Google has officially announced the launch of the Gemini 3.1 Flash Live model, calling it its highest-quality audio and voice model to date. This model will serve as the core engine for the Gemini Live and Search Live features and is now available for preview in Google AI Studio via the Gemini Live API. This upgrade is considered the most significant update to the Gemini Live service so far.
Compared to the previous generation, Gemini 3.1 Flash Live achieves breakthroughs across several key metrics. The new model can more effectively recognize sonic details like pitch and speech rate while significantly reducing latency. Its ability to filter ambient noise is greatly enhanced, allowing it to more precisely isolate a user's voice from background sounds like traffic or television. Users of the Gemini Live mobile app will see tangible improvements: faster response times, fewer awkward pauses, and doubled conversational coherence, making it easier to maintain a train of thought during extended brainstorming sessions. Furthermore, the system can now dynamically adjust the length and tone of its responses to fit the conversation context.
