Gemini 3.8 Live released, tops voice leaderboard with 82.6 points

Beating AI News Flash: Google has released two new real-time voice models, Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking. Extended Thinking is specifically designed for complex tasks, capable of reasoning while speaking and invoking tools in the background without pausing the conversation to wait for results.

On Artificial Analysis's Speech-to-Speech Quality Index, Gemini 3.8 Live Extended Thinking scored 82.6 points, surpassing GPT-Live-1 + Astra's 81.5 points and Grok Voice Think Fast 2.0's 81.3 points to claim the top spot. It also achieved 68.6% on the voice agent task τ-Voice.

The standard version, Gemini 3.8 Live, leans more toward low-cost real-time interaction, supporting automatic switching across 97 languages, real-time visual input, and asynchronous tool calling. Both models are now available in the Gemini API and Google AI Studio, with audio input priced at $0.005 per minute and output at $0.018 per minute.