Google rolled out Gemini 3.1 Flash Live, a lightweight upgrade to its Gemini Live and Search Live services. The new model trims response latency and expands the context window, letting the assistant keep up with longer dialogs. Marketed as the “highest‑quality audio and voice model” to date, it pushes Google’s voice‑first AI ambitions toward faster, more natural rhythm.

Speed gains come with a longer conversational memory – the system can follow a thread twice as long as its predecessor. Multilingual support lets users hear real‑time answers in their preferred language, while refined tonal detection picks up pitch and acoustic nuances. Benchmark tests show higher task‑completion rates, especially in noisy settings, and the Gemini API now lets developers embed these capabilities into custom agents.

Beyond voice, Gemini 3.1 Flash Live adds camera sharing and screen‑sharing hooks, so users can point the phone at an object or a webpage and receive instant explanations. By tightening response times and expanding language reach, the update makes on‑device assistance feel less like a clunky chatbot and more like a conversational partner ready for everyday tasks today.