Group Shots Are Here (and Why Multi-Model AI Photography Is So Hard)
One of the most common requests we've gotten from brands: "Can we put multiple models in one shot?" A lookbook hero with two models in coordinated outfits, a campaign banner with your full seasonal lineup, a lifestyle scene that shows the collection as a whole.
The honest answer, until now, was: not really. And the reason goes deeper than you might think.
Why group shots are genuinely hard for AI
When you ask an AI model to generate multiple people in one image, things fall apart. Faces blur together. Outfits get mixed up between models. One person gets the other's jacket. Poses collapse. This isn't a prompt engineering problem you can fix with better instructions. It's a fundamental limitation of how current image generation works. Every AI tool on the market has the same issue.
We spent a lot of R&D time on this. The approach that actually works is compositing: render each person individually in a clean studio-style shot (preserving their exact face, outfit, and product details), then combine everyone into a shared scene in a second pass. Each generation only handles one person, so nothing gets mixed up. The final composite brings them together with a location, lighting, and natural positioning.
It sounds simple when you describe it in two sentences. The engineering to make it reliable, automatic, and fast was not simple.
How it works for you
You don't see any of that complexity. You create a campaign, add your single shots (each model in their outfit), select which ones to combine, pick a location, and hit generate. Brandmachine handles the two-phase pipeline automatically. You get back a set of group shot results to choose from.
A few constraints worth knowing: group shots work with 2 to 3 people. We validated up to 5, but 4 and above produced inconsistent results, so we capped it at 3 to keep quality reliable. Each person's facial identity is preserved using portrait references throughout the pipeline.
Campaigns got a proper home
This release also restructures how campaigns work. Before, campaign images were floating around in a single view with no real organization. Now, campaigns are proper containers. You open a campaign and see your single shots and group shots together. Single shots appear at the top, group shots below. It mirrors how you actually think about a campaign: individual assets plus the hero/group images that tie them together.
AI Photo Editor, refined
We've also improved the photo editor for fixing those last-mile details (logos, text, small product details that AI doesn't always nail).
Previously, clicking "Fix" gave you one AI attempt. If it wasn't perfect, you tried again. Now, each correction generates multiple variants. You flip through them and pick the best one. We also added blend refinement controls (feathering, grow/shrink, edge sharpness) so you can dial in exactly how the fix blends into the surrounding image. And all of this now runs on GPU-accelerated compositing, so it's noticeably faster and smoother.
Per-perspective styling
One more thing: Product Studio perspectives now support their own styling instructions. If you want specific notes for how a product looks from the front vs. the side vs. a detail shot, you can now write those per perspective. More control, less one-size-fits-all.
What this adds up to
Group shots were the last major gap between what AI photography could do and what brands actually need for production content. A campaign isn't complete with just solo product shots. You need the lifestyle imagery, the hero shots, the multi-model scenes that tell a story. That's now possible, entirely within Brandmachine, without Photoshop compositing or booking an actual group shoot.
Group shots are live now. Create a campaign, add a few single shots, and create a group shot from them.
Group shots work best when your single shots already look great individually. If you haven't set up fashion models and locations yet, start there. The group shot builds on that foundation.