Image to text: extract text from a picture or screenshot online, free
Turn a photo, scan or screenshot into editable text in your browser. Paste or drop an image, pick the language, copy the text. Nothing is uploaded.
100 % in your browser: your files and text never leave your device and nothing is uploaded.
The first image downloads the reader (about 5 to 7 MB) and can take a few seconds. Nothing is uploaded.
How it works
Drop an image, choose it from your files, or just paste a screenshot with Ctrl+V (Cmd+V on Mac). The text is read on your device with an OCR engine, so the image is never uploaded. Pick the language of the text first (it defaults to the page language) for the best result; changing it re-reads the same image. Then edit the text if needed, copy it or download it as a .txt file. Clear, straight, high-contrast text works best; handwriting and photos at an angle are hard for any OCR. The engine is downloaded only when you add an image (about 5 to 7 MB the first time). It uses Tesseract (Apache-2.0) through tesseract.js, with the Tesseract language data (Apache-2.0).
How to use it
- Drop an image on the drop area, choose one from your files, or copy a screenshot and paste it with Ctrl+V (Cmd+V on Mac).
- Check Language of the text. It defaults to the page language; you can pick Spanish, English, Catalan, French, German, Italian or Portuguese. Changing it re-reads the same image.
- Wait a few seconds. The first image downloads the reader (about 5 to 7 MB), and a progress percentage is shown while the text is read.
- Read the Recognised text and the confidence message under it. Correct any mistakes directly in the text box.
- Press Copy to put the text on the clipboard, or Download .txt to save it as a file.
Worked example
You screenshot a paragraph of a scanned letter in Spanish and paste it with Ctrl+V. With the language on Spanish, the text appears after a few seconds with a message such as Confidence 93 %: the text should be accurate. Between 60 and 84 % the message asks you to check for mistakes, and below 60 % it warns of errors. A straight, sharp screenshot of printed text at 1200 px wide normally reads well, while a crooked photo taken in dim light may need a retake.
Frequently asked questions
Is my image uploaded to a server?
No. The text is read on your device with the Tesseract OCR engine running in your browser. Only the engine and the language data are downloaded, the first time. Your image never leaves your computer or phone.
Why is the result full of mistakes?
Check that the Language matches the text, because a wrong language gives nonsense. Then use a sharper, higher resolution image, with straight lines and good contrast between text and background. Handwriting, fancy fonts and photos taken at an angle are hard for any OCR.
Which images and languages work?
PNG, JPG, WebP, BMP and GIF images. The languages are Spanish, English, Catalan, French, German, Italian and Portuguese. PDFs are not read here, only pictures.
Does it keep the layout, tables or columns?
No. You get plain text, line by line. Tables and multiple columns come out as lines of text that you will need to arrange yourself.
Can I use it on my phone?
Yes. Pick a photo from your gallery or take one first, then choose it. The first run needs about 5 to 7 MB of data, so use wifi if you are limited. Reading a large photo can take several seconds on an older phone.