Google adds Live Avatar to Gemini 3.8 Live: the AI now lip-syncs in 97 languages

Google has added a new mode to Gemini 3.8 Live that puts a face on the model. Live Avatar lets users hold a conversation while an animated AI persona answers in real time, syncing its lips to its speech and changing facial expressions as it talks. The feature is available only to Gemini Enterprise customers for now.

97 languages, one avatar

According to Google, the avatar moves between the 97 languages the system supports “without degrading video fidelity or introducing visual drift.” In a video published by the company, the persona speaks English and Japanese, with the mouth animation matching what is being said in both languages. The avatar can also pull up information on screen while it talks.

Preset avatars, or your own

The launch comes about a week after Google revealed the Gemini 3.8 Live model, which the company says “processes visual inputs in near real-time.” Customers will be able to pick from a library of preset avatars, and organisations will also be able to create their own.

SynthID and “respect for identity”

Google says anything Live Avatar produces carries its invisible SynthID watermark, which marks the video as generated, along with safeguards meant to “respect identity.”

The company has not said when the avatar will reach other Gemini tiers.