Extract editable text from images — scanned pages, screenshots or photos. Recognition runs entirely in your browser, so your images never leave your device.
Read the words out of a photograph or scan so you can copy and edit them. The recognition engine runs inside your browser.
No. The recognition engine is downloaded to your browser and the picture itself never leaves your device.
The recognition engine is around 10 MB and is fetched once, then cached. Later uses start immediately.
Clear, straight, well-lit text in a common typeface reads very accurately. Handwriting, skewed photographs and low-contrast scans are much harder and may need correcting.