이미지에서 텍스트로 (OCR)

Drop a PNG, JPG, or WebP image onto the tool — a screenshot, a photo of a page, or a scanned document. It detects every line of text and converts it to editable, selectable characters in a few seconds. You can then copy the text or download it as a .txt file. Everything runs inside your browser, so the image is never uploaded to a server.

It is free to use and needs no account. Because the recognition runs entirely on your device, there is no server cost per scan and no signup wall — just drop an image and get the text back.

The model recognizes 50+ languages, including English, Chinese, Japanese, and Latin-script languages such as Spanish, French, German, and Portuguese. Mixed-language images (for example English and Chinese in the same screenshot) work too.

Yes. The OCR model and runtime are downloaded to your browser once, then every image is processed locally on your own machine. Your pictures and the extracted text never leave your device, which makes this a good fit for sensitive screenshots and documents.

Clear, reasonably high-resolution images of printed or on-screen text give the best results: screenshots, slides, book or document pages, and signs. Very small, blurry, heavily stylized, or handwritten text is harder to read accurately. If a result looks off, try a sharper or larger image.