Skip to main content
POST
https://api.novita.ai
/
v3
/
seedream-3-0-txt2img
Seedream 3.0 Text to Image
curl --request POST \
  --url https://api.novita.ai/v3/seedream-3-0-txt2img \
  --header 'Authorization: <authorization>' \
  --header 'Content-Type: <content-type>' \
  --data '
{
  "prompt": "<string>",
  "model": "<string>",
  "response_format": "<string>",
  "size": "<string>",
  "seed": 123,
  "guidance_scale": 123,
  "watermark": true
}
'
{
  "image_urls": [
    "<string>"
  ],
  "binary_data_base64": [
    "<string>"
  ]
}
Experience fast and efficient image generation with Seedream 3.0, a state-of-the-art text-to-image model that produces high-quality images from textual prompts.
Currently, only the model version seedream-3-0-t2i-250415 is supported for Seedream 3.0.

​
Request Headers

​
Content-Type
string
required
Supports: application/json
​
Authorization
string
required
Bearer authentication format, for example: Bearer {{API Key}}.

​
Request Body

​
prompt
string
required
The text prompt used to generate the image.
​
model
string
The Model ID or inference endpoint (Endpoint ID) to be used for this request. Currently, only seedream-3-0-t2i-250415 for seedream 3.0 is supported.
​
response_format
string
Specifies the format of the generated image returned in the response. Default is url.
Supported values:
  • "url": Returns a downloadable JPEG image link.
  • "b64_json": Returns the image data as a Base64-encoded JSON string.
​
size
string
Specifies the dimensions (width x height in pixels) of the generated image. Must be between [512x512, 2048x2048]. Default is 1024x1024.
Recommended aspect ratios and resolutions:
  • 1024x1024 (1:1)
  • 864x1152 (3:4)
  • 1152x864 (4:3)
  • 1280x720 (16:9)
  • 720x1280 (9:16)
  • 832x1248 (2:3)
  • 1248x832 (3:2)
  • 1512x648 (21:9)
​
seed
integer
Random seed to control the stochasticity of image generation. Range: [-1, 2147483647]. If not specified, a seed will be automatically generated. To reproduce the same output, use the same seed value. Default is -1.
​
guidance_scale
number
Controls how closely the output image aligns with the input prompt. The higher the value, the less freedom the model has, and the stronger the prompt correlation. Range: [1, 10]. Default is 2.5.
​
watermark
boolean
Specifies whether to add a watermark to the generated image. Default is true.
  • false: No watermark.
  • true: Adds a watermark with the label “AI generated” in the bottom-right corner.

​
Response

​
image_urls
string[]
Array of generated image URLs. When response_format is set to "url", this array contains the downloadable image links for the generated images.
​
binary_data_base64
string[]
Array of Base64-encoded image data. When response_format is set to "b64_json", this array contains the generated images as Base64-encoded JSON strings.