Use Cases
Place your own product into studio-quality scenes without a photo shoot.
Short answer
The right way to do AI product photography is not to GENERATE the product but to replace the background of your own photo. Shoot the product on a plain surface, remove the background, then build a new scene with Adobe Firefly's Generative Fill or Canva's background tools. A generated product is not the product you sell — packaging, labels and proportions all change.
Product photography is the most misunderstood use of AI image generation. The expectation: you describe the product and the tool produces its photograph. The reality: what it produces is not your product.
The jar the model generates resembles your jar, but the label is different, the lid has different proportions and your brand is not written on it. In e-commerce that means showing the customer the wrong product, which is not only a commercial problem but a legal one.
The correct approach is the reverse: you shoot the product and AI generates the world around it. That is the method that actually works and is widely used.
Shoot with a phone on a plain white or grey surface in soft light without harsh shadows. This raw photo is the foundation of the whole process; the product's real appearance comes from it.
Separate the product from its surface with Canva's background remover or a similar tool. Clean edges determine how convincing the scene looks in the next step.
Fill in around the product with Firefly's Generative Fill by describing it: "marble counter, soft morning light, blurred kitchen behind". Matching the light direction to the light in your product photo is the single detail that makes the result convincing.
The basic rule of e-commerce law is that the image shown must accurately represent the product sold. A product image generated by AI, however similar, is not the product you sell.
Label text comes out unreadable or wrong, packaging proportions shift, material texture differs. The customer sees the difference on arrival and return rates rise; under consumer regulation you have also used a misleading image.
The second problem is consistency. Try to generate five different photos of the same product and the product changes a little each time. In a catalogue that reads as five different versions of one product.
So AI's correct role in product photography is the scene, not the product: background, light, surface, atmosphere.
When you place a product into a new scene, the reason the image looks fake is almost always the same: the light directions disagree.
If the shadow in your product photo falls to the right, the light in your generated scene has to come from the left. When they disagree, the eye registers "something is wrong" even without noticing why.
State the light direction explicitly in the prompt: "warm directional light from the upper left casting a soft shadow to the lower right". That single sentence is the intervention that most improves believability.
The second detail is contact shadow. Without a shadow where the product meets the surface, it floats. Preserve the product's natural shadow when removing the background, or add a soft one after placing it in the scene.
It can, but what it produces is not your product: the label, packaging proportions and texture all differ. In e-commerce the image shown must accurately represent the product sold, so this approach raises returns and creates regulatory risk. The right route is replacing the background of your own photo with AI.
Adobe Firefly's Generative Fill inside Photoshop is the strongest option and is trained on the safest data set for commercial use. If you do not use Photoshop, Canva's background removal and Magic Expand reach a similar result far more simply.
Almost always because the light does not match. The shadow direction in your product photo and the light direction in your generated scene must agree. State where the light comes from explicitly in the prompt, and make sure there is a soft contact shadow where the product meets the surface.
Prompts are written in English because most tools give markedly better results with English prompts.
seamless studio backdrop in soft beige, a smooth stone pedestal, warm directional light from the upper left casting a soft shadow to the lower right, minimal composition, subtle depth of field, product photography lighting
Works well in these tools: adobe-firefly, canva, flux
Note: Stating the light direction is critical: if the scene's light disagrees with the shadow in your product photo, the image reads as fake.
cosy kitchen counter scene, worn oak surface, blurred window with morning light behind, a linen cloth and a sprig of rosemary at the edge of frame, empty space in the centre for a product, natural colours, 50mm shallow depth of field
Works well in these tools: adobe-firefly, canva, flux
Note: The phrase "empty space in the centre for a product" leaves the area where you will place your product.
Trained on licensed data; for teams that cannot take legal risk on commercial work.
Image generation, design templates and publish-ready output in one interface, with the generated image landing straight into a layout.
The strongest open-weight family for photorealistic people and for text rendered inside images.
Open source; installed locally it is free, uncensored and completely controllable.