{
  "schemaVersion": "1.0",
  "updates": [],
  "companyNews": [],
  "usageGuide": {
    "facts": [
      {
        "key": "limits",
        "label": "使用前看",
        "text": "当前许可基于 Apache 2.0 并有附加条款：与关联方合并月活超过 1 亿或月收入超过 2,000 万美元，须另获商业许可；向第三方提供在线服务还需显著标识使用 MinerU。模型和依赖另行核对许可。",
        "sourceUrls": [
          "https://github.com/opendatalab/MinerU/blob/master/LICENSE.md",
          "https://github.com/opendatalab/MinerU/blob/master/README.md"
        ]
      }
    ]
  },
  "id": "b925",
  "language": "zh",
  "name": "MinerU",
  "originalName": "MinerU",
  "library": "tools",
  "category": "模型开发",
  "categoryKey": "模型开发",
  "categoryKeys": [
    "模型开发"
  ],
  "categories": [
    "模型开发"
  ],
  "entryPoints": [],
  "summary": "用 OCR 和视觉语言模型解析复杂文档，保留页块定位并输出结构化内容。",
  "introduction": "MinerU 4 为文档解析和智能体阅读提供命令行、Python SDK 与本地服务。PDF 和图像可按 Flash、Basic、Standard、Advanced 分层处理，复杂版式可结合 OCR、小模型与视觉语言模型；Office 等原生文档走本地 Flash 路径，并非每种格式都经过 VLM。解析结果可供后续检索、逐页阅读和引用定位使用。它是文档处理工具，不是现成 RAG 问答平台；远程解析需要显式配置。",
  "capabilities": [],
  "company": {
    "name": "MinerU Team",
    "originalName": "MinerU Team"
  },
  "officialUrl": "https://github.com/opendatalab/MinerU",
  "profileUpdatedOn": "2026-10-23",
  "sources": [
    {
      "kind": "product",
      "url": "https://github.com/opendatalab/MinerU"
    },
    {
      "kind": "documentation",
      "url": "https://opendatalab.github.io/MinerU/"
    },
    {
      "kind": "documentation",
      "url": "https://opendatalab.github.io/MinerU/usage/tiers/"
    },
    {
      "kind": "documentation",
      "url": "https://github.com/opendatalab/MinerU/blob/master/README.md"
    }
  ],
  "news": [],
  "links": {
    "interactive": "products/mineru--b925.html",
    "reading": "read/zh/products/mineru--b925.html",
    "markdown": "read/zh/products/mineru--b925.md",
    "json": "read/zh/products/mineru--b925.json"
  }
}