OCR: make a scanned PDF searchable

How to run OCR

1. Choose the scanned PDF (or several JPG/PNG pictures).
2. Pick the language and resolution. For a PDF you can read just some pages.
3. Press "Read the text and save" and wait until every page is done (you can cancel). You get a new PDF whose text can be searched and copied.

Your files stay on your device

This tool opens the file from your device and reads the text with Tesseract, which runs inside your browser (WebAssembly). The program and the language data load from freesabai.com only. Neither your files nor the text that is read are sent anywhere, and nothing is kept in the browser: close the page and it is gone.

Good to know

• Accuracy is limited. Clear printed English reads well. Thai reads well enough to search but is often wrong, especially vowels and tone marks, small or italic print, handwriting, crumpled paper or a crooked scan. It will not match services that read on a server, so check the text before relying on it.
• The new PDF holds the original page picture with the text hidden underneath: it looks the same and can be searched and copied. A misread word stays misread in the file, but it does not change what you see.
• Pages that are read are drawn again from a picture at the resolution you choose, so the file may come out larger or smaller. Pages that already have text are kept as they are (you can turn that off).
• It takes time and works your device hard. Keep this tab open until it finishes; a file with dozens of pages may not finish on an older phone.
• Languages other than Thai and English are not supported yet.
• Files locked with a password need to be unlocked first.

Report a problem or suggest an idea

Found a problem, or want this tool to do more? Tell us on the feedback page. No sign-up needed.