Meta is adding real-time voice conversations and animated avatars to Muse, its personal AI agent. The company announced the rollout on November 5, starting in the United States across mobile, web and Mac.
The update turns a typed exchange into a spoken conversation while Muse continues working on tasks in the background. Users can interrupt a response.
The visual option adds an avatar that reacts while listening and speaking, with expressions and gestures suited to the conversation. Users can keep the session voice-only, display the avatar beside the chat or expand it to full screen. This is a conversation with a generated AI character; Meta’s announcement describes no new person-to-person video-calling service.
A new interface for an agent that acts
Muse was introduced in September as an agent that can work across connected services, including email, travel and shopping. Meta says it can continue longer tasks after the user closes the app and return when approval is needed for sensitive actions.
That makes account access relevant even when the interface feels like an ordinary call. Meta says users choose connected apps and permissions, can disconnect services and can opt out of interactions being used for model training. The company also says conversations and data in the agent’s virtual machine are not shared with its advertising systems.
Eligibility and the avatar technology
In its September 23 technical explanation, Meta states that Muse is for people aged 18 and over. It describes voice and avatar generation using the same speech-token stream to coordinate audio, lip movements and expressions. Meta says generated avatar video includes an invisible Video Seal watermark, and cautions that its research demonstrations do not all represent avatars available inside the app.
The November announcement describes a rollout, without confirming that every eligible US account already has access. Voice support for AI Glasses and Muse Charm remains listed as coming soon, with no launch date in that announcement.
