Meta says US users can now video chat with their Muse AI avatar

Meta said on Thursday that Muse Voice, which lets people place calls and hold video chats with its consumer AI agent Muse, is now available to everyone in the United States. The feature was first shown at Meta's Connect developer conference in September. The rollout was announced in a blog post and on X by Meta's chief AI officer, Alexandr Wang, who posted a demo of his own Muse in action.
The experience is built on two new models the company calls Muse Realtime Avatar and Muse Realtime Voice, which together drive the digital character that talks with the user. The default persona is an animated avatar nicknamed Jolly, and users can reshape it to resemble a person, an animal or another character of their choosing. As the avatar speaks, Meta says it produces subtle expressions and gestures intended to match the tone of the conversation.
Wang's demo video, as described, showed him joking about his habit of posting memes on X and telling his Muse that it may have landed him in trouble with his boss. His avatar, a fluffy white dog, suggested adding disclaimers to memes that could be misread as Meta endorsing something risque.
The feature runs on mobile, web, and Mac today. Meta says real-time conversations will later reach its AI Glasses and an upcoming hardware product it calls Muse Charm. That device is described as a Tamagotchi-like wearable planned for launch before year-end, with a small screen where the digital avatar would appear so users can talk to the agent and have it reply or carry out tasks.
Meta's stated rationale is that giving Muse a body and a face makes the agent feel more personal, and that live conversation could deepen that sense of connection. That is a design bet rather than a technical one: plenty of assistants now handle spoken dialogue, so the differentiator here is chiefly the animated character layer and the emotional framing around it. Whether users form attachment to an avatar, or find it distracting, is the open question the rollout will test.
For developers, the interesting part is the model naming. Splitting "realtime avatar" from "realtime voice" implies separate systems handling speech and the visual performance, which in practice means latency budgets have to cover lip-sync and gesture generation, not just audio. Keeping expressions aligned with conversational tone is a hard synchronization problem, and the quality of that timing is likely what separates a convincing character from an uncanny one. Voice-first assistants from other vendors have generally skipped the persistent on-screen persona, so Meta is competing on presence rather than raw model capability.
The roadmap also matters commercially. Extending the same conversations to AI Glasses and a dedicated wearable suggests Meta wants one agent identity that follows a user across phone, desktop and hardware, rather than separate products per device. If that holds, the avatar becomes a platform surface that third-party experiences could eventually target, though nothing in the announcement describes developer APIs or third-party access.
A likely trade-off is privacy. A always-available animated companion that listens and watches through a camera on a wearable raises questions about when audio and video are captured, stored or used for training. The announcement does not address these points, and they will probably shape how cautious users approach the Charm device.
Why it matters: US users can now talk to Meta's agent face to face, and product teams should read this as Meta positioning an emotional, character-driven assistant rather than a utility chatbot. Developers building voice interfaces may face higher expectations for expressive, low-latency avatars as a baseline. The Muse Charm wearable, if it ships on schedule, would make that agent ambient rather than something you open — an inference about where Meta is heading, based on its stated device plans.
Based on reporting from the original publisher. Visit the source for full context and later updates.
Publisher excerpt
Meta's AI agent can now take calls as a customizable animated avatar, ahead of the launch of a Tamagotchi-like wearable designed to bring Muse to life.