Google rolled out Gemini 3.8 Live and Extended Thinking

Google is rolling out Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking, two voice models built for near real-time dialogue and task execution. Developed by the Gemini Audio Team, both pair spoken conversation with parallel reasoning. Gemini 3.8 Live targets high-volume workloads with cost efficiency, visual grounding and fluid dialogue, while Extended Thinking handles complex, multi-step work.

Extended Thinking leads Artificial Analysis' Speech to Speech Quality Index with 82.6. It scored 68.6% on tau-Voice, 35.1% on Sierra's tau-Voice-banking benchmark and 97.7% on Big Bench Audio. Gemini 3.8 Live placed second in the Speech Agent Arena. On ServiceNow's EVA-Bench, the models pushed the Pareto Frontier for complex workflows by balancing accuracy with conversational quality on the Gemini Enterprise Agent Platform.

Gemini 3.8 Live processes visual context in near real time, switches automatically among 97 supported languages and runs tools or API calls in the background while it keeps speaking. Google's demos show it guiding employee onboarding from on-screen context and playing chess from a camera feed. Extended Thinking reasons and speaks simultaneously, using cues such as “Let me check that...” and progress updates during longer jobs. Demonstrations include turning a hand-drawn wireframe and spoken feedback into React components, coordinating a restaurant booking through asynchronous calls, and creating business plans through speech.

Google is carrying the models across Workspace and Search. Docs Live, Gmail Live and Keep Live support voice-led navigation and drafting, while Search Live can guide troubleshooting through a phone camera. Agora, Fishjam, LangChain, LiveKit, Pipecat, Vercel and Vision Agents use the Gemini Live API, with Salesforce, Genspark and Lumeris also partnering with Google around the models.

Both models are starting to roll out to developers through the Gemini API and Google AI Studio. Gemini 3.8 Live is in private preview for Gemini Enterprise, is coming to Gemini Enterprise for Customer Experience and is available to everyone in Search Live. Extended Thinking is also in enterprise private preview, is coming to Customer Experience and Workspace business customers, and is available to everyone in Gemini Live. Google AI Pro and Ultra subscribers can use it in Docs, while all Google AI subscribers can access it in Gmail and Keep. Google says all audio generated by its AI products carries an imperceptible SynthID watermark, and it has published a model card covering safety and responsibility.

Source