How do I keep the same model in every photo?

Attach the same model reference image every time, along with your garment photo, then paste a pose prompt. The prompts tell ChatGPT to match that person’s face, skin tone, hair and proportions, so the whole lookbook shows one model.

Can I use my own model instead?

Yes. Attach a photo of your own model in place of the reference image. Use a clear, front-facing, full-body photo in even light for the most consistent results.

Which poses suit which products?

Standing and walking poses show fit and length best, seated and leaning poses suit relaxed and casual wear, and the garment and accessory interaction poses draw attention to details like collars, pockets and bags. There are 180 poses in 11 collections.

Fashion pose library

180 poses. One consistent model.

Apparel brands need the same model across a whole lookbook. These prompts use one model reference, so every pose shows the same person wearing your garment.

Standing poses

Walking and movement

Seated poses

Floor and kneeling poses

Leaning poses

Editorial poses

Garment interaction

Accessory interaction

Camera angle variations

Framing variations

Lighting variations

Model reference used by every pose prompt

Start with the model reference

This reference is the visual anchor for the original pose collection, helping a clothing seller maintain one recognisable model across different views. It is useful for coordinated apparel launches and consistent catalogue photography rather than for showing a new product by itself. Attach it in ChatGPT together with your garment photo, clearly assigning identity to the model image and construction to the product image. Check facial features and body proportions after every generation.

Save this model reference image, or use your own model's photo.
In ChatGPT, attach the model reference and a photo of your garment.
Paste any pose prompt from the collections above.

Model reference prompt

Create one realistic full-body fashion model reference photograph using the attached approved model image as the identity source. Preserve the same fictional adult female model's face, warm olive skin tone, brown eyes, shoulder-length straight dark brown hair, slim athletic body proportions and natural skin texture. Present a neutral square front stance, feet hip-width apart, arms relaxed at sides, shoulders level and a calm natural expression. Outfit: an ivory fitted crew-neck short-sleeve t-shirt, charcoal straight-leg tailored trousers and plain white low-top sneakers; retain these unless a replacement product is explicitly supplied. Eye-level front camera with natural perspective, soft frontal-left studio lighting, minimal pale-grey seamless background and floor. Show the complete head, both hands and both feet with clear padding. Keep facial structure, hairline and proportions faithful to the supplied model reference, without beauty-filter smoothing or identity changes. No added text, labels, celebrity resemblance requests, invented logos or watermarks. This reference establishes identity and styling for the individual pose library.

Pose library questions

How do I keep the same model in every photo?: Attach the same model reference image every time, along with your garment photo, then paste a pose prompt. The prompts tell ChatGPT to match that person’s face, skin tone, hair and proportions, so the whole lookbook shows one model.
Can I use my own model instead?: Yes. Attach a photo of your own model in place of the reference image. Use a clear, front-facing, full-body photo in even light for the most consistent results.
Which poses suit which products?: Standing and walking poses show fit and length best, seated and leaning poses suit relaxed and casual wear, and the garment and accessory interaction poses draw attention to details like collars, pockets and bags. There are 180 poses in 11 collections.