Meta is expanding access to Muse Voice, its AI assistant that users can speak to through live voice and video conversations, and the feature is now available to everyone in the US, with support for more devices planned. Unlike conventional voice assistants, Muse appears as a customisable animated character that reacts during conversations, following its unveiling at Meta's Connect developer conference in September. Muse Voice is built around two technologies: Muse Realtime Avatar and Muse Realtime Voice, which together power the assistant's animated appearance and its ability to respond during live conversations. Meta's default character, nicknamed Jolly, is a fluffy animated figure that users can speak to while asking questions or requesting help with tasks, and users can customise the character into something resembling a human, an animal or another fictional design. Meta chief AI officer Alexandr Wang announced the wider availability in a blog post and on X, where he shared a demonstration showing the assistant responding to casual conversation rather than simply waiting for specific instructions. During conversations, the avatar uses facial movements, expressions and gestures to make its responses appear more natural, designed to reflect the mood and tone of what is being discussed. However, Muse does not see the person on the other side of the conversation; it listens to what users say and generates responses based on the ongoing exchange, so its animated reactions are not evidence that it is analysing the user's facial expressions or surroundings.