Read Text from a Scan

Get the words out of a scan, where copying does not work.

Drop files hereor click to choose

    About this converter

    A scanned document contains no text at all — only a picture of a page. That is why selecting and copying does nothing, and why every ordinary text extractor comes back empty.

    This reads the shapes of the letters and works out the words, the same way a person does.

    Frequently asked questions

    How accurate is it?

    Very good on a clean, straight scan of printed text. Less good on faint photocopies, tight columns, unusual fonts or handwriting — handwriting in particular is not supported. Turn on the confidence readout to see how sure it was.

    Which languages does it read?

    English, French, German, Portuguese and Spanish: choose the one the text is in from the menu. Each language other than English is a download of up to about 2 MB the first time you use it. Text in a language not on the list, or with the wrong one chosen, comes back wrong rather than as an error, so it is worth checking the first few lines.

    Do my PDFs get sent anywhere?

    No. Everything happens on your own device, so your documents never leave it. That matters more than usual for PDFs, which tend to be contracts, statements and scans.

    Does it work with password-protected files?

    Not directly. If you know the password, take it off with Unlock PDF first, then use the unlocked copy here.

    Related tools