How to Extract Text from a PDF as Markdown on iPhone (Headings, Lists and Paragraphs Kept)

TL;DR

Import the PDF and PDFExtract reads its text layer, rebuilds headings, lists and paragraph flow, and shows the result as rendered Markdown or source. Copy it or share the .md file. The free version reads the first 5 pages; Pro reads them all.

The problem

Select-all and copy in a PDF viewer gives you text with a hard line break at the end of every line, two columns jumbled into one, page numbers in the middle of sentences and no headings. Pasting that into a note app means ten minutes of cleaning. Online converters fix it but want your document uploaded.

What PDFExtract does

For each page it checks whether the PDF has real text. If it does, it reads every line with its position, not just a flat string, and then reconstructs the structure:

  • Headings, from font size and bold weight (a book set all in 12 pt still marks chapter titles in bold).
  • Lists and paragraphs, merging lines that were only broken by the page width.
  • Running headers and footers that repeat on each page are recognised and kept out of the body.

If a page has no text layer (a scan), it switches to OCR for that page. See the scanned PDF guide.

Steps

  1. Open PDFExtract and tap Import PDF, or send a PDF from Files, Mail or Safari using the Share sheet.
  2. Wait while it extracts. A progress view shows the run; you can cancel.
  3. Open the project. Switch between the rendered preview and the Markdown source.
  4. From the project menu choose Copy Markdown or Share Markdown, which shares a .md file.

Check the result

  • Skim the headings first; a missing heading is the usual sign that the document’s styling doesn’t mark it clearly.
  • Pages that couldn’t be read are listed (“N page(s) could not be read”) and marked in the output so a gap is visible.
  • If something looks wrong, try another Format in Settings. See the output styles guide.

Free and Pro

The first 5 pages of a document are free. Pages after that are not read at all (no battery spent) until you go Pro and run extraction again. Copy and share of the Markdown are free.

FAQ

Does it keep headings and lists? Yes, in the default Structured style. Headings, lists and merged paragraphs are rebuilt from font weight, size and layout. Tables need Pro.

Does it handle two-column pages? It reads native text line by line with positions, which keeps columns in reading order rather than interleaved.

How many pages are free? The first 5 pages of each document. Pro reads every page.

Is my PDF uploaded? No. Text extraction and layout analysis run on the device.

The app PDFExtract →