Turn any document into Markdown.
Drop a PDF, Word doc, PowerPoint, spreadsheet, or image. Get clean, structured Markdown in seconds — ready for Claude, ChatGPT, NotebookLM, and your RAG pipeline.
Drop files here or click to browse
Max 25MB per file · Free: 5/day, 15/day with an account · One file at a time
Files are deleted immediately after conversion. Scanned documents and images use Claude AI for OCR.
Need more? Upgrade to Pro for unlimited conversions + 100MB files.
3,500+ documents converted in the last 30 days — by developers, researchers, and anyone feeding documents to an LLM.
Need to feed social posts into AI tools?
PostToSource turns tweets, threads, Substacks, and public Notion pages into hosted sources for NotebookLM, Claude, and more.
Every format. One tool.
Extract text, tables, and structure from PDFs.
Word (DOCX)
Preserve headings, formatting, and lists.
PowerPoint (PPTX)
Convert slides into structured sections.
Excel (XLSX)
Transform spreadsheets into Markdown tables.
CSV
Turn CSV data into formatted Markdown tables.
JSON
Format JSON data as readable Markdown.
XML
Extract content and hierarchy from XML files.
HTML
Strip tags and keep semantic structure.
EPUB
Convert eBooks with chapter structure intact.
Images
Extract text from screenshots and photos with AI-powered OCR.
How it works
Upload
Drag and drop any supported file — PDF, DOCX, PPTX, and more.
Convert
Our engine processes your file and extracts clean, structured Markdown.
Download
Copy to clipboard or download as a .md file. Ready for any workflow.
Built for
Developers
Convert docs to Markdown for GitHub, blogs, READMEs, or static site generators.
Researchers
Turn papers and slides into AI-ready text for Claude, ChatGPT, or NotebookLM.
AI Power Users
Feed documents into RAG pipelines, vector databases, or Obsidian — or plug our MCP server into Claude and convert without leaving the chat.
Need to feed social posts into AI tools?
PostToSource turns tweets, threads, Substacks, and public Notion pages into hosted sources for NotebookLM, Claude, and more.
Frequently asked questions
Upload any supported file. Our conversion engine processes it and returns clean, structured Markdown. No account needed.
PDF, DOCX, PPTX, XLSX, CSV, JSON, XML, HTML, EPUB, images (JPG, PNG), audio files, and ZIP archives.
Yes — no account needed for the first 5 conversions per day (up to 25MB per file), and a free account raises that to 15 per day. When you hit your limit, Pro unlocks unlimited conversions, 100MB file support, and batch processing of up to 10 files at once. Upgrade to Pro →
Files are deleted immediately after conversion — we never keep your documents. Images and scanned PDFs are OCR'd via Anthropic's Claude API; see our privacy policy for details.
Pro users can batch-convert up to 10 files at once. Free users convert one file at a time.
An LLM asked for “the markdown” retypes your document — models paraphrase, invent table cells, and silently truncate long files. file2markdown runs real extraction engines: exact text, exact tables, the whole document, identical output every time. It also handles what chat models can’t — DOCX/XLSX files behind URLs and scanned PDFs via OCR.
Yes — file2markdown has an MCP server, so Claude, Cursor, and other AI clients can convert documents mid-conversation. The free tier needs no key; Pro API keys unlock unlimited conversions and OCR. Setup takes a minute →
The Markdown Memo
A fortnightly note for lawyers, researchers, accountants, and anyone else drowning in PDFs, scans, and decks. No spam.