From one bottle photo to a PDP hero
Three steps, and one rule: judge the label before you judge the lighting.
- 01
Upload the real bottle
A sharp, evenly lit photo of the actual product. The clearer the label in your input, the better every model holds it.
- 02
Describe one scene
Name the surface, the light, and the mood, for example wet polished marble with soft window light and a premium skincare feel.
- 03
Zoom in on the type
Compare the results at full size and read the label. Anything that smeared or invented copy is out, however pretty the scene.
What you can restage
Keep the bottle and its label, change the world around it.
- Surface and settingWet marble, travertine, brushed steel, sand, or water, the backdrop that signals the price tier.
- Light and moodSoft window light for clean clinical trust, or hard directional light for a bolder editorial look.
- Props and contextWater droplets, botanicals, packaging, or a bare void, depending on whether you are selling ritual or efficacy.
- Shot typeStraight packshot for the listing, macro on the texture, or a full lifestyle scene for paid social.
Which model for which skincare shot
Verdicts from running one serum bottle through all four on an identical prompt. The split was entirely about text.
OpenAI
GPT Image 2
The pick when the full label and claims have to be legible. It held the small print, which makes it the safe choice for a PDP hero or anything carrying regulated copy.
Nano Banana 2
Close behind GPT Image 2 on holding small print, and the better all-rounder if you also want a photoreal lifestyle scene around the bottle.
ByteDance
Seedream 4.5
The most cinematic frosted-glass hero and the best value of the photoreal options. It simplified the label, so use it where only the brand name shows.
Black Forest Labs
FLUX.2 Pro
The cheapest and most editorial, for volume work where you will overlay the label yourself. It malformed the brand letters, so do not trust its text.
Claims, copy, and trust
Renders are high-resolution and licensed for commercial use. Skincare carries a specific duty of care: ingredient lists, volumes, and efficacy claims are regulated copy, so never ship a generated label as-is. Verify it at full size or overlay the real artwork. It also pays commercially, because a shot that flatters the product into something it is not sets up a return, and a faithful bottle protects both the sale and the margin.
The same bottle, four models, one prompt
First-hand renders from our skincare test. Read the labels at full size, that is where the four models separate.
AI skincare photography questions
What beauty brands ask before putting this on a product page.
Can AI render frosted glass and reflections?
Yes, and this was the pleasant surprise of our test. All four models handled the frosted glass, the fill line, the way light scatters through the frost, and the reflection on a wet surface. Glass is no longer the hard part of skincare photography. The hard part moved to the label.
Will it get my label right?
That depends entirely on the model, and it is the single most important thing to check. Cosmetic labels carry a brand name, a product line, a volume, and sometimes ingredient or claims text, and small printed type is exactly what image models smear. In our test GPT Image 2 and Nano Banana 2 held it, Seedream 4.5 simplified it, and FLUX.2 Pro broke it outright. If your label carries regulated copy, treat that as the whole shot rather than a detail, and verify or overlay the real artwork.
Will it keep my exact product?
Yes, and this is the part that matters most. You upload a clean photo of your real product, and the models restyle the scene around it, so the label text, shape, color, and finish stay yours instead of being reinvented. Different models hold fidelity differently on the hard cases (frosted glass, small printed labels, reflective metal), which is exactly why comparing several on your actual product beats trusting one pipeline. Upload the clearest, best-lit shot you have for the most faithful result.
Do shoppers mind AI product imagery?
Enough that it is worth handling deliberately. In Deloitte's 2025 Connected Consumer survey of 3,524 U.S. consumers, 59% said they cannot reliably tell AI-generated content from real, and most want it disclosed. For skincare that cuts both ways: an image that obviously reads as AI erodes trust at the exact moment someone decides to buy. The practical answer is to keep the real product faithful, keep one honest photo on the listing, and use generated scenes for mood rather than for misrepresenting what arrives.
Which model should I use for skincare?
Match it to the shot. GPT Image 2 or Nano Banana 2 when the label and claims must be legible, which covers most PDP heroes. Seedream 4.5 for a premium lifestyle hero where only the brand name shows, at roughly a third of the price. FLUX.2 Pro for cheap, pretty, label-free volume you will finish yourself. The only real mistake is trusting one model's demo over a test on your own bottle.
Can I use several models on one product?
There is no single best model, and that is the point. Different models win at different briefs, so Masonry gives you all of them in one workspace and lets you compare results side by side. For skincare and cosmetics, Nano Banana Pro is the best all-rounder for photoreal results with faithful layout, Seedream is fast and detailed for exploring options, FLUX gives the most control for a consistent style, and GPT Image is great for precise edits. You pick the winner for each shot instead of being locked into one engine.
Related guides
Go deeper on the models and tools behind these shots.
Shoot your skincare line without a studio
Upload one bottle, compare the models on it, and keep the shot whose label you would put in front of a customer.
