Images
Image generation from a text prompt.
Request
http
POST /v1/images/generationsjson
{
"prompt": "Un renard roux dans une foret enneigee, style aquarelle",
"model": null,
"routing_strategy": "auto",
"size": "1024x1024",
"n": 1
}n (number of images, 1 to 4) and size are not supported the same way by every provider — check the capabilities of the chosen model via GET /v1/models?category=image if you set model explicitly.
Response
json
{
"images": ["https://cdn.uploadscenter.com/file_..."],
"model": "gpt-image-1",
"provider": "openai"
}Each URL points to a durably re-hosted file (UploadCenter CDN) — the provider's original temporary URL is never returned as-is.
Providers
OpenAI, Google (Imagen), Black Forest Labs (Flux).