{
  "api_version": "1",
  "updated": "2026-10-26",
  "latest_event": "2026-10-26",
  "rights": "© Frontier Log, all rights reserved.",
  "model": {
    "slug": "qwen3-235b-a22b-thinking-2507",
    "name": "Qwen3-235B-A22B-Thinking-2507",
    "lab": "alibaba",
    "date": "2025-07-25",
    "source": "https://www.alizila.com/alibaba-unveils-new-qwen3-models-for-coding-complexing-reasoning-and-machine-translation/",
    "source_title": "Alibaba Unveils New Qwen3 Models for Coding, Complexing Reasoning and Machine Translation",
    "flagship": false,
    "event": "https://frontierlog.ai/event/2025-07-25-qwen3-235b-a22b-thinking-2507",
    "releases": [],
    "variant_of": null,
    "made_by": "Alibaba",
    "url": "https://frontierlog.ai/model/qwen3-235b-a22b-thinking-2507",
    "results": [
      {
        "id": "qwen3-235b-a22b-thinking-2507:gpqa-diamond:default:1",
        "model": "qwen3-235b-a22b-thinking-2507",
        "model_name": "Qwen3-235B-A22B-Thinking-2507",
        "model_date": "2025-07-25",
        "event": "https://frontierlog.ai/event/2025-07-25-qwen3-235b-a22b-thinking-2507",
        "benchmark": "gpqa-diamond",
        "benchmark_name": "GPQA Diamond",
        "setting": "default",
        "setting_name": "",
        "score": 81.1,
        "unit": "%",
        "printed": "81.1",
        "reported_by": "alibaba",
        "run_by": null,
        "label": "Qwen3-235B-A22B-Thinking-2507",
        "source": "https://www.alizila.com/alibaba-unveils-new-qwen3-models-for-coding-complexing-reasoning-and-machine-translation/",
        "source_title": "Alibaba Unveils New Qwen3 Models for Coding, Complexing Reasoning and Machine Translation",
        "source_date": "2025-07-25",
        "where": "bar chart image in the 'Qwen3-235B: Open-sourced LLM Delivering Massive Performance' section, group 'GPQA', first bar (legend 'Qwen3-235B-A22B-Thinking-2507') value label",
        "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://huggingface.co/Qwen/Qwen3-235B-A22B-Thinking-2507/blob/4cd68849b2a0fadb84866c703b133aa8c8636130/README.md",
            "name": "Qwen3-235B-A22B-Thinking-2507 card",
            "title": "Qwen/Qwen3-235B-A22B-Thinking-2507 model card (revision of 2025-07-25)",
            "lab": "alibaba",
            "date": "2025-07-25",
            "printed": "81.1",
            "where": "Qwen3-235B-A22B-Thinking-2507 HF model card at revision 4cd68849, Performance table, row 'GPQA', column 'Qwen3-235B-A22B-Thinking-2507'; the card's chart image also labels 81.1"
          },
          {
            "url": "https://qwen.ai/blog?id=qwen3-max",
            "name": "Qwen3-Max",
            "title": "Qwen3-Max: Just Scale it",
            "lab": "alibaba",
            "date": "2025-09-24",
            "printed": "81.1",
            "where": "Qwen3-Max blog, Qwen3-Max-Thinking chart, value label on 'GPQA' bar for 'Qwen3-235B-A22B-Thinking-2507'"
          }
        ],
        "source_image": "https://frontierlog.ai/sources/3ebde273acb680b6.png",
        "source_image_original": "https://alizila.oss-us-west-1.aliyuncs.com/uploads/2025/07/0CA37A90-CD03-4037-AF89-F1548BB1C156.png"
      },
      {
        "id": "qwen3-235b-a22b-thinking-2507:aime-2025:no-tools:1",
        "model": "qwen3-235b-a22b-thinking-2507",
        "model_name": "Qwen3-235B-A22B-Thinking-2507",
        "model_date": "2025-07-25",
        "event": "https://frontierlog.ai/event/2025-07-25-qwen3-235b-a22b-thinking-2507",
        "benchmark": "aime-2025",
        "benchmark_name": "AIME 2025",
        "setting": "no-tools",
        "setting_name": "No tools",
        "score": 92.3,
        "unit": "%",
        "printed": "92.3",
        "reported_by": "alibaba",
        "run_by": null,
        "label": "Qwen3-235B-A22B-Thinking-2507",
        "source": "https://www.alizila.com/alibaba-unveils-new-qwen3-models-for-coding-complexing-reasoning-and-machine-translation/",
        "source_title": "Alibaba Unveils New Qwen3 Models for Coding, Complexing Reasoning and Machine Translation",
        "source_date": "2025-07-25",
        "where": "bar chart image in the 'Qwen3-235B' section, group 'AIME25', first bar (legend 'Qwen3-235B-A22B-Thinking-2507') value label",
        "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://huggingface.co/Qwen/Qwen3-235B-A22B-Thinking-2507/blob/4cd68849b2a0fadb84866c703b133aa8c8636130/README.md",
            "name": "Qwen3-235B-A22B-Thinking-2507 card",
            "title": "Qwen/Qwen3-235B-A22B-Thinking-2507 model card (revision of 2025-07-25)",
            "lab": "alibaba",
            "date": "2025-07-25",
            "printed": "92.3",
            "where": "Qwen3-235B-A22B-Thinking-2507 HF model card at revision 4cd68849, Performance table, row 'AIME25', column 'Qwen3-235B-A22B-Thinking-2507'"
          },
          {
            "url": "https://qwen.ai/blog?id=qwen3-max",
            "name": "Qwen3-Max",
            "title": "Qwen3-Max: Just Scale it",
            "lab": "alibaba",
            "date": "2025-09-24",
            "printed": "92.3",
            "where": "Qwen3-Max blog, Qwen3-Max-Thinking chart, value label on 'AIME25' bar for 'Qwen3-235B-A22B-Thinking-2507' (chart compares against 'w/ Python' heavy models, but the 235B bar carries no tool label)"
          }
        ],
        "source_image": "https://frontierlog.ai/sources/3ebde273acb680b6.png",
        "source_image_original": "https://alizila.oss-us-west-1.aliyuncs.com/uploads/2025/07/0CA37A90-CD03-4037-AF89-F1548BB1C156.png"
      },
      {
        "id": "qwen3-235b-a22b-thinking-2507:livecodebench:default:1",
        "model": "qwen3-235b-a22b-thinking-2507",
        "model_name": "Qwen3-235B-A22B-Thinking-2507",
        "model_date": "2025-07-25",
        "event": "https://frontierlog.ai/event/2025-07-25-qwen3-235b-a22b-thinking-2507",
        "benchmark": "livecodebench",
        "benchmark_name": "LiveCodeBench",
        "setting": "default",
        "setting_name": "",
        "score": 74.1,
        "unit": "%",
        "printed": "74.1",
        "reported_by": "alibaba",
        "run_by": null,
        "label": "Qwen3-235B-A22B-Thinking-2507",
        "source": "https://www.alizila.com/alibaba-unveils-new-qwen3-models-for-coding-complexing-reasoning-and-machine-translation/",
        "source_title": "Alibaba Unveils New Qwen3 Models for Coding, Complexing Reasoning and Machine Translation",
        "source_date": "2025-07-25",
        "where": "bar chart image in the 'Qwen3-235B' section, group 'LiveCodeBench v6 (25.02-25.05)', first bar value label",
        "own": true,
        "primary": true,
        "parallel": false,
        "notes": "LiveCodeBench v6, problems from Feb to May 2025",
        "caveat": "",
        "human_verified": false,
        "human_verified_on": null,
        "check": "corroborated",
        "corroborated": true,
        "corroborated_on": "2026-10-26",
        "corroborated_by": [
          {
            "url": "https://huggingface.co/Qwen/Qwen3-235B-A22B-Thinking-2507/blob/4cd68849b2a0fadb84866c703b133aa8c8636130/README.md",
            "name": "Qwen3-235B-A22B-Thinking-2507 card",
            "title": "Qwen/Qwen3-235B-A22B-Thinking-2507 model card (revision of 2025-07-25)",
            "lab": "alibaba",
            "date": "2025-07-25",
            "printed": "74.1",
            "where": "Qwen3-235B-A22B-Thinking-2507 HF model card at revision 4cd68849, Performance table, row 'LiveCodeBench v6 (25.02-25.05)', column 'Qwen3-235B-A22B-Thinking-2507'"
          }
        ],
        "source_image": "https://frontierlog.ai/sources/3ebde273acb680b6.png",
        "source_image_original": "https://alizila.oss-us-west-1.aliyuncs.com/uploads/2025/07/0CA37A90-CD03-4037-AF89-F1548BB1C156.png"
      },
      {
        "id": "qwen3-235b-a22b-thinking-2507:hle:no-tools:1",
        "model": "qwen3-235b-a22b-thinking-2507",
        "model_name": "Qwen3-235B-A22B-Thinking-2507",
        "model_date": "2025-07-25",
        "event": "https://frontierlog.ai/event/2025-07-25-qwen3-235b-a22b-thinking-2507",
        "benchmark": "hle",
        "benchmark_name": "Humanity's Last Exam",
        "setting": "no-tools",
        "setting_name": "No tools",
        "score": 18.2,
        "unit": "%",
        "printed": "18.2",
        "reported_by": "alibaba",
        "run_by": null,
        "label": "Qwen3-235B-A22B-Thinking-2507",
        "source": "https://www.alizila.com/alibaba-unveils-new-qwen3-models-for-coding-complexing-reasoning-and-machine-translation/",
        "source_title": "Alibaba Unveils New Qwen3 Models for Coding, Complexing Reasoning and Machine Translation",
        "source_date": "2025-07-25",
        "where": "bar chart image in the 'Qwen3-235B' section, group 'HLE', first bar (legend 'Qwen3-235B-A22B-Thinking-2507') value label",
        "own": true,
        "primary": true,
        "parallel": false,
        "notes": "text-only subset",
        "caveat": "",
        "human_verified": false,
        "human_verified_on": null,
        "check": "corroborated",
        "corroborated": true,
        "corroborated_on": "2026-10-26",
        "corroborated_by": [
          {
            "url": "https://huggingface.co/Qwen/Qwen3-235B-A22B-Thinking-2507/blob/4cd68849b2a0fadb84866c703b133aa8c8636130/README.md",
            "name": "Qwen3-235B-A22B-Thinking-2507 card",
            "title": "Qwen/Qwen3-235B-A22B-Thinking-2507 model card (revision of 2025-07-25)",
            "lab": "alibaba",
            "date": "2025-07-25",
            "printed": "18.2",
            "where": "Qwen3-235B-A22B-Thinking-2507 HF model card at revision 4cd68849, Performance table, row 'HLE', column 'Qwen3-235B-A22B-Thinking-2507'; printed '18.2#', footnote #: text-only subset"
          }
        ],
        "source_image": "https://frontierlog.ai/sources/3ebde273acb680b6.png",
        "source_image_original": "https://alizila.oss-us-west-1.aliyuncs.com/uploads/2025/07/0CA37A90-CD03-4037-AF89-F1548BB1C156.png"
      },
      {
        "id": "qwen3-235b-a22b-thinking-2507:mmlu-pro:default:1",
        "model": "qwen3-235b-a22b-thinking-2507",
        "model_name": "Qwen3-235B-A22B-Thinking-2507",
        "model_date": "2025-07-25",
        "event": "https://frontierlog.ai/event/2025-07-25-qwen3-235b-a22b-thinking-2507",
        "benchmark": "mmlu-pro",
        "benchmark_name": "MMLU-Pro",
        "setting": "default",
        "setting_name": "",
        "score": 84.4,
        "unit": "%",
        "printed": "84.4",
        "reported_by": "alibaba",
        "run_by": null,
        "label": "Qwen3-235B-A22B-Thinking-2507",
        "source": "https://huggingface.co/Qwen/Qwen3-235B-A22B-Thinking-2507/blob/4cd68849b2a0fadb84866c703b133aa8c8636130/README.md",
        "source_title": "Qwen/Qwen3-235B-A22B-Thinking-2507 model card (revision of 2025-07-25)",
        "source_date": "2025-07-25",
        "where": "model card 'Performance' table, Knowledge block, row 'MMLU-Pro', column 'Qwen3-235B-A22B-Thinking-2507'",
        "own": true,
        "primary": true,
        "parallel": false,
        "notes": "up to 32,768 output tokens",
        "caveat": "",
        "human_verified": true,
        "human_verified_on": "2026-10-26",
        "check": "human",
        "corroborated": false,
        "corroborated_on": null,
        "corroborated_by": [],
        "source_capture": "https://frontierlog.ai/captures/3d049bc43f1609a6.png"
      },
      {
        "id": "qwen3-235b-a22b-thinking-2507:hmmt-2025:no-tools:1",
        "model": "qwen3-235b-a22b-thinking-2507",
        "model_name": "Qwen3-235B-A22B-Thinking-2507",
        "model_date": "2025-07-25",
        "event": "https://frontierlog.ai/event/2025-07-25-qwen3-235b-a22b-thinking-2507",
        "benchmark": "hmmt-2025",
        "benchmark_name": "HMMT February 2025",
        "setting": "no-tools",
        "setting_name": "No tools",
        "score": 83.9,
        "unit": "%",
        "printed": "83.9",
        "reported_by": "alibaba",
        "run_by": null,
        "label": "Qwen3-235B-A22B-Thinking-2507",
        "source": "https://huggingface.co/Qwen/Qwen3-235B-A22B-Thinking-2507/blob/4cd68849b2a0fadb84866c703b133aa8c8636130/README.md",
        "source_title": "Qwen/Qwen3-235B-A22B-Thinking-2507 model card (revision of 2025-07-25)",
        "source_date": "2025-07-25",
        "where": "model card 'Performance' table, Reasoning block, row 'HMMT25', column 'Qwen3-235B-A22B-Thinking-2507'",
        "own": true,
        "primary": true,
        "parallel": false,
        "notes": "up to 81,920 output tokens",
        "caveat": "",
        "human_verified": false,
        "human_verified_on": null,
        "check": "corroborated",
        "corroborated": true,
        "corroborated_on": "2026-10-26",
        "corroborated_by": [
          {
            "url": "https://qwen.ai/blog?id=qwen3-max",
            "name": "Qwen3-Max",
            "title": "Qwen3-Max: Just Scale it",
            "lab": "alibaba",
            "date": "2025-09-24",
            "printed": "83.9",
            "where": "Qwen3-Max blog, Qwen3-Max-Thinking chart, value label on 'HMMT25' bar for 'Qwen3-235B-A22B-Thinking-2507'"
          }
        ],
        "source_capture": "https://frontierlog.ai/captures/3d049bc43f1609a6.png"
      },
      {
        "id": "qwen3-235b-a22b-thinking-2507:tau2-bench:retail:1",
        "model": "qwen3-235b-a22b-thinking-2507",
        "model_name": "Qwen3-235B-A22B-Thinking-2507",
        "model_date": "2025-07-25",
        "event": "https://frontierlog.ai/event/2025-07-25-qwen3-235b-a22b-thinking-2507",
        "benchmark": "tau2-bench",
        "benchmark_name": "τ²-bench",
        "setting": "retail",
        "setting_name": "Retail",
        "score": 71.9,
        "unit": "%",
        "printed": "71.9",
        "reported_by": "alibaba",
        "run_by": null,
        "label": "Qwen3-235B-A22B-Thinking-2507",
        "source": "https://huggingface.co/Qwen/Qwen3-235B-A22B-Thinking-2507/blob/4cd68849b2a0fadb84866c703b133aa8c8636130/README.md",
        "source_title": "Qwen/Qwen3-235B-A22B-Thinking-2507 model card (revision of 2025-07-25)",
        "source_date": "2025-07-25",
        "where": "model card 'Performance' table, Agent block, row 'TAU2-Retail', column 'Qwen3-235B-A22B-Thinking-2507'",
        "own": true,
        "primary": true,
        "parallel": false,
        "notes": "up to 32,768 output tokens",
        "caveat": "",
        "human_verified": true,
        "human_verified_on": "2026-10-26",
        "check": "human",
        "corroborated": false,
        "corroborated_on": null,
        "corroborated_by": [],
        "source_capture": "https://frontierlog.ai/captures/3d049bc43f1609a6.png"
      },
      {
        "id": "qwen3-235b-a22b-thinking-2507:tau2-bench:airline:1",
        "model": "qwen3-235b-a22b-thinking-2507",
        "model_name": "Qwen3-235B-A22B-Thinking-2507",
        "model_date": "2025-07-25",
        "event": "https://frontierlog.ai/event/2025-07-25-qwen3-235b-a22b-thinking-2507",
        "benchmark": "tau2-bench",
        "benchmark_name": "τ²-bench",
        "setting": "airline",
        "setting_name": "Airline",
        "score": 58.0,
        "unit": "%",
        "printed": "58.0",
        "reported_by": "alibaba",
        "run_by": null,
        "label": "Qwen3-235B-A22B-Thinking-2507",
        "source": "https://huggingface.co/Qwen/Qwen3-235B-A22B-Thinking-2507/blob/4cd68849b2a0fadb84866c703b133aa8c8636130/README.md",
        "source_title": "Qwen/Qwen3-235B-A22B-Thinking-2507 model card (revision of 2025-07-25)",
        "source_date": "2025-07-25",
        "where": "model card 'Performance' table, Agent block, row 'TAU2-Airline', column 'Qwen3-235B-A22B-Thinking-2507'",
        "own": true,
        "primary": true,
        "parallel": false,
        "notes": "up to 32,768 output tokens",
        "caveat": "",
        "human_verified": true,
        "human_verified_on": "2026-10-26",
        "check": "human",
        "corroborated": false,
        "corroborated_on": null,
        "corroborated_by": [],
        "source_capture": "https://frontierlog.ai/captures/3d049bc43f1609a6.png"
      },
      {
        "id": "qwen3-235b-a22b-thinking-2507:tau2-bench:telecom:1",
        "model": "qwen3-235b-a22b-thinking-2507",
        "model_name": "Qwen3-235B-A22B-Thinking-2507",
        "model_date": "2025-07-25",
        "event": "https://frontierlog.ai/event/2025-07-25-qwen3-235b-a22b-thinking-2507",
        "benchmark": "tau2-bench",
        "benchmark_name": "τ²-bench",
        "setting": "telecom",
        "setting_name": "Telecom",
        "score": 45.6,
        "unit": "%",
        "printed": "45.6",
        "reported_by": "alibaba",
        "run_by": null,
        "label": "Qwen3-235B-A22B-Thinking-2507",
        "source": "https://huggingface.co/Qwen/Qwen3-235B-A22B-Thinking-2507/blob/4cd68849b2a0fadb84866c703b133aa8c8636130/README.md",
        "source_title": "Qwen/Qwen3-235B-A22B-Thinking-2507 model card (revision of 2025-07-25)",
        "source_date": "2025-07-25",
        "where": "model card 'Performance' table, Agent block, row 'TAU2-Telecom', column 'Qwen3-235B-A22B-Thinking-2507'",
        "own": true,
        "primary": true,
        "parallel": false,
        "notes": "up to 32,768 output tokens",
        "caveat": "",
        "human_verified": true,
        "human_verified_on": "2026-10-26",
        "check": "human",
        "corroborated": false,
        "corroborated_on": null,
        "corroborated_by": [],
        "source_capture": "https://frontierlog.ai/captures/3d049bc43f1609a6.png"
      }
    ]
  }
}