Meta launches Muse voice conversations in the US… avatars also respond with facial expressions and gestures

Meta has begun rolling out real-time voice conversations with its Muse AI agent in the U.S., available on mobile, web, and Mac, with users able to start a session via a phone icon in the chat interface. The feature pairs spoken responses with an on-screen avatar that reacts to the conversation through facial expressions and body gestures, and users can also opt to talk with voice only and no avatar. The rollout marks Meta's push to make its personal AI agent more conversational and multimodal as it competes with rival assistants from OpenAI, Google, and others. Meta has not disclosed usage or revenue figures tied to Muse, leaving adoption metrics as the next thing to watch as the feature expands beyond the initial U.S. release.