Kitvilo

Image to text · Automatic language detection OCR

Extract text from photos and screenshots with an estimated language. You can also choose a language; images are processed on this device.

PNG / JPEG / WebP · 10MiB · 8MP

Result

OCR can contain errors. Compare with the original and correct it. Reloading clears the image and result.

Usage and limits

One PNG/JPEG/WebP, up to 10MiB and 8MP. Supports printed text in 20 languages. Automatic detection is an estimate and can be wrong for short or mixed text. Choose a language and retry if needed. No PDF, handwriting or table reconstruction. The first run downloads the engine and required models; speed depends on the device. Images are not sent to a server and results are not saved automatically.

Sources consulted

Tesseract.js ↗