Google gives Gemini 3.8 Live a face, for enterprise customers only
Live Avatar generates a lip-syncing animated persona while the model talks, in 97 languages, and Google is holding it to Gemini Enterprise at launch.
Google announced Gemini 3.8 Live with Live Avatar on Wednesday, adding a generated face to the live dialogue model it shipped the week before. The feature reaches customers through Gemini Enterprise and the Google Cloud console, and not through the consumer Gemini app.
Google said the avatar delivers precise lip-syncing, natural expressions and fluid turn-taking, and that its speech-to-speech synchronisation is native across 97 languages, with the lip-sync adapting as the language changes. The wording is the company's own. It published no latency figure beyond calling the system near real-time.
Organisations may generate a custom avatar from a reference image, which Google said preserves reference likeness, brand styling or character identity. That path is currently available only through enterprise allowlisting, according to the announcement, and Google did not say when or whether it opens more widely.
Every output carries SynthID, which Google describes as an imperceptible watermark embedded in the audio and video so generated material can be identified. The model can also run tool calls in the background while the conversation continues, the announcement said. Google listed no limitations and no pricing.