Google Expands Gemini With Live AI Avatars and Business Calling

Google is expanding the capabilities of Gemini with two new developments that move its AI assistant further into real-time interaction and task automation. The company has introduced Gemini 3.8 Live with Live Avatar for enterprise customers and is testing a separate feature that allows Gemini to make certain business calls for users.

The two features have different availability and purposes, but both reflect Google’s effort to develop AI that can interact with people and complete tasks rather than simply generate answers.

Gemini 3.8 Live Gets a Visual Presence

Google’s Gemini 3.8 Live with Live Avatar combines real-time conversational AI with generated video. Instead of communicating through audio alone, Gemini can appear as a visual avatar that listens, responds and uses facial expressions while speaking.

Google says the technology is designed for enterprise applications such as customer interactions and interactive demonstrations. It combines audio and visual information during conversations, allowing the avatar to respond to what it sees and hears.

The feature became available through Gemini Enterprise on September 24, meaning it is not a general consumer feature available to every Gemini user.

Support for 97 Languages

Multilingual interaction is another part of Google’s Live Avatar offering.

Google says the system can move between 97 languages while adapting its speech synchronization and facial expressions to the language being spoken. This is built on the multilingual capabilities of the Gemini 3.8 Live models introduced earlier in September.

Businesses can choose from a selection of preset avatars or create customized ones. Google says custom avatars can be generated from a reference image, although this option currently requires enterprise allowlisting.

Google also says AI-generated audio and video from the system includes an invisible SynthID watermark intended to help identify synthetic content.

Gemini Starts Making Business Calls

Google is also experimenting with Call for Me, a feature that allows Gemini to contact businesses on a user’s behalf.

The system can handle certain practical requests, including asking whether a product is available, checking information, making a restaurant reservation or changing an appointment. Gemini can introduce itself, work through automated phone menus, wait on hold and communicate with a business representative.

Users can follow the conversation through a live transcript and take over the call when necessary. The calls are placed using the user’s own phone connection rather than through a separate calling service.

Call for Me Is Still an Experiment

The calling capability should not be confused with a full public rollout of Gemini’s phone-agent technology.

TechCrunch reports that Google is initially testing Call for Me with Pixel 11 owners who have a Gemini subscription in the United States, using the beta version of the Phone app. Google is starting with a limited experiment because real-world conversations can be unpredictable.

This limited availability also means that the feature is not currently a general-purpose calling service for all Gemini users.

Building on Earlier Gemini Live Models

The new avatar experience builds on Gemini 3.8 Live, which Google introduced earlier in September.

Those models were designed for real-time voice interaction and can process visual information while maintaining a conversation. Google also built support for background tool and API calls, allowing an agent to continue communicating while another task is being processed.

Live Avatar adds a visual layer to these capabilities, while Call for Me explores how Gemini can interact with businesses through ordinary phone conversations.

Google’s Long-Term AI Agent Direction

The developments fit into Google’s broader move toward agent-based AI.

Rather than limiting Gemini to answering questions, Google is developing systems that can interpret a request, use tools and perform multiple steps. The company has already been exploring this approach through Gemini’s ability to execute tasks in the background and interact with external services.

The latest announcements show two different applications of that strategy. Live Avatar focuses on enterprise-facing conversations, while Call for Me is aimed at handling routine tasks for individual users.

Safety and User Control

Google has included controls around the experimental calling feature. Users can monitor the conversation through a transcript and intervene when necessary. The feature also identifies itself as an AI system during calls, according to Google’s product documentation.

These controls are important because AI-generated conversations can encounter unexpected responses, unusual requests or situations that require human judgment.

Live Avatar also includes Google’s SynthID watermarking technology for generated audio and video, which is intended to make synthetic media more detectable.

What the New Gemini Features Mean

The latest updates show Google extending Gemini beyond conventional chatbot interactions.

Gemini 3.8 Live with Live Avatar gives businesses a way to build visual, conversational AI experiences, while Call for Me tests whether Gemini can handle routine phone-based tasks for users.

Both features remain subject to availability limits, and the calling system is still experimental. Their broader usefulness will depend on how reliably the AI handles real conversations and how Google expands access over time.

FAQs

What is Gemini 3.8 Live with Live Avatar?

It is an enterprise Gemini feature that combines live conversational AI with a generated visual avatar capable of responding through synchronized audio and video.

How many languages does Live Avatar support?

Google says Live Avatar supports 97 languages, with its visual speech synchronization adapting across supported languages.

What is Call for Me?

Call for Me is an experimental Gemini feature that can contact businesses for users and handle certain tasks such as checking availability, making reservations and changing appointments.

Can users monitor Gemini’s phone calls?

Yes. Users can follow the conversation through a live transcript and take control of the call when needed.

Is Live Avatar available to all Gemini users?

No. Google currently makes Gemini 3.8 Live with Live Avatar available through Gemini Enterprise, so it is positioned as an enterprise offering.

Conclusion

Google’s latest Gemini developments expand the assistant into more interactive and task-oriented territory. Gemini 3.8 Live with Live Avatar brings a visual conversational interface to enterprise applications, while Call for Me is testing whether Gemini can handle selected business calls for users.

The two features are still at different stages of availability, but together they demonstrate Google’s continuing push toward AI systems that can understand conversations, use tools and take actions on behalf of people.