About Us

Why we built it

A lot of everyday text in India is locked inside images: land records, government circulars, old books, newspaper cuttings and photos of notices. Most OCR tools handle English well but treat Indian scripts as an afterthought, or ask you to upload private documents to a server.

We wanted a simple tool that works well for Indian languages, costs nothing, needs no account, and keeps your documents on your own device.

How it works

Text recognition runs inside your web browser using open-source OCR engine. PDFs are opened with Mozilla's PDF.js. Each page is turned into an image and then read by the OCR engine.

Your images and PDFs are never uploaded. The only downloads are the OCR engine and the language model for the language you pick. Your browser fetches these once and then caches them. See our Privacy Policy for details.

Supported languages

Hindi, Marathi, Nepali, Sanskrit, Bengali, Assamese, Tamil, Telugu, Kannada, Malayalam, Gujarati, Punjabi, Odia, Urdu, Tibetan, and English.

What to expect

OCR accuracy depends mostly on the quality of the image. Clear, printed text gives very good results. Blurry photos, decorative fonts, very old print and handwriting give weaker results. You can correct any mistakes in the result box before copying the text.