AI OCR — Image to Text
Pick an image and extract text with on-device OCR.
100% client-side
Pick an image to begin.
How to use
- Pick the language of the text in your image.
- Choose an image file (photo, screenshot, scan).
- Click Extract text — first run downloads the model.
Related AI tools
常見問題
- Is anything uploaded?
- No. The Tesseract.js WebAssembly engine and language model are served from this site and run entirely on your device.
- Why is the first run slow?
- The English language model (~10MB) downloads the first time. After that it stays in your browser cache and subsequent OCRs are fast.
- Which languages work?
- English by default. The same engine supports many languages — you can switch models in the dropdown.