Image Generation

Generate images using OpenAI's gpt-image-1.5 model with support for transparency, custom sizes, and quality settings.

Quick Start

Generate an image:

python scripts/generate_image.py "a futuristic city at sunset" --size 1024x1024

Parameters

Model

• gpt-image-1.5 - State of the art (default)
• gpt-image-1 - Previous generation
• gpt-image-1-mini - Faster, lower cost

Size Options

• 1024x1024 - Square (default)
• 1536x1024 - Landscape
• 1024x1536 - Portrait
• auto - Let model decide

Quality Options

• auto - Model decides (default)
• low - Fast, ~$0.02/image
• medium - Balanced, ~$0.07/image
• high - Detailed, ~$0.19/image

Background Options

• auto - Model decides (default)
• transparent - For icons/logos (requires png/webp output)
• opaque - Solid background

Output Format

• png - Default, supports transparency
• webp - Smaller files, supports transparency
• jpeg - Smallest files, no transparency

Usage Examples

Basic generation:

python scripts/generate_image.py "a cute robot mascot"

Transparent icon:

python scripts/generate_image.py "app icon for a weather app" \
  --background transparent --size 1024x1024 --quality high

Marketing banner (landscape):

python scripts/generate_image.py "modern tech startup banner with blue gradient" \
  --size 1536x1024 --quality high

Portrait artwork:

python scripts/generate_image.py "oil painting of a mountain landscape" \
  --size 1024x1536 --quality high

Output

Images are saved to /home/ec2-user/mega-agent2/generated_images/ with timestamped filenames.

The script returns:

• image_path - Local file path
• revised_prompt - The prompt as interpreted by the model
• size - Actual dimensions