pdfviz

Convert a PDF to Markdown

Markdown keeps reading order and simple structure, not page layout. So the only question that matters is whether your PDF holds real text or just pictures of it. Selectable text converts directly. A scan needs OCR first, and tables and math need a check afterwards either way.

Convert PDF to Markdown

Selectable text or scan: the one check that decides

Open the PDF in any reader and drag across a paragraph. If the text highlights line by line, the file has a text layer and every method below can work on it directly. If the whole page selects as one block, or nothing highlights at all, the page is a picture and there is nothing to extract until OCR guesses the letters.

This split runs through the whole market. Converters that work in the browser treat these as two separate jobs: one mode reads the text layer, another runs OCR on scans. The local tool on this site does the first job only and refuses scans instead of handing back an empty file, which is the honest version of the same split.

The local route: native text in your browser

The PDF to Markdown tool reads the native text layer, rebuilds lines in reading order, and uses font size and font information to infer simple headings, lists and code blocks. Page markers stay in the file as Markdown comments when asked, and repeated headers and footers can be removed when the tool identifies them safely.

Nothing is uploaded; the whole conversion runs in the tab. The generated Markdown is read back and compared with the kept text, and a failed retention check produces no download rather than a broken file. Image-only scans are refused outright, because OCR is separate work: guessed text needs its own accuracy measure and human review.

Copy and paste for a page or two

For a short passage, selecting text and pasting it into notes or an AI tool beats any converter. There is nothing to install and nothing to upload, and you see immediately what survived.

It stops scaling fast. Multi-column pages paste in the wrong order, words stay hyphenated from line breaks, and headings arrive as plain lines with no structure. Fine for a quote, wrong for a document. Past a few pages, run an extraction instead of fighting the clipboard.

Command line extraction for repeatable batches

When many files need the same treatment, the standard pipeline starts with Poppler: its pdftotext utility pulls the text layer out as plain text. Pandoc then moves between text formats on the way toward Markdown.

Be clear about what this pipeline does not do. It extracts the characters that are there; it does not infer document structure from positioned PDF text. Headings, lists and tables come out flat and need rebuilding by hand or with extra tooling. It is the right choice for scripted bulk extraction, and the wrong choice when the structure matters and nobody will review the output.

Scanned pages need OCR before Markdown

OCR guesses letters from pixels, and every Markdown converter for scans is an OCR pass with formatting on top. Well known open tools for the OCR step include Tesseract and OCRmyPDF; several online converters bundle this as a separate mode rather than mixing it into plain extraction.

Budget review time, because guesses carry errors. Small print, unusual fonts, tables and math degrade first, and the mistakes look confident in Markdown exactly as they would on the page. A scan that must be quoted or published needs a human pass over the converted text, not just a conversion.

What the ranking converters offer

Among upload-based options, CloudConvert converts PDF files to Markdown on its servers, with headings, lists and structure preserved where possible. iLovePDF offers a PDF to Markdown converter under its PDF Intelligence section and claims headings, tables, lists and links stay intact. Browser-local alternatives in the current top results include craftmarkdown and JustMarkdown, which make the opposite privacy claim of no uploads and no servers. The middle shape is pdf2md, which presents itself as a browser converter with heading and list detection; it does not advertise an OCR mode for scans. Those are their pages' claims, not measurements made here, and every upload route renders your pages on someone else's servers. These pages were checked in September 2026.

This guide's local terms are narrower on purpose. Native text only, no OCR, nothing uploaded, page markers as Markdown comments, repeat removal when it can be identified safely, and a retention check the output must pass before any download exists. A refused scan is not a missing feature; it is the point where guessing would start.

What breaks in every conversion

A PDF records text at positions on pages while Markdown records a reading order and a small set of structures, so some things never cross over cleanly. Complex tables arrive flattened or misaligned, math and equations lose their notation, multi-column text reads down the wrong column, footnotes detach from their marks, and images and diagrams cross over without their meaning since there is no alt text to carry it.

Plan the check, not just the conversion. Skim the Markdown against the PDF page by page for short documents, or spot check tables, figures and the first and last pages for long ones. Tables that matter deserve their own extraction pass rather than surviving inside a Markdown export by luck.

Questions people ask

How do I know if my PDF needs OCR first?
Drag across a paragraph in any reader. If the text highlights line by line, the file has a text layer and converts directly. If the page selects as one block or nothing highlights, it is a scan and needs OCR before any Markdown step.
Will my tables survive the conversion?
Simple grids often do, complex ones usually arrive flattened or misaligned. Tables that matter deserve a dedicated pass with PDF to CSV rather than surviving inside a Markdown export by luck.
Does the local converter upload my file?
No. The PDF is opened, converted, checked and previewed entirely in the browser tab. It suits confidential research and internal reports for exactly that reason.
What do the page markers look like?
Each page starts with a Markdown comment such as Page 3. The comment keeps the page boundary without appearing as visible content when rendered, or drop markers entirely for one continuous document.
Can I turn the Markdown back into a PDF?
Yes. Markdown to PDF renders Markdown into a verified PDF with selectable text, page numbers and layout options.
Why did I get almost no text out?
Almost always a scan: image-only pages have no text layer to extract. Run OCR first and convert the OCR result, and expect to review it, because guessed text carries errors that look confident.

Do it now

The tool runs in this browser. Your file never leaves the machine, and the result is checked before you download it.

Convert PDF to Markdown

Where to go next