OptionalmodelThe model to use for image generation.
OptionalnThe number of images to generate. Must be between 1 and 10. For dall-e-3, only
n=1 is supported.
OptionalqualityThe quality of the image that will be generated. hd creates images with finer
details and greater consistency across the image. This param is only supported
for dall-e-3.
Optionalresponse_The format in which the generated images are returned. Must be one of url or
b64_json. URLs are only valid for 60 minutes after the image has been
generated.
OptionalsizeThe size of the generated images. Must be one of 256x256, 512x512, or
1024x1024 for dall-e-2. Must be one of 1024x1024, 1792x1024, or
1024x1792 for dall-e-3 models.
OptionalstyleThe style of the generated images. Must be one of vivid or natural. Vivid
causes the model to lean towards generating hyper-real and dramatic images.
Natural causes the model to produce more natural, less hyper-real looking
images. This param is only supported for dall-e-3.
OptionaluserA unique identifier representing your end-user, which can help OpenAI to monitor and detect abuse. Learn more.
A text description of the desired image(s). The maximum length is 1000 characters for
dall-e-2and 4000 characters fordall-e-3.