@tag references that FLUX.2 can anchor objects to. Others provide visual context that gets woven into the scene description as prose.
Categories
Product
Required. The item you want to photograph. The prompt references your first product as@product1 and a second (if uploaded) as @product2.
Upload a clean product photo with a simple or neutral background. The cleaner the source image, the more accurately FLUX.2 reproduces the product’s shape, color, and material.
You can upload up to two products in one generation. This is useful for styling shots — for example, a bag alongside a pair of shoes.
Model
A person to feature with the product. Becomes@model in the prompt. Single image only.
The AI extracts the model’s visible characteristics — skin tone, hair, build, facial features — and uses them to describe the person in the scene. You don’t need to write any of this manually.
Outfit images require a Model image. If you upload an Outfit without a Model, the outfit reference is ignored.
Outfit
Clothing or accessories for the model to wear. References become@outfit1 and @outfit2 if two are uploaded. Requires a Model image to be useful.
Upload the garment flat, on a hanger, or already styled on someone — the AI can interpret any of these. The outfit description gets included in the prompt alongside the model description.
Background
The scene or environment you want the image set in. Single image only. No@tag — the scene context is woven into the prompt as descriptive prose rather than a direct reference.
This doesn’t have to be a photo of an exact location. A mood image, a texture, or any visual that captures the feeling you’re going for works well.
Pose
A body position or gesture reference for the model. Single image only. No@tag — like Background, this becomes prose description in the prompt.
The AI interprets the posture and camera relationship from your reference and describes it in the prompt. You can use a fashion photo, a stock image, or even a rough sketch.
How categories combine
Product only
The AI generates a clean product shot in a neutral or inferred setting. Good for e-commerce cutouts or simple catalog images.
Product + Background
Product placed in a specific scene. The background provides the environment; the AI figures out appropriate placement.
Product + Model + Outfit
Full lifestyle or fashion shot with a styled model. Outfit details appear on the model in the generated scene.
All five categories
Maximum control. You specify the product, who wears it, what they wear, where the shot is set, and how the model is posed.