Image to text (OCR)

Pull the words out of a photo, screenshot, or scan. The recognised text appears below, ready to copy or download.

Tap to choose an image — or drag it here
JPG, PNG, WebP and most other image types.

About this tool

OCR (optical character recognition) scans the shapes in an image and works out which letters they form. This tool uses the Tesseract recognition engine running entirely in the browser: after the engine and its English language data download once (about 10 MB), recognition happens on the device and the image is never uploaded. Clear, straight, high-contrast text recognises best; handwriting and decorative fonts are hard for any OCR and results there will be rough.

Common questions

Which languages are supported? English. Tesseract supports many more, but each language adds a large data download, so this tool ships English only.

Why are some words wrong? OCR reads shapes, not meaning. Blur, glare, low contrast, and text photographed at an angle all cost accuracy — a flat, well-lit screenshot gives the cleanest result. Always proofread anything important.

How this one works: the Tesseract OCR engine and its English language data are downloaded once from the jsDelivr CDN so recognition can run locally. The image itself stays on the device and is never uploaded.