Folio
EditorMergeSplitCompress
Free online PDF tool

OCR PDF

Turn a scanned page — a photograph of text with no letters in it — into one you can search, select and copy from. Folio reads the pixels on a server and lays an invisible layer of the recognised words back over the page, so it looks exactly as it did and now behaves like a real document. This is the one recognition step that needs a server; the file is sent over an encrypted connection and deleted as soon as the result is sent back.

Drop a scanned PDF here
Make a scanned PDF searchable and selectable.

Step by step

How to OCR a PDF

  1. Open the scanned PDF you want to make searchable.
  2. Pick the language it is written in, and send it to be recognised.
  3. Download the searchable PDF — and its text as a .txt file, if you asked for both.

What it does

OCR PDF, in full.

Turn a scanned page — a photograph of text with no letters in it — into one you can search, select and copy from. Folio reads the pixels on a server and lays an invisible layer of the recognised words back over the page, so it looks exactly as it did and now behaves like a real document. This is the one recognition step that needs a server; the file is sent over an encrypted connection and deleted as soon as the result is sent back.

  • Make a scanned page searchable, selectable and copyable
  • Keep the page looking exactly as it did — nothing is redrawn
  • Read English and Bengali, together or apart

This tool needs a server. The file is sent over an encrypted connection, converted, and the copy there is deleted as soon as the result is sent back — unless you are signed in and asked to keep it.

Questions

About this tool

Is this the same as PDF to Word?
No. PDF to Word turns a document into an editable .docx; this keeps it a PDF and adds an invisible layer of the words underneath the picture, so the page still looks the same but can now be searched, selected and copied.
Is my file uploaded?
Yes. Reading the text off a scan needs a server, so the file is sent over an encrypted connection and the copy there is deleted as soon as the result is sent back, unless you are signed in and asked to keep it.
What languages does it read?
Whatever the service has installed — English and Bengali at minimum. The page lists exactly what is available before you send anything.
Will it touch pages that already have text?
Yes — every selected page is recognised, even one that already has text. That's what makes it safe to run this again on a file it already made searchable. Such a page keeps its existing text as well as the new layer, so it reads twice on a search.
How accurate is it?
Good on a clean, straight scan; worse on a faint or crooked one. The result shows a confidence reading, and recognised text can contain mistakes — check anything that matters before you rely on it.

The rest of them

17 more PDF tools.

Edit

Pages

Finish