Google has announced a major update to Gemini Live, its real-time, two-way voice AI, introducing visual guidance, broader app integration, and enhanced voice models aimed at making conversations feel more natural.
Highlights
- Visual Guidance: AI can now highlight relevant objects on-screen with white-bordered rectangles, improving user interaction and clarity.
- Expanded App Integration: Gemini Live now works with Phone, Messages, and Clock, enabling calls, messages, alarms, and reminders during live conversations.
- Enhanced Voice Models: Natural-sounding speech with improved intonation, pitch, rhythm, tone adaptation, tempo control, and character accents for storytelling or narration.
- Device Availability: Launching on Pixel 10 series from August 28, rolling out to other Android devices shortly after, with iOS support arriving later.
- User Benefits: Hands-free multitasking, clearer visual interaction, and more human-like AI conversations compared to competitors.
This update also marks the first time Gemini Live will integrate with Google’s Phone, Messages, and Clock apps, enabling practical tasks like making calls, sending messages, and setting alarms during live interactions.
Visual Guidance Enhances Interactivity
A key feature of the update is visual guidance, which allows Gemini Live to highlight specific objects on the screen in response to user queries.
When multiple objects appear in a video feed, the AI can draw a white-bordered rectangle around the relevant item, helping users follow along more easily.
This feature was first teased at Google I/O 2025 and will debut on the Pixel 10 series on August 28, coinciding with the smartphone’s launch.
Support for other Android devices will roll out the same week, while iOS users can expect the feature at a later, unspecified date. Importantly, access to visual guidance does not require a Google AI Pro or Ultra subscription, making it available to a wide range of users.
Expanded App Integration for Seamless Task Management
Gemini Live’s app support is also expanding. Previously limited to Calendar, Keep, and Tasks, the AI now integrates with Phone, Messages, and Clock, enabling users to manage tasks without leaving the conversation.
Users will be able to,
- Make calls mid-conversation
- Send messages without switching apps
- Set alarms or reminders directly through the AI
Enhanced Voice Models for Natural Conversations
Google has refined Gemini Live’s voice models to make interactions more expressive and human-like.
- Intonation, rhythm, and pitch enhancements for natural-sounding speech
- Tone adaptation, allowing the AI to adjust its voice based on context
- Tempo control, enabling faster or slower speech
- Character accents, useful for storytelling or narration
Users can now manage tasks, explore objects on their screens, or enjoy storytelling with greater clarity and responsiveness.
The rollout begins with the Pixel 10 series on August 28, followed by other Android devices, with iOS support to arrive later.
