Skip to main content
GESTEL lets you combine up to five reference categories — Product, Model, Outfit, Pose, and Style. Not every shoot needs all five. Starting simple and adding references one at a time is the fastest way to understand what each category does.
Start with a product image and one other reference. Add more categories after you see initial results. This makes it easier to understand what each addition changes.

​
Category combinations by shot type

​
Product only

Use a single product image when you want clean product shots across various backgrounds or lighting conditions. The AI generates a fitting scene based on the product’s context. Good for: e-commerce hero shots, catalog images, background exploration.
Single product image input with resulting generated product-on-background output

​
Product + model

Add a model reference when you want a person wearing or holding the product. The AI places the product on the model in a natural way. Good for: apparel listings, lifestyle shots, lookbook pages.

​
Product + model + outfit

Add an outfit reference to specify what the model is wearing beyond the product itself. Useful when the full look matters — for example, showing a handbag styled with a complete outfit. Good for: full lookbook styling, editorial shots, complete outfit reveals.

​
Product + background/style

Use a style reference to define the scene or aesthetic rather than specifying a person. The product appears within that environment. Good for: lifestyle product photography, seasonal campaigns, scene-specific shots.

​
Product + pose

Add a pose reference to control the model’s body position. The AI will use the reference model’s pose while applying your product and model references on top of it. Good for: dynamic shots, action poses, directional editorial looks.
Grid showing different category combinations (product only, product+model, product+model+outfit) with their respective outputs

​
How the AI reads each category

CategoryTag used in promptEffect
Product@product1, @product2The primary subject to feature
Model@modelThe person to place in the scene
Outfit@outfit1, @outfit2Clothing layered onto the model
PoseContext onlyBody position reference
StyleContext onlyScene, environment, or aesthetic
Pose and Style are woven into the generated prompt as context rather than tagged references. They influence the scene without being treated as exact-match subjects.

​
Building up complexity

1

Start with product only

Upload your product photo and generate. This gives you a baseline — you’ll see how the AI interprets your product and what kind of scenes it creates by default.
2

Add one more reference

Add a model or a style reference. Generate again and compare to your baseline. Notice what changed and whether it moved in the direction you wanted.
3

Refine your references

If the model placement looks off, try a cleaner model photo. If the style isn’t matching, try a more specific style reference image.
4

Add remaining references

Once your core combination produces consistent results, add outfit or pose references for more specific control.
Using more references doesn’t always produce better results. If you add many categories and results feel inconsistent, try removing the reference that’s least essential to your shot.