pip — Package Installer — Python

pip — Package Installer

pip is Python's standard package installer. It downloads and installs libraries — anything from small utilities to large frameworks — so you can use them in your own code without writing them from scratch. pip has shipped bundled with Python since version 3.4, so on any modern Python install, it's already available via the pip (or pip3) command.

PyPI: Where Packages Come From

By default, pip downloads packages from the Python Package Index (PyPI), hosted at pypi.org. PyPI is a public registry where anyone can publish a package under a name and version number. When you run pip install requests, pip looks up requests on PyPI, resolves its dependencies, downloads the right files for your platform, and installs them into your environment.

Installing Packages

Bash

pip install requests

This installs the latest version of requests (and any packages it depends on) into your current Python environment.

Installing a Specific Version

Sometimes you need an exact version — to match a production environment, or because a newer release breaks your code. Pin the version with ==:

Bash

pip install requests==2.31.0

Upgrading and Uninstalling

To bring an already-installed package up to its latest version, use -U (short for --upgrade):

Bash

pip install -U requests
# equivalent, spelled out:
pip install --upgrade requests

To remove a package entirely:

Bash

pip uninstall requests

Listing Installed Packages

pip gives you two related but differently-formatted views of what's installed. pip list prints a friendly, human-readable table:

Bash

pip list

pip freeze, on the other hand, prints output in the exact package==version format that a requirements.txt file expects — making it the tool you use to snapshot an environment for reproduction elsewhere:

Bash

pip freeze

The requirements.txt Workflow

Most Python projects track their dependencies in a requirements.txt file. The typical workflow has two halves: capturing the current environment's package versions, and later reinstalling those exact versions somewhere else (a teammate's machine, a CI runner, a production server).

Bash

# Snapshot everything currently installed into requirements.txt
pip freeze > requirements.txt

# Later, on another machine, reproduce that exact set of packages
pip install -r requirements.txt

A typical requirements.txt file looks like this:

Bash

requests==2.31.0
flask==3.0.0
python-dotenv==1.0.0

Note

`pip freeze` captures every package in the current environment, including transitive dependencies you didn't install directly. For larger projects, many teams instead hand-maintain a `requirements.txt` (or use a tool like Poetry or pip-tools) listing only their direct dependencies, letting the tool resolve the rest.

Installing from GitHub or a Local Path

pip isn't limited to PyPI. You can install directly from a Git repository:

Bash

pip install git+https://github.com/user/repo.git

Or from a package sitting in a local directory — useful while developing your own library:

Bash

pip install ./my-local-package

# "editable" install: changes to the source are picked up
# immediately without reinstalling
pip install -e .

The -e (editable) flag is especially handy when you're actively developing a package alongside a project that depends on it — edits to the source take effect immediately, with no reinstall step.

Common pip Commands

Command	Description
pip install <package>	Install the latest version of a package
pip install <package>==<version>	Install a specific version
pip install -U <package>	Upgrade an already-installed package
pip uninstall <package>	Remove an installed package
pip list	Show installed packages in a readable table
pip freeze	Show installed packages in requirements.txt format
pip install -r requirements.txt	Install everything listed in a requirements file
pip install -e .	Install the current directory as an editable package
pip show <package>	Display metadata about an installed package

Tip

Always create and activate a virtual environment before running `pip install`. Without one, packages install into your global (or system) Python, which can quickly lead to version conflicts between unrelated projects. A virtual environment gives each project its own isolated set of installed packages.

Warning

Avoid running `pip install` with `sudo` to "fix" permission errors on your system Python. That installs packages system-wide, can conflict with packages your operating system depends on, and is almost always a sign you should be using a virtual environment instead.