nintegerDefault: 1
Number of images to generate (1-6).
seedinteger
Random seed for reproducibility [0, 2147483647].
Using the same seed helps ensure consistency of generated content.
Note: Results are not guaranteed to be identical for every call even with the same seed.
sizeEnum
Output image resolution in format "width\*height".
Width and height must be in range [512, 2048] pixels.
Default: Maintains aspect ratio similar to input image with resolution close to 1024\*1024.
Possible values: 1024*1024
imageslist<string>* required
Array of input image URLs or Base64-encoded images.
Supports 1-3 images.
For multi-image input, the order defines their sequence.
The aspect ratio of the output image is determined by the last image.
Image requirements:
- Formats: JPG, JPEG, PNG, BMP, TIFF, WEBP, GIF (first frame only)
- Resolution: Width and height should be between 384 and 3072 pixels
- Size: No larger than 10 MB
promptstring* required
Image editing instruction (max 800 characters).
Describes the elements and visual features you want in the generated image.
When editing multiple images, use "Image 1", "Image 2", "Image 3" to refer to corresponding images.
Supports both Chinese and English.
prompt_extendbooleanDefault: true
Enable intelligent prompt rewriting.
When enabled, a large language model optimizes the positive prompt.
Significantly improves results for simple or less descriptive prompts.
safety_filterbooleanDefault: true
Enable content safety filter. Defaults to true. Set to false to disable content safety inspection.
negative_promptstring
Negative prompt describing unwanted content (max 500 characters).
Used to constrain the image generation.
Example: low resolution, error, worst quality, low quality, disfigured, extra fingers, bad proportions.