{
  "api_version": "1",
  "updated": "2026-10-26",
  "latest_event": "2026-10-24",
  "rights": "© Frontier Log, all rights reserved.",
  "model": {
    "slug": "openai-o3",
    "name": "OpenAI o3",
    "lab": "openai",
    "date": "2025-04-16",
    "flagship": false,
    "event": "https://frontierlog.ai/event/2025-04-16-o3-and-o4-mini",
    "releases": [],
    "variant_of": null,
    "made_by": "OpenAI",
    "url": "https://frontierlog.ai/model/openai-o3",
    "results": [
      {
        "id": "openai-o3:swe-bench-verified:default:1",
        "model": "openai-o3",
        "model_name": "OpenAI o3",
        "model_date": "2025-04-16",
        "event": "https://frontierlog.ai/event/2025-04-16-o3-and-o4-mini",
        "benchmark": "swe-bench-verified",
        "benchmark_name": "SWE-bench Verified",
        "setting": "default",
        "setting_name": "",
        "score": 69.1,
        "unit": "%",
        "printed": "69.1",
        "reported_by": "openai",
        "run_by": null,
        "label": "o3",
        "source": "https://openai.com/index/introducing-o3-and-o4-mini",
        "source_title": "Introducing OpenAI o3 and o4-mini",
        "source_date": "2025-04-16",
        "where": "Coding section, bar chart 'SWE-Bench Verified (n=477)', bar labelled 'o3' (3rd bar)",
        "own": true,
        "primary": true,
        "parallel": false,
        "notes": "high reasoning effort; 477 of the 500 verified tasks; 256K context",
        "caveat": "",
        "human_verified": true,
        "human_verified_on": "2026-10-18",
        "check": "human",
        "corroborated": true,
        "corroborated_on": "2026-10-19",
        "corroborated_by": [
          {
            "url": "https://www.anthropic.com/news/claude-4",
            "name": "Claude Code generally available",
            "title": "Introducing Claude 4",
            "lab": "anthropic",
            "date": "2025-05-22",
            "printed": "69.1%",
            "where": "Benchmarks table (HTML, rendered), row 'Agentic coding SWE-bench Verified', column 'OpenAI o3'"
          },
          {
            "url": "https://www.anthropic.com/news/claude-opus-4-1",
            "name": "Claude Opus 4.1",
            "title": "Claude Opus 4.1",
            "lab": "anthropic",
            "date": "2025-08-05",
            "printed": "69.1%",
            "where": "Benchmark table image, row 'Agentic coding SWE-bench Verified', column 'OpenAI o3'"
          },
          {
            "url": "https://www.minimax.io/news/minimax-m25",
            "name": "MiniMax M2.5",
            "title": "MiniMax M2.5: Built for Real-World Productivity.",
            "lab": "minimax",
            "date": "2026-02-12",
            "printed": "69.1%",
            "where": "'SWE-bench Verified Score Evolution' line chart image; point label 'o3 69.1%'"
          },
          {
            "url": "https://www.minimax.io/news/minimaxm1",
            "name": "MiniMax-M1",
            "title": "MiniMax-M1, the World's First Open-Source, Large-Scale, Hybrid-Attention Reasoning Model",
            "lab": "minimax",
            "date": "2025-06-16",
            "printed": "69.1",
            "where": "Full benchmark table image; row 'SWE-bench Verified'; column 'OpenAI-o3 (100K thinking)'"
          },
          {
            "url": "https://z.ai/blog/glm-4.5",
            "name": "GLM-4.5 and GLM-4.5-Air",
            "title": "GLM-4.5: Reasoning, Coding, and Agentic Abililties",
            "lab": "zai",
            "date": "2025-07-28",
            "printed": "69.1",
            "where": "Coding table, row 'SWE-bench Verified' (OpenHands v0.34.0, 100 iterations), column 'o3'"
          }
        ],
        "source_capture": "https://frontierlog.ai/captures/4ae55a93d09f6fa1.png"
      },
      {
        "id": "openai-o3:aime-2024:no-tools:1",
        "model": "openai-o3",
        "model_name": "OpenAI o3",
        "model_date": "2025-04-16",
        "event": "https://frontierlog.ai/event/2025-04-16-o3-and-o4-mini",
        "benchmark": "aime-2024",
        "benchmark_name": "AIME 2024",
        "setting": "no-tools",
        "setting_name": "No tools",
        "score": 91.6,
        "unit": "%",
        "printed": "91.6",
        "reported_by": "openai",
        "run_by": null,
        "label": "o3 (no tools)",
        "source": "https://openai.com/index/introducing-o3-and-o4-mini",
        "source_title": "Introducing OpenAI o3 and o4-mini",
        "source_date": "2025-04-16",
        "where": "'What's changed' section, bar chart 'AIME 2024 / Competition Math', bar labelled 'o3 (no tools)' (3rd bar)",
        "own": true,
        "primary": true,
        "parallel": false,
        "notes": "high reasoning effort",
        "caveat": "",
        "human_verified": true,
        "human_verified_on": "2026-10-18",
        "check": "human",
        "corroborated": true,
        "corroborated_on": "2026-10-19",
        "corroborated_by": [
          {
            "url": "https://www.minimax.io/news/minimaxm1",
            "name": "MiniMax-M1",
            "title": "MiniMax-M1, the World's First Open-Source, Large-Scale, Hybrid-Attention Reasoning Model",
            "lab": "minimax",
            "date": "2025-06-16",
            "printed": "91.6",
            "where": "Full benchmark table image; row 'AIME 2024'; column 'OpenAI-o3 (100K thinking)'"
          }
        ],
        "source_capture": "https://frontierlog.ai/captures/a94b4ebbf6b0463c.png"
      },
      {
        "id": "openai-o3:aime-2025:no-tools:1",
        "model": "openai-o3",
        "model_name": "OpenAI o3",
        "model_date": "2025-04-16",
        "event": "https://frontierlog.ai/event/2025-04-16-o3-and-o4-mini",
        "benchmark": "aime-2025",
        "benchmark_name": "AIME 2025",
        "setting": "no-tools",
        "setting_name": "No tools",
        "score": 88.9,
        "unit": "%",
        "printed": "88.9",
        "reported_by": "openai",
        "run_by": null,
        "label": "o3 (no tools)",
        "source": "https://openai.com/index/introducing-o3-and-o4-mini",
        "source_title": "Introducing OpenAI o3 and o4-mini",
        "source_date": "2025-04-16",
        "where": "'What's changed' section, bar chart 'AIME 2025 / Competition Math', bar labelled 'o3 (no tools)' (3rd bar)",
        "own": true,
        "primary": true,
        "parallel": false,
        "notes": "high reasoning effort",
        "caveat": "",
        "human_verified": true,
        "human_verified_on": "2026-10-18",
        "check": "human",
        "corroborated": true,
        "corroborated_on": "2026-10-19",
        "corroborated_by": [
          {
            "url": "https://www.minimax.io/news/minimaxm1",
            "name": "MiniMax-M1",
            "title": "MiniMax-M1, the World's First Open-Source, Large-Scale, Hybrid-Attention Reasoning Model",
            "lab": "minimax",
            "date": "2025-06-16",
            "printed": "88.9",
            "where": "Full benchmark table image; row 'AIME 2025'; column 'OpenAI-o3 (100K thinking)'"
          }
        ],
        "source_capture": "https://frontierlog.ai/captures/a94b4ebbf6b0463c.png"
      },
      {
        "id": "openai-o3:aime-2025:with-tools:1",
        "model": "openai-o3",
        "model_name": "OpenAI o3",
        "model_date": "2025-04-16",
        "event": "https://frontierlog.ai/event/2025-04-16-o3-and-o4-mini",
        "benchmark": "aime-2025",
        "benchmark_name": "AIME 2025",
        "setting": "with-tools",
        "setting_name": "With tools",
        "score": 98.4,
        "unit": "%",
        "printed": "98.4%",
        "reported_by": "openai",
        "run_by": null,
        "label": "o3",
        "source": "https://openai.com/index/introducing-o3-and-o4-mini",
        "source_title": "Introducing OpenAI o3 and o4-mini",
        "source_date": "2025-04-16",
        "where": "'What's changed' section, paragraph on o4-mini, last sentence (o3 with tool use)",
        "own": true,
        "primary": true,
        "parallel": false,
        "notes": "Python interpreter; pass@1; OpenAI says not to compare with-tools scores to models without tools",
        "caveat": "",
        "human_verified": true,
        "human_verified_on": "2026-10-18",
        "check": "human",
        "corroborated": false,
        "corroborated_on": null,
        "corroborated_by": []
      },
      {
        "id": "openai-o3:aime-2025:with-tools:2",
        "model": "openai-o3",
        "model_name": "OpenAI o3",
        "model_date": "2025-04-16",
        "event": "https://frontierlog.ai/event/2025-04-16-o3-and-o4-mini",
        "benchmark": "aime-2025",
        "benchmark_name": "AIME 2025",
        "setting": "with-tools",
        "setting_name": "With tools",
        "score": 100.0,
        "unit": "%",
        "printed": "100%",
        "reported_by": "openai",
        "run_by": null,
        "label": "o3",
        "source": "https://openai.com/index/introducing-o3-and-o4-mini",
        "source_title": "Introducing OpenAI o3 and o4-mini",
        "source_date": "2025-04-16",
        "where": "'What's changed' section, paragraph on o4-mini, last sentence (o3 with tool use), second figure in the parenthesis",
        "own": true,
        "primary": false,
        "parallel": true,
        "notes": "Python interpreter; majority vote over 8 samples",
        "caveat": "",
        "human_verified": true,
        "human_verified_on": "2026-10-18",
        "check": "human",
        "corroborated": false,
        "corroborated_on": null,
        "corroborated_by": []
      },
      {
        "id": "openai-o3:gpqa-diamond:default:1",
        "model": "openai-o3",
        "model_name": "OpenAI o3",
        "model_date": "2025-04-16",
        "event": "https://frontierlog.ai/event/2025-04-16-o3-and-o4-mini",
        "benchmark": "gpqa-diamond",
        "benchmark_name": "GPQA Diamond",
        "setting": "default",
        "setting_name": "",
        "score": 83.3,
        "unit": "%",
        "printed": "83.3",
        "reported_by": "openai",
        "run_by": null,
        "label": "o3 (no tools)",
        "source": "https://openai.com/index/introducing-o3-and-o4-mini",
        "source_title": "Introducing OpenAI o3 and o4-mini",
        "source_date": "2025-04-16",
        "where": "'What's changed' section, bar chart 'GPQA Diamond / PhD-Level Science Questions', 3rd bar 'o3 (no tools)' (page screenshot (job 20879 capture), part 02)",
        "own": true,
        "primary": true,
        "parallel": false,
        "notes": "high reasoning effort; no tools",
        "caveat": "",
        "human_verified": false,
        "human_verified_on": null,
        "check": "corroborated",
        "corroborated": true,
        "corroborated_on": "2026-10-19",
        "corroborated_by": [
          {
            "url": "https://openai.com/index/introducing-gpt-5-for-developers",
            "name": "OpenAI post",
            "title": "Introducing GPT-5 for developers",
            "lab": "openai",
            "date": "2025-08-07",
            "printed": "83.3%",
            "where": "Intelligence table, row 'GPQA diamond (no tools)', column 'OpenAI o3 (high)'"
          },
          {
            "url": "https://openai.com/index/introducing-gpt-oss",
            "name": "gpt-oss-120b and gpt-oss-20b",
            "title": "Introducing gpt-oss",
            "lab": "openai",
            "date": "2025-08-05",
            "printed": "83.3",
            "where": "GPQA Diamond (without tools) chart, bar 'o3'"
          }
        ],
        "source_capture": "https://frontierlog.ai/captures/392f30ae76e5713f.png"
      },
      {
        "id": "openai-o3:hle:no-tools:1",
        "model": "openai-o3",
        "model_name": "OpenAI o3",
        "model_date": "2025-04-16",
        "event": "https://frontierlog.ai/event/2025-04-16-o3-and-o4-mini",
        "benchmark": "hle",
        "benchmark_name": "Humanity's Last Exam",
        "setting": "no-tools",
        "setting_name": "No tools",
        "score": 20.32,
        "unit": "%",
        "printed": "20.32",
        "reported_by": "openai",
        "run_by": null,
        "label": "o3 (no tools)",
        "source": "https://openai.com/index/introducing-o3-and-o4-mini",
        "source_title": "Introducing OpenAI o3 and o4-mini",
        "source_date": "2025-04-16",
        "where": "'What's changed' section, bar chart 'Humanity's Last Exam', 3rd bar 'o3 (no tools)' (page screenshot (job 20879 capture), part 02)",
        "own": true,
        "primary": true,
        "parallel": false,
        "notes": "high reasoning effort",
        "caveat": "",
        "human_verified": false,
        "human_verified_on": null,
        "check": "corroborated",
        "corroborated": true,
        "corroborated_on": "2026-10-19",
        "corroborated_by": [
          {
            "url": "https://blog.google/products-and-platforms/products/gemini/gemini-2-5-pro-latest-preview/",
            "name": "Google post",
            "title": "Try the latest Gemini 2.5 Pro before general availability.",
            "lab": "google",
            "date": "2025-06-05",
            "printed": "20.3%",
            "where": "Benchmark table image (Gemini 2.5 Pro Preview 06-05); row 'Humanity's Last Exam (no tools)'; column 'OpenAI o3 High'. Footnote: non-Gemini from providers' self-reported numbers unless noted; 'For OpenAI models high level of reasoning is shown where results are available (except for GPQA, AIME 2025, SWE-Bench, FACTS, MMMU)'; where provider numbers unavailable, HLE from agi.safe.ai / Scale leaderboard."
          },
          {
            "url": "https://openai.com/index/introducing-chatgpt-agent",
            "name": "ChatGPT agent",
            "title": "Introducing ChatGPT agent",
            "lab": "openai",
            "date": "2025-07-17",
            "printed": "20.3%",
            "where": "Humanity's Last Exam (Full Set) chart, accuracy pass@1, bar 'OpenAI o3 (no tools)'"
          },
          {
            "url": "https://storage.googleapis.com/deepmind-media/gemini/gemini_v2_5_report.pdf",
            "name": "Gemini 2.5 GA and 2.5 Flash-Lite card",
            "title": "Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilities",
            "lab": "google",
            "date": "2025-06-17",
            "printed": "20.3%",
            "where": "Gemini 2.5 technical report, Table 4 (p.14); row 'Humanity's Last Exam, no tools'; column 'o3 high'. Report text: non-Gemini results from providers' self-reported numbers unless mentioned; 'Results for Humanity's Last Exam results are sourced from Scale's leaderboard'. (sha256 of PDF d10cf3e9...)"
          }
        ],
        "source_capture": "https://frontierlog.ai/captures/392f30ae76e5713f.png"
      },
      {
        "id": "openai-o3:hle:with-tools:1",
        "model": "openai-o3",
        "model_name": "OpenAI o3",
        "model_date": "2025-04-16",
        "event": "https://frontierlog.ai/event/2025-04-16-o3-and-o4-mini",
        "benchmark": "hle",
        "benchmark_name": "Humanity's Last Exam",
        "setting": "with-tools",
        "setting_name": "With tools",
        "score": 24.9,
        "unit": "%",
        "printed": "24.90",
        "reported_by": "openai",
        "run_by": null,
        "label": "o3 (python + browsing** tools)",
        "source": "https://openai.com/index/introducing-o3-and-o4-mini",
        "source_title": "Introducing OpenAI o3 and o4-mini",
        "source_date": "2025-04-16",
        "where": "'What's changed' section, bar chart 'Humanity's Last Exam', 4th bar 'o3 (python + browsing** tools)' (page screenshot (job 20879 capture), part 02)",
        "own": true,
        "primary": true,
        "parallel": false,
        "notes": "high reasoning effort; Python and browsing; known cheating domains blocked; attempts a monitor flags count as wrong",
        "caveat": "",
        "human_verified": false,
        "human_verified_on": null,
        "check": "corroborated",
        "corroborated": true,
        "corroborated_on": "2026-10-19",
        "corroborated_by": [
          {
            "url": "https://openai.com/index/introducing-gpt-oss",
            "name": "gpt-oss-120b and gpt-oss-20b",
            "title": "Introducing gpt-oss",
            "lab": "openai",
            "date": "2025-08-05",
            "printed": "24.9",
            "where": "Humanity's Last Exam chart, bar 'o3 (with tools)'"
          },
          {
            "url": "https://openai.com/index/introducing-chatgpt-agent",
            "name": "ChatGPT agent",
            "title": "Introducing ChatGPT agent",
            "lab": "openai",
            "date": "2025-07-17",
            "printed": "24.9%",
            "where": "Humanity's Last Exam (Full Set) chart, accuracy pass@1, bar 'OpenAI o3 (python + browsing)'"
          }
        ],
        "source_capture": "https://frontierlog.ai/captures/392f30ae76e5713f.png"
      },
      {
        "id": "openai-o3:browsecomp:default:1",
        "model": "openai-o3",
        "model_name": "OpenAI o3",
        "model_date": "2025-04-16",
        "event": "https://frontierlog.ai/event/2025-04-16-o3-and-o4-mini",
        "benchmark": "browsecomp",
        "benchmark_name": "BrowseComp",
        "setting": "default",
        "setting_name": "",
        "score": 49.7,
        "unit": "%",
        "printed": "49.7",
        "reported_by": "openai",
        "run_by": null,
        "label": "o3 with python + browsing",
        "source": "https://openai.com/index/introducing-o3-and-o4-mini",
        "source_title": "Introducing OpenAI o3 and o4-mini",
        "source_date": "2025-04-16",
        "where": "'Instruction following and agentic tool use' section, bar chart 'BrowseComp / Agentic Browsing', 2nd bar 'o3 with python + browsing*' (page screenshot (job 20879 capture), part 04)",
        "own": true,
        "primary": true,
        "parallel": false,
        "notes": "high reasoning effort; Python and browsing",
        "caveat": "",
        "human_verified": false,
        "human_verified_on": null,
        "check": "corroborated",
        "corroborated": true,
        "corroborated_on": "2026-10-19",
        "corroborated_by": [
          {
            "url": "https://openai.com/index/introducing-gpt-5",
            "name": "GPT-5",
            "title": "Introducing GPT-5",
            "lab": "openai",
            "date": "2025-08-07",
            "printed": "49.7%",
            "where": "BrowseComp chart (Agentic search & browsing, accuracy pass@1), bar 'OpenAI o3'"
          },
          {
            "url": "https://openai.com/index/introducing-chatgpt-agent",
            "name": "ChatGPT agent",
            "title": "Introducing ChatGPT agent",
            "lab": "openai",
            "date": "2025-07-17",
            "printed": "49.7%",
            "where": "BrowseComp chart (Agentic Browsing, accuracy pass@1), bar 'OpenAI o3'"
          },
          {
            "url": "https://z.ai/blog/glm-4.5",
            "name": "GLM-4.5 and GLM-4.5-Air",
            "title": "GLM-4.5: Reasoning, Coding, and Agentic Abililties",
            "lab": "zai",
            "date": "2025-07-28",
            "printed": "49.7",
            "where": "Agentic table, row 'BrowseComp', column 'o3'. Tool setup/effort for o3 not stated; no provenance footnote."
          }
        ],
        "source_capture": "https://frontierlog.ai/captures/dc46417a14f5c000.png"
      },
      {
        "id": "openai-o3:hle:no-tools:2",
        "model": "openai-o3",
        "model_name": "OpenAI o3",
        "model_date": "2025-04-16",
        "event": "https://frontierlog.ai/event/2025-04-16-o3-and-o4-mini",
        "benchmark": "hle",
        "benchmark_name": "Humanity's Last Exam",
        "setting": "no-tools",
        "setting_name": "No tools",
        "score": 14.7,
        "unit": "%",
        "printed": "14.7%",
        "reported_by": "openai",
        "run_by": null,
        "label": "OpenAI o3 (no tools)",
        "source": "https://openai.com/index/introducing-gpt-5",
        "source_title": "Introducing GPT-5",
        "source_date": "2025-08-07",
        "where": "'Evaluations' section, bar chart 'Humanity's Last Exam (Full Set)*', 9th bar 'OpenAI o3 (no tools)' (page screenshot (job 20879 capture), part 05)",
        "own": false,
        "primary": false,
        "parallel": false,
        "notes": "rerun on a later version of HLE; full set; effort not stated",
        "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/04f515e270a84d5f.png"
      },
      {
        "id": "openai-o3:hle:with-tools:2",
        "model": "openai-o3",
        "model_name": "OpenAI o3",
        "model_date": "2025-04-16",
        "event": "https://frontierlog.ai/event/2025-04-16-o3-and-o4-mini",
        "benchmark": "hle",
        "benchmark_name": "Humanity's Last Exam",
        "setting": "with-tools",
        "setting_name": "With tools",
        "score": 24.3,
        "unit": "%",
        "printed": "24.3%",
        "reported_by": "openai",
        "run_by": null,
        "label": "OpenAI o3 (python + browser)",
        "source": "https://openai.com/index/introducing-gpt-5",
        "source_title": "Introducing GPT-5",
        "source_date": "2025-08-07",
        "where": "'Evaluations' section, bar chart 'Humanity's Last Exam (Full Set)*', 8th bar 'OpenAI o3 (python + browser)' (page screenshot (job 20879 capture), part 05)",
        "own": false,
        "primary": false,
        "parallel": false,
        "notes": "rerun on a later version of HLE; full set; Python and browser",
        "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/04f515e270a84d5f.png"
      },
      {
        "id": "openai-o3:hle:no-tools:3",
        "model": "openai-o3",
        "model_name": "OpenAI o3",
        "model_date": "2025-04-16",
        "event": "https://frontierlog.ai/event/2025-04-16-o3-and-o4-mini",
        "benchmark": "hle",
        "benchmark_name": "Humanity's Last Exam",
        "setting": "no-tools",
        "setting_name": "No tools",
        "score": 20.2,
        "unit": "%",
        "printed": "20.2%",
        "reported_by": "openai",
        "run_by": null,
        "label": "OpenAI o3 (high)",
        "source": "https://openai.com/index/introducing-gpt-5-for-developers",
        "source_title": "Introducing GPT-5 for developers",
        "source_date": "2025-08-07",
        "where": "intelligence table (columns GPT-5 (high) ... GPT-4.1 nano), row 'HLE[1] (no tools)', column 'OpenAI o3 (high)'",
        "own": false,
        "primary": false,
        "parallel": false,
        "notes": "high reasoning effort",
        "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/c0acb8e4e87e723e.png"
      }
    ]
  }
}