Skip to main content
POST
/
v1
/
images
/
generations
curl --request POST \
  --url https://api.getinfinityblue.com/v1/images/generations \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: application/json' \
  --data '
{
  "model": "gpt-image-2",
  "prompt": "A cute baby sea otter floating on a blue ocean",
  "size": "1024x1024"
}
'
{
  "created": 123,
  "data": [
    {
      "b64_json": "<string>",
      "url": "<string>"
    }
  ],
  "usage": {
    "total_tokens": 123,
    "input_tokens": 123,
    "output_tokens": 123,
    "input_tokens_details": {
      "text_tokens": 123,
      "image_tokens": 123
    }
  }
}

​
Models you can use

Model IDNotes
gpt-image-2GPT Image 2 — flexible sizing and multiple output formats
Snapshot versions (e.g. gpt-image-2-2026-04-21) are currently unavailable. Please use gpt-image-2.

​
Size options (gpt-image-2)

The size parameter defaults to auto and supports flexible dimensions:
  • Maximum edge length: 3840px
  • Both width and height must be multiples of 16
  • Aspect ratio (long side to short side) must not exceed 3:1
  • Total pixel count between 655,360 and 8,294,400
Common presets: 1024x1024, 1536x1024 (landscape), 1024x1536 (portrait), 2048x2048, 2048x1152, 3840x2160, 2160x3840. Custom sizes are also supported (e.g. 1024x768), as long as both dimensions are multiples of 16 and satisfy the constraints above.

​
Streaming

Streaming is not currently supported. The stream and partial_images parameters are accepted but have no effect — the response is always returned as a complete JSON payload.

​
Quality and format

  • quality: Quality tiers are not currently supported; all requests use the auto level. The parameter is accepted but has no observable difference between levels.
  • output_format: png (default) / jpeg / webp
  • output_compression: applies to jpeg / webp only, range 0–100

Authorizations

​
Authorization
string
header
required

Bearer token authentication, format: Authorization: Bearer sk-xxxxxx. Get your API key in the console.

Body

application/json

Image generation request body.

​
model
string
required

Model ID to use for image generation. Pass gpt-image-2. Snapshot versions (e.g. gpt-image-2-2026-04-21) are currently unavailable.

Example:

"gpt-image-2"

​
prompt
string
required

A text description of the desired image. For gpt-image-2, the maximum length is 32,000 characters.

Example:

"A cute baby sea otter floating on a blue ocean"

​
n
integer
default:1

The n parameter is not currently supported. Defaults to n=1, returning one image per request. To generate multiple images, call the endpoint in a loop.

Required range: 1 <= x <= 1
​
size
string
default:auto

Size of the generated image. For gpt-image-2, defaults to auto with flexible dimensions (max edge 3840px, both dimensions multiples of 16, aspect ratio ≤ 3:1, total pixels 655,360–8,294,400). Common presets: 1024x1024, 1536x1024, 1024x1536, 2048x2048, 2048x1152, 3840x2160, 2160x3840.

Examples:

"1024x1024"

"1536x1024"

"auto"

​
background
enum<string>

Background setting for the generated image. For gpt-image-2, defaults to auto; supports opaque or auto. gpt-image-2 does not support transparent — passing it returns a 400 error.

Available options:
opaque,
transparent,
auto
​
moderation
enum<string>

Content moderation level. For gpt-image-2, defaults to auto (standard); low applies a more permissive policy.

Available options:
auto,
low
​
quality
enum<string>

Quality tiers are not currently supported; all requests use the auto level. The parameter is accepted but has no observable difference between levels.

Available options:
low,
medium,
high,
auto
​
stream
boolean
default:false

Streaming is not currently supported. The parameter is accepted but has no effect — the response is always returned as a complete JSON payload.

​
style
string

Image style hint. Only supported by DALL·E 3. For gpt-image-2, the parameter is accepted but has no effect.

​
output_format
enum<string>

Output format of the generated image. For gpt-image-2, defaults to png; options: png, jpeg, webp. Only GPT Image models support this parameter.

Available options:
png,
jpeg,
webp
​
output_compression
integer

Output compression quality (0–100). For gpt-image-2, applies only when output_format is jpeg or webp.

Required range: 0 <= x <= 100
​
partial_images
integer
default:0

Number of intermediate partial frames when streaming (0–3). Depends on streaming support — since streaming is not currently available, this parameter has no effect.

Required range: 0 <= x <= 3
​
response_format
enum<string>
default:b64_json

Format in which the generated image is returned. Defaults to b64_json (Base64-encoded image data); set to url to receive a temporary access URL (download or save promptly).

Available options:
b64_json,
url
​
user
string

A unique identifier for your end user, useful for monitoring and abuse detection.

Response

Image generated successfully

Image generation response body.

​
created
integer

Unix timestamp (seconds) of creation.

Example:

1713833628

​
data
object[]

List of generated images.

​
usage
object

Token usage statistics for the request.