Get a PDF’s content in a structured text format, with its headings, lists and bold text, easy to paste into your notes, a wiki or a conversation with an AI.
How it works
- Select the PDF.
- Click “Convert to Markdown”.
- Copy the text or download the .md file.
Frequently asked questions
What is Markdown?
A simple text format where # marks a heading and ** bold text, supported by most note-taking and documentation tools.
Are tables converted?
Their text is carried over line by line, but not as a Markdown table. For a table you want to work with, use “PDF to Excel” instead.
How do I paste the result into Notion or Obsidian?
Click “Copy text” under the preview, then paste into your page: headings, lists and bold text are recognized. You can also import the downloaded .md file.
What if my PDF is a scan?
PDFRoll tells you it contains no text. Run it through the “OCR PDF” tool first, then run the conversion again.
Which software opens a .md file?
Any text editor, or apps that display formatted Markdown, such as Obsidian, Typora or Visual Studio Code.
My PDF is password-protected. What can I do?
First remove the protection with the “Unlock PDF” tool if you know the password, then start the conversion.
Are my files uploaded to a server?
No. Everything happens in your browser: your documents never leave your device. The tool is free, with no sign-up.
What PDFRoll recognizes
PDFRoll reads the text on each page, groups lines into paragraphs, then translates the formatting into Markdown syntax:
- Headings, detected from text size: the largest become “#” headings, the next ones “##” or “###”.
- Lists, detected from their bullets or numbering, turned into “-” and “1.”.
- Bold and italics within sentences, marked with ** and *.
Characters that have a meaning in Markdown, such as * or _, are escaped so the text displays as is. Check “Add a separator between pages” to insert a --- line at each page break.
For Notion, Obsidian, GitHub or an AI
- Import course material or documentation into Notion or Obsidian.
- Publish a PDF’s content in a README or wiki on GitHub.
- Give a report to ChatGPT, Claude or another AI in a format it reads easily, without the noise of a page layout.
- Keep a lightweight, readable text version of a document.
Once the conversion is done, a preview appears with a “Copy text” button so you can paste the result directly, and you can download the complete .md file. Everything is done in your browser.
Limitations to know
Tables are carried over line by line, not as Markdown tables, and images aren’t carried over. Heading detection depends on text size: in a document where everything is written at the same size, you’ll mostly get paragraphs.
A scanned PDF contains no text, only images of pages: PDFRoll will let you know. Run it through the “OCR PDF” tool first, then convert the result.
Related tools
For plain text without any formatting, use “PDF to text”. For an editable document with its tables, go with “PDF to Word”. And if you only want the gist of a long document, “Summarize a PDF” pulls out the key points.