{
  "api_version": "1",
  "updated": "2026-10-16",
  "latest_event": "2026-10-15",
  "rights": "© Frontier Log, all rights reserved.",
  "benchmark": {
    "slug": "swe-bench-verified",
    "name": "SWE-bench Verified",
    "family": "swe-bench",
    "unit": "%",
    "higher_is_better": true,
    "status": "retired",
    "benchmark_url": "https://www.swebench.com",
    "settings": [
      {
        "setting": "default",
        "name": "",
        "url": "https://frontierlog.ai/benchmark/swe-bench-verified",
        "results": 11
      }
    ]
  },
  "total": 11,
  "results": [
    {
      "id": "claude-3-7-sonnet:swe-bench-verified:default:1",
      "model": "claude-3-7-sonnet",
      "model_name": "Claude 3.7 Sonnet",
      "model_date": "2025-02-24",
      "event": "https://frontierlog.ai/event/2025-02-24-claude-3-7-sonnet",
      "benchmark": "swe-bench-verified",
      "setting": "default",
      "score": 62.3,
      "unit": "%",
      "printed": "62.3%",
      "reported_by": "anthropic",
      "label": "Claude 3.7 Sonnet, no extended thinking",
      "source": "https://www.anthropic.com/news/claude-3-7-sonnet",
      "source_title": "Claude 3.7 Sonnet and Claude Code",
      "source_date": "2025-02-24",
      "where": "benchmark table (image), row 'Agentic coding, SWE-bench Verified', column 'Claude 3.7 Sonnet, No extended thinking', first value",
      "own": true,
      "primary": true,
      "parallel": false,
      "notes": "pass@1 with bash/editor tools plus a 'thinking tool', averaged over 16 trials (footnote 2)",
      "human_verified": true,
      "human_verified_on": "2026-10-16",
      "source_image": "https://frontierlog.ai/sources/7018392ca041e885.png",
      "source_image_original": "https://www-cdn.anthropic.com/images/4zrzovbb/website/654cf6680d32858dfba9af644f8c4a5b04425af1-2600x2360.png"
    },
    {
      "id": "claude-3-7-sonnet:swe-bench-verified:default:2",
      "model": "claude-3-7-sonnet",
      "model_name": "Claude 3.7 Sonnet",
      "model_date": "2025-02-24",
      "event": "https://frontierlog.ai/event/2025-02-24-claude-3-7-sonnet",
      "benchmark": "swe-bench-verified",
      "setting": "default",
      "score": 70.3,
      "unit": "%",
      "printed": "70.3%",
      "reported_by": "anthropic",
      "label": "Claude 3.7 Sonnet, no extended thinking",
      "source": "https://www.anthropic.com/news/claude-3-7-sonnet",
      "source_title": "Claude 3.7 Sonnet and Claude Code",
      "source_date": "2025-02-24",
      "where": "benchmark table (image), row 'Agentic coding, SWE-bench Verified', column 'Claude 3.7 Sonnet, No extended thinking', second value",
      "own": true,
      "primary": false,
      "parallel": true,
      "notes": "internal scoring and custom scaffold on a reduced subset of problems (footnote 2)",
      "human_verified": true,
      "human_verified_on": "2026-10-16",
      "source_image": "https://frontierlog.ai/sources/7018392ca041e885.png",
      "source_image_original": "https://www-cdn.anthropic.com/images/4zrzovbb/website/654cf6680d32858dfba9af644f8c4a5b04425af1-2600x2360.png"
    },
    {
      "id": "gpt-5:swe-bench-verified:default:1",
      "model": "gpt-5",
      "model_name": "GPT-5",
      "model_date": "2025-08-07",
      "event": "https://frontierlog.ai/event/2025-08-07-gpt-5",
      "benchmark": "swe-bench-verified",
      "setting": "default",
      "score": 74.9,
      "unit": "%",
      "printed": "74.9%",
      "reported_by": "openai",
      "label": "GPT-5",
      "source": "https://openai.com/index/introducing-gpt-5",
      "source_title": "Introducing GPT-5",
      "source_date": "2025-08-07",
      "where": "text, opening summary of results",
      "own": true,
      "primary": true,
      "parallel": false,
      "notes": "",
      "human_verified": true,
      "human_verified_on": "2026-10-16"
    },
    {
      "id": "claude-sonnet-4-5:swe-bench-verified:default:1",
      "model": "claude-sonnet-4-5",
      "model_name": "Claude Sonnet 4.5",
      "model_date": "2025-09-29",
      "event": "https://frontierlog.ai/event/2025-09-29-claude-sonnet-4-5",
      "benchmark": "swe-bench-verified",
      "setting": "default",
      "score": 77.2,
      "unit": "%",
      "printed": "77.2%",
      "reported_by": "anthropic",
      "label": "Claude Sonnet 4.5",
      "source": "https://www.anthropic.com/news/claude-sonnet-4-5",
      "source_title": "Introducing Claude Sonnet 4.5",
      "source_date": "2025-09-29",
      "where": "benchmark table (image), row 'Agentic coding, SWE-bench Verified', column 'Claude Sonnet 4.5', first value",
      "own": true,
      "primary": true,
      "parallel": false,
      "notes": "bash and file-editing tools, averaged over 10 trials, no test-time compute, 200K thinking budget (methodology)",
      "human_verified": true,
      "human_verified_on": "2026-10-16",
      "source_image": "https://frontierlog.ai/sources/b6f039f2d22e323a.png",
      "source_image_original": "https://www-cdn.anthropic.com/images/4zrzovbb/website/67081be1ea2752e2a554e49a6aab2731b265d11b-2600x2288.png"
    },
    {
      "id": "claude-sonnet-4-5:swe-bench-verified:default:2",
      "model": "claude-sonnet-4-5",
      "model_name": "Claude Sonnet 4.5",
      "model_date": "2025-09-29",
      "event": "https://frontierlog.ai/event/2025-09-29-claude-sonnet-4-5",
      "benchmark": "swe-bench-verified",
      "setting": "default",
      "score": 82.0,
      "unit": "%",
      "printed": "82.0%",
      "reported_by": "anthropic",
      "label": "Claude Sonnet 4.5",
      "source": "https://www.anthropic.com/news/claude-sonnet-4-5",
      "source_title": "Introducing Claude Sonnet 4.5",
      "source_date": "2025-09-29",
      "where": "benchmark table (image), row 'Agentic coding, SWE-bench Verified', column 'Claude Sonnet 4.5', second value",
      "own": true,
      "primary": false,
      "parallel": true,
      "notes": "with parallel test-time compute",
      "human_verified": true,
      "human_verified_on": "2026-10-16",
      "source_image": "https://frontierlog.ai/sources/b6f039f2d22e323a.png",
      "source_image_original": "https://www-cdn.anthropic.com/images/4zrzovbb/website/67081be1ea2752e2a554e49a6aab2731b265d11b-2600x2288.png"
    },
    {
      "id": "gemini-3-pro:swe-bench-verified:default:1",
      "model": "gemini-3-pro",
      "model_name": "Gemini 3 Pro",
      "model_date": "2025-11-18",
      "event": "https://frontierlog.ai/event/2025-11-18-gemini-3-pro",
      "benchmark": "swe-bench-verified",
      "setting": "default",
      "score": 76.2,
      "unit": "%",
      "printed": "76.2%",
      "reported_by": "google",
      "label": "Gemini 3 Pro",
      "source": "https://blog.google/innovation-and-ai/technology/developers-tools/gemini-3-developers/",
      "source_title": "Start building with Gemini 3",
      "source_date": "2025-11-18",
      "where": "benchmark table (image), row 'SWE-Bench Verified', 'Single attempt', column 'Gemini 3 Pro'",
      "own": true,
      "primary": true,
      "parallel": false,
      "notes": "single attempt",
      "human_verified": true,
      "human_verified_on": "2026-10-16",
      "source_image": "https://frontierlog.ai/sources/3fd7d3190ca58049.gif",
      "source_image_original": "https://storage.googleapis.com/gweb-uniblog-publish-prod/documents/gemini_3_table_final_HLE_Tools_on.gif"
    },
    {
      "id": "claude-opus-4-5:swe-bench-verified:default:1",
      "model": "claude-opus-4-5",
      "model_name": "Claude Opus 4.5",
      "model_date": "2025-11-24",
      "event": "https://frontierlog.ai/event/2025-11-24-claude-opus-4-5",
      "benchmark": "swe-bench-verified",
      "setting": "default",
      "score": 80.9,
      "unit": "%",
      "printed": "80.9%",
      "reported_by": "anthropic",
      "label": "Opus 4.5",
      "source": "https://www.anthropic.com/news/claude-opus-4-5",
      "source_title": "Introducing Claude Opus 4.5",
      "source_date": "2025-11-24",
      "where": "benchmark table (image), row 'Agentic coding, SWE-bench Verified', column 'Opus 4.5'",
      "own": true,
      "primary": true,
      "parallel": false,
      "notes": "no thinking budget, default effort (high), averaged over 5 trials (methodology)",
      "human_verified": true,
      "human_verified_on": "2026-10-16",
      "source_image": "https://frontierlog.ai/sources/790b0c5480216588.png",
      "source_image_original": "https://www-cdn.anthropic.com/images/4zrzovbb/website/52303b11db76017fd0c2f73c7fafa5c752515979-2600x2236.png"
    },
    {
      "id": "gpt-5-2:swe-bench-verified:default:1",
      "model": "gpt-5-2",
      "model_name": "GPT-5.2",
      "model_date": "2025-12-11",
      "event": "https://frontierlog.ai/event/2025-12-11-gpt-5-2",
      "benchmark": "swe-bench-verified",
      "setting": "default",
      "score": 80.0,
      "unit": "%",
      "printed": "80.0%",
      "reported_by": "openai",
      "label": "GPT‑5.2 Thinking",
      "source": "https://openai.com/index/introducing-gpt-5-2",
      "source_title": "Introducing GPT-5.2",
      "source_date": "2025-12-11",
      "where": "coding table, row 'SWE-bench Verified', column 'GPT-5.2 Thinking'",
      "own": true,
      "primary": true,
      "parallel": false,
      "notes": "",
      "human_verified": true,
      "human_verified_on": "2026-10-16"
    },
    {
      "id": "claude-opus-4-6:swe-bench-verified:default:1",
      "model": "claude-opus-4-6",
      "model_name": "Claude Opus 4.6",
      "model_date": "2026-02-05",
      "event": "https://frontierlog.ai/event/2026-02-05-claude-opus-4-6",
      "benchmark": "swe-bench-verified",
      "setting": "default",
      "score": 80.8,
      "unit": "%",
      "printed": "80.8%",
      "reported_by": "anthropic",
      "label": "Opus 4.6",
      "source": "https://www.anthropic.com/news/claude-opus-4-6",
      "source_title": "Claude Opus 4.6",
      "source_date": "2026-02-05",
      "where": "benchmark table (image), row 'Agentic coding, SWE-bench Verified', column 'Opus 4.6'",
      "own": true,
      "primary": true,
      "parallel": false,
      "notes": "averaged over 25 trials (footnote)",
      "human_verified": true,
      "human_verified_on": "2026-10-16",
      "source_image": "https://frontierlog.ai/sources/e6c54904fe1e4f51.png",
      "source_image_original": "https://www-cdn.anthropic.com/images/4zrzovbb/website/f9564dd2f758237bd9dbe775674c4a375aff1e8a-2600x2968.png"
    },
    {
      "id": "gemini-3-1-pro:swe-bench-verified:default:1",
      "model": "gemini-3-1-pro",
      "model_name": "Gemini 3.1 Pro",
      "model_date": "2026-02-19",
      "event": "https://frontierlog.ai/event/2026-02-19-gemini-3-1-pro",
      "benchmark": "swe-bench-verified",
      "setting": "default",
      "score": 80.6,
      "unit": "%",
      "printed": "80.6%",
      "reported_by": "google",
      "label": "Gemini 3.1 Pro, Thinking (High)",
      "source": "https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-3-1-pro/",
      "source_title": "Gemini 3.1 Pro: A smarter model for your most complex tasks",
      "source_date": "2026-02-19",
      "where": "benchmark table (image), row 'SWE-Bench Verified', 'Single attempt', column 'Gemini 3.1 Pro'",
      "own": true,
      "primary": true,
      "parallel": false,
      "notes": "single attempt; Thinking (High)",
      "human_verified": true,
      "human_verified_on": "2026-10-16",
      "source_image": "https://frontierlog.ai/sources/6b612deebae92383.gif",
      "source_image_original": "https://storage.googleapis.com/gweb-uniblog-publish-prod/original_images/gemini_3-1-pro__benchmarks.gif"
    },
    {
      "id": "claude-opus-4-7:swe-bench-verified:default:1",
      "model": "claude-opus-4-7",
      "model_name": "Claude Opus 4.7",
      "model_date": "2026-04-16",
      "event": "https://frontierlog.ai/event/2026-04-16-claude-opus-4-7",
      "benchmark": "swe-bench-verified",
      "setting": "default",
      "score": 87.6,
      "unit": "%",
      "printed": "87.6%",
      "reported_by": "anthropic",
      "label": "Opus 4.7",
      "source": "https://www.anthropic.com/news/claude-opus-4-7",
      "source_title": "Introducing Claude Opus 4.7",
      "source_date": "2026-04-16",
      "where": "benchmark table (image), row 'Agentic coding, SWE-bench Verified', column 'Opus 4.7'",
      "own": true,
      "primary": true,
      "parallel": false,
      "notes": "",
      "human_verified": true,
      "human_verified_on": "2026-10-16",
      "source_image": "https://frontierlog.ai/sources/527cd462f89b6de4.png",
      "source_image_original": "https://www-cdn.anthropic.com/images/4zrzovbb/website/d434d15757c6abac1122af483617741776d5a114-2600x2638.png"
    }
  ]
}