Mirage's Avatar X clones your micro-expressions in 10 seconds
Mirage just launched Avatar X, a new model that creates full digital twins from just 10 seconds of source footage. Unlike existing avatar tools, it preserves micro-expressions like laughing, yawning, and smirking, driving realistic face and body gestures via audio across both horizontal and vertical long-form video.
Mirage just released Avatar X, a new AI avatar model designed to fix the stiff, lifeless look that plagues most digital clones.
Micro-expressions in ten seconds
The headline feature is speed: you only need 10 seconds of source video to generate a digital twin.
Instead of freezing your face in place, Avatar X preserves micro-expressions like laughing, yawning, smirking, and gasping. It also uses your audio track to drive both facial expressions and full-body movement, working across horizontal and vertical video without degrading over long generations.
Why it matters
Creating video avatars used to require studio setups or minutes of high-res training footage. If a 10-second clip can reliably generate realistic body language, the barrier to automated video production drops to near zero.
We think Mirage is targeting the right flaw in current avatars: natural movement beats high-resolution stiffness every time. The big test, however, is consistency. Audio-driven body movement often looks uncanny over longer stretches, so we will see if that 10-second setup holds up under real production workloads.
Sources (1)
- x.com(opens in a new tab)Originalx.com