Models

Google launches Guided Vision for Gemini Live on Android

Google has launched Guided Vision in Gemini Live, a new real-time audio description tool designed to help blind and low-vision Android users interact with their physical surroundings.

The Verge AI11 hrs agoModels
Image: The Verge AI

Google has officially rolled out Guided Vision, a new accessibility feature integrated into Gemini Live for compatible Android devices. The tool leverages the smartphone camera to provide real-time audio descriptions of the user's immediate environment. By sharing their live camera feed with the Gemini assistant, users can receive spoken assistance with reading fine print, identifying nearby objects, and understanding visual details in their surroundings.

The system supports interactive, multi-turn conversations, allowing users to ask follow-up questions about what the camera sees. For example, after the AI helps locate an item, a user can ask Gemini to read its expiration date. To assist users in framing their shots, Google's AI provides real-time audio cues to help align the camera if the target object is out of view. Beyond the Gemini app, Guided Vision is integrated into Google TalkBack and can be mapped to an accessibility shortcut on devices running Android 9 and above, following its initial reveal in recent Pixel updates.

This release positions Google to compete more directly with Apple, which offers a similar VoiceOver Live Recognition feature on the iPhone and Vision Pro. However, Google has issued clear safety guidelines alongside the launch. The company explicitly cautions users not to rely on Guided Vision for critical tasks such as physical navigation, safe-travel guidance, or obstacle detection, emphasizing that the software should not serve as a replacement for traditional mobility aids like canes.

For developers and accessibility practitioners, this integration represents a significant step forward in multimodal AI utility. By combining real-time computer vision with conversational large language models, Google demonstrates how on-device AI can move beyond simple text generation into active, contextual environmental assistance. This sets a new baseline for how mobile operating systems can leverage live video feeds to make daily tasks more accessible.

This is our own summary of reporting by The Verge AI

More in Models