{
  "api_version": "1",
  "updated": "2026-10-26",
  "latest_event": "2026-10-26",
  "rights": "© Frontier Log, all rights reserved.",
  "model": {
    "slug": "claude-3-5-sonnet-upgraded",
    "name": "Claude 3.5 Sonnet (upgraded)",
    "lab": "anthropic",
    "date": "2024-10-22",
    "source": "https://www.anthropic.com/news/3-5-models-and-computer-use",
    "source_title": "Introducing computer use, a new Claude 3.5 Sonnet, and Claude 3.5 Haiku",
    "flagship": true,
    "event": "https://frontierlog.ai/event/2024-10-22-claude-3-5-haiku-new-3-5-sonnet",
    "releases": [],
    "variant_of": null,
    "made_by": "Anthropic",
    "url": "https://frontierlog.ai/model/claude-3-5-sonnet-upgraded",
    "results": [
      {
        "id": "claude-3-5-sonnet-upgraded:gpqa-diamond:default:1",
        "model": "claude-3-5-sonnet-upgraded",
        "model_name": "Claude 3.5 Sonnet (upgraded)",
        "model_date": "2024-10-22",
        "event": "https://frontierlog.ai/event/2024-10-22-claude-3-5-haiku-new-3-5-sonnet",
        "benchmark": "gpqa-diamond",
        "benchmark_name": "GPQA Diamond",
        "setting": "default",
        "setting_name": "",
        "score": 65.0,
        "unit": "%",
        "printed": "65.0%",
        "reported_by": "anthropic",
        "run_by": null,
        "label": "Claude 3.5 Sonnet (new)",
        "source": "https://www.anthropic.com/news/3-5-models-and-computer-use",
        "source_title": "Introducing computer use, a new Claude 3.5 Sonnet, and Claude 3.5 Haiku",
        "source_date": "2024-10-22",
        "where": "benchmark table (image), row 'Graduate level reasoning, GPQA (Diamond)', column 'Claude 3.5 Sonnet (new)'",
        "own": true,
        "primary": true,
        "parallel": false,
        "notes": "0-shot chain of thought",
        "caveat": "",
        "human_verified": false,
        "human_verified_on": null,
        "check": "corroborated",
        "corroborated": true,
        "corroborated_on": "2026-10-26",
        "corroborated_by": [
          {
            "url": "https://assets.anthropic.com/m/1cd9d098ac3e6467/original/Claude-3-Model-Card-October-Addendum.pdf",
            "name": "Upgraded Claude 3.5 Sonnet and 3.5 Haiku card",
            "title": "Model Card Addendum: Claude 3.5 Haiku and Upgraded Claude 3.5 Sonnet",
            "lab": "anthropic",
            "date": "2024-10-22",
            "printed": "65.0%",
            "where": "Table 7 (upgraded Claude 3.5 Sonnet and peer models), row 'GPQA (Diamond)' 0-shot CoT, column 'Claude 3.5 Sonnet (New)' (page 6)"
          },
          {
            "url": "https://www.anthropic.com/news/claude-3-7-sonnet",
            "name": "Claude 3.7 Sonnet",
            "title": "Claude 3.7 Sonnet and Claude Code",
            "lab": "anthropic",
            "date": "2025-02-24",
            "printed": "65.0%",
            "where": "Claude 3.7 Sonnet launch post, benchmark table image, row 'Graduate-level reasoning GPQA Diamond', column 'Claude 3.5 Sonnet (new)' (no shot/CoT setting stated)"
          }
        ],
        "source_image": "https://frontierlog.ai/sources/6214ed8f2bfd7b6f.png",
        "source_image_original": "https://www-cdn.anthropic.com/images/4zrzovbb/website/0eb9a1b7d5db74a6d21500e9f188c83beef3842e-2601x1932.png"
      },
      {
        "id": "claude-3-5-sonnet-upgraded:mmlu-pro:default:1",
        "model": "claude-3-5-sonnet-upgraded",
        "model_name": "Claude 3.5 Sonnet (upgraded)",
        "model_date": "2024-10-22",
        "event": "https://frontierlog.ai/event/2024-10-22-claude-3-5-haiku-new-3-5-sonnet",
        "benchmark": "mmlu-pro",
        "benchmark_name": "MMLU-Pro",
        "setting": "default",
        "setting_name": "",
        "score": 78.0,
        "unit": "%",
        "printed": "78.0%",
        "reported_by": "anthropic",
        "run_by": null,
        "label": "Claude 3.5 Sonnet (new)",
        "source": "https://www.anthropic.com/news/3-5-models-and-computer-use",
        "source_title": "Introducing computer use, a new Claude 3.5 Sonnet, and Claude 3.5 Haiku",
        "source_date": "2024-10-22",
        "where": "benchmark table (image), row 'Undergraduate level knowledge, MMLU Pro', column 'Claude 3.5 Sonnet (new)'",
        "own": true,
        "primary": true,
        "parallel": false,
        "notes": "0-shot chain of thought",
        "caveat": "",
        "human_verified": false,
        "human_verified_on": null,
        "check": "corroborated",
        "corroborated": true,
        "corroborated_on": "2026-10-26",
        "corroborated_by": [
          {
            "url": "https://assets.anthropic.com/m/1cd9d098ac3e6467/original/Claude-3-Model-Card-October-Addendum.pdf",
            "name": "Upgraded Claude 3.5 Sonnet and 3.5 Haiku card",
            "title": "Model Card Addendum: Claude 3.5 Haiku and Upgraded Claude 3.5 Sonnet",
            "lab": "anthropic",
            "date": "2024-10-22",
            "printed": "78.0%",
            "where": "Table 7, row 'MMLU Pro' 0-shot CoT, column 'Claude 3.5 Sonnet (New)'"
          }
        ],
        "source_image": "https://frontierlog.ai/sources/6214ed8f2bfd7b6f.png",
        "source_image_original": "https://www-cdn.anthropic.com/images/4zrzovbb/website/0eb9a1b7d5db74a6d21500e9f188c83beef3842e-2601x1932.png"
      },
      {
        "id": "claude-3-5-sonnet-upgraded:aime-2024:no-tools:1",
        "model": "claude-3-5-sonnet-upgraded",
        "model_name": "Claude 3.5 Sonnet (upgraded)",
        "model_date": "2024-10-22",
        "event": "https://frontierlog.ai/event/2024-10-22-claude-3-5-haiku-new-3-5-sonnet",
        "benchmark": "aime-2024",
        "benchmark_name": "AIME 2024",
        "setting": "no-tools",
        "setting_name": "No tools",
        "score": 16.0,
        "unit": "%",
        "printed": "16.0%",
        "reported_by": "anthropic",
        "run_by": null,
        "label": "Claude 3.5 Sonnet (new)",
        "source": "https://www.anthropic.com/news/3-5-models-and-computer-use",
        "source_title": "Introducing computer use, a new Claude 3.5 Sonnet, and Claude 3.5 Haiku",
        "source_date": "2024-10-22",
        "where": "benchmark table (image), row 'High school math competition, AIME 2024', column 'Claude 3.5 Sonnet (new)'",
        "own": true,
        "primary": true,
        "parallel": false,
        "notes": "0-shot chain of thought",
        "caveat": "",
        "human_verified": false,
        "human_verified_on": null,
        "check": "corroborated",
        "corroborated": true,
        "corroborated_on": "2026-10-26",
        "corroborated_by": [
          {
            "url": "https://assets.anthropic.com/m/1cd9d098ac3e6467/original/Claude-3-Model-Card-October-Addendum.pdf",
            "name": "Upgraded Claude 3.5 Sonnet and 3.5 Haiku card",
            "title": "Model Card Addendum: Claude 3.5 Haiku and Upgraded Claude 3.5 Sonnet",
            "lab": "anthropic",
            "date": "2024-10-22",
            "printed": "16.0%",
            "where": "Table 7, row 'AIME 2024' 0-shot CoT, column 'Claude 3.5 Sonnet (New)'"
          },
          {
            "url": "https://www.anthropic.com/news/claude-3-7-sonnet",
            "name": "Claude 3.7 Sonnet",
            "title": "Claude 3.7 Sonnet and Claude Code",
            "lab": "anthropic",
            "date": "2025-02-24",
            "printed": "16.0%",
            "where": "Claude 3.7 Sonnet launch post, benchmark table image, row 'High school math competition AIME 2024', column 'Claude 3.5 Sonnet (new)' (methodology note: pass@1 averaged over several trials, up to 16 for AIME)"
          }
        ],
        "source_image": "https://frontierlog.ai/sources/6214ed8f2bfd7b6f.png",
        "source_image_original": "https://www-cdn.anthropic.com/images/4zrzovbb/website/0eb9a1b7d5db74a6d21500e9f188c83beef3842e-2601x1932.png"
      },
      {
        "id": "claude-3-5-sonnet-upgraded:swe-bench-verified:default:1",
        "model": "claude-3-5-sonnet-upgraded",
        "model_name": "Claude 3.5 Sonnet (upgraded)",
        "model_date": "2024-10-22",
        "event": "https://frontierlog.ai/event/2024-10-22-claude-3-5-haiku-new-3-5-sonnet",
        "benchmark": "swe-bench-verified",
        "benchmark_name": "SWE-bench Verified",
        "setting": "default",
        "setting_name": "",
        "score": 49.0,
        "unit": "%",
        "printed": "49.0%",
        "reported_by": "anthropic",
        "run_by": null,
        "label": "Claude 3.5 Sonnet (new)",
        "source": "https://www.anthropic.com/news/3-5-models-and-computer-use",
        "source_title": "Introducing computer use, a new Claude 3.5 Sonnet, and Claude 3.5 Haiku",
        "source_date": "2024-10-22",
        "where": "benchmark table (image), row 'Agentic coding, SWE-bench Verified', column 'Claude 3.5 Sonnet (new)'",
        "own": true,
        "primary": true,
        "parallel": false,
        "notes": "",
        "caveat": "",
        "human_verified": false,
        "human_verified_on": null,
        "check": "corroborated",
        "corroborated": true,
        "corroborated_on": "2026-10-26",
        "corroborated_by": [
          {
            "url": "https://assets.anthropic.com/m/1cd9d098ac3e6467/original/Claude-3-Model-Card-October-Addendum.pdf",
            "name": "Upgraded Claude 3.5 Sonnet and 3.5 Haiku card",
            "title": "Model Card Addendum: Claude 3.5 Haiku and Upgraded Claude 3.5 Sonnet",
            "lab": "anthropic",
            "date": "2024-10-22",
            "printed": "49.0%",
            "where": "Table 2 'SWE-bench Verified results', column 'Claude 3.5 Sonnet (New)'; also sentence 'state-of-the-art pass@1 performance on SWE-bench Verified of 49.0%'"
          },
          {
            "url": "https://www.anthropic.com/news/claude-3-7-sonnet",
            "name": "Claude 3.7 Sonnet",
            "title": "Claude 3.7 Sonnet and Claude Code",
            "lab": "anthropic",
            "date": "2025-02-24",
            "printed": "49.0%",
            "where": "Claude 3.7 Sonnet launch post, benchmark table image, row 'Agentic coding SWE-bench Verified', column 'Claude 3.5 Sonnet (new)'"
          },
          {
            "url": "https://www.anthropic.com/engineering/swe-bench-sonnet",
            "name": "Anthropic post",
            "title": "Claude SWE-Bench Performance",
            "lab": "anthropic",
            "date": "2025-01-06",
            "printed": "49%",
            "where": "Anthropic engineering post 'Raising the bar on SWE-bench Verified with Claude 3.5 Sonnet' (2025-01-06), HTML table row 'SWE-bench Verified score', column 'Claude 3.5 Sonnet (new)'; also sentences 'achieved 49% on SWE-bench Verified'. Printed at integer precision (49% vs 49.0%)"
          }
        ],
        "source_image": "https://frontierlog.ai/sources/6214ed8f2bfd7b6f.png",
        "source_image_original": "https://www-cdn.anthropic.com/images/4zrzovbb/website/0eb9a1b7d5db74a6d21500e9f188c83beef3842e-2601x1932.png"
      },
      {
        "id": "claude-3-5-sonnet-upgraded:aime-2024:no-tools:2",
        "model": "claude-3-5-sonnet-upgraded",
        "model_name": "Claude 3.5 Sonnet (upgraded)",
        "model_date": "2024-10-22",
        "event": "https://frontierlog.ai/event/2024-10-22-claude-3-5-haiku-new-3-5-sonnet",
        "benchmark": "aime-2024",
        "benchmark_name": "AIME 2024",
        "setting": "no-tools",
        "setting_name": "No tools",
        "score": 27.6,
        "unit": "%",
        "printed": "27.6%",
        "reported_by": "anthropic",
        "run_by": null,
        "label": "Claude 3.5 Sonnet (New)",
        "source": "https://assets.anthropic.com/m/1cd9d098ac3e6467/original/Claude-3-Model-Card-October-Addendum.pdf",
        "source_title": "Model Card Addendum: Claude 3.5 Haiku and Upgraded Claude 3.5 Sonnet",
        "source_date": "2024-10-22",
        "where": "Table 7 (evaluation results), row 'AIME 2024', second line 'Maj@64 0-shot CoT', column 'Claude 3.5 Sonnet (New)'",
        "own": true,
        "primary": false,
        "parallel": true,
        "notes": "majority vote over 64 samples; 0-shot chain of thought",
        "caveat": "",
        "human_verified": true,
        "human_verified_on": "2026-10-26",
        "check": "human",
        "corroborated": false,
        "corroborated_on": null,
        "corroborated_by": []
      }
    ]
  }
}
