FLUX.2 Virtual Try-On

FLUX.2 Virtual Try-On transfers a garment from one image onto a person in another image. It is a specialized clothing-editing workflow built around the person and garment assets rather than a general text-to-image request. A written instruction defines the intended wardrobe change while the paired images establish the visual material.

The route is useful for producing a dressed-person visualization without preparing a mask or removing the background first. It keeps the try-on operation within one image-editing request and returns a finished raster image. Additional controls let teams tune edit strength, composition shape, and inference settings when the default balance is not enough.

Highlights

• Two-image virtual try-on workflow. The exact workflow is defined around a person image and a garment image. This gives the model both the wearer context and the clothing reference in the same request.
• Specialized garment transfer. The route uses a clothing-focused FLUX.2 LoRA rather than treating the task as unrestricted image manipulation. Its purpose is transferring the referenced garment onto the depicted person.
• No manual masking or background removal. The documented dual-input workflow eliminates those preparation steps. Teams can start from the two visual assets instead of producing a separate cutout or mask.
• Broader clothing-editing foundation. FLUX.2 documentation covers outfit changes, accessories, garment recoloring, and clothing-style adaptation. The virtual try-on specialization places garment transfer within that established editing family.

Pricing

When to Use

Prompt Guide

Provide the person image first and the garment image second. Write a concise instruction that names the clothing transfer and any visible treatment you need; the images carry the primary identity and garment information.

Images:
1. <person image URL>
2. <garment image URL>
Prompt: Put the garment from image 2 on the person in image 1. <optional visible styling direction>

Example request:

{
  "images": [
    "https://example.com/person.jpg",
    "https://example.com/garment.jpg"
  ],
  "prompt": "Put the garment from image 2 on the person in image 1.",
  "lora_scale": 1,
  "aspect_ratio": "4:5",
  "output_format": "png"
}

Technical Specs

Related

• FLUX.2 Pro Edit — use the sibling editing route for broader image transformations beyond virtual try-on.