Design the model yourself, from age and ethnicity to hair and eye color. Dress it in your own garments, then keep that face through lookbook sets, extra angles and short clips.
An AI fashion model in ZenCreator is a person generated from settings you pick, not a photograph of a real model, and you reuse that one face for every look in a collection. You cast it once in Face Generator and carry it through dressing, sets, angles and clips.
Your brand decides the face, the garments it wears, the settings each shoot uses and the image formats. What the images cannot give you is a measured fit: drape is generated, so they show styling, not measurements. Workflows for other teams are collected in AI Creative Studio.
Six stages take an AI model in your clothes from a first portrait to finished frames and clips. Each names its tool and one limit worth knowing.
1Face Generator
2Image Editor
3Photoshoot
4Variations
5Head & Face Swap
6Image to Video
Runs in: Face Generator
Inputs: gender, age, ethnicity, hair and eye color.
A run returns one to four portraits. Keep the sharpest front-facing one as the face reference for every later stage.
Runs in: Image Editor with the Nano Banana 2 model
Inputs: the portrait, a garment photo and, if pose matters, a full-body photo.
With Batch Mode off, the model combines as many as 14 references into one image. For a complex look, settle the pose in one run and swap the clothing in the next.
Runs in: Photoshoot
Inputs: the portrait, plus the dressed full-body shot with its face cropped out to steer outfit, framing and stance.
Choose categories and a count for each. Every run uses a single aspect ratio, and frames arrive resized to the dimensions you set.
Runs in: Variations
Inputs: the frames worth expanding.
Each frame yields up to ten new angles or crops with no prompt, and its color grade and light stay matched. One ratio and one resolution, 1K or 2K, apply to the whole batch.
Runs in: Head & Face Swap
Inputs: your portrait as the source and the chosen frames as targets.
Run it as the last image step, on up to 1,000 frames per batch. It opens after your first credit pack, and a slight change in pose can occur.
Runs in: Image Upscaler or Image to Video
Inputs: your final selects.
SeedVR2 3B Base enlarges to as much as 8K without deliberately redrawing the face, and the Kling 2.1 model in Image to Video makes a clip of 5 or 10 seconds from each image.
Face Generator turns the settings below into portraits. Settle on one face per campaign before any garment is shot.
Six jobs a collection usually needs, with the route through ZenCreator tools for each.
Run your dressed model through Photoshoot for a set across several categories; the tool keeps the face recognizable from frame to frame. Plan portrait pages and wide banners as separate runs, because each run holds one aspect ratio.
Route: Photoshoot, then Variations
Give Image Editor the portrait and a flat-lay photo, and describe how the outfit should sit. To change clothes in one photo you already have, use the AI clothes changer; to dress a person you photographed, see virtual try-on.
Route: Image Editor, references combined
Reference-To-Image with Seedream 5 Pro, in Scene mode with Keep Pose, takes pose, outfit and background from an existing photo and gives the frame your model's identity, with up to 1,000 source photos in one batch. Use photos you hold the rights to, with consent from anyone shown.
Route: Reference-To-Image, Scene mode
Cast two to four faces, then send the same frames through Head & Face Swap once per face. Each batch carries a single face, since swapping several faces inside one photo is still in development.
Route: Face Generator, then Head & Face Swap
Variations builds a swipeable sequence from one strong frame without a prompt. Five to seven per frame keep a carousel tight and the story easy to follow.
Route: Variations, then Head & Face Swap
Animate a finished frame with the Kling 2.1 model in Image to Video, using start and end frames to control how the motion begins and ends. Each image in a batch of up to 100 renders its own clip.
Route: Image to Video
The same AI-generated face in four outfits, each made through a different route described above.
Four nearby jobs sound alike. Find the row that matches yours and go where that job is handled.
All four routes sit in one ZenCreator account and draw on one credit balance, so a single collection can mix them.
Have a garment photo ready? Cast a face first, then dress it in the same account.
Design your model in Face GeneratorGenerated images handle some fashion jobs reliably and others less so. Go through both lists before a look goes live.
Dimeris Ltd is the company responsible for ZenCreator: about us.
No. The face comes from the traits you set in Face Generator; it is not a photo of an existing model. Because the license is non-exclusive, another user could end up with a similar face, so present the model as an AI-generated character rather than a real, one-of-a-kind person, and do not invent a biography for it.
A clear, evenly lit photo that shows the whole garment, typically as a JPG or PNG. Add a sharp face portrait so the identity holds, and a full-body photo when pose and proportions matter. For a complicated look, adjust the pose in one generation and the clothing in the next instead of asking for both at once.
Yes. Cast two to four faces, then put the same finished frames through Head & Face Swap once for each face, because one batch carries one face. To place several people in a single scene, Character swap in Reference-To-Image drops one to four character photos into a source image. Anyone real in a source photo must have consented.
The tool needs Trusted Status, which your account receives automatically once its first credit pack is paid for; the same status also opens some models. If you have bought a pack and the tool is still missing, refresh the page or sign out and back in, then contact support if nothing changes.
In credits. Before anything runs, each tool displays what one image or clip costs and what the whole run will cost; credits are taken when a task begins. A task that fails for a platform-side reason typically has its credits returned without you asking. Credits have no expiry date, a single balance pays for any tool, and they come in packs rather than a subscription; see Pricing.
Start with up to four candidate faces in Face Generator, keep the one your brand will wear, and take it into Photoshoot.