Why convert a PDF to Markdown
A PDF stores where each word sits on a page. It does not store what the words mean. When you paste one into an AI tool, the tool spends many tokens on page furniture, broken lines and odd spacing, and still loses the structure.
Markdown keeps the structure that matters, such as headings, lists, links and emphasis, and drops the rest. The same document usually costs far fewer tokens, and the model can see which line is a heading and which is body text.
How to convert a PDF
Drop a PDF on the page, or choose one. The converter reads it on your computer and shows the Markdown straight away, with an estimate of the tokens it will cost.
- Choose a text PDF of up to the size limit shown on the page.
- Switch bold and italic on or off, and choose whether page breaks are marked.
- Copy the Markdown, or download it as a .md file.
- Read the warnings under the result. They tell you when a page could not be read.
Your file stays on your computer
The conversion runs inside this page. The file is read with your browser, handed to a PDF reader in the same tab, and turned into text there. Nothing is uploaded and nothing is stored, so you can use it on documents that must stay private.
What the converter cannot do
A scanned page is a picture of text, so there are no words to read. The converter flags those pages instead of guessing. It does not run optical character recognition. Complex tables and multi-column layouts may come out in reading order rather than as a grid, so check them before you rely on them.
Frequently asked questions
Is it safe to convert a confidential PDF here?
Yes. The file is processed entirely in your browser and is never sent to a server. Close the tab and nothing remains.
Why does Markdown use fewer tokens than a PDF?
A PDF pasted as text carries layout noise, such as page numbers, running headers and hard line breaks. Markdown keeps only the words and the light structure, so the same content takes fewer tokens.
Can it read scanned PDFs?
No. A scan is an image, and this tool does not do OCR. It marks those pages with a warning so you know the text is missing.
Does it keep tables and images?
It keeps text, headings, lists, links and emphasis. Images are left out, and tables are read as text in order. Check any table-heavy page against the original.