OCR adds a text layer you can select, search and copy under the scan, without changing a pixel of how the page looks. The page is never sent anywhere. OCR is one of two paid tools here: 1,000 credits ($5) once, together with watermarking.
A scanned PDF is a picture of a page. You can read it, but you cannot search it, select a sentence or copy a figure out of it, because the file holds pixels and no text.
OCR reads the words in the picture and adds them as a searchable, selectable text layer under the scan. The page looks exactly as it did — not a pixel of it changes — but find-in-document now works, and text can be selected and copied.
Almost everything in OpenPdfEdit is free. OCR is one of the two exceptions: it and watermarking are Supporter tools, and they unlock together, once, for 1,000 credits.
It is not a subscription and not a charge per use. Credits never expire, and the unlock is remembered by your account, so reinstalling or moving to a new device does not charge you again. The account exists only to make the unlock follow you; everything else works signed out.
The page being recognised is never sent anywhere. Recognition happens on your machine like every other edit.
In the web app, the recogniser and its language data are downloaded once — about 3 MB, from openpdfedit.com itself rather than a third-party CDN — and then cached, so OCR works offline afterwards. The desktop app has the recogniser built in and fetches nothing at all.
Twelve languages ship with it, including Simplified and Traditional Chinese, Japanese, Korean, Arabic and Russian.
In the web app, in a browser tab, and in the desktop app for macOS and Windows.
Not in the Chrome and Edge extension. It is the only tool the extension does not have: the recogniser is a one-time download, and that is exactly the kind of remote code a Manifest V3 extension may not fetch.
Search the document, select and copy text, and export it to Markdown. A scan exports as paragraphs rather than headings and lists: nothing in a picture of a page says which line was a heading.
A scan can also be redacted without OCR — the covered pixels are overwritten — and that is free.
In a browser tab, with nothing to install. The scan stays on your machine.
Guides: How to convert a PDF to Markdown · How to compare two PDFs · All guides
No. OCR and watermarking are the two paid Supporter tools. They unlock together, once, for 1,000 credits ($5). It is not a subscription and not a charge per use, and credits never expire. Every other tool is free.
No. Recognition runs on your own device and the page being recognised is never sent anywhere. In the web app the recogniser is downloaded once, about 3 MB from openpdfedit.com itself, and then works offline. The desktop app has it built in.
Twelve languages ship with it, including Simplified and Traditional Chinese, Japanese, Korean, Arabic and Russian.
No. OCR adds a searchable, selectable text layer under the scan without changing a pixel of how the page looks.
No. OCR is in the web app and the desktop app. It is the only tool the extension does not have, because a Manifest V3 extension may not fetch the recogniser as remote code.