Drop any document, get clean Markdown back.

t0md converts PDFs to Markdown that humans can read and AI agents can actually use. No sign-up. Files are processed in memory and never stored.

PDF live Microsoft Word live HTML live PowerPoint live EPUB soon XLSX soon Image (OCR) soon URL soon
Markdown preview

Why convert PDFs to Markdown?

PDFs lock content into a layout that's hard to search, hard to diff, and hard for an LLM to read. Markdown is plain text — every editor, every chat model, every static site generator understands it natively.

Feed it to an LLM

Markdown is the format Claude, ChatGPT, Gemini and Grok already speak. Convert a research paper or a legal contract once and quote it forever.

Take notes that travel

Drop a converted PDF straight into Obsidian, Notion, Bear, or a GitHub README. No more screenshots, no more copy-paste mangling.

Diff and version

Markdown plays nicely with git. Track contract revisions, compare drafts, blame a paragraph — all the things PDFs make impossible.

Built for agents

An MCP server ships with t0md. Claude Code, Cursor, Windsurf and any MCP-compatible agent can convert PDFs mid-conversation.

For AI agents — MCP server

t0md exposes a remote MCP server at https://t0md.com/mcp/sse. Add it to Claude Code once and any agent in any session can convert PDFs to Markdown without leaving the chat.

claude mcp add --transport http t0md https://t0md.com/mcp

It's free while we figure out demand. Read more on the MCP page.

FAQ

How do I convert a PDF to Markdown?

Drop the PDF on the box above. The conversion runs server-side, the file is held in memory only, and you can copy the Markdown or download a .md file. No account required.

Which formats are supported?

PDF, Microsoft Word, HTML and PowerPoint are live today. EPUB, XLSX, image OCR and URL-to-Markdown are queued up — the pills above mark them as soon until they ship.

How clean is the Markdown?

Headings, bullet lists, numbered lists and paragraph breaks are preserved. Tables and complex layouts are best-effort — PDFs are visual, not semantic, so we apply heuristics. For most academic papers, books and reports, the output is "drop into an LLM" clean.

Are my files stored?

No. Uploads are converted in memory and dropped immediately. The temporary download links from the MCP server expire after 10 minutes.

Is there an API?

Yes — POST a multipart form to /convert with field file and you get JSON back: {markdown, filename, pages, bytes}. The MCP server wraps the same endpoint for agents.

Will t0md stay free?

The website is free and will stay free. The MCP server is free for now. We'll likely introduce paid tiers for high-volume API and MCP traffic — the basic browser conversion isn't going behind a paywall.