{
  "schema_version": "2.0",
  "record_type": "dataset",
  "tier": "verified",
  "slug": "medbench",
  "name": "MedBench（上海人工智能实验室）",
  "summary": {
    "zh": "许可证声明：无法核实（NOASSERTION）；本站处理：待确认，暂只链接原出处。许可证无法核实：未找到数据许可；仓库 404，平台需登录；确认前只链接原出处。",
    "en": "License: unverifiable (NOASSERTION); Mdexa hosting: pending. License unverifiable: Data license not found; repositories returned 404 and the platform requires login; link only until resolved."
  },
  "systems": [
    "modern-medicine"
  ],
  "domains": [
    "clinical-medicine"
  ],
  "content_types": [
    "benchmark-exam"
  ],
  "content_description": {
    "en": "Benchmark (multi-specialty comprehensive evaluation)",
    "zh": "评测集（多专科综合评测）"
  },
  "languages": [
    "zh"
  ],
  "modalities": [
    "text"
  ],
  "size": {
    "raw": "v1：300,901题、43个临床专科（搜索摘要）；v4：70万+专家编制任务（论文称）；5.0 称近100万条试题、开源7万+条（来自素材）",
    "measures": [
      {
        "value": 300901,
        "unit": "questions",
        "qualifier": "claimed",
        "scope": "v1 (search snippet)"
      },
      {
        "value": 43,
        "unit": "categories",
        "qualifier": "claimed",
        "scope": "v1 clinical specialties"
      },
      {
        "value": 700000,
        "unit": "tasks",
        "qualifier": "at-least",
        "scope": "v4 (paper)"
      },
      {
        "value": 1000000,
        "unit": "questions",
        "qualifier": "approx",
        "scope": "v5.0 (source material)"
      },
      {
        "value": 70000,
        "unit": "questions",
        "qualifier": "at-least",
        "scope": "v5.0 open-sourced (source material)"
      }
    ]
  },
  "publisher": "上海人工智能实验室 / OpenMedLab",
  "year": {
    "value": 2024,
    "raw": "v1 论文2024；v4 论文2025（arXiv 2511.14439）"
  },
  "expert_review": {
    "en": "The paper says reviewed by clinicians from 500+ institutions (v4, from source material)",
    "zh": "论文称500+机构临床医生审核（v4，来自素材）"
  },
  "license": {
    "legacy_status": "pending",
    "spdx": "NOASSERTION",
    "commercial_use": "unverified",
    "derivatives": "unverified",
    "redistribution": "unverified",
    "rationale": {
      "en": "Data license not found; repositories returned 404 and the platform requires login",
      "zh": "未找到数据许可；仓库 404，平台需登录"
    },
    "evidence": [
      {
        "location": "search-snippet",
        "location_detail": {
          "en": "Search results showed no data license [Note: No URL recorded, so no snapshot exists]",
          "zh": "搜索结果未显示数据许可［备注：未记录网址，无快照］"
        },
        "excerpt": null,
        "spdx": null,
        "url": null,
        "verified_on": null,
        "snapshot_id": null,
        "observed": "search-snippet",
        "secondhand": true
      },
      {
        "location": "paper",
        "location_detail": {
          "en": "SciOpen version of the paper is CC BY 4.0, covering the paper only [Note: No URL recorded, so no snapshot exists]",
          "zh": "SciOpen 版论文为 CC BY 4.0，仅覆盖论文［备注：未记录网址，无快照］"
        },
        "excerpt": null,
        "spdx": "CC-BY-4.0",
        "url": null,
        "verified_on": null,
        "snapshot_id": null,
        "observed": "secondary-source",
        "secondhand": true
      }
    ]
  },
  "license_status": "unverifiable",
  "hosting": "pending",
  "hosting_reason": {
    "zh": "许可证无法核实：未找到数据许可；仓库 404，平台需登录；确认前只链接原出处。",
    "en": "License unverifiable: Data license not found; repositories returned 404 and the platform requires login; link only until resolved."
  },
  "signoff": null,
  "risk_flags": [],
  "provenance": {
    "source_description": {
      "en": "The paper says expert-authored with clinician review",
      "zh": "论文称专家编制与临床医生审核"
    },
    "upstream": []
  },
  "known_issues": [
    {
      "en": "The evaluation platform requires login to obtain data; github.com/open-compass/MedBench and OpenMEDLab/MedBench both returned 404 this time",
      "zh": "评测平台需登录获取数据，github.com/open-compass/MedBench 与 OpenMEDLab/MedBench 两个地址本次均返回404"
    },
    {
      "en": "Only part of the full set is open-sourced",
      "zh": "开源比例仅为全量的一部分"
    }
  ],
  "preflight": {
    "status": "not-run"
  },
  "access": {
    "mdexa_download": null,
    "external_urls": [
      "https://arxiv.org/pdf/2407.10990",
      "https://arxiv.org/abs/2511.14439"
    ],
    "notes": null
  },
  "citation": {
    "text": null,
    "bibtex": null,
    "references": [
      "arXiv:2407.10990",
      "arXiv:2511.14439"
    ]
  },
  "verification_level": "unverified",
  "last_verified": "2026-11-12",
  "tags": [
    {
      "facet": "source_type",
      "value": "expert-written",
      "basis": "inferred",
      "inferred_from": "card-source",
      "note": {
        "en": "Card: paper says expert-authored (from source material); low confidence.",
        "zh": "卡片称论文称专家编制（来自素材），低置信"
      }
    },
    {
      "facet": "intended_use",
      "value": "evaluation",
      "basis": "author-stated",
      "excerpt": "Title: MedBench v4: A Robust and Scalable Benchmark for Evaluating Chinese Medical Language Models, Multimodal Models, and Intelligent Agents",
      "snapshot_id": "s01"
    },
    {
      "facet": "expert_involvement",
      "value": "review-unspecified",
      "basis": "inferred",
      "inferred_from": "card-expert-review",
      "note": {
        "en": "Card expert_review; method unspecified.",
        "zh": "卡片专家审核描述来自素材，方式未说明"
      }
    },
    {
      "facet": "specialty",
      "value": "multi-specialty",
      "basis": "author-stated",
      "excerpt": "d-based benchmarking infrastructure comprising over 700,000 expert-curated tasks spanning 24 primary and 91 secondary specialties, with dedicated tracks for LLMs, multimodal models, and agents. Items undergo multi-stage",
      "snapshot_id": "s01"
    }
  ],
  "changelog": [
    {
      "date": "2026-11-12",
      "change": {
        "en": "Card converted from an earlier manual catalog; no new facts added; reclassified under taxonomy v2.",
        "zh": "由早期人工整理的目录转换生成，未新增事实；按分类体系 v2 重新归类。"
      }
    },
    {
      "date": "2026-11-12",
      "change": {
        "en": "License status split into license_status (what the license says) and hosting (what Mdexa does); the earlier conclusion is kept for traceability.",
        "zh": "把许可证状态拆成 license_status（许可证怎么写）与 hosting（本站怎么处理）；原结论保留以便追溯。"
      }
    },
    {
      "date": "2026-11-12",
      "change": {
        "en": "Added verbatim license excerpts and snapshot ids from captured page snapshots; no conclusion changed.",
        "zh": "补入来自已抓取页面快照的许可证逐字摘录与快照编号；未改动任何结论。"
      }
    },
    {
      "date": "2026-11-13",
      "change": {
        "en": "Added dataset-level tags: 5 (1 stated by the authors with excerpts, 4 inferred from the content). No conclusion changed.",
        "zh": "新增数据集标签 5 个（作者声明 1 个，附逐字摘录；据内容推断 4 个）；未改动任何结论。"
      }
    },
    {
      "date": "2026-11-13",
      "change": {
        "en": "Removed the stored size-tier tag; the tier is now computed from the size field when pages are built.",
        "zh": "移除存储的规模档位标签；档位改为生成页面时由 size 字段计算。"
      }
    },
    {
      "date": "2026-11-13",
      "change": {
        "en": "Added specialty tags: multi-specialty(author-stated).",
        "zh": "新增专科标签：multi-specialty(author-stated)。"
      }
    }
  ],
  "_links": {
    "text": "在本库继续查找。本条目是 109 条已处理并核实的记录之一（数据集、数据资源与开源模型）。查找其他条目：调用 `https://mcp.mdexa.com/v1/datasets?q=<关键词>`，或连接 MCP 服务 `https://mcp.mdexa.com/mcp`；每分钟 20 次以内无需密钥。每行返回不到 1 KB；空结果即\"数据库未收录\"。核对一份清单：一次批量核对（免费账号）即返回报告。说明：/agents.md。相似条目：CMB (Chinese Medical Benchmark) (https://mdexa.com/zh/datasets/cmb/)；DiagnosisArena (https://mdexa.com/zh/datasets/diagnosisarena/)；MedMCQA (https://mdexa.com/zh/datasets/medmcqa/)。",
    "records": 109,
    "search": {
      "template": "https://mcp.mdexa.com/v1/datasets?q={words}",
      "example": "https://mcp.mdexa.com/v1/datasets?q=%E7%BB%93%E5%90%88%E4%BA%B2%E5%92%8C%E5%8A%9B"
    },
    "mcp": "https://mcp.mdexa.com/mcp",
    "anonymous_calls_per_minute": 20,
    "guide": "https://mdexa.com/zh/agents.md",
    "faq": "https://mdexa.com/zh/developers/faq/",
    "similar": [
      {
        "name": "CMB (Chinese Medical Benchmark)",
        "url": "https://mdexa.com/zh/datasets/cmb/",
        "json": "https://mdexa.com/zh/datasets/cmb.json"
      },
      {
        "name": "DiagnosisArena",
        "url": "https://mdexa.com/zh/datasets/diagnosisarena/",
        "json": "https://mdexa.com/zh/datasets/diagnosisarena.json"
      },
      {
        "name": "MedMCQA",
        "url": "https://mdexa.com/zh/datasets/medmcqa/",
        "json": "https://mdexa.com/zh/datasets/medmcqa.json"
      }
    ]
  }
}