Images

images

Methods

Create Image Variation ->
post/images/variations

Creates a variation of a given image.

Create Image Edit ->
post/images/edits

Creates an edited or extended image given an original image and a prompt.

Create Image ->
post/images/generations

Creates an image given a prompt.

Security
Bearer Auth

Example: Authorization: Bearer My API Key

Body parameters
prompt: string

A text description of the desired image(s). The maximum length is 1000 characters for dall-e-2 and 4000 characters for dall-e-3.

model?: string |

The model to use for image generation.

n?: number

The number of images to generate. Must be between 1 and 10. For dall-e-3, only n=1 is supported.

quality?: "standard" | "hd"

The 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.

response_format?: "url" | "b64_json"

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.

size?: "256x256" | "512x512" | "1024x1024" | 2 more...

The 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.

style?: "vivid" | "natural"

The 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.

user?: string

A unique identifier representing your end-user, which can help OpenAI to monitor and detect abuse. Learn more.

Response fields
Request example
200Example

Domain types

Image = { b64_json, revised_prompt, url }
ImageModel = "dall-e-2" | "dall-e-3"
ImagesResponse = { created, data }