Face, still, motion.
Three steps, and each one depends on the last. Once you can see the chain, everything else in the studio makes sense.
- 1
Cast a face, and lock it
Upload a photo or describe someone, pick one of four stylized looks, and choose from the faces that come back. That face is locked — every look and every clip is generated from it, which is what makes the character the same person each time.
- 2
Make a look
A look is your character in a place: a preset, or a photo of your own room, your own outfit, your own product. It's a still image, and it costs cents rather than credits — so try a few.
- 3
Animate the look into a clip
Choose a look, pick a format, write about twenty words. The clip animates that still, with your character's voice, synced. Two to three minutes later it's in the Clips tab, ready to post.
A clip animates the still it starts from. It can keep an object steady, move a person, change an expression — but it cannot add something that isn't already in that first frame. If your character should be holding a coffee, make a look where they're holding it, then animate that.
The studio tells you this the moment it matters, and offers to make the look for you.
What we won't do
Output is stylized, always — 3D, anime, claymation or illustrated, never a photorealistic double of a real person. Every clip is signed with content credentials and declared as AI-generated when you publish it. Reference photos of rooms, products and clothing must not contain people, and photo casting requires you to confirm the photo is of you or that you have permission.