Extract Text from Image (OCR)
Upload an image containing text, pick its language and extract the content in seconds. Recognition runs entirely on your device.
Your image is not sent to any server — all recognition happens on your device.
Your files never leave your device. This tool downloads its language model the first time; after that, it works offline.
How it works
OCR (optical character recognition) is the process of looking at the pixels of an image and deciding which of them form letters. The engine used here works with a neural network trained per language — which is why picking the right one changes the outcome: the model uses that language's vocabulary and patterns to disambiguate lookalike shapes, such as a zero and the letter O. Image quality matters more than anything else: sharp, high-contrast, straight text is recognised almost perfectly, while a blurry, tilted or shadowed photo drops accuracy noticeably. On the first run the browser downloads the engine and the chosen language pack, which takes a few seconds; after that it is cached and later extractions start immediately.