Meta rolls out Muse Realtime Avatar in the Muse app in the US
Meta said Muse Realtime Avatar, previewed at Connect, is rolling out in the Muse app in the US. The technology turns Muse Realtime Voice conversations into expressive, interactive avatars for talking with personal agents.
Meta announced on the AI at Meta account on X on November 6, 2026 (Singapore time) that Muse Realtime Avatar, previewed during Connect, is rolling out in the Muse app in the US starting today. Meta says it gives people a more natural way to engage with their personal agents.
What it is
According to Meta's September 23 research blog post, Muse Realtime Avatar turns Muse Realtime Voice into avatars with facial expressions, hand gestures and full-body movement. It is conditioned on reference media, so a photographic portrait, a full-body illustration, or even animals and everyday objects can be brought into a live conversation. Voice and video come from the same token stream, which keeps speech, lip motion and expression in sync.
Figures Meta reports
- 448×768 portrait video at 25 frames per second.
- About 870 ms of latency from the end of a user's turn to the first byte of the voice-and-video response.
- Generated video carries an invisible Meta Video Seal watermark for traceability.
The blog also says raters in Meta's own study preferred Muse Realtime Avatar overall over Runway Characters and HeyGen LiveAvatar. Meta notes that examples in the blog illustrate model capability and do not all reflect avatars available in the Muse app. Muse is for users aged 18+.