Lower latency and improved precision define the latest update to Gemini 3.1 Flash Live. Google DeepMind optimized the voice model to make interactions feel more fluid. This incremental update targets the lag and reliability issues common in real-time audio AI. Developers can now build more responsive voice interfaces with reduced response delays.