Appearance

Image generation ​

LibertAI offers image generation models with competitive pricing. Generate high-quality images from text prompts via two compatible endpoints — Stable Diffusion WebUI and OpenAI's image generation format.

Loading models & pricing…

API endpoints ​

Stable Diffusion WebUI API (sdapi) ​

Compatible with the Stable Diffusion WebUI format.

  • Endpoint: POST https://api.libertai.io/sdapi/v1/txt2img
json
{
  "model": "z-image-turbo",
  "prompt": "a cute cat, anime style",
  "width": 512,
  "height": 512
}
json
{ "images": ["<base64_encoded_image>"] }

OpenAI-compatible API ​

Compatible with OpenAI's image generation format.

  • Endpoint: POST https://api.libertai.io/v1/images/generations
json
{
  "model": "z-image-turbo",
  "prompt": "a cute cat, anime style",
  "size": "512x512"
}
json
{ "data": [{ "b64_json": "<base64_encoded_image>" }] }

Usage examples ​

sh
curl -s -X POST "https://api.libertai.io/sdapi/v1/txt2img" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -d '{
    "model": "z-image-turbo",
    "prompt": "a cute cat, anime style",
    "width": 512,
    "height": 512
  }' \
  | jq -r '.images[0]' | base64 -d > output.png

# OpenAI-compatible
curl -s -X POST "https://api.libertai.io/v1/images/generations" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -d '{
    "model": "z-image-turbo",
    "prompt": "a cute cat, anime style",
    "size": "512x512"
  }' \
  | jq -r '.data[0].b64_json' | base64 -d > output.png

Parameters ​

Required ​

  • model — model ID, e.g. z-image-turbo
  • prompt — text description of the image

sdapi optional parameters ​

  • width, height — image size in pixels (default: 1024)
  • steps — diffusion steps (default: 9 — more steps = higher quality, slower)
  • seed — seed for deterministic generation
  • n — number of images (default: 1, max: 4)
  • remove_background — strip the background post-generation (default: false)

OpenAI-compatible optional parameters ​

  • size — WIDTHxHEIGHT (e.g. 512x512, default 1024x1024)
  • n — number of images (default: 1, max: 4)
  • remove_background — strip the background post-generation (default: false)

Direct model interaction ​

You can call the image model's CRN host directly using GET /libertai/models for discovery and the same /sdapi/v1/txt2img or /v1/images/generations paths on the host with your LibertAI API key. The pattern, trade-offs, and code samples are documented once in the text guide: Direct model interaction.

See also ​