Guide
Getting structured text out of a PDF for notes and docs
A PDF doesn’t know what a paragraph is. Inside the file, text is stored as short pieces placed at exact positions on a page: a line here, half a line there, a page number at the bottom. Copying from a PDF therefore gives you broken lines, stray headers and hyphenated words. This tool rebuilds the structure (headings, paragraphs and lists) and writes it as clean Markdown that drops straight into a wiki, a docs site or a notes app.
How the structure is recovered
- Lines. Text pieces that share a baseline are joined into a line, in reading order.
- Headings. The most common font size in the document is taken as the body text. Larger sizes become headings: the largest is
#, the next##, and so on down to####. A short bold line at body size that doesn’t end with punctuation, the typical sub-heading in reports, becomes the lowest level. - Paragraphs. Lines close together with the same size are joined. A word broken with a hyphen at the end of a line is put back together, and a sentence that runs over a page break continues in the same paragraph.
- Lists. Lines starting with a bullet or a number become list items; indented lines under an item are added to it.
- Clean-up. Lines repeated at the top or bottom of most pages, such as a company name, a document title or “Page 3 of 12”, are removed.
Reading the result
Use Preview to check the structure and Source to see the Markdown. The counters at the top tell you how many headings, paragraphs and lists were found; a document with 0 headings probably uses one font size throughout, and a scan shows a warning that no text was found.
Mark where each page starts inserts invisible comments like <!-- page 7 -->, handy when you need to cite page numbers later. Turn off Remove running headers if a document has meaningful text in its margins.
Where it works best
- Reports, manuals and policies exported from Word or Google Docs.
- Academic papers, apart from their tables and formulas.
- E-books and articles saved as PDF.
Two-column layouts are read line by line across the page and can mix columns; magazine-style PDFs need checking.
What is not converted
Markdown output covers text structure only. Tables come through as lines of text, so use PDF to CSV or PDF to Excel for them. Images are left out; save them with Extract PDF Images. Inline bold and links inside paragraphs are not carried over.
A worked example: a manual into a docs site
- Open the 60-page product manual.
- The counters show 48 headings, 312 paragraphs and 41 lists.
- In the preview, check that the chapter titles are
#and sections##. - Download the .md file and split it into pages for your docs site.
Related tools
For a web page instead of Markdown, use PDF to HTML. For plain text without structure, PDF to Text. For an editable Word document, PDF to Word. To write Markdown with a live preview, open the Markdown Editor.
On a phone
The converter also works on Android and iPhone: open the page in your browser, pick the PDF from Files or Downloads, switch to the source view and use Copy to paste the Markdown straight into a notes app. Long documents take longer to analyse on a phone, so for a 300-page book a computer is more comfortable.
Checked against the current version of the PDF to Markdown on .