How to OCR a scanned PDF
- Drop a scanned PDF on the box.
- Choose the document's language, and a second language if it is mixed.
- Tick Also create a searchable PDF if you want a PDF in which the text can be selected, searched and copied.
- Press Recognise text. Each page is rendered and recognised; the progress bar shows how far it has got. Then copy the text or download the text file and searchable PDF.
Why make PDFs searchable?
A scanned PDF is just a stack of pictures: you cannot search it, copy a sentence from it or have it read aloud by a screen reader. Optical character recognition adds an invisible text layer behind each page image, so the document looks the same but behaves like a digital file. That makes archives of letters, contracts, receipts and old reports findable with Ctrl+F and in your computer's search, and makes them accessible.
Recognition uses Tesseract, an open-source OCR engine, running in your browser with WebAssembly. It works best on clean, straight, printed pages scanned at 200–300 dpi. Large documents take a while, because every page is processed on your own device rather than on a server; in exchange, your documents stay private.
FAQ
How long does OCR take?
Usually a few seconds per page on a modern computer, longer on phones and for dense pages.
Is the searchable PDF the same size?
It can be larger, because each page is stored as an image plus a text layer. Use Compress PDF if needed.
Which languages are supported?
More than 30, including English, Spanish, French, German, Chinese, Japanese, Arabic and Hindi.