Image to text
We recognize text in a photo, screenshot or scan and give it to you ready to copy. It works right in your browser, images are not uploaded to a server.
Changed the language? Click "Recognize again". The first run takes longer: the recognition engine and language files are downloaded, then they come from the cache.
Your images stay with you. Recognition runs on the open-source Tesseract library right in your browser. We never see your images or text, so you can safely process documents, receipts and chats.
Text from an image in half a minute
- 1Choose a languagethe page language and English are already selected
- 2Add an imagedrag it in, choose a file or paste a screenshot with Ctrl+V
- 3Get the textfix it if needed, copy it or download a .txt
How to get an accurate result
Printed text works best: screenshots, scans, photos of documents and screens. Shoot straight, without glare and in good light, with the text filling most of the frame. We enlarge small images before recognition. Handwriting and text on busy backgrounds are recognized poorly.
The language choice matters
Select the languages that appear in the image: Russian and English, Japanese and English, and so on. Without the right language, letters turn into similar-looking symbols. You cannot select more than three languages at once: recognition gets slower and less accurate.
Where it comes in handy
Copy text from a screenshot or a video, pull bank details from a scanned contract, move a table from a photo into a document, translate a label on packaging. You can paste the recognized text straight into a translator or a word counter.
FAQ
Didn't find an answer? Message us and we'll help.
No. Recognition runs right in your browser: only the program itself and the language files are downloaded from our server. Your images and text are not sent anywhere.
Russian, English, Spanish, Portuguese, German, French, Italian, Turkish, Ukrainian and Japanese. You can select up to three languages at once if the image has text in several languages.
On the first run the browser downloads the recognition program (about 4 MB) and a file for each selected language (1-3 MB). After that they come from the cache and an image is recognized in a few seconds.
Poorly. The tool is made for printed text: documents, screenshots, books, signs. Neat hand-printed letters are sometimes recognized, normal handwriting almost never.
Check that the right language is selected and try a sharper photo: straight, without glare, with large text. Crop away everything around the text. You can fix small mistakes right in the result box before copying.
Only images here. If a PDF already has a text layer, you can simply select and copy it. For a scanned PDF, take a screenshot of the page or save it as an image and add it here.
Try also
Also free, no sign-up