Meta is making its Muse AI experience feel more like a face-to-face conversation. The company is rolling out real-time voice interactions and expressive animated avatars in the Muse app, allowing users to speak naturally with their AI companions, interrupt responses, and watch their avatars react during conversations.

The update builds on a feature Meta previewed at Connect 2026. With the rollout now underway, Muse users can interact with their AI creations through voice instead of relying entirely on text prompts.

Talk naturally with your AI Muse

The new voice experience is designed to make conversations feel more spontaneous. Rather than carefully typing prompts, users can simply talk through their thoughts, ask questions, or explain what they want to accomplish.

Meta says the feature can be useful for everyday activities such as brainstorming ideas, planning trips, preparing for job interviews, rehearsing presentations, and working through a pitch.

For example, someone planning a vacation can start with a rough idea about a destination, preferred accommodation, or the type of food they want to try. Muse can follow the conversation, help develop those ideas, and work on tasks while the user continues talking.

The hands-free approach also means users can interact with Muse while walking, preparing dinner, or doing other activities that make typing inconvenient.

More importantly, conversations happen in real time. Users can hear responses almost immediately and interrupt their Muse whenever they want to add something or change the direction of the discussion. This makes the interaction feel less like exchanging messages with a chatbot and more like having an ongoing conversation.

Expressive avatars bring Muse to life

Voice is only part of the update. Meta is also introducing expressive avatars that react to conversations with facial expressions and gestures.

The avatars respond while users are speaking and adjust their expressions and movements to match the tone of the interaction. This gives users a visual indication that their Muse is following along, rather than presenting every response as plain text or audio.

Users can also decide how prominently the avatar appears. Meta offers three viewing options: voice-only conversations, an avatar displayed alongside the conversation, and a full-screen avatar experience for a more face-to-face interaction.

There is also plenty of room for personalization. Meta says users can choose from hundreds of voices, including styles inspired by a New Yorker or even a pirate. The selected voice can be changed whenever users want, allowing them to give their Muse a different personality depending on their preferences.

The experience is powered by Meta’s Muse Realtime Voice and Muse Realtime Avatar technologies, which are designed to keep conversations fluid while synchronizing spoken responses with the avatar’s expressions and gestures.

Muse voice rollout starts in the US

Meta is rolling out voice conversations and expressive avatars in the Muse app in the United States first. The features are available across mobile, web, and Mac versions of the app.

To start a voice conversation, users simply need to tap the phone icon in the top-right corner of the Muse app.

Meta has also confirmed that voice conversations with Muse are coming to its AI glasses and Muse Charm in the future, although the company hasn’t provided a specific release date for those integrations.

The update is another step toward making AI assistants feel less like traditional chatbots. By combining real-time voice, interruptible conversations, and animated avatars, Meta is moving Muse closer to an interactive digital companion that users can talk to naturally.