Skip to main content
POST
Generate images from references
For specific content requests, select the Special-Ns resource group.

Authorizations

Authorization
string
header
required

Bearer authentication header of the form Bearer <token>, where <token> is your auth token.

Body

application/json
model
enum<string>
required

Model ID.

Available options:
seedream-5-0-lite-260128
Example:

"seedream-5-0-lite-260128"

prompt
string
required

Text instruction for image generation or editing.

Example:

"Create four coherent cinematic storyboard frames of an astronaut repairing a spacecraft."

image
required

One reference image URL/Base64 data URI or an array of references. Up to 14 reference images are supported. The number of references plus generated images must not exceed 15.

Example:
size
string

Resolution tier 2K, 3K, or 4K, or a supported explicit pixel size.

Example:

"2K"

seed
integer<int32>

Random seed used to make repeated requests more consistent.

sequential_image_generation
enum<string>
default:disabled

Set to auto to allow coherent batch image output; disabled returns one image.

Available options:
auto,
disabled
sequential_image_generation_options
object

Batch output options. Used when sequential_image_generation is auto.

response_format
enum<string>
default:url

Return generated images as temporary URLs or Base64 data. URLs expire after 24 hours.

Available options:
url,
b64_json
output_format
enum<string>
default:jpeg

Generated image file format.

Available options:
png,
jpeg
watermark
boolean
default:true

Whether to add an AI-generated watermark.

optimize_prompt_options
object

Prompt optimization configuration.

Response

Images generated successfully.

model
string

Model used for generation.

Example:

"seedream-5-0-260128"

created
integer

Creation time as a Unix timestamp in seconds.

Example:

1784880000

data
object[]

Generated image results.

usage
object