Google has announced the Live Avatar feature for Gemini 3.8 Live. According to the September 24 announcement, the system pairs voice responses with low-latency video, enabling it to generate lip movements and facial expressions for the speaking character. Initial access is being offered through Gemini Enterprise; the feature targets enterprise use cases such as customer service and interactive product demonstrations.
Live Avatar can process audio and visual inputs simultaneously, keep conversations going during tool calls, and adapt lip movements and facial expressions to speech in 97 languages.
Why it matters
This development takes enterprise AI interactions beyond voice-only responses to an experience that also incorporates visual communication. As the way users interact with systems in customer service and product demonstrations changes, facial expressions and lip movements create a new layer that could affect how responses are perceived. The simultaneous processing of visual and audio inputs, along with the ability to keep conversations from being interrupted during tool calls, allows the feature to be adapted to ongoing conversations rather than one-time responses. Although support for 97 languages expands its reach, the fact that the initial rollout is limited to Gemini Enterprise leaves open which user groups the feature will reach and to what extent.
Background
Google is not a new name in the FikirPilot archive: we have published 59 news articles mentioning the name in the last 90 days; the latest is dated September 24, 2026.