XNews

Gemini 3.8 Live Audio Models Launch

Gemini 3.8 Live Audio Models Launch

Google announced Gemini 3.8 Live and 3.8 Live Extended Thinking, new state-of-the-art live audio models with frontier pricing and performance. The models support 97 languages with seamless switching and async tool calls. A first-of-its-kind Gemini Audio experience event will be held in San Francisco.

Key insights

Model Specifications and Capabilities

Gemini 3.8 Live and 3.8 Live Extended Thinking are described as new state-of-the-art live audio models. They support 97 languages and can seamlessly switch between them. The models also include async tool calls, enabling more flexible and concurrent interactions. Pricing is positioned at the frontier level, matching high performance with premium cost.

In-Person Gemini Audio Experience Event

Google is hosting its first-ever Gemini Audio experience event in San Francisco the following week. The event is intended to showcase the new live audio models and allow hands-on interaction. The official announcement encourages interested parties to attend and learn more about the technology.

Source posts