Skip to main content
GPT Image 1 is OpenAI’s autoregressive image generation model, available through the Starrise AI API. It generates high-quality images from text prompts, with precise text rendering and multi-style generation support.

​
Key Capabilities

  • Text-to-image — Generate images from natural language descriptions
  • Multi-style — Photorealistic, illustration, anime, vector, 3D, data visualization
  • Text rendering — Precise typographic output, ideal for posters, UI mockups, and labels
  • Transparent background — Supports generating images with transparent backgrounds
  • Batch generation — Generate up to 4 images per request

​
Output Specs

PropertyValues
Size1024x1024, 1536x1024, 1024x1536, auto
Qualitylow, medium, high
Formatpng, webp, jpeg
Backgroundtransparent, opaque

​
Quick Example

curl https://ai.alad.com/v1/images/generations \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-image-1",
    "prompt": "A cute cat floating in the Milky Way wearing a spacesuit",
    "size": "1024x1024",
    "quality": "high"
  }'

​
Parameters

ParameterTypeRequiredDescription
modelstringYesMust be gpt-image-1
promptstringYesImage description text
nintegerNoNumber of images to generate (1–4), default 1
sizestringNo1024x1024, 1536x1024, 1024x1536, auto, default auto
qualitystringNolow, medium, high, default medium
output_formatstringNopng, webp, jpeg, default png
backgroundstringNotransparent or opaque, default opaque
moderationstringNoauto or low, default auto
output_compressionintegerNoCompression level for jpeg/webp format (0–100)

API Reference

View the interactive API Playground for GPT Image 1.