Introducing Gemini 3.8 Live and 3.8 Live Extended Thinking

Google has introduced Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking. These new voice models enable simultaneous speech processing and tool execution. The technology incorporates SynthID audio watermarking and supports multi-language auto-detection. These updates provide the AI with background thinking capabilities to enhance performance. The announcement highlights the integration of the Gemini Spark, focusing on advanced voice interactions and the ability to process complex tasks in real-time across various languages for improved user experiences.