Scribe pulls the words out of a screenshot, a scanned page, or a photo of a sign — the kind of text you can see but can't select or copy. Drop or paste an image, pick its language, and it reads the letters right here in your browser and hands you plain text you can edit, copy, or save. The image never leaves your device; only the one-time recognition model is fetched.

Image with text in it

Drop an image here.

PNG · JPG · WebP · or paste with Ctrl/⌘+V · no upload

Language of the text

Sharpen small text High-contrast (black & white) Layout

Recognition happens on your device — the picture is never uploaded. The first time you use a language, its recognition model (a few megabytes) is downloaded once from a public code CDN and cached by your browser for next time. Accuracy depends on the image: clear, straight, high-contrast text reads best; handwriting, heavy stylisation, and blurry photos are harder. Everything runs on your device.