Introducing Gemini 3.8 Live with Live Avatar

In a follow-up to last week's Gemini 3.8 Live launch, software engineers from the Gemini Audio Team are pleased to announce the introduction of Gemini 3.8 Live with Live Avatar. This feature brings near real-time visual presence to the native live dialogue models, creating an experience in which the model listens, sees, and speaks with a dynamic visual persona.

The Live Avatar feature combines precise lip-syncing, natural expressions, and fluid turn-taking to enable enterprises to expand their virtual offerings more interactively. It can be used for customer service or interactive walkthroughs, making digital exchanges richer and more accessible.

Gemini 3.8 Live with Live Avatar is currently available in Gemini Enterprise. The feature supports a range of characters, each with a distinct look, voice, and expressive presence. Conversational presence is enhanced by the feature's advanced reasoning, which allows it to trigger tool calls and fetch data in the background while maintaining an uninterrupted conversational flow.

The Live Avatar feature is not limited by languages, as it features native multilingual speech-to-speech synchronization. It can seamlessly transition across 97 languages without degrading video fidelity or introducing visual drift. Custom avatars can also be created by developers using their own high-quality reference images.

Safety and responsible deployment are at the forefront of Gemini 3.8 Live with Live Avatar. All output generated by the AI products is watermarked with SynthID, a watermark that is imperceptible and woven directly into the audio and video output. This helps ensure AI-generated content remains detectable to minimize misinformation and misattribution. For more information on the comprehensive approach to safety and responsible deployment, the model card is available.