Extract text from images, screenshots, and scans locally using OCR.
Drop an image here or click to select
Your file never leaves your device · No signup · No daily limit
Up to 48 MB per file
First use loads the OCR engine. Later use can work from your browser cache.
No. Recognition runs with Tesseract compiled to WebAssembly in your browser, so the image and the extracted text stay on your device. Unusual for OCR, which is normally a cloud service.
Twelve: English, Turkish, German, French, Spanish, Portuguese, Italian, Polish, Russian, Arabic, Japanese and Korean. Pick the language before you run it so the engine loads the right model; for a page that mixes in English, tick “Also read English words”. Each language downloads once, only when you pick it.
OCR is not perfect on low-quality or stylized input. A sharper, well-lit, straight-on image of printed text gives the best results.
