Image Generation Overview

The image generation endpoint is compatible with the OpenAI Images API. Integrated image model families are:

Model Family	Representative Model	Highlights	Docs
GPT-Image	gpt-image-2-pro	Strong instruction following, accurate text rendering	OpenAI-compatible format
Gemini Image	gemini-3-pro-image	Chat-based image editing, cost-effective	OpenAI-compatible format / Gemini native
Grok Imagine	grok-imagine-image / -quality / -2.0	Text/image-to-image, 1K/2K, quality tiers on 2.0	OpenAI-compatible format
Seedream	dola-seedream-5.0-pro / seedream-5.0-lite	Text-to-image, dual resolution-band pricing, strong Chinese text rendering	OpenAI-compatible format

Generic Endpoint

POST $URL/v1/images/generations

Generic Example

curl $URL/v1/images/generations \
  -H "Authorization: Bearer $YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-image-2-pro",
    "prompt": "A cyberpunk cat wearing sunglasses walking down a neon-lit street",
    "n": 1,
    "size": "1024x1024"
  }'

Common Parameters

Parameter	Type	Required	Description
model	string	Yes	Model name
prompt	string	Yes	Image description
n	integer	No	Number of images, default 1
size	string	No	Size; varies by model
quality	string	No	Quality, e.g. standard / hd

New image models will get a dedicated page by model family when launched.