Extract Text from Images
Optical character recognition (OCR) reads the words in a picture. Upload a photo of a page, a screenshot, a notice or a scan, choose the language and get editable text. Recognition runs on your device with Tesseract, so your image is never uploaded.
What you can do
- English, Nepali (Devanagari), Hindi, or English and Nepali together.
- Improve-contrast mode for faint, shadowed or uneven photos.
- Edit the result, copy it or save it as a .txt file.
- Works with photos, screenshots and HEIC pictures.
Good to know
- Language data is downloaded the first time you use a language, so the first run takes longer.
- Accuracy is best on sharp, straight, high-contrast text. Handwriting and decorative fonts are hard for OCR.
- Always proofread the output, especially numbers and names.