Optional
modelThe model to use for image generation.
Optional
nThe number of images to generate. Must be between 1 and 10. For dall-e-3
, only
n=1
is supported.
Optional
qualityThe 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
.
Optional
response_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.
Optional
sizeThe 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.
Optional
styleThe 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
.
Optional
userA 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-2
and 4000 characters fordall-e-3
.