Google launches Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking, its "most advanced live dialogue models yet", to more effectively enable voice agents

Voice agents now understand and act on complex, multi-step requests in real-time without interrupting your conversation.

Google DeepMind has released two new advanced AI models, Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking, designed to significantly improve voice agent capabilities. Gemini 3.8 Live offers fluid, intelligent conversations with real-time visual and language support, capable of handling 97 languages mid-conversation and executing background tasks without interrupting dialogue. Gemini 3.8 Live Extended Thinking is built for high-complexity tasks, featuring enhanced multi-step reasoning and simultaneous processing of tasks and speech, as demonstrated by its performance on benchmarks like Artificial Analysis' Speech to Speech Quality Index and τ-Voice. These models are now accessible via the Gemini API, Google Workspace, and the Gemini app, aiming to make voice interactions more natural and collaborative for both developers and end-users.

These new Gemini models represent a significant leap in making AI interactions feel more natural and intuitive, particularly for voice agents. Gemini 3.8 Live's ability to process visual input and switch languages seamlessly, coupled with its background task execution, means voice assistants can now handle interruptions and perform actions without breaking the flow of conversation. This advancement aims to bridge the gap between human dialogue and machine processing, making AI a more collaborative partner for complex tasks.

The introduction of Gemini 3.8 Live and 3.8 Live Extended Thinking signals a maturing market for AI-powered voice agents, moving beyond simple command-and-control to sophisticated, contextual interactions. For developers and enterprises, these models offer robust tools for building production-ready voice interfaces, potentially lowering the barrier to entry for advanced conversational AI. The focus on high performance, cost-efficiency, and benchmarks like the Speech Agent Arena indicates intense competition and a push towards more capable and reliable AI assistants across various platforms.