Meta lets everyone in the US video-call its Muse AI agent as an animated avatar

Meta made voice and video calls with its consumer AI agent, Muse, available to everyone in the US on November 5, 2026, TechCrunch reported. The feature runs on two new models, Muse Realtime Avatar and Muse Realtime Voice, and works on mobile, web and Mac, with Meta’s AI Glasses and a forthcoming “Muse Charm” device to follow. The default avatar is an animated character nicknamed Jolly, which users can change to look like a person, an animal or another character.

Why it matters: real-time talking avatars have mostly been sold to businesses as presenters and support agents. Meta is making one the default interface of a mass-market assistant.

What does the avatar do?

Meta says the avatar uses “subtle expressions and gestures” matched to the tone of the conversation, and users can interrupt it mid-reply. Muse can hear the user but cannot see them. Meta’s chief AI officer, Alexandr Wang, announced the rollout with a demo in which his Muse appeared as a white dog. The feature was first shown at Meta’s Connect conference in September.

Where is it going next?

Meta plans to launch a small, Tamagotchi-like wearable for Muse before the end of the year, on whose screen the avatar will appear, according to TechCrunch. It has not given a date for the Muse Charm. We covered Google’s real-time avatar product in Gemini Live Avatar in September, and the wider shift in AI avatars as the consumer face of AI.