Create Image

Generate pixel art images from text prompts. The system automatically selects the optimal AI model for your task.

A selection of sprites generated with Create Image — characters, objects, and environments.

Endpoint

POST /api/generations/image

Parameters

Max Sizes by Plan

Example

curl -X POST https://api.spritesynth.com/api/generations/image \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "description": "a pixel art wizard character with a blue robe",
    "image_size": "128x128",
    "seed": 42
  }'

Response

{
  "id": "gen_abc123",
  "status": "completed",
  "asset": {
    "url": "https://cdn.spritesynth.com/assets/gen_abc123.png",
    "width": 128,
    "height": 128
  },
  "credits_cost": 1,
  "duration_ms": 1200
}

Prompt Tips

Watch: How different prompt styles affect the output. Specific descriptions beat vague ones.

• Be specific — "32x32 warrior with red armor and sword" beats "a character"
• Specify size — Include pixel dimensions in your prompt for best results
• Use negative prompts — Exclude unwanted elements: "blurry, realistic, photorealistic"
• Use seeds — Save good seeds for reproducible results