Copy text that was trapped in an image
A screenshot of an error message, a photo of a book page, a scanned letter — drop it above and the text becomes selectable, editable, and copyable. Recognition runs locally with a progress bar; typical screenshots take a few seconds.
Because nothing is uploaded, it's safe for documents you'd never put into a random OCR site: IDs, contracts, and medical paperwork.
Frequently asked questions
Is my image uploaded for OCR?
No — this is rare for OCR tools. The recognition engine (Tesseract, compiled to WebAssembly) runs inside your browser. The engine itself downloads once on first use, but your image never leaves your device.
How accurate is it?
Very good on clean sources: screenshots, scans, and well-lit photos of printed text. Accuracy drops with handwriting, stylized fonts, or blurry photos. The output is editable, so you can fix stray characters before copying.
What languages are supported?
English, Spanish, French, German, Portuguese, and Italian — pick the language before dropping the image so the right recognition model is used.