Google has quietly rolled out a new "Memory" feature for Gemini Live, significantly boosting the AI assistant's contextual awareness. The update is currently rolling out in phases to compatible Android devices, including Galaxy smartphones and tablets.

This new upgrade allows Gemini Live to access and leverage past conversations, enabling it to deliver highly personalized responses. Additionally, the AI can now pull relevant data from linked apps and services within a user's ecosystem. Following a limited release after last month's Google I/O 2026 developer conference, the feature is now expanding to select users in the US.

According to Google, Gemini Live is now deeply integrated into the chat interface, allowing users to seamlessly transition between voice and text inputs at any time. The assistant can also provide real-time visual information, such as location data from Google Maps, weather cards, or images directly from Google Photos.

Gaining memory access puts Gemini Live nearly on par with the standard Gemini model's capabilities. What sets the Live version apart, however, is its ability to directly scan on-screen content or utilize the device's camera to analyze the surrounding environment. This allows the AI to answer questions and provide highly accurate, context-aware support on the fly.