 Astra (max)",
      "score": 58.18,
      "secondary": {
        "configurationRank": 1,
        "agentHarness": "Codex",
        "reasoningEffort": "max",
        "runDate": "2026-09-03",
        "nTasks": 66,
        "nTrials": 330,
        "successes": 192,
        "ciMinus": 2.79,
        "ciPlus": 2.79,
        "passAt2": 0.6485,
        "passAt3": 0.6788,
        "passAt4": 0.697,
        "passAt5": 0.7121,
        "totalCostUsd": 3267.18,
        "costUsdPerTrialTask": 9.9,
        "totalTokens": 1529778322,
        "outputTokens": 23988992,
        "cachedInputTokens": 1443351702,
        "uncachedInputTokens": 1505789330,
        "avgTrialDurationSec": 2796.3,
        "entryId": "5c537be4-7fc3-449b-8bfc-ceb9061c2535",
        "entryUpdatedAt": "2026-09-10T21:58:00.001022+00:00"
      },
      "source": "https://www.tbench.ai/leaderboard",
      "checkedAt": "2026-10-01",
      "note": "Official Terminal-Bench 4.0 leaderboard row; score belongs to this harness × model × effort configuration."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "claude-fable-5.1",
      "variantLabel": "Claude Code - Fable 5.1 (max)",
      "score": 57.88,
      "secondary": {
        "configurationRank": 2,
        "agentHarness": "Claude Code",
        "reasoningEffort": "max",
        "runDate": "2026-09-01",
        "nTasks": 66,
        "nTrials": 330,
        "successes": 191,
        "ciMinus": 3.76,
        "ciPlus": 3.76,
        "passAt2": 0.7,
        "passAt3": 0.747,
        "passAt4": 0.7727,
        "passAt5": 0.7879,
        "totalCostUsd": 6243.5,
        "costUsdPerTrialTask": 18.92,
        "totalTokens": 2746221560,
        "outputTokens": 63121568,
        "cachedInputTokens": 2467056906,
        "uncachedInputTokens": 2683099992,
        "avgTrialDurationSec": 3893.5,
        "entryId": "c741608e-c94e-417d-b7a8-e67111a0c887",
        "entryUpdatedAt": "2026-09-03T02:56:08.149542+00:00"
      },
      "source": "https://www.tbench.ai/leaderboard",
      "checkedAt": "2026-10-01",
      "note": "Official Terminal-Bench 4.0 leaderboard row; score belongs to this harness × model × effort configuration."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "gpt-6-astra",
      "variantLabel": "Codex - GPT-6 Astra (xhigh)",
      "score": 57.88,
      "secondary": {
        "configurationRank": 2,
        "agentHarness": "Codex",
        "reasoningEffort": "xhigh",
        "runDate": "2026-09-03",
        "nTasks": 66,
        "nTrials": 330,
        "successes": 191,
        "ciMinus": 2.72,
        "ciPlus": 2.72,
        "passAt2": 0.6424,
        "passAt3": 0.6697,
        "passAt4": 0.6848,
        "passAt5": 0.697,
        "totalCostUsd": 2350.51,
        "costUsdPerTrialTask": 7.12,
        "totalTokens": 1201907334,
        "outputTokens": 14950287,
        "cachedInputTokens": 1140730194,
        "uncachedInputTokens": 1186957047,
        "avgTrialDurationSec": 2208.5,
        "entryId": "16db8ad5-84aa-4588-b660-1ce68c0d45e2",
        "entryUpdatedAt": "2026-09-10T21:57:57.437008+00:00"
      },
      "source": "https://www.tbench.ai/leaderboard",
      "checkedAt": "2026-10-01",
      "note": "Official Terminal-Bench 4.0 leaderboard row; score belongs to this harness × model × effort configuration."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "gpt-6-astra",
      "variantLabel": "Codex - GPT-6 Astra (high)",
      "score": 57.88,
      "secondary": {
        "configurationRank": 2,
        "agentHarness": "Codex",
        "reasoningEffort": "high",
        "runDate": "2026-09-03",
        "nTasks": 66,
        "nTrials": 330,
        "successes": 191,
        "ciMinus": 2.97,
        "ciPlus": 2.97,
        "passAt2": 0.6545,
        "passAt3": 0.6864,
        "passAt4": 0.703,
        "passAt5": 0.7121,
        "totalCostUsd": 2269.42,
        "costUsdPerTrialTask": 6.88,
        "totalTokens": 1208224034,
        "outputTokens": 13703943,
        "cachedInputTokens": 1151219647,
        "uncachedInputTokens": 1194520091,
        "avgTrialDurationSec": 2113.7,
        "entryId": "3475050c-bf5e-4261-a3f6-0af5350af13f",
        "entryUpdatedAt": "2026-09-10T21:57:54.867871+00:00"
      },
      "source": "https://www.tbench.ai/leaderboard",
      "checkedAt": "2026-10-01",
      "note": "Official Terminal-Bench 4.0 leaderboard row; score belongs to this harness × model × effort configuration."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "claude-fable-5.1",
      "variantLabel": "Claude Code - Fable 5.1 (xhigh)",
      "score": 57.88,
      "secondary": {
        "configurationRank": 2,
        "agentHarness": "Claude Code",
        "reasoningEffort": "xhigh",
        "runDate": "2026-09-01",
        "nTasks": 66,
        "nTrials": 330,
        "successes": 191,
        "ciMinus": 3.36,
        "ciPlus": 3.36,
        "passAt2": 0.6758,
        "passAt3": 0.7152,
        "passAt4": 0.7394,
        "passAt5": 0.7576,
        "totalCostUsd": 4872.04,
        "costUsdPerTrialTask": 14.76,
        "totalTokens": 2343414000,
        "outputTokens": 50426931,
        "cachedInputTokens": 2140665130,
        "uncachedInputTokens": 2292987069,
        "avgTrialDurationSec": 3202.1,
        "entryId": "9e1c5e11-748f-4f2e-911a-070a396d5bf8",
        "entryUpdatedAt": "2026-09-17T23:52:40.302826+00:00"
      },
      "source": "https://www.tbench.ai/leaderboard",
      "checkedAt": "2026-10-01",
      "note": "Official Terminal-Bench 4.0 leaderboard row; score belongs to this harness × model × effort configuration."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "claude-fable-5.1",
      "variantLabel": "Claude Code - Fable 5.1 (high)",
      "score": 54.55,
      "secondary": {
        "configurationRank": 6,
        "agentHarness": "Claude Code",
        "reasoningEffort": "high",
        "runDate": "2026-09-01",
        "nTasks": 66,
        "nTrials": 330,
        "successes": 180,
        "ciMinus": 3.44,
        "ciPlus": 3.44,
        "passAt2": 0.647,
        "passAt3": 0.697,
        "passAt4": 0.7303,
        "passAt5": 0.7576,
        "totalCostUsd": 3985.38,
        "costUsdPerTrialTask": 12.08,
        "totalTokens": 2233621484,
        "outputTokens": 37429044,
        "cachedInputTokens": 2064726577,
        "uncachedInputTokens": 2196192440,
        "avgTrialDurationSec": 3017.7,
        "entryId": "4643b23c-53e8-47ae-a3bd-9abc900e9588",
        "entryUpdatedAt": "2026-09-17T23:52:40.302826+00:00"
      },
      "source": "https://www.tbench.ai/leaderboard",
      "checkedAt": "2026-10-01",
      "note": "Official Terminal-Bench 4.0 leaderboard row; score belongs to this harness × model × effort configuration."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "gpt-6-astra",
      "variantLabel": "Codex - GPT-6 Astra (medium)",
      "score": 54.24,
      "secondary": {
        "configurationRank": 7,
        "agentHarness": "Codex",
        "reasoningEffort": "medium",
        "runDate": "2026-09-03",
        "nTasks": 66,
        "nTrials": 330,
        "successes": 179,
        "ciMinus": 2.66,
        "ciPlus": 2.66,
        "passAt2": 0.603,
        "passAt3": 0.6318,
        "passAt4": 0.6515,
        "passAt5": 0.6667,
        "totalCostUsd": 1914.8,
        "costUsdPerTrialTask": 5.8,
        "totalTokens": 1051885207,
        "outputTokens": 10770605,
        "cachedInputTokens": 1003874758,
        "uncachedInputTokens": 1041114602,
        "avgTrialDurationSec": 1885.9,
        "entryId": "f3c3d5a6-6424-4acb-bfcc-615c3f79f6dd",
        "entryUpdatedAt": "2026-09-10T21:57:52.562258+00:00"
      },
      "source": "https://www.tbench.ai/leaderboard",
      "checkedAt": "2026-10-01",
      "note": "Official Terminal-Bench 4.0 leaderboard row; score belongs to this harness × model × effort configuration."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "claude-fable-5.1",
      "variantLabel": "Claude Code - Fable 5.1 (medium)",
      "score": 53.94,
      "secondary": {
        "configurationRank": 8,
        "agentHarness": "Claude Code",
        "reasoningEffort": "medium",
        "runDate": "2026-09-01",
        "nTasks": 66,
        "nTrials": 330,
        "successes": 178,
        "ciMinus": 3.39,
        "ciPlus": 3.39,
        "passAt2": 0.6379,
        "passAt3": 0.6788,
        "passAt4": 0.7,
        "passAt5": 0.7121,
        "totalCostUsd": 2832.9,
        "costUsdPerTrialTask": 8.58,
        "totalTokens": 1620460859,
        "outputTokens": 25991800,
        "cachedInputTokens": 1498585169,
        "uncachedInputTokens": 1594469059,
        "avgTrialDurationSec": 2528.3,
        "entryId": "83e6bf7e-aeb4-4116-9f30-ca8de96603fb",
        "entryUpdatedAt": "2026-09-17T23:52:40.302826+00:00"
      },
      "source": "https://www.tbench.ai/leaderboard",
      "checkedAt": "2026-10-01",
      "note": "Official Terminal-Bench 4.0 leaderboard row; score belongs to this harness × model × effort configuration."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "claude-opus-5",
      "variantLabel": "Claude Code - Opus 5 (xhigh)",
      "score": 53.94,
      "secondary": {
        "configurationRank": 8,
        "agentHarness": "Claude Code",
        "reasoningEffort": "xhigh",
        "runDate": "2026-07-24",
        "nTasks": 66,
        "nTrials": 330,
        "successes": 178,
        "ciMinus": 3.17,
        "ciPlus": 3.17,
        "passAt2": 0.6258,
        "passAt3": 0.6621,
        "passAt4": 0.6848,
        "passAt5": 0.697,
        "totalCostUsd": 6086.22,
        "costUsdPerTrialTask": 18.44,
        "totalTokens": 6897141993,
        "outputTokens": 59162574,
        "cachedInputTokens": 6631366566,
        "uncachedInputTokens": 6837979419,
        "avgTrialDurationSec": 4514.7,
        "entryId": "f7844abc-3109-4d5f-91ad-481df7b9fac2",
        "entryUpdatedAt": "2026-09-17T23:52:40.302826+00:00"
      },
      "source": "https://www.tbench.ai/leaderboard",
      "checkedAt": "2026-10-01",
      "note": "Official Terminal-Bench 4.0 leaderboard row; score belongs to this harness × model × effort configuration."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "claude-opus-5",
      "variantLabel": "Claude Code - Opus 5 (max)",
      "score": 51.82,
      "secondary": {
        "configurationRank": 10,
        "agentHarness": "Claude Code",
        "reasoningEffort": "max",
        "runDate": "2026-07-24",
        "nTasks": 66,
        "nTrials": 330,
        "successes": 171,
        "ciMinus": 3.39,
        "ciPlus": 3.39,
        "passAt2": 0.6167,
        "passAt3": 0.6576,
        "passAt4": 0.6818,
        "passAt5": 0.697,
        "totalCostUsd": 5969.11,
        "costUsdPerTrialTask": 18.09,
        "totalTokens": 6527147637,
        "outputTokens": 66012156,
        "cachedInputTokens": 6271909956,
        "uncachedInputTokens": 6461135481,
        "avgTrialDurationSec": 4792.8,
        "entryId": "d71ac3d0-da36-49cb-89d9-323136e77111",
        "entryUpdatedAt": "2026-09-03T00:09:44.281605+00:00"
      },
      "source": "https://www.tbench.ai/leaderboard",
      "checkedAt": "2026-10-01",
      "note": "Official Terminal-Bench 4.0 leaderboard row; score belongs to this harness × model × effort configuration."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "gpt-6-astra",
      "variantLabel": "Codex - GPT-6 Astra (low)",
      "score": 50.61,
      "secondary": {
        "configurationRank": 11,
        "agentHarness": "Codex",
        "reasoningEffort": "low",
        "runDate": "2026-09-03",
        "nTasks": 66,
        "nTrials": 330,
        "successes": 167,
        "ciMinus": 2.75,
        "ciPlus": 2.75,
        "passAt2": 0.5712,
        "passAt3": 0.6015,
        "passAt4": 0.6212,
        "passAt5": 0.6364,
        "totalCostUsd": 1557.3,
        "costUsdPerTrialTask": 4.72,
        "totalTokens": 889780710,
        "outputTokens": 7988337,
        "cachedInputTokens": 851115630,
        "uncachedInputTokens": 881792373,
        "avgTrialDurationSec": 1683.7,
        "entryId": "b3ad58f3-b311-4d4d-875b-3158cf0309d6",
        "entryUpdatedAt": "2026-09-10T21:57:49.894535+00:00"
      },
      "source": "https://www.tbench.ai/leaderboard",
      "checkedAt": "2026-10-01",
      "note": "Official Terminal-Bench 4.0 leaderboard row; score belongs to this harness × model × effort configuration."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "claude-opus-5",
      "variantLabel": "Claude Code - Opus 5 (high)",
      "score": 50.3,
      "secondary": {
        "configurationRank": 12,
        "agentHarness": "Claude Code",
        "reasoningEffort": "high",
        "runDate": "2026-07-24",
        "nTasks": 66,
        "nTrials": 330,
        "successes": 166,
        "ciMinus": 3.73,
        "ciPlus": 3.73,
        "passAt2": 0.6227,
        "passAt3": 0.6803,
        "passAt4": 0.7182,
        "passAt5": 0.7424,
        "totalCostUsd": 4662.27,
        "costUsdPerTrialTask": 14.13,
        "totalTokens": 5482033334,
        "outputTokens": 47599774,
        "cachedInputTokens": 5303103991,
        "uncachedInputTokens": 5434433560,
        "avgTrialDurationSec": 3822.2,
        "entryId": "0b2908ee-3035-447c-9876-e68fecb3f413",
        "entryUpdatedAt": "2026-09-17T23:52:40.302826+00:00"
      },
      "source": "https://www.tbench.ai/leaderboard",
      "checkedAt": "2026-10-01",
      "note": "Official Terminal-Bench 4.0 leaderboard row; score belongs to this harness × model × effort configuration."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "claude-opus-5",
      "variantLabel": "Claude Code - Opus 5 (medium)",
      "score": 44.85,
      "secondary": {
        "configurationRank": 13,
        "agentHarness": "Claude Code",
        "reasoningEffort": "medium",
        "runDate": "2026-07-24",
        "nTasks": 66,
        "nTrials": 330,
        "successes": 148,
        "ciMinus": 3.83,
        "ciPlus": 3.83,
        "passAt2": 0.5742,
        "passAt3": 0.6379,
        "passAt4": 0.6727,
        "passAt5": 0.697,
        "totalCostUsd": 3191.59,
        "costUsdPerTrialTask": 9.67,
        "totalTokens": 3668060107,
        "outputTokens": 33478510,
        "cachedInputTokens": 3541133339,
        "uncachedInputTokens": 3634581597,
        "avgTrialDurationSec": 3292.9,
        "entryId": "8c102a88-0789-4ce7-b503-bd39e84e157b",
        "entryUpdatedAt": "2026-09-17T23:52:40.302826+00:00"
      },
      "source": "https://www.tbench.ai/leaderboard",
      "checkedAt": "2026-10-01",
      "note": "Official Terminal-Bench 4.0 leaderboard row; score belongs to this harness × model × effort configuration."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "claude-fable-5",
      "variantLabel": "Claude Code - Fable 5 (max)",
      "score": 44.55,
      "secondary": {
        "configurationRank": 14,
        "agentHarness": "Claude Code",
        "reasoningEffort": "max",
        "runDate": "2026-06-09",
        "nTasks": 66,
        "nTrials": 330,
        "successes": 147,
        "ciMinus": 3.85,
        "ciPlus": 3.85,
        "passAt2": 0.5727,
        "passAt3": 0.6318,
        "passAt4": 0.6636,
        "passAt5": 0.6818,
        "totalCostUsd": 7265.01,
        "costUsdPerTrialTask": 22.02,
        "totalTokens": 3785220369,
        "outputTokens": 58619905,
        "cachedInputTokens": 3564065093,
        "uncachedInputTokens": 3726600464,
        "avgTrialDurationSec": 4202.8,
        "entryId": "36c077e0-4879-4444-b315-8532d66401d6",
        "entryUpdatedAt": "2026-09-03T00:09:00.046808+00:00"
      },
      "source": "https://www.tbench.ai/leaderboard",
      "checkedAt": "2026-10-01",
      "note": "Official Terminal-Bench 4.0 leaderboard row; score belongs to this harness × model × effort configuration."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "claude-fable-5.1",
      "variantLabel": "Claude Code - Fable 5.1 (low)",
      "score": 43.33,
      "secondary": {
        "configurationRank": 15,
        "agentHarness": "Claude Code",
        "reasoningEffort": "low",
        "runDate": "2026-09-01",
        "nTasks": 66,
        "nTrials": 330,
        "successes": 143,
        "ciMinus": 3.61,
        "ciPlus": 3.61,
        "passAt2": 0.5455,
        "passAt3": 0.6,
        "passAt4": 0.6303,
        "passAt5": 0.6515,
        "totalCostUsd": 2358.72,
        "costUsdPerTrialTask": 7.15,
        "totalTokens": 1334012516,
        "outputTokens": 19935011,
        "cachedInputTokens": 1225657226,
        "uncachedInputTokens": 1314077505,
        "avgTrialDurationSec": 2290.8,
        "entryId": "e18d02ab-0973-4d86-872f-a353c8a7f344",
        "entryUpdatedAt": "2026-09-17T23:52:40.302826+00:00"
      },
      "source": "https://www.tbench.ai/leaderboard",
      "checkedAt": "2026-10-01",
      "note": "Official Terminal-Bench 4.0 leaderboard row; score belongs to this harness × model × effort configuration."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "glm-5.3",
      "variantLabel": "Claude Code - GLM-5.3 (max)",
      "score": 41.82,
      "secondary": {
        "configurationRank": 16,
        "agentHarness": "Claude Code",
        "reasoningEffort": "max",
        "runDate": "2026-08-14",
        "nTasks": 66,
        "nTrials": 330,
        "successes": 138,
        "ciMinus": 3.23,
        "ciPlus": 3.23,
        "passAt2": 0.5076,
        "passAt3": 0.547,
        "passAt4": 0.5667,
        "passAt5": 0.5758,
        "totalCostUsd": 2727.63,
        "costUsdPerTrialTask": 8.27,
        "totalTokens": 8677039373,
        "outputTokens": 68660047,
        "cachedInputTokens": 8444041600,
        "uncachedInputTokens": 8608379326,
        "avgTrialDurationSec": 5831,
        "entryId": "d72e8775-f4f1-4313-99e7-35b1cb499f24",
        "entryUpdatedAt": "2026-09-03T00:08:03.780431+00:00"
      },
      "source": "https://www.tbench.ai/leaderboard",
      "checkedAt": "2026-10-01",
      "note": "Official Terminal-Bench 4.0 leaderboard row; score belongs to this harness × model × effort configuration."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "grok-4.7",
      "variantLabel": "Grok Build - Grok 4.7 (xhigh)",
      "score": 37.58,
      "secondary": {
        "configurationRank": 17,
        "agentHarness": "Grok Build",
        "reasoningEffort": "xhigh",
        "runDate": "2026-09-21",
        "nTasks": 66,
        "nTrials": 330,
        "successes": 124,
        "ciMinus": 3.54,
        "ciPlus": 3.54,
        "passAt2": 0.4833,
        "passAt3": 0.5394,
        "passAt4": 0.5758,
        "passAt5": 0.6061,
        "totalCostUsd": 3683.287146,
        "costUsdPerTrialTask": 11.16,
        "totalTokens": 5485862389,
        "outputTokens": 22977067,
        "cachedInputTokens": 5207370880,
        "uncachedInputTokens": 255514442,
        "avgTrialDurationSec": 5700.9,
        "entryId": "84b39f56-fe3c-46c3-919f-3b67b73f9b49",
        "entryUpdatedAt": "2026-09-21T21:28:10.802412+00:00"
      },
      "source": "https://www.tbench.ai/leaderboard",
      "checkedAt": "2026-10-01",
      "note": "Official Terminal-Bench 4.0 leaderboard row; score belongs to this harness × model × effort configuration."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "gpt-5.6-sol",
      "variantLabel": "Codex - GPT-5.6 Sol (max)",
      "score": 37.27,
      "secondary": {
        "configurationRank": 18,
        "agentHarness": "Codex",
        "reasoningEffort": "max",
        "runDate": "2026-06-26",
        "nTasks": 66,
        "nTrials": 330,
        "successes": 123,
        "ciMinus": 3.78,
        "ciPlus": 3.78,
        "passAt2": 0.4955,
        "passAt3": 0.55,
        "passAt4": 0.5818,
        "passAt5": 0.6061,
        "totalCostUsd": 2541.7,
        "costUsdPerTrialTask": 7.7,
        "totalTokens": 4409948969,
        "outputTokens": 23451415,
        "cachedInputTokens": 4316454466,
        "uncachedInputTokens": 4386497554,
        "avgTrialDurationSec": 2388.4,
        "entryId": "0e349bde-b264-494d-a853-fada9c696192",
        "entryUpdatedAt": "2026-09-03T00:09:13.502808+00:00"
      },
      "source": "https://www.tbench.ai/leaderboard",
      "checkedAt": "2026-10-01",
      "note": "Official Terminal-Bench 4.0 leaderboard row; score belongs to this harness × model × effort configuration."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "claude-opus-5",
      "variantLabel": "Claude Code - Opus 5 (low)",
      "score": 34.85,
      "secondary": {
        "configurationRank": 19,
        "agentHarness": "Claude Code",
        "reasoningEffort": "low",
        "runDate": "2026-07-24",
        "nTasks": 66,
        "nTrials": 330,
        "successes": 115,
        "ciMinus": 3.94,
        "ciPlus": 3.94,
        "passAt2": 0.4818,
        "passAt3": 0.547,
        "passAt4": 0.5879,
        "passAt5": 0.6212,
        "totalCostUsd": 2393.88,
        "costUsdPerTrialTask": 7.25,
        "totalTokens": 2721178403,
        "outputTokens": 23835411,
        "cachedInputTokens": 2619174445,
        "uncachedInputTokens": 2697342992,
        "avgTrialDurationSec": 2947.3,
        "entryId": "03fe4f10-b683-427b-a820-61e38145d28b",
        "entryUpdatedAt": "2026-09-17T23:52:40.302826+00:00"
      },
      "source": "https://www.tbench.ai/leaderboard",
      "checkedAt": "2026-10-01",
      "note": "Official Terminal-Bench 4.0 leaderboard row; score belongs to this harness × model × effort configuration."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "claude-opus-4.8",
      "variantLabel": "Claude Code - Opus 4.8 (max)",
      "score": 23.64,
      "secondary": {
        "configurationRank": 20,
        "agentHarness": "Claude Code",
        "reasoningEffort": "max",
        "runDate": "2026-05-28",
        "nTasks": 66,
        "nTrials": 330,
        "successes": 78,
        "ciMinus": 3.56,
        "ciPlus": 3.56,
        "passAt2": 0.3455,
        "passAt3": 0.4045,
        "passAt4": 0.4424,
        "passAt5": 0.4697,
        "totalCostUsd": 6481.26,
        "costUsdPerTrialTask": 19.64,
        "totalTokens": 6415823126,
        "outputTokens": 88813309,
        "cachedInputTokens": 6136275149,
        "uncachedInputTokens": 6327009817,
        "avgTrialDurationSec": 5163,
        "entryId": "952b4217-421f-4d14-849b-9968fcb2063b",
        "entryUpdatedAt": "2026-09-03T00:09:38.524394+00:00"
      },
      "source": "https://www.tbench.ai/leaderboard",
      "checkedAt": "2026-10-01",
      "note": "Official Terminal-Bench 4.0 leaderboard row; score belongs to this harness × model × effort configuration."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "gpt-5.6-terra",
      "variantLabel": "Codex - GPT-5.6 Terra (max)",
      "score": 21.52,
      "secondary": {
        "configurationRank": 21,
        "agentHarness": "Codex",
        "reasoningEffort": "max",
        "runDate": "2026-06-26",
        "nTasks": 66,
        "nTrials": 330,
        "successes": 71,
        "ciMinus": 3.25,
        "ciPlus": 3.25,
        "passAt2": 0.3061,
        "passAt3": 0.3636,
        "passAt4": 0.4061,
        "passAt5": 0.4394,
        "totalCostUsd": 1733.52,
        "costUsdPerTrialTask": 5.25,
        "totalTokens": 5683805798,
        "outputTokens": 33224833,
        "cachedInputTokens": 5560907776,
        "uncachedInputTokens": 5650580965,
        "avgTrialDurationSec": 2514.6,
        "entryId": "e53da412-5e92-408c-b369-c767924c7c1c",
        "entryUpdatedAt": "2026-09-03T00:09:20.912429+00:00"
      },
      "source": "https://www.tbench.ai/leaderboard",
      "checkedAt": "2026-10-01",
      "note": "Official Terminal-Bench 4.0 leaderboard row; score belongs to this harness × model × effort configuration."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "grok-4.6",
      "variantLabel": "Grok Build - Grok 4.6 (high)",
      "score": 20.3,
      "secondary": {
        "configurationRank": 22,
        "agentHarness": "Grok Build",
        "reasoningEffort": "high",
        "runDate": "2026-08-12",
        "nTasks": 66,
        "nTrials": 330,
        "successes": 67,
        "ciMinus": 3.09,
        "ciPlus": 3.09,
        "passAt2": 0.2848,
        "passAt3": 0.3303,
        "passAt4": 0.3636,
        "passAt5": 0.3939,
        "totalCostUsd": 3591.58,
        "costUsdPerTrialTask": 10.88,
        "totalTokens": 4001923428,
        "outputTokens": 32618221,
        "cachedInputTokens": 3877638528,
        "uncachedInputTokens": 3969305207,
        "avgTrialDurationSec": 2351.5,
        "entryId": "26354542-edc0-40cd-8f8d-9fe6fbe92ac3",
        "entryUpdatedAt": "2026-09-03T00:09:32.707079+00:00"
      },
      "source": "https://www.tbench.ai/leaderboard",
      "checkedAt": "2026-10-01",
      "note": "Official Terminal-Bench 4.0 leaderboard row; score belongs to this harness × model × effort configuration."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "gemini-3.8-flash",
      "variantLabel": "mini-SWE-agent - Gemini 3.8 Flash (high)",
      "score": 19.09,
      "secondary": {
        "configurationRank": 23,
        "agentHarness": "mini-SWE-agent",
        "reasoningEffort": "high",
        "runDate": "2026-09-02",
        "nTasks": 66,
        "nTrials": 330,
        "successes": 63,
        "ciMinus": 3.36,
        "ciPlus": 3.36,
        "passAt2": 0.2879,
        "passAt3": 0.353,
        "passAt4": 0.4,
        "passAt5": 0.4394,
        "totalCostUsd": 1828.77,
        "costUsdPerTrialTask": 5.54,
        "totalTokens": 17189729495,
        "outputTokens": 68056484,
        "cachedInputTokens": 16692887408,
        "uncachedInputTokens": 17121673011,
        "avgTrialDurationSec": 1995.1,
        "entryId": "06850434-507d-4dbd-b74a-09d52519ee35",
        "entryUpdatedAt": "2026-09-03T18:03:18.272136+00:00"
      },
      "source": "https://www.tbench.ai/leaderboard",
      "checkedAt": "2026-10-01",
      "note": "Official Terminal-Bench 4.0 leaderboard row; score belongs to this harness × model × effort configuration."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "gpt-5.6-luna",
      "variantLabel": "Codex - GPT-5.6 Luna (max)",
      "score": 17.27,
      "secondary": {
        "configurationRank": 24,
        "agentHarness": "Codex",
        "reasoningEffort": "max",
        "runDate": "2026-06-26",
        "nTasks": 66,
        "nTrials": 330,
        "successes": 57,
        "ciMinus": 2.85,
        "ciPlus": 2.85,
        "passAt2": 0.2424,
        "passAt3": 0.2848,
        "passAt4": 0.3121,
        "passAt5": 0.3333,
        "totalCostUsd": 346.67,
        "costUsdPerTrialTask": 1.05,
        "totalTokens": 11557463640,
        "outputTokens": 61282672,
        "cachedInputTokens": 11305212851,
        "uncachedInputTokens": 11496180968,
        "avgTrialDurationSec": 4088.3,
        "entryId": "51c6d76e-5baa-48c1-97b3-88656c620eef",
        "entryUpdatedAt": "2026-09-03T00:09:06.717559+00:00"
      },
      "source": "https://www.tbench.ai/leaderboard",
      "checkedAt": "2026-10-01",
      "note": "Official Terminal-Bench 4.0 leaderboard row; score belongs to this harness × model × effort configuration."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "grok-4.5",
      "variantLabel": "Grok Build - Grok 4.5 (high)",
      "score": 12.42,
      "secondary": {
        "configurationRank": 25,
        "agentHarness": "Grok Build",
        "reasoningEffort": "high",
        "runDate": "2026-07-16",
        "nTasks": 66,
        "nTrials": 330,
        "successes": 41,
        "ciMinus": 2.62,
        "ciPlus": 2.62,
        "passAt2": 0.1833,
        "passAt3": 0.2227,
        "passAt4": 0.2515,
        "passAt5": 0.2727,
        "totalCostUsd": 2094.11,
        "costUsdPerTrialTask": 6.35,
        "totalTokens": 3402536479,
        "outputTokens": 32389328,
        "cachedInputTokens": 3248613376,
        "uncachedInputTokens": 3370147151,
        "avgTrialDurationSec": 3162.1,
        "entryId": "0eed5a0d-96b6-491b-aeb7-cd5e4700c2d7",
        "entryUpdatedAt": "2026-09-03T00:09:26.760784+00:00"
      },
      "source": "https://www.tbench.ai/leaderboard",
      "checkedAt": "2026-10-01",
      "note": "Official Terminal-Bench 4.0 leaderboard row; score belongs to this harness × model × effort configuration."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "claude-sonnet-5",
      "variantLabel": "Claude Code - Sonnet 5 (max)",
      "score": 12.42,
      "secondary": {
        "configurationRank": 25,
        "agentHarness": "Claude Code",
        "reasoningEffort": "max",
        "runDate": "2026-06-30",
        "nTasks": 66,
        "nTrials": 330,
        "successes": 41,
        "ciMinus": 3.06,
        "ciPlus": 3.06,
        "passAt2": 0.2045,
        "passAt3": 0.2621,
        "passAt4": 0.3091,
        "passAt5": 0.3485,
        "totalCostUsd": 9603.86,
        "costUsdPerTrialTask": 29.1,
        "totalTokens": 21562180742,
        "outputTokens": 116598866,
        "cachedInputTokens": 21027162464,
        "uncachedInputTokens": 21445581876,
        "avgTrialDurationSec": 6510.3,
        "entryId": "8180b9e4-4990-43d0-b906-cd8b43adaaa2",
        "entryUpdatedAt": "2026-09-03T00:09:50.318045+00:00"
      },
      "source": "https://www.tbench.ai/leaderboard",
      "checkedAt": "2026-10-01",
      "note": "Official Terminal-Bench 4.0 leaderboard row; score belongs to this harness × model × effort configuration."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "gemini-3.7-flash",
      "variantLabel": "mini-SWE-agent - Gemini 3.7 Flash (high)",
      "score": 11.21,
      "secondary": {
        "configurationRank": 27,
        "agentHarness": "mini-SWE-agent",
        "reasoningEffort": "high",
        "runDate": "2026-08-13",
        "nTasks": 66,
        "nTrials": 330,
        "successes": 37,
        "ciMinus": 2.45,
        "ciPlus": 2.45,
        "passAt2": 0.1636,
        "passAt3": 0.2,
        "passAt4": 0.2303,
        "passAt5": 0.2576,
        "totalCostUsd": 1261.87,
        "costUsdPerTrialTask": 3.82,
        "totalTokens": 11137189802,
        "outputTokens": 52134001,
        "cachedInputTokens": 10736925351,
        "uncachedInputTokens": 11085055801,
        "avgTrialDurationSec": 1688.9,
        "entryId": "14f4da14-b86a-4dca-92d7-11178d4fe064",
        "entryUpdatedAt": "2026-09-03T02:56:10.037161+00:00"
      },
      "source": "https://www.tbench.ai/leaderboard",
      "checkedAt": "2026-10-01",
      "note": "Official Terminal-Bench 4.0 leaderboard row; score belongs to this harness × model × effort configuration."
    }
  ],
  "notFound": [
    {
      "boardId": "terminal_bench_4",
      "model": "claude-opus-5.5",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "claude-sonnet-5.5",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "composer-2.5",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "deepseek-v4-flash",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "deepseek-v4-flash-fast",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "deepseek-v4-flash-vision-exp",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "deepseek-v4-pro",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "deepseek-v4.1-flash",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "deepseek-v4.1-flash-fast",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "gemini-3.6-flash",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "glm-5",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "glm-5.1",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "glm-5.2",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "glm-5.2-fast",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "glm-5.3-flash",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "glm-5.3-flashx",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "gpt-5.5",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "gpt-6-luna",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "gpt-6-sol",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "gpt-6.1-sol",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "hy3",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "hy4-preview",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "inkling",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "inkling-small",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "kimi-k2.5",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "kimi-k2.6",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "kimi-k2.7-code",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "kimi-k2.7-code-highspeed",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "kimi-k3",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "longcat-2.0",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "mimo-v2.5",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "mimo-v2.5-pro",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "mimo-v2.6-flash",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "mimo-v2.6-pro",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "mimo-v2.6-pro-ultraspeed",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "minimax-m2.5",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "minimax-m2.7",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "minimax-m3",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "minimax-m3.1-flash-preview",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "mistral-medium-3.5",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "muse-spark-1.2",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "muse-spark-1.2-contributor",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "muse-spark-1.3",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "muse-spark-1.3-contributor",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "nemotron-3-ultra",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "qwen3.6-max-preview",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "qwen3.6-plus",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "qwen3.7-flash",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "qwen3.7-max",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "qwen3.7-plus",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "qwen3.8-27b",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "qwen3.8-flash",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "qwen3.8-max",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "qwen3.8-max-0902",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "qwen3.8-omni-flash",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "step-3.5-flash",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "step-3.7-flash",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "step-5-preview",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    },
    {
      "boardId": "terminal_bench_4",
      "model": "swe-2",
      "reason": "No official tbench.ai row on 2026-10-01 (27 rows)."
    }
  ]
}