OCR PDF

Add a searchable text layer to a scanned PDF, without changing how it looks.

Drop a PDF here

Pages appear straight away — nothing is sent anywhere.

Read on this device. No account, no upload.

Settings

Nothing starts until you press the button, and nothing is sent when it does.

Pages

All pages—Every page in the document, in order.First page only—Check the settings on one page before converting a long document.Odd pages—Pages 1, 3, 5 and on: the fronts of a double-sided scan.Even pages—Pages 2, 4, 6 and on: the backs of a double-sided scan.

Recognition languages

EnglishIn use4.1 MB
German1.5 MB
Spanish2.3 MB

Recognition needs at least one language.

Adding a language downloads nothing. Running the job fetches 4.1 MB of language data once, from this site, and the device keeps it for next time.

What you get

Searchable PDFRecommended.pdfThe scan looks identical and its text becomes selectable and findable.Text only.txtThe recognized words on their own, without the page images.

Advanced

Detect page rotationLosslessChecks each page's rotation and reads a sideways or upside-down scan the right way up. A page that needs turning is read twice, so it takes about twice as long.

Every page reports how sure recognition was. Pages that average below 95% are flagged in the result instead of being presented as correct.

What OCR can read

Add a file to start

Runs in this tab. Your file is not uploaded.

Your account