How to Create a Product Bundle Image From Separate Photos With AI

If you sell a gift set, multipack or product bundle, you may already have good photographs of every item but no image showing the complete set together. That can mean arranging another shoot just to create one hero image.

Current AI image-combining tools can shorten that job. Adobe Firefly lets you upload two or more reference images and describe how they should be combined into one scene. Photoroom also offers a Photo Composition workflow specifically for merging catalogue products into bundle, gift-set and campaign images.

The advantage is that you can start from photographs of the real products instead of generating a fictional bundle from text. The risk is that AI can make several correct source products look beautifully coordinated while quietly changing a label, count, clasp or proportion. Treat each source photo as product evidence and the composite as something that still needs checking.

Step 1: Prepare one clean photo of every included item

Use genuine images of the exact SKUs that will be sold together. Choose simple views where shape, colour and identifying details are easy to see. If the set contains a bracelet, gift box and pouch, photograph all three rather than asking AI to invent the packaging around the bracelet.

Try to make the source photos broadly compatible. Similar camera height and neutral lighting help. They do not need to come from the same shoot, but wildly different perspectives create more work because the model has to reconcile both scale and viewpoint.

Step 2: Record quantity and relative scale before generating

Write down what the bundle actually contains and how the products compare in size. AI is good at making a balanced composition, but visual balance can conflict with physical scale.

A small jewellery pouch should not become almost as large as its presentation box just because the layout looks better. A 30 ml bottle should not grow to the height of a 100 ml bottle. If quantity matters, state it explicitly: one bracelet, one box, one pouch. If two identical units are included, say two.

Real measurements are useful when scale matters commercially. You do not need an engineering drawing, but a few known dimensions give you something concrete to check against afterwards.

Step 3: Upload the references and assign each one a role

In Adobe Firefly's current image-combiner workflow, open Edit Image, upload two or more reference images, then describe how the inputs should come together. Adobe says the tool can use multiple uploads as composition and style references and can harmonise depth, lighting, shadows and mood.

For ecommerce, make each image's role explicit. For example: “Image 1 is the exact bracelet, image 2 is the exact gift box and image 3 is the exact pouch. Create one gift-set hero image containing exactly one of each. Preserve each product's shape, colour, markings and proportions. Arrange them naturally on a neutral cream surface with soft studio lighting.”

That is safer than “make a luxury jewellery gift set using these photos”, which gives the model more freedom to reinterpret what the set should contain.

Step 4: Make the first composition visually simple

Do not begin with ribbons, flowers, candles, hands and dramatic reflections. Start with a simple surface, one clear light direction and enough space around every item to inspect it.

A plain first pass lets you check the bundle itself. Did every item survive? Are quantities correct? Do proportions make sense? Are labels, engravings, clasps and packaging details unchanged? Add campaign styling only after those checks pass.

Photoroom's Photo Composition API follows the same product-first idea at catalogue scale. It combines two or more existing product images and exposes controls for positioning, scale, shadows and background. Photoroom specifically positions it for bundle hero shots, gift sets and campaign creative.

Step 5: Check each SKU separately

Review the final composite like several product photos sharing one canvas. Compare product one with source one, then product two with source two. Do not approve the image merely because the group looks convincing.

For jewellery, check clasp shape, chain links, stone count, engraving, metal finish and dimensions. For cosmetics or packaged goods, inspect logos, labels, caps and quantities. Count the products and make sure nothing has been duplicated or omitted.

Pay extra attention to overlaps. If one item covers another, AI may invent the hidden edge or merge the objects. When that happens, a clearer arrangement is usually safer than repeatedly repairing an ambiguous overlap.

Step 6: Add seasonal styling after the neutral bundle is approved

Once the neutral composition is accurate, adapt it for marketing. Add ribbon, change the surface, create space for copy or produce a vertical social version. Keep the approved bundle image as a reference so later generations have a stable starting point.

Adobe currently says its Firefly image combiner can be tried on a free tier with daily generations. Adobe also says output from Adobe Firefly models can be used commercially, while source images still need to be yours to use or otherwise cleared. Partner models available inside Firefly can have different terms, so check those separately.

Sometimes normal layers are safer than generative AI

If every item already has a clean transparent cutout and exact fidelity matters more than realistic integration, placing those cutouts manually in Canva, Photoshop or another layer-based editor can be safer. The product pixels remain unchanged and only the layout moves.

AI is more useful when separate products need shared lighting, shadows, depth or a natural scene that would otherwise require more compositing work. The trade-off is that the model may reinterpret parts of the inputs while making the scene coherent.

Use the least generative method that solves the problem. If drag-and-drop layers are enough, use them. If the set needs to feel physically photographed together, AI composition can save meaningful setup time, but SKU review becomes mandatory.

Why this is useful for jewellery sellers

Jewellery bundles are a natural use case because sellers often photograph the core piece separately from boxes, pouches, cards or complementary items. A bracelet and necklace set may already have strong individual images but no coordinated gift-set hero shot.

Lustra Studio can help organise and create jewellery-led product imagery, while a composition workflow can bring several approved assets together. The real references define what is being sold. AI handles how those products are presented together.

The takeaway

You do not necessarily need another photoshoot to create a bundle or gift-set image. Start with one accurate photograph of every included SKU, record the true quantities and relative scale, combine the references in a simple scene, then verify each item individually before adding creative styling.

AI is useful because it can make separately photographed products feel like they belong in one scene. It should not decide what the bundle contains. The strongest bundle image is one that looks coordinated while every item still matches what the customer will receive.

Sources: Adobe Firefly, “AI image combiner”, accessed 29 August 2026; Adobe Firefly documentation on combining reference images and commercial use, accessed 29 August 2026; Photoroom, Photo Composition API, accessed 29 August 2026.