Image Generation CLI

A command line interface for generating images. The intial focus will be on using the OpenAI API and gpt-image-2.

Install

Install with uv so image-gen is available on your PATH. From a clone of this repository:

uv tool install --editable .

Or install directly from GitHub:

uv tool install git+https://github.com/bwbensonjr/image-gen-cli

uv places the launcher in its tool-bin directory (typically ~/.local/bin). If image-gen is not found after install, run uv tool update-shell once and restart your shell, or check the location with uv tool dir --bin.

OPENAI_API_KEY (and optionally OPENAI_BASE_URL) must be set in your shell environment when invoking image-gen.

To upgrade after pulling new commits (non-editable installs):

uv tool upgrade image-gen-cli

Behavior

• Accepts an overall prompt describing what is wanted
• Optionally it can accept other inputs including
  • Markdown or text files
  • Example images
• An output file name specification (defaults to "output-image")
• The output is one or more images with names like "output-image-1.png" using the output file name spec.

Technical Guidelines

• Use OpenAI Python API which will use environment variables for OpenAI key and base URL.
• Use Python argparse.
• Use clig.dev command line guidelines cached locally as docs/cli_guidelines.md.

OpenAI gpt-image-2 Use

The recommendation for using gpt-image-2 from the OpenAI API is shown in this example:

prompt = "Generate an image of Otter hugging a Penguin"
output_file = "image_output.png"
client = OpenAI() 
response = client.responses.create(
    model="gpt-5.5",
    input=prompt,
    tools=[{"type": "image_generation"}],
)
# Save the image to a file
image_data = [
    output.result
    for output in response.output
    if output.type == "image_generation_call"
]
if image_data:
    image_base64 = image_data[0]
    with open(output_file, "wb") as f:
        f.write(base64.b64decode(image_base64))