Latent Consistency (SDXL & SDv1.5) Text to Image

{
  "images": [
    {
      "url": "",
      "content_type": "image/jpeg"
    }
  ],
  "num_inference_steps": 4
}