MinerU
MinerU parses PDFs, Office files, images, and web URLs into structured outputs for agent and RAG workflows. Developers and research teams can use the extracted text, tables, formulas, and chemistry data in downstream applications.
One of 140 tools in Research
Who it's for
- Developers building agent workflows who need document parsing with MCP support.
- Teams preparing mixed-format documents for RAG pipelines or knowledge bases.
- Scientific and chemistry research teams extracting formulas, molecule structures, and reaction information from papers.
How it fits your workflow
Provide MinerU with a PDF, Word, PowerPoint, Excel file, image, or web URL. It extracts text and visual content while retaining context and layout structure.
Use the resulting Markdown, JSON, CSV, HTML, LaTeX, or MathML in an agent workflow or RAG data pipeline. The page names MCP, Dify, Notion, and OpenClaw among its agent-related support and export options.
Pricing
The vendor doesn't publish prices. Check with MinerU directly.
Prices checked on Oct 17, 2026 from the vendor's site. They can change; confirm before you buy.
Key features
- Multi-format document parsing
- Parses PDF, Word, PowerPoint, Excel, image, and web URL content, extracting text and images while preserving their contextual relationships.
- Structured machine-readable output
- Produces high-fidelity Markdown and JSON, with CSV, HTML, Markdown, LaTeX, and MathML outputs described for specific extraction tasks.
- Table reconstruction
- Extracts and structures complex tables, including rotated tables, tables spanning pages, and merged cells.
- Formula extraction
- Converts long, multi-line, and nested mathematical formulas into LaTeX or MathML.
- Agent workflow support
- Supports MCP and named agent ecosystems, with export options for Dify and Notion.
- Chemistry-paper parsing
- Provides molecule detection and structure-diagram recognition, extracts chemical reactions and key elements such as reactants and conditions, and links molecule identifiers across images and text.
Works with
- MCP
- Dify
- Notion
- OpenClaw
Alternatives to consider
-
Monkt
Choose it if you need batch document processing, custom JSON schemas, or REST API integration, which its description specifically lists.
-
Convert Anything to Markdown
Choose it if your main requirement is converting URLs, PDFs, Office files, HTML, and images into plain Markdown through a low-latency API.
Getting started
- Choose online use or download the MinerU client from the website.
- Provide a supported document or web URL for parsing.
- For agent integration, consult the Agent access guide.