Google has finally given its AI a face, and it can speak in 97 languages without its expression falling apart. The company’s new Gemini 3.8 Live update introduces a feature called Live Avatar, which animates a digital persona that lip-syncs and shows different facial expressions during conversations. The catch: it’s only available to Gemini Enterprise customers right now.
The feature is designed for business use, and it marks a significant step in how companies present AI-driven customer service. Instead of a disembodied voice, the avatar acts as a visible representative of the brand. That matters for companies that want their AI to feel more human — or at least more present.
What the Live Avatar Does
The avatar can transition between the 97 languages it supports “without degrading video fidelity or introducing visual drift,” according to Google. That means the digital face can change its mouth shape and facial expressions smoothly as it switches from one language to another.
One video shared by Google shows its Live Avatar talking in English and Japanese, with its mouth animation lining up with what it’s saying in both languages. It can also pull up information on-screen while it talks, adding context to the conversation.
The avatar is not a static image. It moves, it changes expression, and it matches what it is saying. That is the core of the announcement, and it is the part that makes the feature worth noting.
Gemini 3.8 Live and the Timing
The launch of Live Avatar comes about a week after Google revealed its new Gemini 3.8 Live model. That model processes visual inputs in near real-time.
The timing suggests a pattern. Google announces a new underlying model, then builds a visible layer on top of it. The public sees the visible result first, even if the underlying technology took longer to develop.
The Preset Library and Custom Avatars
Google will offer a library of preset avatars for customers to choose from. Organizations can also create their own avatars to match their branding. That flexibility is part of why the feature is aimed at enterprise customers rather than individual users.
The company says output from its Live Avatar comes with its invisible SynthID watermark alongside safeguards to “respect identity.” The watermark is described as invisible, meaning it does not visibly alter the appearance of the avatar on screen.
Why 97 Languages Matter
Supporting 97 languages is a large number, and the fact that the avatar can switch between them without visual drift is the technical achievement here.
Google’s claim that the avatar can transition “without degrading video fidelity or introducing visual drift” suggests it has solved the problem of language switching. Whether the claim holds up in practice remains to be seen, but the company has shown the capability in action.
The Enterprise Focus
The fact that Live Avatar is only available to Gemini Enterprise customers tells you who Google thinks will use this first. Individual users are not getting access yet, which means the feature is aimed squarely at businesses.
For those businesses, the benefit is clear. A visible avatar makes the interaction feel more personal, even if the person on screen is not a real human.
What This Means for Customers
The immediate effect is that customer interactions with enterprise brands could become more engaging. An avatar that moves and speaks naturally is more likely to hold attention than a static chatbot.
There is also a practical benefit. The avatar can show information on-screen while it talks, which means it can explain something visually as it describes it verbally. That combination of audio and visual cues could make complex explanations easier to follow.
The Limitations So Far
The biggest limitation is availability. Individual users cannot access this feature yet, and there is no announced timeline for a consumer release. That means the first wave of Live Avatars will appear in corporate settings, not on personal devices.
The other limitation is the scope of the avatar’s capabilities. While it can lip-sync and show different facial expressions, the feature is presented as a conversational companion, and its full range of actions remains unexplained.
The Broader Trend
Google is positioning its avatar as a tool for enterprise customer engagement. The company has said it will allow organizations to create their own avatars, which suggests the feature is being built for customization.
The feature is live, the technology works, and the market is watching to see how it plays out.
- The avatar can lip-sync and show different facial expressions.
- It supports 97 languages without visual drift.
- It can pull up information on-screen while it talks.
- It comes with the invisible SynthID watermark.
- It is only available to Gemini Enterprise customers.
What is still unknown is whether individual users will ever get access. For now, the answer is no — but the technology is clearly being built with customization in mind, and the company has shown it works in action.
See the a run of 28 images at The Verge.
Get the Notebook.
The day's best stories and every fresh verdict, in plain English, in your inbox by seven. One email a day, no more.