Skip to main content
MarkdownLLMs.txt

GPT-Image-2

In this page's examples, {BASE_URL} is https://api.uniall.ai.

Use gpt-image-2 for text-to-image and image editing through the OpenAI-compatible image endpoints. This model uses size to control both dimensions and aspect ratio. It does not support aspect_ratio.

Endpoint​

POST /v1/images/generations
POST /v1/images/edits

Use /v1/images/generations for text-to-image. Use /v1/images/edits when a reference image is included.

Authentication​

Authorization: Bearer sk-***
Content-Type: application/json

Request Parameters​

ParameterTypeRequiredDescription
modelstringYesUse gpt-image-2.
promptstringYesImage generation or editing prompt.
imagestring/objectConditionalReference image for editing.
imagesarrayConditionalMultiple reference images when supported.
sizestringYesOutput dimensions, such as 1024x1024.
qualitystringNolow, medium, or high. Default: medium.
output_formatstringNopng, jpeg, or webp.
backgroundstringNoauto (model chooses from the prompt), opaque (opaque background), or transparent (transparent background). Omitted or null: auto.
nintegerNoOutput image count. Default: 1; range: 1 to 8.
response_formatstringNoPrefer url.

Background Control​

Background control is pending release. Pass background at the top level of the request body.

  • auto: the model chooses the background from the prompt. This is the default when omitted or set to null.
  • opaque: request an opaque background.
  • transparent: request a transparent background. Set output_format to png or webp; if omitted, it defaults to png.

An invalid background value or a transparent/jpeg combination returns HTTP 400 during request validation, before generation starts.

Applies to both synchronous generation and editing endpoints on this page. Existing GPT Image 2 async image tasks and image-chat compatibility requests use the same top-level field. Image results remain URL-based.

Text-To-Image Example​

curl -X POST "{BASE_URL}/v1/images/generations" \
  -H "Authorization: Bearer sk-***" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-image-2",
    "prompt": "An orange cat sticker, full subject visible, transparent background",
    "size": "1024x768",
    "quality": "low",
    "background": "transparent",
    "output_format": "webp",
    "n": 1,
    "response_format": "url"
  }'

Image Editing Example​

curl -X POST "{BASE_URL}/v1/images/edits" \
  -H "Authorization: Bearer sk-***" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-image-2",
    "prompt": "Keep the product shape and remove the background to make it transparent",
    "image": "https://example.com/input.png",
    "size": "1024x1024",
    "quality": "medium",
    "background": "transparent",
    "output_format": "png",
    "n": 1,
    "response_format": "url"
  }'

Response Example​

{
  "created": 1778688000,
  "data": [
    {
      "url": "https://example.com/generated.png"
    }
  ]
}

Common Errors​

  • Missing size.
  • Passing aspect_ratio; use size instead.
  • Passing quality outside low, medium, or high.
  • Requesting n outside 1 to 8.
  • Using an unsupported size for the current account.