Seedream5.5 - Layer Decomposition

This reference documents the ModelFold gateway.

Endpoint

POST /images/generations

Full URL: https://modelfold.ai/v1/images/generations

Authentication: every request must include the Authorization: Bearer <API_KEY> header.

Input parameters for the layer decomposition task

Request Parameters

Body (application/json)

Parameter	Type	Required	Description
model	string	Yes	The model name used to execute the layer decomposition task. Required field. - Must be seedream/5-5-layer-decomposition for this endpoint Default："seedream/5-5-layer-decomposition"。 Allowed values：seedream/5-5-layer-decomposition。 Example：seedream/5-5-layer-decomposition。
image_url	string	Yes	Required. URL of the image to decompose. - Exactly 1 image, publicly accessible URL - Supported formats: JPEG, PNG, WEBP - Max 10 MB Example：https://modelfold.ai/examples/seedream-5-5/1794318739537_P4dNQrYe.png。
prompt	string	No	Optional. At most 2000 characters. Optional. Describes which elements to separate, 0–2000 characters. - When provided, the model separates the described elements - When empty, the model identifies the main elements automatically - You can use <bbox>x1 y1 x2 y2</bbox> to mark element positions, with coordinates normalized to 0–1000 minLength: 0. maxLength: 2000. Example：Separate the title text <bbox>179 58 809 197</bbox> and the parrot <bbox>330 274 641 991</bbox> into independent layers。
size	string	No	Output size: auto, 1K, 1.5K or 2K. Default auto. auto or omitted keeps the input size, clamped to 1K–2K.
max_images	integer	No	Number of outputs, 2–100. Default 30 (one base image plus layers). The example uses 2.
output_format	string	No	Format of the base image. Defaults to png. Only affects the base image; separated layers are always PNG (transparent background). Default："png"。 Allowed values：png / jpeg。 Example：png。
text_recognition	boolean	No	Whether to recognize text in the layers. Defaults to false. Default：false。 Example：false。
nsfw_checker	boolean	No	Defaults to false. When false, no additional content filtering is applied and results are returned directly by the model.

Request Example

curl -X POST "https://modelfold.ai/v1/images/generations" \
  -H "Authorization: Bearer $MODELFOLD_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "seedream/5-5-layer-decomposition",
  "image_url": "https://modelfold.ai/examples/seedream-5-5/1794318739537_P4dNQrYe.png",
  "prompt": "Separate the title text <bbox>179 58 809 197</bbox> and the parrot <bbox>330 274 641 991</bbox> into independent layers",
  "size": "1K",
  "output_format": "png",
  "max_images": 2,
  "text_recognition": false
}'

{
  "model": "seedream/5-5-layer-decomposition",
  "image_url": "https://modelfold.ai/examples/seedream-5-5/1794318739537_P4dNQrYe.png",
  "prompt": "Separate the title text <bbox>179 58 809 197</bbox> and the parrot <bbox>330 274 641 991</bbox> into independent layers",
  "size": "1K",
  "output_format": "png",
  "max_images": 2,
  "text_recognition": false
}

Response

{
  "ok": true,
  "created": 1792134000,
  "data": [
    {
      "url": "https://modelfold.ai/uploads/outputs/result.png"
    }
  ],
  "usage": {}
}