Feeding a document to an LLM, pasting it into a wiki, or checking it into a repo alongside your code all want the same thing: structure, not a wall of plain text. This reads the actual headings, bold and italic text, and lists out of your PDF and writes them back as real Markdown.
How to pdf to markdown
Upload your PDFChoose the file you want converted.
Choose page markersTurn on markers if you want to know where each page began in the source document.
ConvertWe read the document's own font sizes and styles to tell headings from body text, bold from plain, and list items from paragraphs.
DownloadGrab the .md file. Your PDF is gone the instant your download is ready.
What you can do
- Feed a document's contents into an LLM or RAG pipeline with its structure intact
- Turn a PDF spec or report into something you can paste into a wiki or README
- Check a document into a repository as version-controlled Markdown instead of an opaque PDF
- Pull structured notes out of a PDF you were sent to edit or extend
- Convert a PDF-only manual into something a static site generator can build from
What happens to your file
Your files are never stored. Nothing is parked, queued, or kept — the moment your download is ready, it's gone, and no one ever sees it.
Plenty of online tools store your file on their servers and delete it later. PDF Cubby doesn't keep it at all — there's nothing left behind to clean up afterward.
Features
- Headings detected from font size - the document's own largest text becomes # and ##, not a guess
- Bold and italic preserved, read directly from the font used in the PDF
- Bulleted and numbered lists come through as real Markdown lists
- Optional page markers so you can trace Markdown back to its source page
- Free, no sign-up, no watermark
- Never stored - files are gone the moment you are done
Tips
- Getting nothing back? The PDF is a scan - run OCR first, then convert.
- Just need the plain words, no structure? PDF to Text is the simpler tool for that.
- Need the file back as a document instead? Go the other way with HTML to PDF once you've edited the Markdown as HTML.
Frequently asked questions
- How are headings detected?
- By font size. The most common text size in the document is treated as body text, and distinctly larger sizes become heading levels, largest first. A PDF with no size variation has no detected headings - the whole thing renders as paragraphs, which is the honest result for a document that never actually had a heading structure.
- Will tables come through properly?
- Table cells extract as text, but this tool does not rebuild Markdown table syntax (pipes and dashes) - for anything table-heavy, check the result before relying on it.
- Does it run any AI model to do this?
- No. It reads the PDF's own font sizes and style flags directly - no model, nothing sent anywhere for analysis.
- Why did I get an empty file?
- Because the PDF has no text in it - it's a scan, so the words are pictures. Run it through OCR first and the text becomes extractable.
- Do I need an account?
- No. Converting to Markdown is free with no sign-up.
- Are my files kept anywhere?
- No. Your PDF is read and discarded the instant your download is ready.