OCR PDF
Make scanned PDFs searchable and copyable – recognition runs entirely on your device.
- Stays on your device
- No limits
- No watermark
- No account
How can I check that nothing is uploaded?
Your files are processed in this browser tab. This page has no function that sends files to a server, and its security rules (Content Security Policy) forbid connections to other servers.
Check it yourself: load this page, then switch on flight mode or disconnect Wi-Fi and run the tool. It still works – because everything happens on your device.
Good to know
How it works – step by step
Choose PDF
Pick one or more scanned PDFs. You can unlock password-protected files right here.
Choose the language
Set whether the text is English, German or both, and limit the pages if you like.
Recognise text
The pages are read one after another. Then save the searchable PDF and copy the text or save it as TXT.
Why FaLe?
The scan stays as it is
An invisible text layer is placed over each page. The image is not recompressed – the PDF looks exactly as before.
No upload
Text recognition runs in your browser. Your documents never leave your device – not even for recognition.
Thought through
Pages that already contain text are skipped. Pages scanned sideways or upside down are detected when they read poorly.
Offline afterwards
The recognition data is downloaded once the first time. After that, text recognition also works without internet.
| FaLe Tools | Typical online tools | |
|---|---|---|
| Where are files processed? | In your browser, on your device | Usually on the provider’s server – the file is uploaded |
| Account needed? | No | Often for larger files or more runs |
| Limits | No artificial limits | Often daily limits or size caps in the free version |
| Watermark | Never | With some providers in the free version |
| Works offline | Yes, after the first visit | No |
| Ads and tracking | None | Often banner ads and analytics services |
Frequently asked questions
What does “searchable” mean?
After recognition you can search the PDF for words, select text and copy it. The pages look the same; the text sits invisibly over the scan.
Which languages are recognised?
English and German, also mixed. Other languages using the Latin alphabet partly work, but not reliably.
How accurate is it?
Very good on clean scans at 300 dpi. Handwriting, very small print, photos with strong shadows or tables with many lines are recognised less well. Check important parts of the result – you can correct the text before copying it.
How long does it take?
A few seconds per page, depending on your device. The first time, loading the recognition data (about 5–7 MB) adds to that. You can cancel at any time.
Are my files uploaded?
No. The PDF is read in your browser and text recognition runs on your device. Nothing is sent to a server.
What happens to pages that already contain text?
By default they are skipped and kept unchanged, so no text is duplicated. You can switch this off.
More features in the FaLe PDF app
Sign, fill in forms, scan with text recognition and more than 40 other tools – free, no account and offline, too.