Google launches Gemini 3.8 Live series for real-time conversation

Google announced the launch of Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking, calling them its most advanced real-time conversational models yet. The first is built for scale and cost efficiency, while the second targets complex tasks with stronger multi-step reasoning.

Gemini 3.8 Live Extended Thinking leads in multiple benchmarks, ranking first on Artificial Analysis speech-to-speech quality index (82.6%), achieving 68.6% on τ-Voice agent tasks, and scoring 97.7% on Big Bench Audio. Gemini 3.8 Live ranks second on Speech Agent Arena with strong cost efficiency.

Gemini 3.8 Live processes visual input in near real-time, auto-detects and switches among 97 languages, and handles background tool and API calls. Extended Thinking allows reasoning and speaking simultaneously. All AI-generated audio includes SynthID invisible watermarks.

Both models are rolling out now via Gemini API and Google AI Studio, with enterprise access through Gemini Enterprise.