Convert PDF to Markdown with the structure intact

Yes — that is the point of this page. Headings become Markdown headings, bullet lists become dashes, bold and italic are kept, and tables come out as pipe tables. It is not a text dump with the line breaks rearranged.

Because the structure is machine-readable. It is the format that documentation sites, static site generators, note apps and language models all read natively, and it stays readable as-is if none of those are involved.

They come out as plain dashes. Word stores the numbering separately from the text, so the numbers themselves are not part of what we can carry across. Renumbering them in your editor takes a moment.

No, and deliberately not. Slides are floating text boxes with no reading order attached, so any Markdown we produced would be a guess about what comes after what — and it would be wrong often enough to read as scrambled. Convert the deck to Word first if you need the text.

Yes. A PDF with no text layer goes through text recognition first, automatically, and then the Markdown is built from the recognised text.

No. Markdown references images as separate files, and this returns a single .md file, so pictures are dropped rather than linked to something that is not there. Convert to Word if the images matter.

Up to 10 MB per file. A PDF longer than 20 pages is converted up to its first 20 pages — 100 on a paying account — and the result tells you so.

Because it is doing more. The document is laid out properly before the Markdown is written, which is what makes a heading a heading — the plain text export skips that step and skips the structure with it.

No. Your first conversion runs without signing up.

Yes. Ask for both in the sentence — for example, convert it and translate it — and you get one finished file back.