wheel

A wheel is the standard built distribution format for Python packages: a ZIP-format archive with the .whl extension whose contents only need to be moved into the right location on the target system to be installed. The .whl file format was introduced to make Python package installation faster and more reliable than installing from source distributions (.tar.gz files).

Wheels are essentially ZIP files with a special naming convention and internal structure. They contain the package’s Python code, any compiled extensions, and metadata.

Key advantages of wheels include:

  • Faster installation, because the build step already happened, so installers just unpack the archive and copy files into place
  • No compiler needed on the installation system
  • Consistent installations across different systems
  • Reduced likelihood of installation errors

Wheels come in two varieties:

  • Pure Python wheels, which carry the compatibility tag py3-none-any (as in mypkg-1.0-py3-none-any.whl) and work on any Python 3 installation
  • Platform-specific wheels that contain compiled code built for a particular operating system, CPU architecture, and Python ABI. Wheels built against the stable ABI carry the abi3 tag and stay compatible across multiple CPython versions

Package maintainers build wheels by running a build frontend like build, which delegates the work to a build backend such as setuptools, and users typically install them automatically through pip, which preferentially chooses wheels over source distributions when available. The wheel tool can inspect, unpack, and repack .whl files directly.

What Are Python Wheels and Why Should You Care?

Tutorial

What Are Python Wheels and Why Should You Care?

In this tutorial, you'll learn what Python wheels are and why you should care as both a developer and end user of Python packages. You'll see how the wheel format has gained momentum over the last decade and how it has made the package installation process faster and more stable.

intermediate devops

For additional information on related topics, take a look at the following resources:

By Dan Bader • Updated Aug. 22, 2026
Python Glossary Share Feedback