The five
Each one gets a bible — about thirty photographs of her, every angle — and then a trained lock, so any new picture of her is really her.
Mira — the pilot
Her bible is finished: 24 photographs made for her, plus the 5 she came from. Clean files, no filters on any of them — the finish goes on later, when a picture becomes a post.
Who she is, locked
Early twenties. Long straight black hair, sleek and wet-look, swept back. Strong straight dark brows, almond dark eyes, high cheekbones, full natural lips. Warm olive-gold skin, deep sun tan. A striking hourglass figure on a visibly trained, athletic frame — very full bust, narrow toned waist, strong curved hips and toned thighs, defined shoulders and arms. Serene and self-assured. Her gaze goes to the camera or up to the sun.
What breaks her
- Wavy or curly hair, dry loose hair, bangs
- A big teeth smile as her default — the soft closed-lip smile is her warmest
- Pale or cool-toned skin
- Heavy glam makeup
- A petite or willowy build
Where she stands
Bible complete — 29 photographs ready for training. The training pack is zipped and ready to send. Her lock gets trained on fal the moment the fal balance is topped up; the training itself runs about $5. After that, a new photograph of her costs about four cents, and a short video of her about fifty cents.
Download her training pack (19 MB)
The photographs she was built from
Build record, not a description of her: the Lucienne shoot photographs her bible started from.
How this works
- House faces, not a replacement. The models carry collection drops, the lookbook, and suits that are not made yet. Real creators wearing real suits stays the engine that sells.
- A model is a person, not an outfit. Her bible rotates suits on purpose — the colour and the cut change from shot to shot — so what gets learned is her, never a swimsuit. She can wear anything in the catalogue, or anything that does not exist yet.
- Clean files, always. Model pictures are made with no filters on them — no grain, nothing. Filters are a finish, added later, picture by picture.
- Nothing baked in. Pictures are made on routes that leave the files clean. No hidden marks in the pixels.
- The label is set by hand. On Instagram, the made-with-AI label gets switched on at posting time, post by post. Nothing about it lives in the files.
- Judged like everything else. Sends and shares per reach first, then watch time, then likes. If the posts do not travel, the models do not stay.