Gemini 3.8 Live launches: #1 voice model at $0.005/min with async background thinking

_philschmid · x · 2026-09-16

Google released Gemini 3.8 Live and 3.8 Live Extended Thinking, continuing its real-time voice agent push after last month's 3.5 Transcribe.

Available via the Gemini API in AI Studio, Google Search, and the Gemini App, plus partner plugins for LiveKit, Pipecat, LangChain, and Vercel.

Related event: Google Launches Gemini 3.8 Live and Live Extended Thinking Realtime Voice Models(11 posts)→

Original post →

More from Models

Models channel →