Skip to main content

Install and Run Python Applications in Isolated Environments

Project description

pipx — Install and Run Python Applications in Isolated Environments

image PyPI version


Source Code:

For comparison to other tools including pipsi, see Comparison to Other Tools.

Install pipx


It is not recommended to install pipx via pipx. If you'd like to do this anyway, take a look at the pipx-in-pipx project and read about the limitations there.

On macOS

brew install pipx
pipx ensurepath
sudo pipx ensurepath --global # optional to allow pipx actions with --global argument

Upgrade pipx with brew update && brew upgrade pipx.

On Linux

  • Ubuntu 23.04 or above
sudo apt update
sudo apt install pipx
pipx ensurepath
sudo pipx ensurepath --global # optional to allow pipx actions with --global argument
  • Fedora:
sudo dnf install pipx
pipx ensurepath
sudo pipx ensurepath --global # optional to allow pipx actions with --global argument
  • Arch:
sudo pacman -S python-pipx
pipx ensurepath
sudo pipx ensurepath --global # optional to allow pipx actions with --global argument
  • Using pip on other distributions:
python3 -m pip install --user pipx
python3 -m pipx ensurepath
sudo pipx ensurepath --global # optional to allow pipx actions with --global argument

Upgrade pipx with python3 -m pip install --user --upgrade pipx.

On Windows

scoop install pipx
pipx ensurepath

Upgrade pipx with scoop update pipx.

  • install via pip (requires pip 19.0 or later)
# If you installed python using Microsoft Store, replace `py` with `python3` in the next line.
py -m pip install --user pipx

It is possible (even most likely) the above finishes with a WARNING looking similar to this:

WARNING: The script pipx.exe is installed in `<USER folder>\AppData\Roaming\Python\Python3x\Scripts` which is not on PATH

If so, go to the mentioned folder, allowing you to run the pipx executable directly. Enter the following line (even if you did not get the warning):

.\pipx.exe ensurepath

This will add both the above mentioned path and the %USERPROFILE%\.local\bin folder to your search path. Restart your terminal session and verify pipx does run.

Upgrade pipx with py -m pip install --user --upgrade pipx.

Using pipx without installing (via zipapp)

You can also use pipx without installing it. The zipapp can be downloaded from Github releases and you can invoke it with a Python 3.8+ interpreter:

python pipx.pyz ensurepath

Use with pre-commit

pipx has pre-commit support.

Shell completions

Shell completions are available by following the instructions printed with this command:

pipx completions

For more details, see the installation instructions.

Overview: What is pipx?

pipx is a tool to help you install and run end-user applications written in Python. It's roughly similar to macOS's brew, JavaScript's npx, and Linux's apt.

It's closely related to pip. In fact, it uses pip, but is focused on installing and managing Python packages that can be run from the command line directly as applications.

How is it Different from pip?

pip is a general-purpose package installer for both libraries and apps with no environment isolation. pipx is made specifically for application installation, as it adds isolation yet still makes the apps available in your shell: pipx creates an isolated environment for each application and its associated packages.

pipx does not ship with pip, but installing it is often an important part of bootstrapping your system.

Where Does pipx Install Apps From?

By default, pipx uses the same package index as pip, PyPI. pipx can also install from all other sources pip can, such as a local directory, wheel, git url, etc.

Python and PyPI allow developers to distribute code with "console script entry points". These entry points let users call into Python code from the command line, effectively acting like standalone applications.

pipx is a tool to install and run any of these thousands of application-containing packages in a safe, convenient, and reliable way. In a way, it turns Python Package Index (PyPI) into a big app store for Python applications. Not all Python packages have entry points, but many do.

If you would like to make your package compatible with pipx, all you need to do is add a console scripts entry point. If you're a poetry user, use these instructions. Or, if you're using hatch, try this.


pipx enables you to

  • Expose CLI entrypoints of packages ("apps") installed to isolated environments with the install command. This guarantees no dependency conflicts and clean uninstalls!
  • Easily list, upgrade, and uninstall packages that were installed with pipx
  • Run the latest version of a Python application in a temporary environment with the run command

Best of all, pipx runs with regular user permissions, never calling sudo pip install (you aren't doing that, are you? 😄).

Walkthrough: Installing a Package and its Applications With pipx

You can globally install an application by running

pipx install PACKAGE

This automatically creates a virtual environment, installs the package, and adds the package's associated applications (entry points) to a location on your PATH. For example, pipx install pycowsay makes the pycowsay command available globally, but sandboxes the pycowsay package in its own virtual environment. pipx never needs to run as sudo to do this.


>> pipx install pycowsay
  installed package pycowsay 2.0.3, Python 3.10.3
  These apps are now globally available
    - pycowsay
done! ✨ 🌟 ✨

>> pipx list
venvs are in /home/user/.local/share/pipx/venvs
apps are exposed on your $PATH at /home/user/.local/bin
   package pycowsay 2.0.3, Python 3.10.3
    - pycowsay

# Now you can run pycowsay from anywhere
>> pycowsay mooo
< mooo >
            (__)\       )\/\
                ||----w |
                ||     ||

Installing from Source Control

You can also install from a git repository. Here, black is used as an example.

pipx install git+
pipx install git+  # branch of your choice
pipx install git+  # git hash
pipx install  # install a release

The pip syntax with egg must be used when installing extras:

pipx install "git+[jupyter]"

Inject a package

If an application installed by pipx requires additional packages, you can add them with pipx inject. For example, if you have ipython installed and want to add the matplotlib package to it, you would use:

pipx inject ipython matplotlib

You can inject multiple packages by specifying them all on the command line, or by listing them in a text file, with one package per line, or a combination. For example:

pipx inject ipython matplotlib pandas
# or:
pipx inject ipython -r useful-packages.txt

Walkthrough: Running an Application in a Temporary Virtual Environment

This is an alternative to pipx install.

pipx run downloads and runs the above mentioned Python "apps" in a one-time, temporary environment, leaving your system untouched afterwards.

This can be handy when you need to run the latest version of an app, but don't necessarily want it installed on your computer.

You may want to do this when you are initializing a new project and want to set up the right directory structure, when you want to view the help text of an application, or if you simply want to run an app in a one-off case and leave your system untouched afterwards.

For example, the blog post How to set up a perfect Python project uses pipx run to kickstart a new project with cookiecutter, a tool that creates projects from project templates.

A nice side benefit is that you don't have to remember to upgrade the app since pipx run will automatically run a recent version for you.

Okay, let's see what this looks like in practice!

pipx run APP [ARGS...]

This will install the package in an isolated, temporary directory and invoke the app. Give it a try:

> pipx run pycowsay moo

< moo >
   \   ^__^
    \  (oo)\_______
       (__)\       )\/\
           ||----w |
           ||     ||

Notice that you don't need to execute any install commands to run the app.

Any arguments after the application name will be passed directly to the application:

> pipx run pycowsay these arguments are all passed to pycowsay!

< these arguments are all passed to pycowsay! >
   \   ^__^
    \  (oo)\_______
       (__)\       )\/\
           ||----w |
           ||     ||

Ambiguous arguments

Sometimes pipx can consume arguments provided for the application:

> pipx run pycowsay --py

usage: pipx run [-h] [--no-cache] [--pypackages] [--spec SPEC] [--verbose] [--python PYTHON]
                [--system-site-packages] [--index-url INDEX_URL] [--editable] [--pip-args PIP_ARGS]
                app ...
pipx run: error: ambiguous option: --py could match --pypackages, --python

To prevent this put double dash -- before APP. It will make pipx to forward the arguments to the right verbatim to the application:

> pipx run -- pycowsay --py

< --py >
   \   ^__^
    \  (oo)\_______
       (__)\       )\/\
           ||----w |
           ||     ||

Re-running the same app is quick because pipx caches Virtual Environments on a per-app basis. The caches only last a few days, and when they expire, pipx will again use the latest version of the package. This way you can be sure you're always running a new version of the package without having to manually upgrade.

Package with multiple apps, or the app name doesn't match the package name

If the app name does not match the package name, you can use the --spec argument to specify the PACKAGE name, and provide the APP to run separately:

pipx run --spec PACKAGE APP

For example, the esptool package doesn't provide an executable with the same name:

>> pipx run esptool
'esptool' executable script not found in package 'esptool'.
Available executable scripts: - usage: 'pipx run --spec esptool [arguments?]' - usage: 'pipx run --spec esptool [arguments?]' - usage: 'pipx run --spec esptool [arguments?]' - usage: 'pipx run --spec esptool [arguments?]'

You can instead run the executables that this package provides by using --spec:

pipx run --spec esptool
pipx run --spec esptool
pipx run --spec esptool
pipx run --spec esptool

Note that the .py extension is not something you append to the executable name. It is part of the executable name, as provided by the package. This can be anything. For example, when working with the pymodbus package:

>> pipx run pymodbus[repl]
'pymodbus' executable script not found in package 'pymodbus'.
Available executable scripts:
    pymodbus.console - usage: 'pipx run --spec pymodbus pymodbus.console [arguments?]'
    pymodbus.server - usage: 'pipx run --spec pymodbus pymodbus.server [arguments?]'
    pymodbus.simulator - usage: 'pipx run --spec pymodbus pymodbus.simulator [arguments?]'

You can run the executables like this:

pipx run --spec pymodbus[repl] pymodbus.console
pipx run --spec pymodbus[repl] pymodbus.server
pipx run --spec pymodbus[repl] pymodbus.simulator

Running a specific version of a package

The PACKAGE argument above is actually a requirement specifier. Therefore, you can also specify specific versions, version ranges, or extras. For example:

pipx run mpremote==1.20.0
pipx run --spec esptool==4.6.2
pipx run --spec 'esptool>=4.5'
pipx run --spec 'esptool >= 4.5'

Notice that some requirement specifiers have to be enclosed in quotes or escaped.

Running from Source Control

You can also run from a git repository. Here, black is used as an example.

pipx run --spec git+ black
pipx run --spec git+ black  # branch of your choice
pipx run --spec git+ black  # git hash
pipx run --spec black # install a release

Running from URL

You can run .py files directly, too.

pipx run
pipx is working!


That's it! Those are the most important commands pipx offers. To see all of pipx's documentation, run pipx --help or see the docs.


"Thanks for improving the workflow that pipsi has covered in the past. Nicely done!"
Jannis Leidel, PSF fellow, former pip and Django core developer, and founder of the Python Packaging Authority (PyPA)

"My setup pieces together pyenv, poetry, and pipx. [...] For the things I need, it’s perfect."

"I'm a big fan of pipx. I think pipx is super cool."
Michael Kennedy, co-host of PythonBytes podcast in episode 139


pipx was inspired by pipsi and npx. It was created by Chad Smith and has had lots of help from contributors. The logo was created by @IrishMorales.

pipx is maintained by a team of volunteers (in alphabetical order)


Issues and Pull Requests are definitely welcome! Check out Contributing to get started. Everyone who interacts with the pipx project via codebase, issue tracker, chat rooms, or otherwise is expected to follow the PSF Code of Conduct.

Project details

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

pipx-1.6.0.tar.gz (307.3 kB view hashes)

Uploaded Source

Built Distribution

pipx-1.6.0-py3-none-any.whl (77.8 kB view hashes)

Uploaded Python 3

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page