Models · Wednesday, September 16, 2026
Google launches Gemini 3.8 Live dialogue models
The File desk · Sep 16, 2026, 2:28 PM UTC
Status
Confirmed from Google’s blog post dated Sept. 15, 2026.
Google introduced two live-dialogue models on Sept. 15. Gemini 3.8 Live is built for scale and cost efficiency, with fluid conversation and visual grounding. Gemini 3.8 Live Extended Thinking is built for harder tasks, with more multi-step reasoning while the talk continues. The company says both can run tools in the background while the conversation keeps going, and that 3.8 Live can switch among 97 languages mid-conversation. Rollout started the same day in the Gemini API and Google AI Studio. Consumer paths include Search Live for 3.8 Live, and Gemini Live plus Workspace Live surfaces for Extended Thinking (with subscription tiers noted on the post). Audio outputs carry SynthID watermarking. Google points readers to the model card for safety details.
Google introduced two live-dialogue models on Sept. 15.
Gemini 3.8 Live is built for scale and cost efficiency, with fluid conversation and visual grounding.
Gemini 3.8 Live Extended Thinking is built for harder tasks, with more multi-step reasoning while the talk continues.
The company says both can run tools in the background while the conversation keeps going, and that 3.8 Live can switch among 97 languages mid-conversation.
Rollout started the same day in the Gemini API and Google AI Studio.
Consumer paths include Search Live for 3.8 Live, and Gemini Live plus Workspace Live surfaces for Extended Thinking, with subscription tiers noted on the post.
Audio outputs carry SynthID watermarking. Google points readers to the model card for safety details.
What is still unknown or disputed
- Independent re-runs of the cited Speech-to-Speech and voice-agent benchmarks are not in this post.
Primary sources
Every claim in this story is drawn from the documents below. If a fetch failed, that is recorded on the card.
Source 1
Introducing Gemini 3.8 Live and 3.8 Live Extended Thinking
Google · September 15, 2026
Today, we’re introducing two new models… Gemini 3.8 Live… Gemini 3.8 Live Extended Thinking… rolling out starting today in the Gemini API and Google AI Studio.
https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-3-8-live-gemini-3-8-live-extended-thinking/