Everything happens in your browser. Nothing is uploaded.

OCR PDF

Recognize text in scans so you can search, select and copy it.

How it works

  1. Choose a scanned PDF or image.
  2. Select Recognize text. The first time, your browser loads the English text model (about 10 MB) from this site.
  3. Download the searchable PDF, or copy the recognized text.

Private by design

This tool runs entirely in your browser. Your files are never uploaded to a server, and the page is technically blocked from sending them anywhere.

Security and privacy

Questions and answers

Is OCR done on a server?

No. Text recognition runs in your browser, on your device. The first time, your browser loads the English text model (about 10 MB) from this site; your pages are never uploaded.

Which languages does it recognize?

English only for now. Text in other languages may be recognized poorly or not at all.

Will my PDF look different?

No. The recognized text is laid invisibly over each page, so the pages look the same, but you can now search, select and copy the text.

Can I use photos or scanned images instead of a PDF?

Yes. Select Choose images under the file picker. The images are combined into one searchable PDF.

How accurate is it?

It depends on the scan. Clear, upright pages in good light give the best results, so check important text before you rely on it.

Why does it take a while?

Each page is recognized on your own device, so the speed depends on its processor. Long documents take longer, and you can follow the progress page by page.

Guides

Related tools