Google adds Live Avatar to Gemini 3.8 Live, pairing real-time speech with generated video of a persona

Google has released Gemini 3.8 Live with Live Avatar. The feature pairs the company's native live-dialogue models with near real-time video generation, so users see an animated persona while they talk to the model. The launch follows last week's release of Gemini 3.8 Live. Google says Live Avatar provides precise lip-syncing, natural expressions and fluid turn-taking, and supports a range of characters, each with its own look, voice and expressive presence. The Verge reports that the avatar can move between the 97 languages it supports. Google is aiming the feature at businesses, citing customer service and interactive walkthroughs as uses. It is available now only through Gemini Enterprise. The release puts a visual layer on conversational AI in customer-facing work, where companies have mostly used text or voice. The same week, Google began testing a feature that lets Gemini place phone calls to businesses for some US Pixel users. Google's post does not include independent measurements of latency or accuracy, or details on how businesses will disclose to customers that they are talking to a generated persona.