Apply a garment or full outfit reference to a person image for fashion and ecommerce ad creatives while keeping the person identity, face, and pose as consistent as possible.
Describe how the clothing should be worn, styled, layered, or fitted. The person image keeps its original aspect ratio.
Upload the person image to dress. Large uploads are automatically optimized; under 1 megapixel is recommended for faster results.
Upload a garment packshot or one pre-composed full outfit image. Large references are automatically optimized; around 0.5 megapixels is recommended.
For multiple clothing items, combine them into one full outfit reference image before upload.
Sign in to spend tokens
Styling Instructions is required
Your generated images will appear here
FLUX Virtual Try-On dresses a person photo in a garment or full outfit from a reference image — keeping the face, pose, and body proportions intact — so fashion and e-commerce brands can produce on-model creatives without a photoshoot. On Prizmad it's 10 tokens per run.
FLUX Virtual Try-On is an image tool that applies a clothing reference to a person image. You upload two pictures — the person to dress and the garment to put on them — add styling instructions, and the model renders the person wearing that clothing while keeping their identity, face, and pose as consistent as possible. The output keeps the person image's original aspect ratio.
It handles two kinds of clothing references. Packshot mode takes a single garment image — the flat product photo you already have from your supplier or catalog. Full Outfit Composite mode takes one pre-composed image containing a complete look; if you want to apply multiple items at once, you combine them into one outfit reference before uploading. The styling prompt controls how the clothing is worn: tucked in, layered open over a shirt, fitted, with realistic fabric folds under clean e-commerce lighting.
On Prizmad, FLUX Virtual Try-On uses the same token balance as , video, voiceover, , and music. Uploads are automatically optimized for the model, there's no API key or separate account to manage, and every image you produce is yours to use commercially.
Turn flat supplier packshots into on-model imagery for product pages — no photographer, studio, or sample shipping required.
Dress the same model in different garments to produce a full set of ad variants for testing, all with a consistent face and pose.
Generate polished catalog photography — realistic fabric folds, premium lighting — from one person photo and your garment library.
Apply a complete pre-composed outfit in one pass to show how pieces work together as a styled look.
Convert generic supplier photos into unique on-model visuals so your store doesn't look like every other listing selling the same product.
Keeps the person's face, pose, and body proportions as consistent as possible while swapping in the new clothing — the model still looks like your model.
Use a single garment packshot straight from your catalog, or a pre-composed full-outfit image to dress the person in a complete look at once.
Styling instructions control how the clothing is worn — tucked, open, layered — plus the lighting and photography style of the final shot.
Large uploads are automatically optimized before processing; under 1 megapixel for the person image and around 0.5 megapixels for the garment reference is recommended for faster results.
A quality-passes slider (1–4, default 4) trades speed for detail — higher values can improve fabric and fit rendering but take longer.
Set a seed to reproduce a similar result with the same inputs — useful when you've approved a look and want consistent variations.
Upload the person image — the model or customer photo you want to dress. JPG, PNG, WebP, and AVIF are accepted; large files are optimized automatically, and under 1 megapixel is recommended for faster results.
Upload the clothing reference: a garment packshot or one pre-composed full-outfit image (around 0.5 megapixels recommended).
Set the reference type — Packshot / Single Garment for one item, or Full Outfit Composite for a complete look. To apply multiple items, combine them into one outfit image before uploading.
Write styling instructions: how the clothing should be worn, styled, layered, or fitted, plus the shot style — for example, a natural front-facing studio shot with clean e-commerce lighting.
Optional: adjust the quality passes (default 4), set a seed for repeatable results, and pick JPEG or PNG output in advanced settings.
Click Generate. Results typically arrive in about 1–3 minutes, keep the person image's original aspect ratio, and land in your asset library.
Each FLUX Virtual Try-On run costs 10 tokens, regardless of reference type or quality passes. Tokens come from your token balance.
The tool is unlocked on the same token balance you use for AI avatars, video, voiceover, and music — no separate provider account, no API key setup, no per-seat photo-studio software fees.
At 10 tokens per image, testing ten garments on the same model costs 100 tokens — a fraction of a single studio session. If tokens run out mid-campaign, a one-off top-up covers it without an upgrade.
No — each run consumes 10 tokens from your Prizmad balance. There's no separate account or API key required; the tool is available for Prizmad tokens alongside the platform's other image and video tools.
Yes. Images you generate with FLUX Virtual Try-On on Prizmad are yours to use commercially — in paid ads, product listings, lookbooks, and email campaigns. Make sure you have the rights to the person photo you upload.
10 tokens per run — the cost doesn't change with reference type, quality passes, or output format.
That's the point of the tool: it applies the garment while keeping the person's identity, face, pose, and body proportions as consistent as possible. You can reinforce this in the styling prompt — for example, 'preserve the face, pose, and body proportions'.
Yes, via the Full Outfit Composite reference type — combine the items into one pre-composed outfit image before uploading, and the tool applies the complete look in a single run. For one item, use the Packshot / Single Garment type.
JPG, PNG, WebP, and AVIF are accepted, and large uploads are automatically optimized. For faster results, keep the person image under 1 megapixel and the garment reference around 0.5 megapixels. The output keeps the person image's original aspect ratio.
General image tools like Seedream 5.0 Pro or Ideogram 4.0 generate a new person and scene from a prompt, which means the model and garment are both synthetic. FLUX Virtual Try-On works from your actual photos — your real model and your real garment packshot — and its job is to combine them faithfully. Use it when the specific person and the specific product both need to stay recognizable.
Typically about 1–3 minutes. Runs are asynchronous, so you can queue several garments against the same model photo and collect the results from your asset library.
Published 2026-07-16 · Last updated 2026-07-16