Google gives its Gemini AI a face with new Live Avatar feature

Google has rolled out an update to Gemini 3.8 Live that introduces "Live Avatar," an animated AI persona capable of holding real-time conversations with users. The avatar synchronizes its lip movements with spoken responses and displays a range of facial expressions as the conversation unfolds.

According to Google, Live Avatar can switch between all 97 languages the model supports without any loss of video fidelity or visual drift. A demonstration video released by Google shows the avatar speaking in both English and Japanese, with mouth movements accurately matching the spoken words in each language. Beyond speech, the avatar is also able to display on-screen information while continuing to talk.

Enterprise-only for now

The Live Avatar feature is currently available exclusively to Gemini Enterprise customers rather than the general public. Its release comes roughly a week after Google unveiled its new Gemini 3.8 Live model, which the company says processes visual inputs in near real-time.

Google plans to offer customers a library of preset avatars to choose from, while also giving organizations the option to build their own custom avatars. The company states that all output generated by Live Avatar carries its invisible SynthID watermark, along with additional safeguards designed to protect identity.