PDF to Markdown (Extract Text to .md)

Headings, lists, code blocks and links recognized - headers and footers removed, converted locally

Want the content of a PDF in your notes, a docs site, or a prompt for ChatGPT or Claude? This tool extracts the text and turns it into Markdown (.md): heading levels (from the PDF bookmarks when it has them, otherwise guessed from font size and bold), paragraphs, ordered and unordered lists, code blocks, bold / italic and hyperlinks - and it strips the headers, footers and page numbers that repeat on every page.

Copy the result or download it as a .md file and paste it into Obsidian, Notion, Typora, GitHub, or any static docs site. Common two-column layouts are read left column first, then right, instead of mixing the columns line by line; paragraphs split across pages or columns are rejoined where possible, and words hyphenated at the end of a line are put back together.

Let’s be upfront about the limits: this extracts text and structure. Images are not exported; tables are not rebuilt as Markdown tables (the cell text comes out in reading order); math is not converted to LaTeX, and vertical or rotated text is skipped. Headings and lists are inferred from the layout, so read the result through once. Scanned PDFs (no text layer) should go through this site’s Batch OCR first. The difference from online converters: your file is processed inside your browser and never uploaded to a server.

How to use

Open the PDF (the converter appears automatically)
Choose whether to drop headers, footers and page numbers, and whether to mark page breaks with a rule or a comment
Click "Convert", wait while text is extracted page by page, then check the summary and preview
Click "Copy", or download the .md file

FAQ

Does it turn tables into Markdown tables?

Not yet: tables are not rebuilt - the text of each cell comes out as paragraphs in reading order, and you tidy it into a table yourself. If the table is a screenshot or a scan, this site’s table recognition can export it to Excel.

Are images and formulas kept?

No. Only text is extracted: images are not exported (use "PDF to JPG" to save pages as images), and math is not converted to LaTeX - its characters come out as they appear in the PDF.

Can it convert scanned PDFs?

A scan has no text data, so converting it directly gives an empty result - the summary tells you how many pages have no text layer. Run Batch OCR on this site first to embed a text layer, then convert.

How are headings detected?

If the PDF has bookmarks and most of their titles can be matched to text on the pages they point to, the bookmark levels become the # headings. Otherwise headings are inferred from the layout: lines clearly larger than the body text, or short lines set entirely in bold. It is a guess, so a few mistakes may need a manual fix.

Are lists and code blocks recognized?

Bullets written as text (• ◦ - and similar) and numbering (1. , 1) , (1), ①) become lists, nested by indentation; lines in a monospace font become code blocks. Some PDFs - for example web pages printed to PDF by a browser - draw the bullet dots as shapes instead of text; those bullets cannot be detected and the list is merged into an ordinary paragraph.

PDF to Word or PDF to Markdown?

Pick PDF to Word if you want to keep editing and formatting in Word or WPS. Pick Markdown if the content is headed for notes, a docs site, a code repository, or an AI assistant: it is plain text with no style files, and easy to compare with version-control tools.