TextSnap Offline OCR

TextSnap Offline OCR turns printed words in camera views and existing images into editable text on your Android device.

Scan your way
• Point the camera at a document and align it inside the scan frame.
• Import a picture from the system photo picker.
• Try the built-in sample to see the workflow without supplying a document.

Five offline recognition models
Choose Latin, Chinese, Devanagari, Japanese, or Korean recognition before scanning. The selected ML Kit model is bundled with the app, so recognition can run on the device without uploading the image to an OCR service.

Useful text actions
Review and edit recognized text before using it. Copy the result to the clipboard, share it through compatible Android apps, or save it to your private on-device scan history. Reopen a saved scan when you need it and clear the history at any time.

Built around privacy
Recognition runs locally. Saved scan text stays in this app's private storage until you clear it or uninstall the app. Camera access is requested only when the camera scanner needs it, and gallery access uses Android's system picker.

Recognition notes
Results depend on lighting, focus, image resolution, print clarity, text orientation, and the selected script model. Mixed scripts, decorative fonts, handwriting, curved pages, glare, and blurred images may reduce accuracy. Review important text before relying on or sharing it.

Internet access is used by the advertising SDK to request ads; it is not required for the bundled text-recognition models.