Google Gemini 3.8 Live Introduces Real-time Virtual Avatar

Google has introduced a new Live Avatar feature for its Gemini 3.8 Live platform, adding a visual dimension to real-time conversations. This technology combines video generation with voice, aiming to provide businesses with more interactive virtual services.

The Live Avatar feature generates dynamic visual representations capable of listening, observing, and communicating. It utilizes precise lip-syncing, natural facial expressions, and seamless conversation flow to create more immersive and engaging experiences. The AI model can process both visual and audio data, leading to richer dialogues.

Leveraging Gemini's advanced reasoning capabilities, Live Avatar can handle complex tasks without interrupting the conversation flow. It can perform asynchronous tool or API calls in the background, retrieving data while maintaining a smooth interaction with the user.

The technology supports native multilingual voice synchronization across 97 languages, adapting lip movements and expressions without compromising video fidelity or introducing visual shifts. Businesses can also customize avatars to match their brand identity, though this feature is currently available to select enterprise clients.

Google has also implemented security measures by embedding invisible watermarks in all AI-generated content using SynthID. This helps in identifying AI-produced material and mitigating issues of misinformation. The Live Avatar feature is now available to Gemini Enterprise users via API.