Free PDF to Markdown Converter
Convert PDF files into clean Markdown with headings, lists and tables intact and formulas rendered as LaTeX. Free, no sign-up, up to 20 MB, private by design.
How to use it
- 1
Upload your PDF
Any text-based PDF up to 20 MB — papers, manuals, reports, specs. Documents exported from Word, LaTeX or Google Docs convert most faithfully; scans vary with image quality.
- 2
Decide on images
Keep images on and the download bundles a ZIP — the .md file plus an images/ folder of page snapshots, so every figure reference resolves on disk. Switch off for clean text-only Markdown headed to notes or an LLM pipeline.
- 3
Convert and check the spine
Headings map to #/##/### following the document's own hierarchy, lists stay lists, tables become pipe tables, and formulas come back as LaTeX wrapped in $...$ or $$...$$.
- 4
Paste where it lives
The output drops straight into Notion, Obsidian, a GitHub README or a docs site — Markdown that was actually structured, not a text dump with decorations bolted on.
PDF is a print format wearing a document's clothes. It knows where every glyph sits on the page, but nothing about what a heading is, where a list starts, or why that block of numbers counts as a table. Markdown is the opposite — pure structure, no layout. Converting between them well is exactly as fiddly as it sounds, which is why so many converters hand you back a wall of text with the headings baked in.
This one reads the document's actual hierarchy and rebuilds it in Markdown. Chapter and section titles become #, ## and ### at the right depth, following how the source is organized rather than font sizes alone. Bullet and numbered lists come back as lists. Tables convert to pipe tables with their columns intact — the comparison matrix in section four stays a matrix instead of collapsing into comma soup. Mathematical formulas render as LaTeX: inline math wrapped in $...$, display equations in $$...$$, ready for MathJax, KaTeX, Obsidian or GitHub.
The images option covers the two main workflows. With images on, each figure becomes a Markdown reference pointing at a snapshot of the page it appears on — images/page-7.jpg for a figure on page seven — and the download button packages a ZIP with the .md file and the complete images/ folder inside, so every reference resolves the moment you unzip it next to the document. With images off, you get clean text-only Markdown — figures skipped, captions kept — which is usually what you want when the output is headed for a note-taking app, a docs site, or a pipeline feeding a language model.
The use cases run from mundane to technical: dropping an academic paper into Obsidian without retyping the tables, turning a vendor's PDF spec into a README, migrating legacy documentation into a static site generator, or pre-processing reports so downstream tools receive structured text instead of raw PDF bytes.
Files up to 20 MB are accepted, no account required, and the privacy posture matches every tool on this site. Files are processed in memory and deleted within 15 minutes. Nothing is stored, nothing trains a model, and there is no account to tie a document to — so contracts, internal reports and unpublished manuscripts pass through without leaving copies.
FAQ
Is the PDF to Markdown converter free?
Yes — no account, no page quota, no watermarked output. The 30-requests-per-hour rate limit is there for bots, not for you.
What happens to my PDF after conversion?
Files are processed in memory and deleted within 15 minutes. We never store documents and never use them for training — which matters for the contracts, client reports and unpublished drafts that pass through here.
Do scanned PDFs work?
Text-based PDFs — anything exported from Word, LaTeX or Google Docs — convert cleanly. Pure image scans depend on how well the model can read the pages; expect some cleanup if the scan is crooked or low-contrast.
How are formulas and tables handled?
Formulas become LaTeX: inline math wrapped as $x^2$, display equations as $$...$$ blocks. Tables convert to GitHub-style pipe tables with their original rows and columns preserved.
What does the images option actually do?
With images on, every figure becomes a Markdown reference like images/page-7.jpg pointing at a snapshot of the page it sits on, and the download button gives you a ZIP containing the .md plus that whole images/ folder — references resolve the moment you unzip. With images off, figures are skipped and captions kept as italic lines.
Is there a page limit?
The cap is 20 MB per file, which covers most papers and reports. Genuinely enormous documents — several hundred pages — may have their later pages compressed to stay within processing limits.