GitHub PyPI conda-forge
Read the Docs Build

defopt is a lightweight, no-effort argument parser.

defopt will:

  • Allow functions to be run from code and the command line without modification.

  • Reward you for documenting your functions.

  • Save you from writing, testing and maintaining argument parsing code.

defopt will not:

  • Modify your functions in any way.

  • Allow you to build highly complex or customized command line tools.

If you want total control over how your command line looks or behaves, try docopt, click or argh. If you just want to write Python code and leave the command line interface up to someone else, defopt is for you.


Once you have written and documented your function, simply pass it to and you’re done.

import defopt

# Use type hints:
def main(greeting: str, *, count: int = 1):
    Display a friendly greeting.

    :param greeting: Greeting to display
    :param count: Number of times to display the greeting
    for _ in range(count):

# ... or document parameter types in the docstring:
def main(greeting, *, count=1):
    Display a friendly greeting.

    :param str greeting: Greeting to display
    :param int count: Number of times to display the greeting
    for _ in range(count):

if __name__ == '__main__':

Descriptions of the parameters and the function itself are used to build an informative help message.

$ python -h
usage: [-h] [-c COUNT] greeting

Display a friendly greeting.

positional arguments:
  greeting              Greeting to display

optional arguments:
  -h, --help            show this help message and exit
  -c COUNT, --count COUNT
                        Number of times to display the greeting
                        (default: 1)

Your function can now be called identically from Python and the command line.

>>> from test import main
>>> main('hello!', count=2)
$ python hello! --count 2


defopt was developed with the following guiding principles in mind:

  1. The interface can be fully understood in seconds. If it took any longer, your time would be better spent learning a more flexible tool.

  2. Anything you learn applies to the existing ecosystem. The exact same docstrings used by defopt are also used by Sphinx’s autodoc extension to generate documentation, and by your IDE to do type checking. Chances are you already know everything you need to know to use defopt.

  3. Everything is handled for you. If you’re using defopt, it’s because you don’t want to write any argument parsing code at all. You can trust it to build a logically consistent command line interface to your functions with no configuration required.

  4. Your Python functions are never modified. Type conversions are only ever applied to data originating from the command line. When used in code, duck-typing still works exactly as you expect with no surprises.


For source code, examples, questions, feature requests and bug reports, visit the GitHub repository.


Documentation is hosted on Read the Docs.