Extract text from images
Recognize printed text in photos in your browser, then edit and copy the result.
Add a document photo or an image of a sign and start in printed text mode. Input supports JPG, PNG, WebP, AVIF and HEIC/HEIF. Select files, drop them or paste an image. An error appears if the image cannot be decoded.
Select the languages present in the image. The engine and selected language data download when first needed. The image itself stays in your browser and is not sent to a server.
Compare the original line boxes with the text. Edit the result or apply text cleanup, then copy or save as TXT. Blurry photos, tilted text, vertical writing and complex tables can produce errors.
How to use
- Select or drop an image, or paste an image from the clipboard.
- Check printed text mode and the text languages, then start extraction.
- Compare and edit the result, then copy it or save as TXT.
What to check
Try cropping to the relevant area or rotating the image. Adjustments cannot restore missing text information or guarantee accuracy.
Frequently asked questions
Can I add several images?
You can select multiple files. They share recognition settings, so check the languages if the files differ.
Can I edit the result?
Yes. Edit the result text directly. Copy and TXT export use the edited text.
Is confidence the actual accuracy?
No. It is an engine estimate, not a measured proportion of correct answers. Some recognition methods do not provide confidence.