Gemini 3.8 live with live avatar enhances real-time conversational AI

Google has launched Gemini 3.8 Live with Live Avatar, a feature that integrates real-time visual presence into its conversational AI. The update, available in Gemini Enterprise, allows users to engage in more natural and intuitive interactions by combining live dialogue with low-latency video streaming. The Live Avatar feature enables enterprises to offer more interactive virtual services, such as customer support and guided walkthroughs, through precise lip-syncing and natural expressions. The technology supports near real-time video generation and speech, creating a dynamic visual persona that listens, sees, and speaks. It also features asynchronous tool calling, allowing the system to perform complex tasks like hotel check-ins without disrupting the conversation. Gemini 3.8 Live with Live Avatar is multilingual, supporting seamless transitions across 97 languages without affecting video quality. Custom avatars can be created from reference images, with branding and identity preserved, though currently only available through enterprise allowlisting. All outputs are watermarked with SynthID to ensure transparency and prevent misinformation.