- How does it know what to replace?
- The AI compares your scene with the product photo and replaces the item of the same kind, so your mug takes the place of the cup someone is holding. A clear product photo on a plain background works best.
- Will the lighting and shadows match?
- Yes. The AI draws the new product with the scene's light, perspective and shadows instead of pasting the reference in, and keeps the rest of the photo as it is.
- Same shape vs different shape: does it matter?
- Same shape (tube to tube, bottle to bottle, can to can) gives the strongest results. Different shapes work too, since the AI redraws the item rather than filling an outline.
- What about people in the scene?
- If a person is holding the product, the swap keeps their hands and pose, and the rest of the person stays the same.
- Why are my variants different?
- Each variant is a separate attempt. They all match your reference product, but ambient details (reflections, micro-shadows, exact label-text positioning) vary. Pick the one that lands cleanest.