Try MiniMax H3 (Text-to-Video) (Image-to-Video) from MiniMax →
Sign InStart Building

Task History

Click to see output list

No tasks yet

Go to Models
Image GenerationActive

pruna / p-image-try-on

p-image-try-on

bypruna

Pruna’s p-image-try-on creates virtual try-on images by applying one or more garment photos to a person photo. Generate up to 11 outfit variants in one run.

Image to ImageFast Inference
Model ID
p-image-try-on
Provider
pruna
Updated
1784809085
wiro playground—pruna/p-image-try-on
Reset to defaults
0 / 1
Maximum 1 image allowed
Drop image to upload

OR

Click to browse your device

Supports: JPG, JPEG, PNG, GIF, WEBP, HEIC

Required

Delete All
0 / 11
Maximum 11 image allowed
Drop image to upload

OR

Click to browse your device

Supports: JPG, JPEG, PNG, GIF, WEBP, HEIC

Required. Max 11 images allowed, if more then 11 images are send then only first 11 will be processed.

Sample outputs
Sample 1
Sample 2
Sample 3
Sample 4
Updated 1784809085
## Overview P-image-try-on by Pruna is an image-to-image virtual try-on model. You provide a person photo and one or more garment images. It generates new images that show the person wearing each garment. This helps teams preview looks without a full photoshoot. It also helps shoppers see outfits on a real person photo. ## What you can build - A virtual fitting-room widget for product pages. - On-model image generation for catalogs from flat-lay garment photos. - Outfit A/B tests for ads by generating multiple clothing options on the same person photo. - Styling previews for creators who want quick wardrobe mockups. - Batch try-on workflows that return multiple outfit results from one submission. ## Inputs - One person image. Use a clear, well-lit photo where the full outfit area is visible. - One to 11 garment images. If you submit more than 11, only the first 11 are processed. - The output image format you want (JPG, PNG, or WebP). - An output quality value from 0 to 100 when you choose JPG or WebP. - An option to keep the original person image dimensions, or to resize the final result back to the model’s capped person-image size. - An optional numeric seed for repeatable results. If you leave it at zero, the run randomizes the seed. ## Outputs - A set of generated images that show the person wearing the provided garment(s). - The platform returns the images in the format you selected (JPG, PNG, or WebP). - When you submit multiple garment images, you typically receive one result per garment image, up to 11 outputs. ## Limitations - The model processes a maximum of 11 garment images per run. - Output size behavior depends on whether you choose to preserve the input size. If you don’t, the result may be resized to a capped size. - Results depend heavily on input photo quality. Low-resolution images, heavy compression, blur, or harsh shadows can reduce realism. - Complex poses, strong occlusions (arms crossing the torso), and busy backgrounds can cause garment artifacts. - Garment images with cluttered backgrounds, folds that hide key features, or unclear silhouettes can reduce fit accuracy. ## Safety & compliance - Only upload person photos you have rights to use and consent to process. - Don’t use copyrighted garment imagery unless you have permission. - Don’t use the model to create misleading images for fraud, impersonation, or harassment. - Follow your platform’s content rules for sexual content, minors, and non-consensual imagery.

API quick start

Run p-image-try-on with a single API call.

POST https://api.wiro.ai/v1/Run/pruna/p-image-try-on
{
  "prompt": "Your prompt here",
  "inputImagePerson": "https://your-cdn.com/input.png",
  "inputImageGarments": "https://your-cdn.com/input.png",
  "inputImagePose": "https://your-cdn.com/input.png"
}
View full API docs