Principle:Openai Openai python Image Generation Parameters
| Knowledge Sources | |
|---|---|
| Domains | Computer_Vision, Image_Generation |
| Last Updated | 2026-02-15 00:00 GMT |
Overview
A configuration pattern that defines the parameters controlling image generation and editing including model selection, dimensions, quality, format, and streaming options.
Description
Image generation parameters determine the characteristics of generated images. Key decisions include model selection (DALL-E 2, DALL-E 3, GPT-Image-1), output dimensions, quality level, response format (URL or base64), output file format (PNG, JPEG, WebP), and streaming partial image previews. Edit operations add parameters for source images, masks, and input fidelity.
Understanding parameter interactions is essential: different models support different size options, quality levels, and features (e.g., only GPT-Image-1 supports transparent backgrounds and streaming).
Usage
Use this principle when configuring image generation or editing requests. Select parameters based on the use case: high quality for production images, standard for drafts, streaming for interactive previews.
Theoretical Basis
Parameter configuration follows a TypedDict Schema pattern:
# Generation parameters
params = {
"prompt": description,
"model": "gpt-image-1", # Model selection
"size": "1024x1024", # Dimensions
"quality": "high", # Quality level
"response_format": "b64_json", # URL or base64
"output_format": "png", # File format
"background": "transparent", # Transparency support
"n": 1, # Number of images
"partial_images": 3, # Streaming previews
}
# Edit parameters extend with:
edit_params = {
"image": source_image,
"mask": mask_image, # DALL-E 2 only
"input_fidelity": "high", # Source image fidelity
}