Google Gemini Now Supports Live Avatars for Enterprise Agents

Google’s Gemini chatbot can now show animated avatars in its conversational “Gemini Live” mode. The new avatars are powered by the new Gemini 3.8 Live dialogue model that Google launched last week, but they’re only available for enterprise agents.

“With precise lip-syncing, natural expressions, and fluid turn-taking, Live Avatar enables enterprises to expand their virtual offerings more interactively. Whether providing engaging customer service or delivering interactive walkthroughs, it transforms digital exchanges into richer, more accessible experiences,” Google said today.

Through Gemini Enterprise, organizations will be able to customize the appearance of these Live Avatars, which can look like real humans or animated characters. Each avatar can have a distinct voice and unique facial expressions, and they can perform tasks in the background all while carrying on with a conversation with a customer.

Live Avatars are also capable of switching between languages right in the middle of a conversation. 97 languages are currently supported, and switching languages mid-conversation won’t interrupt lip-syncing.

Google said that “strict safeguards” are in place to “respect identity and keep AI-generated content transparent.” All audio and video generated by Live Avatars will also have an invisible SynthID watermark.

While Google releases new Gemini 3.x models at a regular pace, Koray Kavukcuoglu, the head of Google’s DeepMind division, recently told The Information that the release of a flagship Gemini 4 was getting close. “Our intention is to, like — as soon as possible — to release an early post-training output because we see the results and we are excited,” the exec said.