Recognition runs entirely in your browser — the image is not sent to any server and works without internet/CDN.
Drag an image here or click to choose
You can also paste from the clipboard (Ctrl+V). PNG, JPG, WEBP, BMP.
0 Selected
Enable "Math / equations" together with a text language (e.g. English + Math / equations) for best results.
The tool recognizes printed text from images (screenshots, photos, scans) and converts it into
editable text. Recognition runs entirely on your device via Tesseract.js (WebAssembly): the image is never
uploaded anywhere, so it is suitable for confidential documents and works even offline after the first load.
Cyrillic (Russian, Ukrainian) and Latin (English) scripts are supported. The "Auto" mode recognizes mixed
Cyrillic and Latin text at once. For tables keep the "Preserve layout" option enabled — the engine keeps
inter-word spaces so columns stay aligned.
How to use: drag an image into the upload area (or paste it from the clipboard with Ctrl+V), pick a language
and press "Recognize". The resulting text can be copied or downloaded as a .txt file.
Comments