Google has generally released Gemini 3.8 Live with Live Avatar in Gemini Enterprise, bringing live dialogue together with near-real-time video generation for business users.
The system creates animated personas that can lip-sync and maintain expressions across 97 languages. This places spoken interaction and generated video in the same enterprise product experience.
Avatar options for businesses
Businesses can choose from preset faces. With allowlisted access, they can also generate an avatar using a reference . The release does not specify how broad that allowlist is or when wider access may become available.
The pipeline includes asynchronous tool execution and multimodal visual processing. It also applies SynthID watermarking to all output, adding an identification layer to generated content.
Google is pitching the release as a combination of native live-dialogue models and live visual generation. The reported release date is September 24, 2026.
Conclusion
Gemini 3.8 Live with Live Avatar is now generally released in Gemini Enterprise, with multilingual lip-syncing, selectable avatar creation, visual processing, tool execution and SynthID watermarking among its stated features.
Frequently Asked Questions
Q. What is Gemini 3.8 Live with Live Avatar?
It is a Gemini Enterprise capability that combines live dialogue models with near-real-time video generation and animated personas.
Q. How many languages does the avatar support?
The avatars can lip-sync and maintain expressions across 97 languages.
Q. Can businesses create a custom avatar?
Yes. Businesses can use preset faces, while reference-avatar generation is available through allowlisted access.
Q. Does the system support visual inputs?
Yes. The pipeline includes multimodal visual processing.
Q. Does Google watermark the generated output?
Yes. SynthID watermarking is included on all output.
