Updated

Muse Realtime Avatar

Meta introduced Muse Realtime Avatar, an embodiment model that turns Muse Realtime Voice into expressive interactive avatars with synchronized lip movement and gestures. It powers live video conversations in the Muse app, which Meta says are rolling out, and generated video is watermarked with Video Seal.

By No Code MBA · Facts checked

Provider
Meta
Announcement date
Status at last check
Limited access or preview
Model family
Muse
Capabilities
Video, Audio

What changed

Muse Realtime Avatar animates a reference image (photo, illustration, animal or object) as a live conversational avatar driven by Muse Realtime Voice. It is an audio-driven Diffusion Transformer that generates video in short causal chunks from the same speech token stream as the voice, keeping lips and expression in sync. Meta reports 448x768 video at 25 fps, about 870 ms from end of user turn to first response, and rater preference of 78% over Runway Characters and 88% over HeyGen LiveAvatar.

Access and availability

It powers live video conversations in Meta's Muse app for users 18 and older, which Meta describes as rolling out. No developer API was mentioned.

Official access documentation

What you can build

Run a short live conversation with a reference photo or illustration and check whether the avatar keeps consistent appearance and lip sync across turns before building a branded character around it.

Limitations

There is no API, and the rollout timing and regions are not stated. Meta notes its examples show model capability and not all reflect avatars available in the app, and output resolution is 448x768. All generated video carries an invisible Video Seal watermark.

Sources and release history