Google Launches Gemini 3.1 Flash Live: Voice AI's Major Upgrade

Google has officially announced Gemini 3.1 Flash Live, calling it the "highest-quality audio and voice model" to date. This model is designed to provide core upgrades for Gemini Live and Search Live, and is now available for preview through the Gemini Live API in Google AI Studio.

This upgrade brings several significant improvements: the model can more effectively recognize acoustic details such as tone and rhythm, and offers lower latency compared to the previous generation 2.5 Flash Native Audio. It performs better in distinguishing relevant speech from environmental noise (like traffic or TV sounds), with enhanced background noise filtering capabilities. In the Gemini Live apps on Android and iOS, 3.1 Flash Live achieves faster response speeds, reducing awkward pauses in conversations, and can double the conversation thread tracking time, helping to maintain coherent thinking during extended brainstorming sessions. Additionally, Gemini Live can now dynamically adjust response length and tone to match real-time situations.

Google is also rolling out the Search Live feature globally in over 200 countries or all languages and regions where AI Mode is available, based on this model. This move marks the further popularization and optimization of Google's voice interaction technology.

This upgrade not only enhances the real-time performance and accuracy of voice AI but also lays the foundation for more natural and fluent human-machine dialogue experiences, indicating that the potential of smart assistants in daily applications will be further expanded.