GPT Image 2.5
Use gpt-image-2.5-flare or gpt-image-2.5-sunburst with the OpenAI-compatible Images API to generate images. Model availability depends on your account access.
Endpoint and Models
http
POST /v1/images/generations
Authorization: Bearer YOUR_API_KEY
Content-Type: application/jsonmodel | Description |
|---|---|
gpt-image-2.5-flare | Flare model |
gpt-image-2.5-sunburst | Sunburst model |
Request Parameters
| Parameter | Required | Description |
|---|---|---|
model | Yes | One of the model IDs above |
prompt | Yes | Description of the image to generate |
size | No | Output dimensions in WIDTHxHEIGHT format, for example 1024x1024 |
quality | No | auto, low, medium, or high; if omitted, the model chooses |
n | No | Number of images to generate; the example uses 1 |
The standard Images API uses size for dimensions, not resolution, aspect_ratio, or image_size. GPT Image models return base64-encoded images; do not send response_format.
cURL Example
Change model to gpt-image-2.5-sunburst to use Sunburst.
bash
curl -X POST https://cubicspaces.cloud/v1/images/generations \
-H "Content-Type: application/json" \
-H "Authorization: Bearer YOUR_API_KEY" \
-d '{
"model": "gpt-image-2.5-flare",
"prompt": "A small blue circle centered on a plain white background",
"size": "1024x1024",
"quality": "low",
"n": 1
}'Response Shape
json
{
"created": 1780000000,
"data": [
{ "b64_json": "iVBORw0KGgo..." }
],
"usage": {
"input_tokens": 18,
"output_tokens": 196,
"output_tokens_details": { "image_tokens": 196 }
}
}Read and base64-decode data[0].b64_json to get the image. usage reports token usage for this request; values vary by request. For reference-image editing, upload an image file to /v1/images/edits. Do not send image_urls to the generation endpoint.