Meta Rolls Out Voice and Expressive Avatars for Muse in the US
Meta is rolling out Voice and expressive avatars within the Muse app in the United States across mobile, web, and Mac. Users start a conversation by tapping the phone icon; AI glasses and Muse Charm will follow.
Meta is rolling out Voice and expressive avatars in the Muse app across the United States, on mobile, web, and Mac. Announced on November 5, 2026 on about.fb.com and confirmed by the official AI at Meta and Muse accounts on X, the product moves from a Connect 2026 demo to concrete availability inside the app. To talk to your Muse, just tap the phone icon in the top right corner.
Voice, Avatar, Full Screen
Meta offers three ways to use the feature: voice only, avatar beside the conversation, or full screen for a face-to-face setup. The avatar reacts while you speak and while it replies, featuring expressions and gestures designed to match the tone. You can interrupt Muse, and Meta claims Muse keeps working while you speak—useful for planning a trip, rehearsing an interview, or brainstorming hands-free.
On the personalization side, Meta promises hundreds of voices: accents, styles, and personas that can be changed to match your mood. The underlying tech is dubbed Muse Realtime Voice and Muse Realtime Avatar.
What Was Shown in September
At Connect 2026 (September), Meta previewed these real-time conversations. The research post from September 23, 2026 detailed the stack: 448×768 portrait streaming at 25 frames/s, latency around 870 ms (end of user turn → first byte of synchronized voice+video response), the Meta Video Seal invisible watermark, and access restricted to users 18 and older. Meta also reported a rater preference against Runway Characters and HeyGen LiveAvatar under its internal protocol—a Meta-reported result that has not been confirmed elsewhere. Those technical figures come from the preview; today’s news is the rollout inside the Muse app in the US.
What Comes Next
Voice and avatars are launching in the US first. Meta subsequently announced voice conversations on AI Glasses and Muse Charm, though without a set date. The exact pace of the US rollout (immediate for all accounts or rolled out in waves) is not detailed beyond “rolling out now.”
Our take
Muse is moving beyond Connect demo status into an embodied voice product, right as ChatGPT, Gemini, and others are aggressively pushing real-time voice. Meta’s differentiating angle is the expressive avatar piped directly into the voice stream, rather than simply offering a microphone interface. For creators, the immediate appeal remains limited as long as the feature stays US-only and centered around the Muse personal agent; the real test will be perceived quality outside scripted demos and its deployment to smart glasses. Latency benchmarks and reported preferences over Runway/HeyGen remain Meta’s own September figures—helpful as ballpark estimates, but by no means an independent verdict.
Sources
- Meta — Give Your Muse a Voice (November 5, 2026) Primary source
- AI at Meta (@AIatMeta) on X — Muse Realtime Avatar rollout in the United States (November 5, 2026) Primary source
- Muse (@Muse) on X — Introducing Voice (November 5, 2026) Primary source
- Meta AI Research — Bringing Your Muse to Life (September 23, 2026, technical preview) Secondary source — September 2026 research post: latency, resolution, watermark, and preference study. This is not the November 5 availability announcement.
A primary source comes directly from the company, lab or organization concerned. A secondary source reports the information without being its author: we flag it as such. Spotted an error? See our corrections policy.
