OCR a scanned PDF

Recognise the words on scanned pages and photos of documents.

or drop it anywhere

How to OCR a scanned PDF

  1. Open your scan

    Pick the PDF, or a photo, which becomes a PDF.

  2. Recognise

    Press Recognise; the words are read on your device, using its graphics chip where it can.

  3. Edit or convert

    Click any word to edit it, or convert the document to Word, Excel or text.

OCR PDF in your browser

A scanned page is a picture of text: it cannot be searched, copied or edited. OCR, optical character recognition, finds the words in the picture and adds them to the page as text. OnePDF runs PaddleOCR's recognition models in your browser, on your own device, in about 50 languages, among them English, Chinese, Japanese, French, Spanish, German and Portuguese.

Once recognised, the words can be searched and copied, edited like any other text, and converted with PDF to Word, Excel or Text. Photos of documents work too: a picture opens as a PDF first.

Why use OnePDF to OCR a scanned PDF

  • Modern recognition

    PaddleOCR's PP-OCRv6 models, which read English, most European languages, Chinese and Japanese.

  • Edit the scan itself

    Recognised words can be retyped right on the scanned page, keeping the paper behind them.

  • Private

    Recognition runs in your browser; the scan is never uploaded to an OCR service.

Questions about OCR PDF

Which languages does OCR support?

About 50: English, Chinese, Japanese and dozens of Latin-script languages such as French, Spanish, German and Portuguese. The models suited to your browser's languages are chosen for you. Letters with dots below them, as in Yoruba, are not recognised yet.

Does OCR run on a server?

No. The recognition models run in your browser, on your own device: text recognition never uses a server, signed in or not.

Can I convert a scan to Word after OCR?

Yes. Once the words are recognised, PDF to Word, Excel and text all use them.

More questions

Guides