Overview
Generate images from text prompts using the command line.The CLI has not yet picked up the Python defaults from PR #6262:
--model still defaults to dall-e-3 (retired by OpenAI on 2026-05-12) and --quality to standard (only standard/hd accepted). Because quality is now forwarded to every model, and GPT image models don’t accept standard, the CLI cannot generate with current OpenAI models yet. Use the Python API, which defaults to gpt-image-2.5-flare and sends quality only when set, until the CLI is updated.Commands
Basic Image Generation
praisonai images generate "A sunset over mountains"
With Model Selection
praisonai images generate "A blue circle" --model <model>
With Size
praisonai images generate "A landscape" --size 1536x1024
With Quality
praisonai images generate "A detailed portrait" --quality hd
Save to File
praisonai images generate "A cat" --output cat.png
Options
| Option | Short | Default | Description |
|---|---|---|---|
--model | -m | dall-e-3 (retired) | Model to use |
--size | -s | 1024x1024 | Image size |
--quality | -q | standard | standard or hd (always sent) |
--output | -o | output.png | Output file path |
--n | 1 | Number of images |
Examples
# Generate with the default model
praisonai images generate "A futuristic cityscape at night"
# Edit an image with a mask (uses image_edit's default, gpt-image-2.5-flare)
praisonai images edit photo.png "Make the cup blue" --mask mask.png -o edited.png