MinerU document parser MCP server

This is a small helper that lets your AI read documents that are not plain text. You hand it a PDF, a Word file, a PowerPoint deck, an image or a spreadsheet, and it turns the content into Markdown, which is simple formatted text your AI can actually read. It is handy for anyone who deals with reports, papers, contracts or slide decks and wants their AI to summarize or answer questions about them.

What this MCP server does

You ask your AI to parse a document, either a file on your computer or a link to a file online. Your AI passes that request to this helper, which uploads the file to the MinerU service. MinerU reads the document and converts it into Markdown, keeping tables and formulas where it can. The result comes back to your AI, which can then summarize it, answer questions or save the text for you. For single files you often see the Markdown right in the chat; for batches you get files saved on your computer.

Click to zoom

What you can do with it

• Convert a PDF into Markdown your AI can read
• Parse Word, PowerPoint and image files into text
• Extract just a page range, like pages 3 to 5
• Parse a file straight from a web link
• Handle several files at once, each with its own pages
• List the OCR languages MinerU supports
• Save the parsed Markdown to a folder you choose

Try asking your AI

• “Parse the 3rd to 5th pages of this PDF into markdown: /Users/me/Documents/report.pdf”
• “Extract the contents of this paper: https://arxiv.org/pdf/2509.22186”
• “Parse /Users/me/contract.docx and save the markdown to /Users/me/output”
• “Use the pipeline model to parse this Korean file /Users/me/notes.pdf”

What it gives back to you

For a single file, you usually get the Markdown text right in the chat, plus a path to the saved file and a download link for a zip. For batches, the results are saved to disk and you get file details instead of the full text. If the text is very long, it is saved locally and you get a path to it. Your AI can then summarize, quote or analyze whatever came back.

Before you start

Who it's for

Anyone who works with PDFs, Word files, slides or scanned images and wants their AI to read and summarize them, like researchers, analysts, students and office teams.