Launching today
pdfmarkdown.app
Turn PDFs into AI-ready Markdown, right in your browser
9 followers
Turn PDFs into AI-ready Markdown, right in your browser
9 followers
Turn PDFs into AI-ready Markdown without uploading your files to a server. Tables, formulas, and images stay usable, while a side-by-side view lets you check the output against the original before trusting it. It also handles scanned and encrypted PDFs. Free, no signup, and everything runs in your browser.
Hey PH 👋
I’m Jerome, the indie maker behind pdfmarkdown.app.
I use Obsidian to manage much of my personal knowledge. Markdown notes are easy to search, link, edit, and reuse with AI. PDFs are the awkward exception.
Research papers, books, reports, and manuals often arrive as PDFs, but they don’t fit naturally into a Markdown knowledge base. Every small question means parsing the whole PDF again and burning more tokens.
I kept looking for a better way to convert them.
Some open-source tools are genuinely powerful, but they require a command line and a fresh setup on every computer. Most aren’t directly accessible from a phone, and I couldn’t easily hand them to a non-technical family member.
Simpler web tools often upload the PDF to a server. That may be fine for a public paper, but not for medical records, tax documents, or anything else I want to keep private.
👉 So I built pdfmarkdown.app: a PDF-to-Markdown tool that runs directly in your browser.
What makes it different:
→ Keep the document, not just the words
It reconstructs tables, formulas, and reading order instead of flattening everything into plain text. Images are cropped, kept in place, and bundled with the Markdown for easy Obsidian import.
→ See it before your AI does
The PDF and Markdown stay side-by-side and scroll in sync in either direction. Click any part of the Markdown to jump to its source in the PDF, making it faster to verify the result before using it in your knowledge base or AI workflow.
→ Private without the setup
Everything runs on your device, including OCR for scanned pages. There’s no server upload, account, installation, or command line. Open the site on a computer or phone and start converting.
It also handles encrypted PDFs and CJK documents, including Chinese, Japanese, and Korean text.
The hardest part wasn’t extracting words. PDFs store where content is drawn, not how it should be read. Reconstructing columns, tables, formulas, and image placement took far more work than I expected.
pdfmarkdown.app is free to use. I’d especially love feedback from Obsidian users, researchers, and anyone building a personal knowledge base.
Try one of the real examples on the homepage, or give it the hardest PDF you have.
I’d love to hear what you’re bringing into your knowledge base and how the conversion holds up.
Ran a 40-page scanned PDF through it and the table extraction came out surprisingly clean. Love that nothing leaves my laptop, the side-by-side view sealed the deal for me.
Love that everything runs locally, gives me real peace of mind for client docs. One thing I'd love to see is batch processing, so I could drop in a folder of PDFs and get a zip of markdown files back instead of converting them one at a time. That would make it way more useful for my research workflow.