{
  "api_version": "1",
  "updated": "2026-10-26",
  "latest_event": "2026-10-24",
  "rights": "© Frontier Log, all rights reserved.",
  "model": {
    "slug": "claude-sonnet-4-6",
    "name": "Claude Sonnet 4.6",
    "lab": "anthropic",
    "date": "2026-02-17",
    "source": "https://www.anthropic.com/news/claude-sonnet-4-6",
    "source_title": "Introducing Sonnet 4.6",
    "flagship": false,
    "event": "https://frontierlog.ai/event/2026-02-17-claude-sonnet-4-6",
    "releases": [],
    "variant_of": null,
    "made_by": "Anthropic",
    "url": "https://frontierlog.ai/model/claude-sonnet-4-6",
    "results": [
      {
        "id": "claude-sonnet-4-6:hle:no-tools:1",
        "model": "claude-sonnet-4-6",
        "model_name": "Claude Sonnet 4.6",
        "model_date": "2026-02-17",
        "event": "https://frontierlog.ai/event/2026-02-17-claude-sonnet-4-6",
        "benchmark": "hle",
        "benchmark_name": "Humanity's Last Exam",
        "setting": "no-tools",
        "setting_name": "No tools",
        "score": 33.2,
        "unit": "%",
        "printed": "33.2%",
        "reported_by": "anthropic",
        "run_by": null,
        "label": "Sonnet 4.6",
        "source": "https://www.anthropic.com/news/claude-sonnet-4-6",
        "source_title": "Introducing Sonnet 4.6",
        "source_date": "2026-02-17",
        "where": "benchmark table (image), row 'Multidisciplinary reasoning, Humanity's Last Exam (HLE)', column 'Sonnet 4.6', 'without tools'",
        "own": true,
        "primary": true,
        "parallel": false,
        "notes": "",
        "caveat": "",
        "human_verified": true,
        "human_verified_on": "2026-10-19",
        "check": "human",
        "corroborated": true,
        "corroborated_on": "2026-10-19",
        "corroborated_by": [
          {
            "url": "https://www.anthropic.com/claude-sonnet-4-6-system-card",
            "name": "Claude Sonnet 4.6 card",
            "title": "Claude Sonnet 4.6 System Card",
            "lab": "anthropic",
            "date": "2026-02-17",
            "printed": "33.2%",
            "where": "Table 2.1.A, HLE No tools row, Sonnet 4.6 column"
          }
        ],
        "source_image": "https://frontierlog.ai/sources/d694e6e89a414801.png",
        "source_image_original": "https://www-cdn.anthropic.com/images/4zrzovbb/website/10b2602771d21378cd6d76628a081c8a76dcf216-2600x2960.png"
      },
      {
        "id": "claude-sonnet-4-6:hle:with-tools:1",
        "model": "claude-sonnet-4-6",
        "model_name": "Claude Sonnet 4.6",
        "model_date": "2026-02-17",
        "event": "https://frontierlog.ai/event/2026-02-17-claude-sonnet-4-6",
        "benchmark": "hle",
        "benchmark_name": "Humanity's Last Exam",
        "setting": "with-tools",
        "setting_name": "With tools",
        "score": 49.0,
        "unit": "%",
        "printed": "49.0%",
        "reported_by": "anthropic",
        "run_by": null,
        "label": "Sonnet 4.6",
        "source": "https://www.anthropic.com/news/claude-sonnet-4-6",
        "source_title": "Introducing Sonnet 4.6",
        "source_date": "2026-02-17",
        "where": "benchmark table (image), row 'Multidisciplinary reasoning, Humanity's Last Exam (HLE)', column 'Sonnet 4.6', 'with tools'",
        "own": true,
        "primary": true,
        "parallel": false,
        "notes": "web search, web fetch, code execution, programmatic tool calling; compaction at 50K up to 3M tokens; max effort",
        "caveat": "",
        "human_verified": true,
        "human_verified_on": "2026-10-19",
        "check": "human",
        "corroborated": true,
        "corroborated_on": "2026-10-19",
        "corroborated_by": [
          {
            "url": "https://www.anthropic.com/claude-sonnet-4-6-system-card",
            "name": "Claude Sonnet 4.6 card",
            "title": "Claude Sonnet 4.6 System Card",
            "lab": "anthropic",
            "date": "2026-02-17",
            "printed": "49.0%",
            "where": "Table 2.1.A, HLE With tools row, Sonnet 4.6 column"
          }
        ],
        "source_image": "https://frontierlog.ai/sources/d694e6e89a414801.png",
        "source_image_original": "https://www-cdn.anthropic.com/images/4zrzovbb/website/10b2602771d21378cd6d76628a081c8a76dcf216-2600x2960.png"
      },
      {
        "id": "claude-sonnet-4-6:browsecomp:default:1",
        "model": "claude-sonnet-4-6",
        "model_name": "Claude Sonnet 4.6",
        "model_date": "2026-02-17",
        "event": "https://frontierlog.ai/event/2026-02-17-claude-sonnet-4-6",
        "benchmark": "browsecomp",
        "benchmark_name": "BrowseComp",
        "setting": "default",
        "setting_name": "",
        "score": 74.7,
        "unit": "%",
        "printed": "74.7%",
        "reported_by": "anthropic",
        "run_by": null,
        "label": "Sonnet 4.6",
        "source": "https://www.anthropic.com/news/claude-sonnet-4-6",
        "source_title": "Introducing Sonnet 4.6",
        "source_date": "2026-02-17",
        "where": "benchmark table (image), row 'Agentic search, BrowseComp', column 'Sonnet 4.6'",
        "own": true,
        "primary": true,
        "parallel": false,
        "notes": "web search, web fetch, programmatic tool calling; compaction at 50K up to 10M tokens; max effort, no thinking",
        "caveat": "",
        "human_verified": true,
        "human_verified_on": "2026-10-19",
        "check": "human",
        "corroborated": true,
        "corroborated_on": "2026-10-19",
        "corroborated_by": [
          {
            "url": "https://www.anthropic.com/claude-sonnet-4-6-system-card",
            "name": "Claude Sonnet 4.6 card",
            "title": "Claude Sonnet 4.6 System Card",
            "lab": "anthropic",
            "date": "2026-02-17",
            "printed": "74.72%",
            "where": "Changelog, March 6, 2026: 'Updated highest single-agent BrowseComp score (74.72% -> 74.01%)'; the original value"
          }
        ],
        "source_image": "https://frontierlog.ai/sources/d694e6e89a414801.png",
        "source_image_original": "https://www-cdn.anthropic.com/images/4zrzovbb/website/10b2602771d21378cd6d76628a081c8a76dcf216-2600x2960.png"
      },
      {
        "id": "claude-sonnet-4-6:browsecomp:default:2",
        "model": "claude-sonnet-4-6",
        "model_name": "Claude Sonnet 4.6",
        "model_date": "2026-02-17",
        "event": "https://frontierlog.ai/event/2026-02-17-claude-sonnet-4-6",
        "benchmark": "browsecomp",
        "benchmark_name": "BrowseComp",
        "setting": "default",
        "setting_name": "",
        "score": 74.01,
        "unit": "%",
        "printed": "74.01%",
        "reported_by": "anthropic",
        "run_by": null,
        "label": "Claude Sonnet 4.6",
        "source": "https://www.anthropic.com/claude-sonnet-4-6-system-card",
        "source_title": "Claude Sonnet 4.6 System Card",
        "source_date": "2026-02-17",
        "where": "Section 2.20.1.1 'BrowseComp', first paragraph after the thinking note",
        "own": true,
        "primary": false,
        "parallel": false,
        "notes": "card update of 6 Mar 2026 after improved cheating detection; no thinking, max effort",
        "caveat": "",
        "human_verified": true,
        "human_verified_on": "2026-10-19",
        "check": "human",
        "corroborated": false,
        "corroborated_on": null,
        "corroborated_by": []
      },
      {
        "id": "claude-sonnet-4-6:arc-agi-2:default:1",
        "model": "claude-sonnet-4-6",
        "model_name": "Claude Sonnet 4.6",
        "model_date": "2026-02-17",
        "event": "https://frontierlog.ai/event/2026-02-17-claude-sonnet-4-6",
        "benchmark": "arc-agi-2",
        "benchmark_name": "ARC-AGI-2",
        "setting": "default",
        "setting_name": "",
        "score": 58.3,
        "unit": "%",
        "printed": "58.3%",
        "reported_by": "anthropic",
        "run_by": null,
        "label": "Sonnet 4.6",
        "source": "https://www.anthropic.com/news/claude-sonnet-4-6",
        "source_title": "Introducing Sonnet 4.6",
        "source_date": "2026-02-17",
        "where": "benchmark table (image), row 'Novel problem-solving, ARC-AGI-2', column 'Sonnet 4.6'",
        "own": true,
        "primary": true,
        "parallel": false,
        "notes": "max effort, 120K thinking budget",
        "caveat": "",
        "human_verified": true,
        "human_verified_on": "2026-10-19",
        "check": "human",
        "corroborated": true,
        "corroborated_on": "2026-10-19",
        "corroborated_by": [
          {
            "url": "https://www.anthropic.com/claude-sonnet-4-6-system-card",
            "name": "Claude Sonnet 4.6 card",
            "title": "Claude Sonnet 4.6 System Card",
            "lab": "anthropic",
            "date": "2026-02-17",
            "printed": "58.3%",
            "where": "Table 2.1.A, ARC-AGI-2 (Verified) row, Sonnet 4.6 column (footnote 4: max effort; 60.42% at high effort)"
          }
        ],
        "source_image": "https://frontierlog.ai/sources/d694e6e89a414801.png",
        "source_image_original": "https://www-cdn.anthropic.com/images/4zrzovbb/website/10b2602771d21378cd6d76628a081c8a76dcf216-2600x2960.png"
      },
      {
        "id": "claude-sonnet-4-6:gpqa-diamond:default:1",
        "model": "claude-sonnet-4-6",
        "model_name": "Claude Sonnet 4.6",
        "model_date": "2026-02-17",
        "event": "https://frontierlog.ai/event/2026-02-17-claude-sonnet-4-6",
        "benchmark": "gpqa-diamond",
        "benchmark_name": "GPQA Diamond",
        "setting": "default",
        "setting_name": "",
        "score": 89.9,
        "unit": "%",
        "printed": "89.9%",
        "reported_by": "anthropic",
        "run_by": null,
        "label": "Sonnet 4.6",
        "source": "https://www.anthropic.com/news/claude-sonnet-4-6",
        "source_title": "Introducing Sonnet 4.6",
        "source_date": "2026-02-17",
        "where": "benchmark table (image), row 'Graduate-level reasoning, GPQA Diamond', column 'Sonnet 4.6'",
        "own": true,
        "primary": true,
        "parallel": false,
        "notes": "",
        "caveat": "",
        "human_verified": true,
        "human_verified_on": "2026-10-19",
        "check": "human",
        "corroborated": true,
        "corroborated_on": "2026-10-19",
        "corroborated_by": [
          {
            "url": "https://www.anthropic.com/claude-sonnet-4-6-system-card",
            "name": "Claude Sonnet 4.6 card",
            "title": "Claude Sonnet 4.6 System Card",
            "lab": "anthropic",
            "date": "2026-02-17",
            "printed": "89.9%",
            "where": "Table 2.1.A, GPQA Diamond row, Sonnet 4.6 column (section 2.9 sentence also: 89.9%, averaged over 10 trials)"
          }
        ],
        "source_image": "https://frontierlog.ai/sources/d694e6e89a414801.png",
        "source_image_original": "https://www-cdn.anthropic.com/images/4zrzovbb/website/10b2602771d21378cd6d76628a081c8a76dcf216-2600x2960.png"
      },
      {
        "id": "claude-sonnet-4-6:swe-bench-pro:default:1",
        "model": "claude-sonnet-4-6",
        "model_name": "Claude Sonnet 4.6",
        "model_date": "2026-02-17",
        "event": "https://frontierlog.ai/event/2026-02-17-claude-sonnet-4-6",
        "benchmark": "swe-bench-pro",
        "benchmark_name": "SWE-bench Pro",
        "setting": "default",
        "setting_name": "",
        "score": 58.1,
        "unit": "%",
        "printed": "58.1%",
        "reported_by": "anthropic",
        "run_by": null,
        "label": "Sonnet 4.6",
        "source": "https://www.anthropic.com/news/claude-sonnet-5",
        "source_title": "Introducing Claude Sonnet 5",
        "source_date": "2026-06-30",
        "where": "benchmark table (image), row 'Agentic coding, SWE-bench Pro', column 'Sonnet 4.6'",
        "own": false,
        "primary": true,
        "parallel": false,
        "notes": "",
        "caveat": "The source doesn't say whether this is the public or the commercial set",
        "human_verified": true,
        "human_verified_on": "2026-10-19",
        "check": "human",
        "corroborated": true,
        "corroborated_on": "2026-10-19",
        "corroborated_by": [
          {
            "url": "https://www.anthropic.com/claude-sonnet-5-system-card",
            "name": "Claude Sonnet 5 card",
            "title": "Claude Sonnet 5 System Card",
            "lab": "anthropic",
            "date": "2026-06-30",
            "printed": "58.1",
            "where": "Table 8.1.A, SWE-bench Pro row, Sonnet 4.6 column"
          }
        ],
        "source_image": "https://frontierlog.ai/sources/849d6a908f0d46ac.png",
        "source_image_original": "https://www-cdn.anthropic.com/images/4zrzovbb/website/9941d610909f28a504e16dd5af823df172ec6035-2600x1234.png"
      },
      {
        "id": "claude-sonnet-4-6:hle:no-tools:2",
        "model": "claude-sonnet-4-6",
        "model_name": "Claude Sonnet 4.6",
        "model_date": "2026-02-17",
        "event": "https://frontierlog.ai/event/2026-02-17-claude-sonnet-4-6",
        "benchmark": "hle",
        "benchmark_name": "Humanity's Last Exam",
        "setting": "no-tools",
        "setting_name": "No tools",
        "score": 34.6,
        "unit": "%",
        "printed": "34.6%",
        "reported_by": "anthropic",
        "run_by": null,
        "label": "Sonnet 4.6",
        "source": "https://www.anthropic.com/news/claude-sonnet-5",
        "source_title": "Introducing Claude Sonnet 5",
        "source_date": "2026-06-30",
        "where": "benchmark table (image), row 'Multidisciplinary reasoning, Humanity's Last Exam', column 'Sonnet 4.6', 'no tools'",
        "own": false,
        "primary": false,
        "parallel": false,
        "notes": "re-graded with an updated grader model",
        "caveat": "",
        "human_verified": true,
        "human_verified_on": "2026-10-19",
        "check": "human",
        "corroborated": true,
        "corroborated_on": "2026-10-19",
        "corroborated_by": [
          {
            "url": "https://www.anthropic.com/claude-sonnet-5-system-card",
            "name": "Claude Sonnet 5 card",
            "title": "Claude Sonnet 5 System Card",
            "lab": "anthropic",
            "date": "2026-06-30",
            "printed": "34.6",
            "where": "Table 8.1.A, Humanity's Last Exam No tools row, Sonnet 4.6 column"
          }
        ],
        "source_image": "https://frontierlog.ai/sources/849d6a908f0d46ac.png",
        "source_image_original": "https://www-cdn.anthropic.com/images/4zrzovbb/website/9941d610909f28a504e16dd5af823df172ec6035-2600x1234.png"
      },
      {
        "id": "claude-sonnet-4-6:hle:with-tools:2",
        "model": "claude-sonnet-4-6",
        "model_name": "Claude Sonnet 4.6",
        "model_date": "2026-02-17",
        "event": "https://frontierlog.ai/event/2026-02-17-claude-sonnet-4-6",
        "benchmark": "hle",
        "benchmark_name": "Humanity's Last Exam",
        "setting": "with-tools",
        "setting_name": "With tools",
        "score": 46.8,
        "unit": "%",
        "printed": "46.8%",
        "reported_by": "anthropic",
        "run_by": null,
        "label": "Sonnet 4.6",
        "source": "https://www.anthropic.com/news/claude-sonnet-5",
        "source_title": "Introducing Claude Sonnet 5",
        "source_date": "2026-06-30",
        "where": "benchmark table (image), row 'Multidisciplinary reasoning, Humanity's Last Exam', column 'Sonnet 4.6', 'with tools'",
        "own": false,
        "primary": false,
        "parallel": false,
        "notes": "re-graded with an updated grader model",
        "caveat": "",
        "human_verified": true,
        "human_verified_on": "2026-10-19",
        "check": "human",
        "corroborated": true,
        "corroborated_on": "2026-10-19",
        "corroborated_by": [
          {
            "url": "https://www.anthropic.com/claude-sonnet-5-system-card",
            "name": "Claude Sonnet 5 card",
            "title": "Claude Sonnet 5 System Card",
            "lab": "anthropic",
            "date": "2026-06-30",
            "printed": "46.8",
            "where": "Table 8.1.A, Humanity's Last Exam With tools row, Sonnet 4.6 column"
          }
        ],
        "source_image": "https://frontierlog.ai/sources/849d6a908f0d46ac.png",
        "source_image_original": "https://www-cdn.anthropic.com/images/4zrzovbb/website/9941d610909f28a504e16dd5af823df172ec6035-2600x1234.png"
      },
      {
        "id": "claude-sonnet-4-6:browsecomp:default:3",
        "model": "claude-sonnet-4-6",
        "model_name": "Claude Sonnet 4.6",
        "model_date": "2026-02-17",
        "event": "https://frontierlog.ai/event/2026-02-17-claude-sonnet-4-6",
        "benchmark": "browsecomp",
        "benchmark_name": "BrowseComp",
        "setting": "default",
        "setting_name": "",
        "score": 76.2,
        "unit": "%",
        "printed": "76.2",
        "reported_by": "anthropic",
        "run_by": null,
        "label": "Claude Sonnet 4.6",
        "source": "https://www.anthropic.com/claude-sonnet-5-system-card",
        "source_title": "Claude Sonnet 5 System Card",
        "source_date": "2026-06-30",
        "where": "Section 8.1 evaluation summary, Table 8.1.A (page 115), row 'BrowseComp', column 'Claude Sonnet 4.6'",
        "own": false,
        "primary": false,
        "parallel": false,
        "notes": "",
        "caveat": "",
        "human_verified": true,
        "human_verified_on": "2026-10-19",
        "check": "human",
        "corroborated": false,
        "corroborated_on": null,
        "corroborated_by": [],
        "source_capture": "https://frontierlog.ai/captures/83869c127dd1cbc7.png"
      },
      {
        "id": "claude-sonnet-4-6:arc-agi-2:default:2",
        "model": "claude-sonnet-4-6",
        "model_name": "Claude Sonnet 4.6",
        "model_date": "2026-02-17",
        "event": "https://frontierlog.ai/event/2026-02-17-claude-sonnet-4-6",
        "benchmark": "arc-agi-2",
        "benchmark_name": "ARC-AGI-2",
        "setting": "default",
        "setting_name": "",
        "score": 60.4,
        "unit": "%",
        "printed": "60.4%",
        "reported_by": "anthropic",
        "run_by": null,
        "label": "Claude Sonnet 4.6 (high effort)",
        "source": "https://www.anthropic.com/news/claude-sonnet-4-6",
        "source_title": "Introducing Sonnet 4.6",
        "source_date": "2026-02-17",
        "where": "Footnotes, 'ARC-AGI-2' paragraph, the high-effort score",
        "own": true,
        "primary": false,
        "parallel": false,
        "notes": "high effort, 120K thinking budget; run by the ARC Prize Foundation on its private set",
        "caveat": "",
        "human_verified": false,
        "human_verified_on": null,
        "check": "corroborated",
        "corroborated": true,
        "corroborated_on": "2026-10-19",
        "corroborated_by": [
          {
            "url": "https://www.anthropic.com/claude-sonnet-4-6-system-card",
            "name": "Claude Sonnet 4.6 card",
            "title": "Claude Sonnet 4.6 System Card",
            "lab": "anthropic",
            "date": "2026-02-17",
            "printed": "60.42%",
            "where": "Table footnote 4 under the evaluation summary: 'with high effort, Claude Sonnet 4.6 achieves a score of 60.42%'; 2.7 text: '60.42% on ARC-AGI-2 with 120k thinking tokens and High effort'"
          }
        ]
      }
    ]
  }
}