{
  "api_version": "1",
  "updated": "2026-10-26",
  "latest_event": "2026-10-24",
  "rights": "© Frontier Log, all rights reserved.",
  "model": {
    "slug": "claude-opus-5-5",
    "name": "Claude Opus 5.5",
    "lab": "anthropic",
    "date": "2026-09-22",
    "source": "https://www.anthropic.com/claude-opus-5-5",
    "source_title": "Introducing Claude Opus 5.5",
    "flagship": true,
    "event": "https://frontierlog.ai/event/2026-09-22-claude-opus-5-5",
    "releases": [],
    "variant_of": null,
    "made_by": "Anthropic",
    "url": "https://frontierlog.ai/model/claude-opus-5-5",
    "results": [
      {
        "id": "claude-opus-5-5:hle:no-tools:1",
        "model": "claude-opus-5-5",
        "model_name": "Claude Opus 5.5",
        "model_date": "2026-09-22",
        "event": "https://frontierlog.ai/event/2026-09-22-claude-opus-5-5",
        "benchmark": "hle",
        "benchmark_name": "Humanity's Last Exam",
        "setting": "no-tools",
        "setting_name": "No tools",
        "score": 64.4,
        "unit": "%",
        "printed": "64.4",
        "reported_by": "anthropic",
        "run_by": null,
        "label": "Claude Opus 5.5",
        "source": "https://www.anthropic.com/claude-opus-5-5-system-card",
        "source_title": "Claude Opus 5.5 System Card",
        "source_date": "2026-09-22",
        "where": "Section 8.1 evaluation summary table, row 'Humanity's Last Exam, No tools', column 'Claude Opus 5.5'",
        "own": true,
        "primary": true,
        "parallel": false,
        "notes": "",
        "caveat": "",
        "human_verified": true,
        "human_verified_on": "2026-10-16",
        "check": "human",
        "corroborated": true,
        "corroborated_on": "2026-10-19",
        "corroborated_by": [
          {
            "url": "https://www.anthropic.com/claude-haiku-5-5-system-card",
            "name": "Claude Haiku 5.5 card",
            "title": "Claude Haiku 5.5 System Card",
            "lab": "anthropic",
            "date": "2026-10-07",
            "printed": "64.4%",
            "where": "Figure 8.8.1.B HLE (no tools) test-time compute scaling (p.120), Claude Opus 5.5 \"Max\" point label"
          },
          {
            "url": "https://www.anthropic.com/claude-sonnet-5-5-system-card",
            "name": "Claude Sonnet 5.5 card",
            "title": "Claude Sonnet 5.5 System Card",
            "lab": "anthropic",
            "date": "2026-09-28",
            "printed": "64.4",
            "where": "Table 8.1.A (p.109), row Humanity's Last Exam / No tools, column Claude Opus 5.5"
          }
        ],
        "source_capture": "https://frontierlog.ai/captures/8a49bd8c526c6c66.png"
      },
      {
        "id": "claude-opus-5-5:terminal-bench-4-0:default:1",
        "model": "claude-opus-5-5",
        "model_name": "Claude Opus 5.5",
        "model_date": "2026-09-22",
        "event": "https://frontierlog.ai/event/2026-09-22-claude-opus-5-5",
        "benchmark": "terminal-bench-4-0",
        "benchmark_name": "Terminal-Bench 4.0",
        "setting": "default",
        "setting_name": "",
        "score": 66.4,
        "unit": "%",
        "printed": "66.4%",
        "reported_by": "anthropic",
        "run_by": null,
        "label": "Opus 5.5",
        "source": "https://www.anthropic.com/claude-opus-5-5",
        "source_title": "Introducing Claude Opus 5.5",
        "source_date": "2026-09-22",
        "where": "benchmark table, row 'Terminal-Bench 4.0', column 'Opus 5.5'",
        "own": true,
        "primary": true,
        "parallel": false,
        "notes": "xhigh effort; standard error ±2.6 points",
        "caveat": "",
        "human_verified": true,
        "human_verified_on": "2026-10-16",
        "check": "human",
        "corroborated": true,
        "corroborated_on": "2026-10-19",
        "corroborated_by": [
          {
            "url": "https://www.anthropic.com/claude-haiku-5-5-system-card",
            "name": "Claude Haiku 5.5 card",
            "title": "Claude Haiku 5.5 System Card",
            "lab": "anthropic",
            "date": "2026-10-07",
            "printed": "66.4%",
            "where": "Section 8.4 (p.115-116) sentence: \"Claude Opus 5.5 scored 66.4%\" (xhigh effort)"
          },
          {
            "url": "https://www.anthropic.com/claude-opus-5-5-system-card",
            "name": "Claude Opus 5.5 card",
            "title": "Claude Opus 5.5 System Card",
            "lab": "anthropic",
            "date": "2026-09-22",
            "printed": "66.4",
            "where": "Table 8.1.A (p.174), row Terminal-Bench 4.0, column Claude Opus 5.5 (caption: reported at xhigh effort)"
          },
          {
            "url": "https://www.anthropic.com/claude-sonnet-5-5-system-card",
            "name": "Claude Sonnet 5.5 card",
            "title": "Claude Sonnet 5.5 System Card",
            "lab": "anthropic",
            "date": "2026-09-28",
            "printed": "66.4%",
            "where": "Section 8.5 (p.113) sentence: \"Claude Opus 5.5 scored 66.4%\" (xhigh effort)"
          },
          {
            "url": "https://www.anthropic.com/claude-sonnet-5-5",
            "name": "Claude Sonnet 5.5",
            "title": "Introducing Claude Sonnet 5.5",
            "lab": "anthropic",
            "date": "2026-09-28",
            "printed": "66.4%",
            "where": "HTML benchmark table under 'Performance', row Terminal-Bench 4.0, column Opus 5.5 (footnote 1: Xhigh effort)"
          },
          {
            "url": "https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-4-argon/",
            "name": "Gemini 4 Argon",
            "title": "Gemini 4 Argon: our next era of frontier intelligence",
            "lab": "google",
            "date": "2026-09-30",
            "printed": "66.4%",
            "where": "Benchmark table image; row 'Terminal-bench 4.0'; column claude-opus-5-5"
          }
        ],
        "source_capture": "https://frontierlog.ai/captures/ca3ad55d7ec61f3c.png"
      },
      {
        "id": "claude-opus-5-5:terminal-bench-science-0-1:default:1",
        "model": "claude-opus-5-5",
        "model_name": "Claude Opus 5.5",
        "model_date": "2026-09-22",
        "event": "https://frontierlog.ai/event/2026-09-22-claude-opus-5-5",
        "benchmark": "terminal-bench-science-0-1",
        "benchmark_name": "Terminal-Bench Science 0.1",
        "setting": "default",
        "setting_name": "",
        "score": 58.7,
        "unit": "%",
        "printed": "58.7%",
        "reported_by": "anthropic",
        "run_by": null,
        "label": "Opus 5.5",
        "source": "https://www.anthropic.com/claude-opus-5-5",
        "source_title": "Introducing Claude Opus 5.5",
        "source_date": "2026-09-22",
        "where": "benchmark table, row 'Terminal-Bench-Science 0.1', column 'Opus 5.5'",
        "own": true,
        "primary": true,
        "parallel": false,
        "notes": "max effort; standard error ±3.5 to 5 points",
        "caveat": "",
        "human_verified": true,
        "human_verified_on": "2026-10-16",
        "check": "human",
        "corroborated": true,
        "corroborated_on": "2026-10-19",
        "corroborated_by": [
          {
            "url": "https://www.anthropic.com/claude-haiku-5-5-system-card",
            "name": "Claude Haiku 5.5 card",
            "title": "Claude Haiku 5.5 System Card",
            "lab": "anthropic",
            "date": "2026-10-07",
            "printed": "58.7%",
            "where": "Section 8.5 (p.116) sentence: \"Claude Opus 5.5 scored 58.7%\""
          },
          {
            "url": "https://www.anthropic.com/claude-opus-5-5-system-card",
            "name": "Claude Opus 5.5 card",
            "title": "Claude Opus 5.5 System Card",
            "lab": "anthropic",
            "date": "2026-09-22",
            "printed": "58.7",
            "where": "Table 8.1.A (p.174), row Terminal-Bench-Science 0.1, column Claude Opus 5.5"
          },
          {
            "url": "https://www.anthropic.com/claude-sonnet-5-5-system-card",
            "name": "Claude Sonnet 5.5 card",
            "title": "Claude Sonnet 5.5 System Card",
            "lab": "anthropic",
            "date": "2026-09-28",
            "printed": "58.7%",
            "where": "Section 8.6 (p.114) sentence: \"Claude Opus 5.5 scored 58.7%\""
          }
        ],
        "source_capture": "https://frontierlog.ai/captures/ca3ad55d7ec61f3c.png"
      },
      {
        "id": "claude-opus-5-5:automationbench:default:1",
        "model": "claude-opus-5-5",
        "model_name": "Claude Opus 5.5",
        "model_date": "2026-09-22",
        "event": "https://frontierlog.ai/event/2026-09-22-claude-opus-5-5",
        "benchmark": "automationbench",
        "benchmark_name": "AutomationBench",
        "setting": "default",
        "setting_name": "",
        "score": 40.0,
        "unit": "%",
        "printed": "40.0%",
        "reported_by": "anthropic",
        "run_by": null,
        "label": "Opus 5.5",
        "source": "https://www.anthropic.com/claude-opus-5-5",
        "source_title": "Introducing Claude Opus 5.5",
        "source_date": "2026-09-22",
        "where": "benchmark table, row 'AutomationBench', column 'Opus 5.5'",
        "own": true,
        "primary": true,
        "parallel": false,
        "notes": "run by Zapier, without fallback models",
        "caveat": "",
        "human_verified": true,
        "human_verified_on": "2026-10-16",
        "check": "human",
        "corroborated": true,
        "corroborated_on": "2026-10-19",
        "corroborated_by": [
          {
            "url": "https://www.anthropic.com/claude-opus-5-5-system-card",
            "name": "Claude Opus 5.5 card",
            "title": "Claude Opus 5.5 System Card",
            "lab": "anthropic",
            "date": "2026-09-22",
            "printed": "40.0",
            "where": "Table 8.1.A (p.174), row AutomationBench, column Claude Opus 5.5"
          },
          {
            "url": "https://www.anthropic.com/claude-sonnet-5-5-system-card",
            "name": "Claude Sonnet 5.5 card",
            "title": "Claude Sonnet 5.5 System Card",
            "lab": "anthropic",
            "date": "2026-09-28",
            "printed": "40.0%",
            "where": "Section 8.14.6 (p.137) sentence: \"(40.0% without the re-run, as reported in the Claude Opus 5.5 System Card)\""
          }
        ],
        "source_capture": "https://frontierlog.ai/captures/ca3ad55d7ec61f3c.png"
      },
      {
        "id": "claude-opus-5-5:hle:with-tools:1",
        "model": "claude-opus-5-5",
        "model_name": "Claude Opus 5.5",
        "model_date": "2026-09-22",
        "event": "https://frontierlog.ai/event/2026-09-22-claude-opus-5-5",
        "benchmark": "hle",
        "benchmark_name": "Humanity's Last Exam",
        "setting": "with-tools",
        "setting_name": "With tools",
        "score": 67.7,
        "unit": "%",
        "printed": "67.7%",
        "reported_by": "anthropic",
        "run_by": null,
        "label": "Opus 5.5",
        "source": "https://www.anthropic.com/claude-opus-5-5",
        "source_title": "Introducing Claude Opus 5.5",
        "source_date": "2026-09-22",
        "where": "benchmark table, row 'Humanity's Last Exam', 'with tools', column 'Opus 5.5'",
        "own": true,
        "primary": true,
        "parallel": false,
        "notes": "max effort",
        "caveat": "",
        "human_verified": true,
        "human_verified_on": "2026-10-16",
        "check": "human",
        "corroborated": true,
        "corroborated_on": "2026-10-19",
        "corroborated_by": [
          {
            "url": "https://www.anthropic.com/claude-haiku-5-5-system-card",
            "name": "Claude Haiku 5.5 card",
            "title": "Claude Haiku 5.5 System Card",
            "lab": "anthropic",
            "date": "2026-10-07",
            "printed": "67.7%",
            "where": "Figure 8.8.1.A HLE (with tools) reasoning effort scaling (p.119), Claude Opus 5.5 \"Max\" point label"
          },
          {
            "url": "https://www.anthropic.com/claude-opus-5-5-system-card",
            "name": "Claude Opus 5.5 card",
            "title": "Claude Opus 5.5 System Card",
            "lab": "anthropic",
            "date": "2026-09-22",
            "printed": "67.7",
            "where": "Table 8.1.A (p.174), row Humanity's Last Exam / With tools, column Claude Opus 5.5"
          },
          {
            "url": "https://www.anthropic.com/claude-sonnet-5-5-system-card",
            "name": "Claude Sonnet 5.5 card",
            "title": "Claude Sonnet 5.5 System Card",
            "lab": "anthropic",
            "date": "2026-09-28",
            "printed": "67.7",
            "where": "Table 8.1.A (p.109), row Humanity's Last Exam / With tools, column Claude Opus 5.5"
          },
          {
            "url": "https://www.anthropic.com/claude-sonnet-5-5",
            "name": "Claude Sonnet 5.5",
            "title": "Introducing Claude Sonnet 5.5",
            "lab": "anthropic",
            "date": "2026-09-28",
            "printed": "67.7%",
            "where": "HTML benchmark table under 'Performance', row Humanity's Last Exam (with tools), column Opus 5.5"
          }
        ],
        "source_capture": "https://frontierlog.ai/captures/ca3ad55d7ec61f3c.png"
      },
      {
        "id": "claude-opus-5-5:frontiercode-1-1:main:1",
        "model": "claude-opus-5-5",
        "model_name": "Claude Opus 5.5",
        "model_date": "2026-09-22",
        "event": "https://frontierlog.ai/event/2026-09-22-claude-opus-5-5",
        "benchmark": "frontiercode-1-1",
        "benchmark_name": "FrontierCode 1.1",
        "setting": "main",
        "setting_name": "Main set",
        "score": 54.4,
        "unit": "%",
        "printed": "54.4%",
        "reported_by": "anthropic",
        "run_by": null,
        "label": "Opus 5.5",
        "source": "https://www.anthropic.com/claude-opus-5-5",
        "source_title": "Introducing Claude Opus 5.5",
        "source_date": "2026-09-22",
        "where": "benchmark table, row 'FrontierCode v1.1 (Main)', column 'Opus 5.5'",
        "own": true,
        "primary": false,
        "parallel": false,
        "notes": "max effort",
        "caveat": "",
        "human_verified": true,
        "human_verified_on": "2026-10-16",
        "check": "human",
        "corroborated": true,
        "corroborated_on": "2026-10-19",
        "corroborated_by": [
          {
            "url": "https://www.anthropic.com/claude-opus-5-5-system-card",
            "name": "Claude Opus 5.5 card",
            "title": "Claude Opus 5.5 System Card",
            "lab": "anthropic",
            "date": "2026-09-22",
            "printed": "54.4",
            "where": "Table 8.1.A (p.174), row FrontierCode v1.1 (Main), column Claude Opus 5.5"
          },
          {
            "url": "https://www.anthropic.com/claude-sonnet-5-5-system-card",
            "name": "Claude Sonnet 5.5 card",
            "title": "Claude Sonnet 5.5 System Card",
            "lab": "anthropic",
            "date": "2026-09-28",
            "printed": "54.4",
            "where": "Table 8.1.A (p.109), row FrontierCode v1.1 (Main), column Claude Opus 5.5"
          },
          {
            "url": "https://www.anthropic.com/claude-sonnet-5-5",
            "name": "Claude Sonnet 5.5",
            "title": "Introducing Claude Sonnet 5.5",
            "lab": "anthropic",
            "date": "2026-09-28",
            "printed": "54.4%",
            "where": "HTML benchmark table under 'Performance', row FrontierCode 1.1 (Main), column Opus 5.5"
          }
        ],
        "source_capture": "https://frontierlog.ai/captures/ca3ad55d7ec61f3c.png"
      },
      {
        "id": "claude-opus-5-5:swe-bench-pro:default:1",
        "model": "claude-opus-5-5",
        "model_name": "Claude Opus 5.5",
        "model_date": "2026-09-22",
        "event": "https://frontierlog.ai/event/2026-09-22-claude-opus-5-5",
        "benchmark": "swe-bench-pro",
        "benchmark_name": "SWE-bench Pro",
        "setting": "default",
        "setting_name": "",
        "score": 89.9,
        "unit": "%",
        "printed": "89.9",
        "reported_by": "anthropic",
        "run_by": null,
        "label": "Claude Opus 5.5",
        "source": "https://www.anthropic.com/claude-opus-5-5-system-card",
        "source_title": "Claude Opus 5.5 System Card",
        "source_date": "2026-09-22",
        "where": "Section 8.1 Evaluation summary, Table 8.1.A, row 'SWE-bench Pro', column 'Claude Opus 5.5'",
        "own": true,
        "primary": true,
        "parallel": false,
        "notes": "adaptive thinking, max effort; average of 5 trials",
        "caveat": "The source doesn't say whether this is the public or the commercial set",
        "human_verified": true,
        "human_verified_on": "2026-10-17",
        "check": "human",
        "corroborated": true,
        "corroborated_on": "2026-10-19",
        "corroborated_by": [
          {
            "url": "https://www.anthropic.com/claude-sonnet-5-5-system-card",
            "name": "Claude Sonnet 5.5 card",
            "title": "Claude Sonnet 5.5 System Card",
            "lab": "anthropic",
            "date": "2026-09-28",
            "printed": "89.9",
            "where": "Table 8.1.A (p.109), row SWE-Bench Pro, column Claude Opus 5.5"
          }
        ],
        "source_capture": "https://frontierlog.ai/captures/8a49bd8c526c6c66.png"
      },
      {
        "id": "claude-opus-5-5:deepswe-1-1:default:1",
        "model": "claude-opus-5-5",
        "model_name": "Claude Opus 5.5",
        "model_date": "2026-09-22",
        "event": "https://frontierlog.ai/event/2026-09-22-claude-opus-5-5",
        "benchmark": "deepswe-1-1",
        "benchmark_name": "DeepSWE v1.1",
        "setting": "default",
        "setting_name": "",
        "score": 74.2,
        "unit": "%",
        "printed": "74.2%",
        "reported_by": "anthropic",
        "run_by": null,
        "label": "Claude Opus 5.5",
        "source": "https://www.anthropic.com/claude-opus-5-5-system-card",
        "source_title": "Claude Opus 5.5 System Card",
        "source_date": "2026-09-22",
        "where": "Section 8.3 'DeepSWE v1.1', paragraph",
        "own": true,
        "primary": true,
        "parallel": false,
        "notes": "average of 5 trials",
        "caveat": "",
        "human_verified": true,
        "human_verified_on": "2026-10-17",
        "check": "human",
        "corroborated": true,
        "corroborated_on": "2026-10-19",
        "corroborated_by": [
          {
            "url": "https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-4-argon/",
            "name": "Gemini 4 Argon",
            "title": "Gemini 4 Argon: our next era of frontier intelligence",
            "lab": "google",
            "date": "2026-09-30",
            "printed": "74.2%",
            "where": "Benchmark table image; row 'DeepSWE v1.1'; column claude-opus-5-5"
          }
        ]
      },
      {
        "id": "claude-opus-5-5:automationbench:default:2",
        "model": "claude-opus-5-5",
        "model_name": "Claude Opus 5.5",
        "model_date": "2026-09-22",
        "event": "https://frontierlog.ai/event/2026-09-22-claude-opus-5-5",
        "benchmark": "automationbench",
        "benchmark_name": "AutomationBench",
        "setting": "default",
        "setting_name": "",
        "score": 42.5,
        "unit": "%",
        "printed": "42.5",
        "reported_by": "anthropic",
        "run_by": null,
        "label": "Claude Opus 5.5",
        "source": "https://www.anthropic.com/claude-sonnet-5-5-system-card",
        "source_title": "Claude Sonnet 5.5 System Card",
        "source_date": "2026-09-28",
        "where": "Section 8.1, Table 8.1.A, row 'AutomationBench', column 'Claude Opus 5.5'",
        "own": false,
        "primary": false,
        "parallel": false,
        "notes": "run by Zapier; refused tasks re-run with default fallbacks; private set, release 1.0.6",
        "caveat": "",
        "human_verified": true,
        "human_verified_on": "2026-10-20",
        "check": "human",
        "corroborated": false,
        "corroborated_on": null,
        "corroborated_by": [],
        "source_capture": "https://frontierlog.ai/captures/53b68efd4b35b2e3.png"
      },
      {
        "id": "claude-opus-5-5:frontiercode-1-1:extended:1",
        "model": "claude-opus-5-5",
        "model_name": "Claude Opus 5.5",
        "model_date": "2026-09-22",
        "event": "https://frontierlog.ai/event/2026-09-22-claude-opus-5-5",
        "benchmark": "frontiercode-1-1",
        "benchmark_name": "FrontierCode 1.1",
        "setting": "extended",
        "setting_name": "Extended set",
        "score": 63.6,
        "unit": "%",
        "printed": "63.6%",
        "reported_by": "anthropic",
        "run_by": null,
        "label": "Claude Opus 5.5",
        "source": "https://www.anthropic.com/claude-opus-5-5-system-card",
        "source_title": "Claude Opus 5.5 System Card",
        "source_date": "2026-09-22",
        "where": "Section 8.4 'FrontierCode', paragraph ranking Opus 5.5, last sentence (max-effort scores), Extended figure",
        "own": true,
        "primary": false,
        "parallel": false,
        "notes": "max effort; run by Cognition in Claude Code",
        "caveat": "",
        "human_verified": true,
        "human_verified_on": "2026-10-20",
        "check": "human",
        "corroborated": false,
        "corroborated_on": null,
        "corroborated_by": []
      },
      {
        "id": "claude-opus-5-5:frontiercode-1-1:extended:2",
        "model": "claude-opus-5-5",
        "model_name": "Claude Opus 5.5",
        "model_date": "2026-09-22",
        "event": "https://frontierlog.ai/event/2026-09-22-claude-opus-5-5",
        "benchmark": "frontiercode-1-1",
        "benchmark_name": "FrontierCode 1.1",
        "setting": "extended",
        "setting_name": "Extended set",
        "score": 65.3,
        "unit": "%",
        "printed": "65.3%",
        "reported_by": "anthropic",
        "run_by": null,
        "label": "Claude Opus 5.5",
        "source": "https://www.anthropic.com/claude-opus-5-5-system-card",
        "source_title": "Claude Opus 5.5 System Card",
        "source_date": "2026-09-22",
        "where": "Section 8.4 'FrontierCode', paragraph ranking Opus 5.5, sentence on FrontierCode (Extended)",
        "own": true,
        "primary": true,
        "parallel": false,
        "notes": "medium effort; run by Cognition in Claude Code",
        "caveat": "",
        "human_verified": true,
        "human_verified_on": "2026-10-20",
        "check": "human",
        "corroborated": false,
        "corroborated_on": null,
        "corroborated_by": []
      },
      {
        "id": "claude-opus-5-5:frontiercode-1-1:main:2",
        "model": "claude-opus-5-5",
        "model_name": "Claude Opus 5.5",
        "model_date": "2026-09-22",
        "event": "https://frontierlog.ai/event/2026-09-22-claude-opus-5-5",
        "benchmark": "frontiercode-1-1",
        "benchmark_name": "FrontierCode 1.1",
        "setting": "main",
        "setting_name": "Main set",
        "score": 54.6,
        "unit": "%",
        "printed": "54.6%",
        "reported_by": "anthropic",
        "run_by": null,
        "label": "Claude Opus 5.5",
        "source": "https://www.anthropic.com/claude-opus-5-5-system-card",
        "source_title": "Claude Opus 5.5 System Card",
        "source_date": "2026-09-22",
        "where": "Section 8.4 'FrontierCode', paragraph ranking Opus 5.5, sentence on Opus 5.5's highest FrontierCode scores, the Main figure",
        "own": true,
        "primary": true,
        "parallel": false,
        "notes": "medium effort; run by Cognition in Claude Code",
        "caveat": "",
        "human_verified": true,
        "human_verified_on": "2026-10-20",
        "check": "human",
        "corroborated": false,
        "corroborated_on": null,
        "corroborated_by": []
      },
      {
        "id": "claude-opus-5-5:osworld-2-1:partial:1",
        "model": "claude-opus-5-5",
        "model_name": "Claude Opus 5.5",
        "model_date": "2026-09-22",
        "event": "https://frontierlog.ai/event/2026-09-22-claude-opus-5-5",
        "benchmark": "osworld-2-1",
        "benchmark_name": "OSWorld 2.1",
        "setting": "partial",
        "setting_name": "Full set, partial score",
        "score": 81.8,
        "unit": "%",
        "printed": "81.8%",
        "reported_by": "anthropic",
        "run_by": null,
        "label": "Opus 5.5",
        "source": "https://www.anthropic.com/claude-opus-5-5",
        "source_title": "Introducing Claude Opus 5.5",
        "source_date": "2026-09-22",
        "where": "benchmark table (HTML), row 'Computer use / OSWorld 2.1', column 'Opus 5.5', the value tagged 'partial'",
        "own": true,
        "primary": true,
        "parallel": false,
        "notes": "max effort; mean of 5 runs; September 10, 2026 task files; the system card labels it OSWorld 2.0",
        "caveat": "",
        "human_verified": false,
        "human_verified_on": null,
        "check": "corroborated",
        "corroborated": true,
        "corroborated_on": "2026-10-20",
        "corroborated_by": [
          {
            "url": "https://www.anthropic.com/claude-opus-5-5-system-card",
            "name": "Claude Opus 5.5 card",
            "title": "Claude Opus 5.5 System Card",
            "lab": "anthropic",
            "date": "2026-09-22",
            "printed": "81.8%",
            "where": "section 8.13.3 OSWorld 2.0 (p. 206), sentence: 'Claude Opus 5.5 achieved a partial score of 81.8% and a strict pass rate of 48.7%'; also 8.1 Evaluation summary, row 'OSWorld 2.0 (partial/strict)', column Claude Opus 5.5: '81.8/48.7'"
          },
          {
            "url": "https://www.anthropic.com/claude-sonnet-5-5",
            "name": "Claude Sonnet 5.5",
            "title": "Introducing Claude Sonnet 5.5",
            "lab": "anthropic",
            "date": "2026-09-28",
            "printed": "81.8%",
            "where": "Benchmark table, row 'Computer use / OSWorld 2.1', column Opus 5.5: '81.8%' with 'partial' under it"
          },
          {
            "url": "https://www.anthropic.com/claude-sonnet-5-5-system-card",
            "name": "Claude Sonnet 5.5 card",
            "title": "Claude Sonnet 5.5 System Card",
            "lab": "anthropic",
            "date": "2026-09-28",
            "printed": "81.8%",
            "where": "section 8.13.3 OSWorld 2.1 (p. 130), sentence: 'Claude Opus 5.5 scored 81.8% partial and 48.7% strict pass'; also 8.1 Evaluation summary, row 'OSWorld 2.1', column Claude Opus 5.5: '81.8'"
          }
        ],
        "source_capture": "https://frontierlog.ai/captures/ca3ad55d7ec61f3c.png"
      },
      {
        "id": "claude-opus-5-5:osworld-2-1:offline-partial:1",
        "model": "claude-opus-5-5",
        "model_name": "Claude Opus 5.5",
        "model_date": "2026-09-22",
        "event": "https://frontierlog.ai/event/2026-09-22-claude-opus-5-5",
        "benchmark": "osworld-2-1",
        "benchmark_name": "OSWorld 2.1",
        "setting": "offline-partial",
        "setting_name": "Offline set, partial score",
        "score": 87.2,
        "unit": "%",
        "printed": "87.2%",
        "reported_by": "anthropic",
        "run_by": null,
        "label": "Claude Opus 5.5",
        "source": "https://www.anthropic.com/claude-haiku-5-5-system-card",
        "source_title": "Claude Haiku 5.5 System Card",
        "source_date": "2026-10-07",
        "where": "section 8.9.3 OSWorld 2.1 offline subset, text (PDF page 128), the Claude Opus 5.5 partial score",
        "own": false,
        "primary": true,
        "parallel": false,
        "notes": "82-task offline subset, no internet access; max effort; mean of 5 attempts",
        "caveat": "",
        "human_verified": true,
        "human_verified_on": "2026-10-20",
        "check": "human",
        "corroborated": false,
        "corroborated_on": null,
        "corroborated_by": [],
        "source_capture": "https://frontierlog.ai/captures/78b2fca2e8e996c4.png"
      },
      {
        "id": "claude-opus-5-5:osworld-2-1:offline-strict:1",
        "model": "claude-opus-5-5",
        "model_name": "Claude Opus 5.5",
        "model_date": "2026-09-22",
        "event": "https://frontierlog.ai/event/2026-09-22-claude-opus-5-5",
        "benchmark": "osworld-2-1",
        "benchmark_name": "OSWorld 2.1",
        "setting": "offline-strict",
        "setting_name": "Offline set, strict pass rate",
        "score": 53.2,
        "unit": "%",
        "printed": "53.2%",
        "reported_by": "anthropic",
        "run_by": null,
        "label": "Claude Opus 5.5",
        "source": "https://www.anthropic.com/claude-haiku-5-5-system-card",
        "source_title": "Claude Haiku 5.5 System Card",
        "source_date": "2026-10-07",
        "where": "section 8.9.3 OSWorld 2.1 offline subset, text (PDF page 128), the Claude Opus 5.5 strict pass rate",
        "own": false,
        "primary": true,
        "parallel": false,
        "notes": "82-task offline subset, no internet access; max effort; mean of 5 attempts",
        "caveat": "",
        "human_verified": true,
        "human_verified_on": "2026-10-20",
        "check": "human",
        "corroborated": false,
        "corroborated_on": null,
        "corroborated_by": [],
        "source_capture": "https://frontierlog.ai/captures/78b2fca2e8e996c4.png"
      },
      {
        "id": "claude-opus-5-5:osworld-2-1:strict:1",
        "model": "claude-opus-5-5",
        "model_name": "Claude Opus 5.5",
        "model_date": "2026-09-22",
        "event": "https://frontierlog.ai/event/2026-09-22-claude-opus-5-5",
        "benchmark": "osworld-2-1",
        "benchmark_name": "OSWorld 2.1",
        "setting": "strict",
        "setting_name": "Full set, strict pass rate",
        "score": 48.7,
        "unit": "%",
        "printed": "48.7%",
        "reported_by": "anthropic",
        "run_by": null,
        "label": "Claude Opus 5.5",
        "source": "https://www.anthropic.com/claude-sonnet-5-5-system-card",
        "source_title": "Claude Sonnet 5.5 System Card",
        "source_date": "2026-09-28",
        "where": "section 8.13.3 OSWorld 2.1, text (PDF page 130), the Claude Opus 5.5 strict pass rate",
        "own": false,
        "primary": true,
        "parallel": false,
        "notes": "max effort; mean of 5 runs; September 10, 2026 task files",
        "caveat": "",
        "human_verified": true,
        "human_verified_on": "2026-10-20",
        "check": "human",
        "corroborated": false,
        "corroborated_on": null,
        "corroborated_by": [],
        "source_capture": "https://frontierlog.ai/captures/ed9b8a21a7f60448.png"
      }
    ]
  }
}