Bria Virtual Try-On Image

This family does one job: it dresses a person in clothes you already have pictures of. The order of the pictures is what tells the model which is which — the first one you hand over is the person photo, and every picture after it is a garment or an accessory to put on them. There is no output shape to pick: the result takes its proportions from the person photo, so crop that picture before you upload if the frame should end up different.

What a run costs does not change with how many garments you add, so an outfit can be tried as a whole rather than piece by piece. A seed can be set, so when a result is worth coming back to, note it along with the pictures and the wording. The prompt is required; use it for what the pictures do not show, such as how a piece should be worn, rather than describing the person again.

• Size: 1K · Matches input image
• Reference images: 2–4 reference images · first the person photo, then garments

How the image side of the catalog works as a whole, and what it allows, is on the uncensored AI image generator page; changing a picture you already have is covered by the uncensored image editor.

Send vellria-bria-virtual-try-on as the model field to POST /v1/generate/image, with your API key as a Bearer token. The request below uses this model's default settings from the catalog and returns a generation id to poll.

curl https://vellria.com/v1/generate/image \
  -H "Authorization: Bearer vll_..." \
  -H "Content-Type: application/json" \
  -d '{
    "model": "vellria-bria-virtual-try-on",
    "prompt": "…",
    "size": "1K",
    "image_urls": [
      "https://…",
      "https://…"
    ]
  }'

Every parameter, the polling call and the upload endpoint are in the API documentation; what the image endpoints allow is on the NSFW image API page.