OCR – Make scans searchable

Detects text in scanned PDFs and adds an invisible text layer over the pages. Layout is preserved.

This conversion needs a server. Your file is sent over an encrypted connection, processed immediately, and deleted right after — within 15 minutes at the latest as a safety net.

A scanned or image-based PDF becomes searchable: an invisible text layer is placed over the images. Layout stays intact.

How to

  1. 1

    Upload PDF

    Max 80 MB (signed in: 95 MB). Partially-OCR'd documents are also accepted.

  2. 2

    Pick language

    12 European languages or "Automatic" (combines the top 6). More languages = slower OCR.

  3. 3

    Run OCR

    Server runs Tesseract OCR, adds a text layer, returns the finished PDF.

Tips

  • If you only expect one language, pick it — faster and more accurate than "Automatic".
  • Existing text layers are skipped (`--skip-text`) — saves time on mixed documents.

Privacy and limits

Server processing. 10 jobs per 10 minutes (25 when signed in), max 80 / 95 MB. ~2-15 seconds per page depending on DPI and language. The time limit scales with file size (about 11 seconds per MB, at least 2 minutes, at most 15 — 20 when signed in).

Frequently asked about this tool

Is image quality preserved?
Yes. ocrmypdf places the text layer UNDER the original image — visual appearance stays 1:1.
Why does OCR struggle with my doc?
Poor recognition usually comes from low resolution (<200 DPI), skewed scans or heavy compression. Try a higher-res original.

Related tools