Appearance, voice, persona, knowledge and boundaries in one record.
ENSEMBLE STUDIO
Bring a character into the world.
LIVE NOW
Featured performers
Idle, listening, speaking and emotional carry before the first word.
Browser video today; the same performance contract drives MetaHumans.
BUILD A BACHELOR
Create a character who can carry a story.
- 1Appearance
- 2Voice
- 3Persona
- 4Knowledge
- 5Finalize
MY CAST
Characters you direct.
EXPLORE
Walk into someone else’s story.
Only creator-shared characters appear here. Public profiles keep AI disclosure, report controls and their sharing boundary.
MEASURED, NOT GUESSED
Audience and performance.
Instrumentation contract
Every public build should emit privacy-minimized product events into the existing HQ analytics envelope: acquisition source, onboarding completion, character build step, generation outcome, room wait, session start, first response, interruption, report, share and return.
No invented dashboard
This preview reports only values the local runtime can measure. Retention, conversion and usage charts remain empty until the production identity and event pipeline are connected and reviewed.
ENSEMBLE API
A character system, not a face endpoint.
One record. Every representation.
Resolve the same character into generated video, LiveKit or Unreal without handing their memory and persona to the face vendor.
https://ensemble.fermata.me/v1/sessions{
"character": "willa",
"representation": "generated_video",
"room": "my-livekit-room"
}Bring your own voice audio.
Interrupt without waiting.
Know when performance lands.
Continue the story beat.