Metamodel
September 24, 2026·Source

Muse Realtime Avatar

Meta's model that turns Muse's voice into a live, talking video character built from any reference image, unveiled at Meta Connect 2026.

Muse Realtime Avatar announcement image
  • Give it a reference image, whether a photo, an illustration, an animal, or an object, and it renders that character talking, gesturing, and shifting posture in continuous video
  • It runs on the same stream as Muse Realtime Voice, so speech, lip movement, and facial expression stay in sync
  • The first voice-and-video reply arrives about 870 milliseconds after you stop talking, and a session can run as long as the conversation does
  • Every video it produces carries an invisible watermark marking it as AI-generated
  • In Meta's own tests, raters preferred it over Runway Characters and HeyGen LiveAvatar in two-to-three minute conversations
  • Meta says it is in the Muse app for users 18 and older, while TechCrunch's Connect coverage described the avatar video chat as coming soon, so expect a staggered rollout

Best for

Face-to-face style video conversations with Muse instead of voice or text alone.

When to use it

A conversation would land better with a visible, expressive character on screen than with a voice.

Who benefits

Muse app users 18 and older, plus anyone tracking where AI video characters are heading for customer-facing work.

Join the newsletter to get ahead with AI.

I share strategies, new modules, free guides, and tools to get you better at using AI in your work.

Leading a team? See team programs →