PDF to Markdown - Convert PDF files to clean Markdown in your browser.

Yes. PDFtoMD.im is free to use, requires no account, and does not place a watermark in the Markdown. The current browser limits are 40 MB, 100 pages for digital PDFs, and 25 pages that require OCR.

No. PDF.js reads the PDF inside your browser, and the Markdown remains in this tab until you copy, download, clear, or reload it. The document is not sent to a conversion API.

Yes, for English scanned pages. When a page contains very little selectable text, the converter can render that page locally and run Tesseract OCR. OCR is slower and less certain than extracting an existing text layer, so scanned output should be reviewed beside the PDF.

It detects aligned text cells and recommends Markdown for regular rectangular tables or HTML for irregular rows. Every detected table can also be changed to CSV or plain text. A confidence note appears when the structure is uncertain.

Yes. Use the per-table Download CSV button or download the ZIP package, which places every detected table in a separate CSV file. CSV is useful for spreadsheets, while Markdown or HTML is usually better inside a document.

The source pages remain visible in the side-by-side PDF preview, but this version does not extract and embed raster images into the Markdown file. This avoids misleading image order and oversized downloads. Keep the original PDF when figures are essential.

Selectable formula characters may carry over, but PDF drawing instructions do not contain the original LaTeX source. Complex equations, superscripts, and matrix layouts need manual review. The converter does not claim equation reconstruction.

Usually, yes. Headings, paragraphs, lists, code-like text, tables, and page-break comments make the result easier to inspect and chunk than raw PDF text. Remove navigation pages, references, or sensitive content before sending the result to any external AI service.

A PDF stores characters at coordinates rather than as semantic paragraphs. Multi-column pages, floating captions, overlapping text boxes, and unusual font encodings can make reading order ambiguous. Use the original panel to compare and edit the Markdown before download.

A PDF can be up to 40 MB and 100 pages. Automatic OCR is limited to the first 25 pages that need it because rendering and recognition use significant device memory and battery. These are browser safety limits, not paid plan restrictions.