Does OCR replace the page image?

You still see the scan. The text layer sits on top so search and copy work. If recognition is wrong, the image is still the visual source of truth.

Printed type is the intended input. Cursive handwriting will miss. Treat those pages as images.

Which language codes?

Use engine codes such as eng. Add a second code only for pages that actually mix languages.

Not installedPlugin

37 tools

Make scanned PDFs searchable

Input PDF*

Drop files here or click to browse

PDF files only

Languages

Deskew pages