Google DeepMind introduced Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking, describing them as its most advanced live dialogue models to date. The company said the models bring major upgrades in intelligence and parallel reasoning, making voice collaboration more intuitive and better suited to executing complex tasks.
The two models target different needs. Gemini 3.8 Live is positioned for scale and cost efficiency, pairing conversational intelligence with fluid dialogue and visual grounding. Gemini 3.8 Live Extended Thinking is aimed at high-complexity work, offering increased intelligence and multi-step reasoning while reasoning and speaking simultaneously.
According to the announcement, Gemini 3.8 Live Extended Thinking captured the top overall spot on Artificial Analysis' Speech to Speech Quality Index with a score of 82.6. It also led on agentic task completion with 68.6% on τ-Voice and 35.1% on Sierra's τ-Voice-banking benchmark, and scored 97.7% on Big Bench Audio. Gemini 3.8 Live secured second place in the Speech Agent Arena and was highlighted for its cost-effectiveness.
On the capability side, Gemini 3.8 Live processes visual inputs in near real-time and automatically detects and transitions between 97 supported languages mid-conversation. It can execute tools and API calls in the background while the conversation continues. The Extended Thinking variant uses early verbal cues such as "Let me check that…" and live progress narration to keep users informed during multi-step background tasks.
The models are rolling out through the Gemini API, Google AI Studio, Google Workspace, and the Gemini app. Gemini 3.8 Live is available to developers in the Gemini API and Google AI Studio, and in private preview in Gemini Enterprise. Gemini 3.8 Live Extended Thinking is in private preview in Gemini Enterprise, with broader availability planned for Google Workspace business customers, Gemini Live, and Google AI subscribers in Docs, Gmail, and Keep.
Google also said all audio generated by its AI products is watermarked with SynthID, an imperceptible watermark woven into audio output to help detect AI-generated content and prevent misinformation. Developer platforms including Agora, Fishjam, LiveKit, Pipecat, Vercel, and Vision Agents support the Gemini Live API, while Salesforce, Genspark, and Lumeris were named as partners.