Skip to content
API Reference

Z-Image

Z-Image generates an image from a text prompt. Two models are available: z-image-spicy, and z-image-spicy-pro with a larger maximum canvas.

POST https://llm.siraya.ai/v1/images/generations

The finished image comes back in the same response — no polling. Generation typically takes 10–60 seconds.

Supported Models

Model ID Width × height Default size prompt_extend default
z-image-spicy 256–1536 px per side 1024 × 1536 true (surcharged)
z-image-spicy-pro 256–2560 px per side 1024 × 1024 false
Authorization string Required
Your API Key (e.g., Bearer <API_KEY>).

Body

model string Required
z-image-spicy or z-image-spicy-pro.
prompt string Required
A text description of the image.
width / height integer
Output dimensions in pixels: 256–1536 on z-image-spicy, 256–2560 on z-image-spicy-pro. Defaults: 1024 × 1536 (spicy), 1024 × 1024 (pro).
size string
A standard WxH size such as 1024x1024, filling width and height in one field. When both are sent, size wins.
prompt_extend boolean
Rewrites the prompt before generation. On by default on z-image-spicy, where each rewritten request is surcharged; off by default on z-image-spicy-pro.
seed integer
Random seed for reproducibility. Omit it, or pass 0, for a random seed. 
curl https://llm.siraya.ai/v1/images/generations \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer <API_KEY>" \
  -d '{
    "model": "z-image-spicy-pro",
    "prompt": "A lighthouse on a cliff at dusk, oil painting",
    "width": 2048,
    "height": 1152
  }'

import requests

url = "https://llm.siraya.ai/v1/images/generations"
headers = {
    "Content-Type": "application/json",
    "Authorization": "Bearer <API_KEY>"
}
data = {
    "model": "z-image-spicy-pro",
    "prompt": "A lighthouse on a cliff at dusk, oil painting",
    "width": 2048,
    "height": 1152
}

response = requests.post(url, headers=headers, json=data)
print(response.json())

Example Response

{
  "created": 1792557768,
  "data": [
    { "url": "https://.../result_00.png" }
  ]
}

Behavior worth knowing

One image per request. Each call returns a single image in data.

prompt_extend costs extra on z-image-spicy. It is on by default there and each rewritten request is surcharged — send "prompt_extend": false if you don't need it. On z-image-spicy-pro it is off unless you turn it on.

Errors

A request-time refusal is final — retrying the same body won't help.

Status Meaning
400 Parameter validation — a dimension outside the model's range, or a missing prompt
401 Invalid API key
404 Unrecognized model code