P Image Try On: Dress Anyone From One Photo

Every new product means another photo shoot. A new colorway, a different body type, a fresh seasonal drop, and you're booking a model, a studio, a photographer, and a day you don't have. Then the returns come in because the jacket looked different online than it did on a real person, and you start the cycle again. Product imagery is the quiet tax on every fashion store, and it scales badly with exactly the thing you want to grow: more products.

P Image Try On takes a swing at that whole problem. Built by Pruna, it places any garment onto a photo of a person and generates a realistic try-on image, keeping the person's identity, body shape, and pose intact. The interesting part isn't that it edits clothes onto someone. It's that it does it well enough to put on a product page, which is where most try-on tools quietly fall apart.

The Real Problem With Product Photography

Here's the question that decides how fast a catalog can move: are you photographing products, or are you photographing every product on every body in every colorway, forever?

Because that second list is the real one. A single garment isn't one image. It's that garment on a few body types, in each available color, in a couple of poses, maybe in a lifestyle setting and a clean studio one. Multiply that across a catalog and the shoot schedule becomes the bottleneck on the whole business. You're not limited by how many products you can source. You're limited by how many you can photograph.

That's the gap P Image Try On is shaped to close. Start from one photo of a person, bring in the garments you want them wearing, and generate the try-on. The shoot that used to gate everything becomes a single source image you reuse. You stop rationing product views and start producing them.

What P Image Try On Actually Is

P Image Try On is an image-to-image virtual try-on model from Pruna's P-Image family, built specifically for the try-on job rather than general image editing. You give it a photo of a person and one or more garment references, and it returns a single realistic image of that person wearing the clothes, with their face, body shape, and pose preserved.

That specialization is the point. A general editing model will happily warp a face or invent a new body to make the clothes fit. P Image Try On is tuned to keep the human and change the wardrobe, which is exactly the constraint e-commerce needs, because the whole value is that it's still recognizably the same model across your catalog. On Eachlabs it runs as an image-to-image step you can call on its own or drop into a larger workflow.

How P Image Try On Works

The inputs are refreshingly simple. You supply a person image and between one and eleven garment reference images, and the model fits those garments onto the person. The references are flexible: a flat lay of the product, a photo of the item already worn by someone else, or a set of separate garment shots all work. You're pointing at the clothes, and the model handles putting them on the body.

What it protects is as important as what it changes. The subject's facial features, likeness, body shape, and pose stay put, so the result reads as the same person in different clothes rather than a new person who happens to be wearing your product. That consistency is what makes it usable across a whole product line instead of one lucky image.

Up to Eleven Garments in One Pass

The multi-garment support is where this stops being a toy. You can feed up to eleven garment references into a single generation, which covers a full outfit, top, bottom, jacket, accessories, or a set of options to combine. Instead of dressing a person one item at a time and stitching results together, you assemble the look in one call.

For a catalog, that's the difference between a tedious manual process and something you can actually automate. A complete styled look comes out of one generation, built from the product references you already have.

It Keeps the Person, Not Just the Clothes

Identity preservation sounds like a detail until you've watched a try-on tool quietly swap someone's face. For influencer and model reuse, it's the entire game. A creator previewing a sponsored outfit needs to still look like themselves. A store building a catalog needs the same fit model across forty products so the line feels coherent.

P Image Try On is built around that constraint, holding the subject's likeness and pose while replacing the garments. Show the same base portrait wearing your spring line, then your summer line, and it's recognizably one model throughout. That continuity is what turns a pile of one-off edits into a catalog that looks like it was shot on purpose.

Where P Image Try On Pays Off in E-commerce

This is where the model earns its place, so be specific about it. A fashion store can render one garment on several body types from a single base portrait per type, giving shoppers a truer sense of fit without booking a shoot for each. Colorways and seasonal variants become generations instead of calendar entries. A new drop can have full imagery on launch day rather than a week later when the studio frees up.

It stretches past the catalog, too. Influencers and creators can preview brand outfits for campaigns while keeping their own look consistent across posts. Designers and stylists can test silhouettes and colorways on a single fit model and iterate in minutes instead of fittings. And teams building virtual fitting rooms can put the try-on behind an app, so a shopper uploads a selfie and sees the product on themselves, which is the kind of feature that moves conversion and cuts returns. The common thread is that all of it used to require a camera and a calendar, and now it requires a photo and a prompt.

Using P Image Try On on Eachlabs

On Eachlabs the flow matches the inputs. You provide the person image, add your garment references, up to eleven of them, choose an output format and quality, and run it. Results come back quickly, on the order of seconds, so trying a few garment combinations stays interactive rather than turning into a wait.

The quiet advantage is that it sits behind the same single interface as every other model on Eachlabs, so try-on can be one step in a bigger pipeline. Generate or clean up a base portrait, run the try-on, then pass the result to a background or upscaling step, all in one workflow. For catalog work that runs at volume, that chaining is what makes it practical rather than a nice demo.

Getting Better Results Out of P Image Try On

Start with a clean person photo. Front-facing or three-quarter views, the subject centered, well lit, with the clothing area unobstructed, give the model the most to work with. Bags, crossed arms, and props that cover the torso are what cause warped garments.

Bring high-quality garment references. The clearer the fabric, pattern, and silhouette in your reference, the more faithfully it transfers. A crisp flat lay beats a dim, folded photo every time.

Keep a neutral background and minimal clutter. The less the model has to separate the person from the scene, the more attention lands on the clothing, which is the part you care about.

Don't mix conflicting references. Keep one strong reference per garment slot rather than two very different versions of the same item, and reuse a single high-quality garment reference across products to keep color and pattern consistent.

The Honest Limitations

I don't want this to read like a brochure with the rough edges sanded off, so here's the straight version.

It's a single-image model, not a physics engine. It won't perfectly simulate how heavy fabric drapes under every pose, and it doesn't do video or time-consistent try-on across frames. Tricky materials are genuinely tricky: metallic, translucent, or highly complex patterned fabrics can come back imperfect, especially under unusual lighting.

Wrapping Up

The shift P Image Try On represents is the kind that changes a cost structure, not just a workflow. When dressing a person in a garment is a single generation from one photo, product imagery stops being a shoot you schedule and becomes something you produce on demand: more body types, more colorways, more looks, all recognizably the same model, ready the day the product is.

If your catalog is growing faster than your studio can keep up, P Image Try On is worth a real try. You can run it on Eachlabs right now: bring a clean person photo, add your garment references, and see a product-page-ready try-on come back in seconds.

Frequently Asked Questions

How many garments can P Image Try On handle at once?

Up to eleven garment references in a single generation, which is enough for a full styled look or a set of options to combine. You supply one person image and your garment references, from a flat lay, a worn photo, or separate product shots, and the model fits them onto the person in one pass.

Does P Image Try On keep the original person looking like themselves?

Yes, that's the core of it. The model preserves the subject's facial features, body shape, and pose while replacing the clothing, so the result is the same recognizable person in different garments. That consistency is what makes it usable across a whole catalog or for an influencer who needs to stay on-brand across posts.

Why is P Image Try On useful for e-commerce specifically?

Because it removes the per-product shoot. A store can show one garment on several body types, generate every colorway and seasonal variant without a studio day, and even power a virtual fitting room where shoppers see items on themselves. Truer fit previews tend to lift conversion and cut returns, and the imagery is ready on launch day instead of a week later.