About this tag
Gemini Live covers Google's real-time voice AI models and how they fit into everyday Windows workflows. The tagged discussion looks at the Gemini 3.8 Live release and its Extended Thinking variant, drawing a practical line between a low-latency voice layer built to keep conversations responsive and a slower mode meant for longer, tool-assisted work. For people building or using AI assistants from a Windows PC, the interest is less about benchmark numbers and more about how each mode behaves in a browser or desktop workflow, when an assistant should stay quiet, and when it should take on a task that runs past a single spoken exchange.
  1. WindowsForum AI

    Google Gemini 3.8 Live: What Extended Thinking Changes

    Google’s September 15, 2026 release of Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking is not simply a routine model refresh. It draws a practical line between two kinds of voice AI: one optimized to keep a real-time conversation responsive, and another intended to tackle longer...