Visionauta is an accessibility app that uses artificial intelligence to help blind and low-vision people in their daily lives.
Key features:
π Text Reading
Point the camera at any document, packaging or sign and the app reads the content aloud. Works offline.
π€ Ask AI
Ask questions about what the camera is seeing. What product is this? What's the expiry date? Describe this scene. The AI answers aloud.
π Object Finder
Say "help me find my phone" and the app continuously searches for the object, alerting with vibration when found.
π΅ Money Recognition
Automatically identifies Real, Dollar, Euro and other banknotes.
π Electronic Magnifier
Enlarge text with different contrast modes for low vision, including white on black, yellow on blue and more.
π¦ Flashlight
Illuminate environments with vibration reminder.
β¨ New in version 3.0
Interactive tutorial with 7 pages to learn all app features.
Visionauta was developed with the independence and autonomy of visually impaired people in mind.
βΏ Visionauta Magic and the accessibility service
Visionauta Magic uses Android's AccessibilityService API for two things, and only with your permission:
1) Describing unlabeled images. When your screen reader lands on an image, icon or card with no description, Visionauta crops just that element and speaks a description of it. It needs the accessibility service because that is the only way to know where the screen reader focus is and to read the visual content on screen.
2) Running tasks you asked for. You ask out loud, Visionauta shows a plan on screen, and nothing happens until you tap "Confirm and run". Only then does it open apps, tap buttons, type and scroll for you β narrating every step. You can stop it at any moment with the STOP ACTIONS button, the two-finger gesture, or by saying "stop".
Visionauta Magic is off by default. You turn it on in the app settings and then in Android's Accessibility settings, and you can turn it off the same way at any time. The service does not collect or share your screen content for any other purpose: the image crop is sent only to generate the description you asked for and is not stored on our servers. Keyboards, protected windows and password screens are ignored.
π Available in 16 languages
Portuguese, English, Spanish, French, Korean, German, Italian, Japanese, Arabic, Russian, Simplified Chinese, Turkish, Vietnamese, Hindi, Indonesian and Bengali. The interface, AI, commands, OCR and online voices adapt to your language. The on-device Natural Voice remains available in the five languages supported by its model.
AI-powered visual assistant for visually impaired people.