Image to text, also called OCR, turns a photo or a scan into text you can copy. It is the quick way out when the words you need are stuck in a picture: a letter you photographed on the kitchen table, a screenshot of a chat, a receipt, a page from a book, a scanned PDF from the bank, a sign or a menu.
Drop the file and get plain text back. Columns come out in reading order, tables come out row by row, and words that can't be read are marked with […] instead of being guessed. Handwriting works when a person could read it easily; messy notes or faded ink may come back with gaps. With old documents, the spelling can come out slightly modernized, so compare names and dates with the original.
It takes photos and screenshots in JPG, PNG, WebP and HEIC from an iPhone. Large photos and HEIC files are converted to JPG in your browser first, so less data goes up. You can also extract text from PDF and Word files up to 15 MB; with a scanned PDF, up to 15 pages are read. If a PDF already has real text in it, that text is taken directly.
For the best result, take the photo straight on, fill the frame with the page and avoid shadows and glare. A photo takes 2 or 3 seconds; a scanned PDF of 3 pages took about 8 seconds in our tests.
The file is processed on our server and deleted from there after 24 hours. The extracted text is saved in your library, which lives only in this browser.
Once you have the text, you can keep going without uploading anything again: translate a letter you can't read, fix the spelling of a note you typed up, or reply to a message that came as a screenshot. If you need a whole PDF translated as a file rather than its text, the PDF translator is the better fit.