pyappdist

Ship your Python app as a native installer straight from pyproject.toml. If it installs with pip, it almost certainly ships with pyappdist.

📖 Documentation: https://pyappdist.readthedocs.io/

pyappdist bridges the Python packaging ecosystem and native application distribution. It reads your application's pyproject.toml and builds setup packages for Windows, macOS, and Linux:

• Windows: MSI / MSIX
• macOS: DMG / .app bundle, PKG installer, or a self-extracting installer
• Linux: self-extracting installer

For example, to build a Windows MSI, configure pyproject.toml like this:

[tool.pyappdist]
name = "My App"
python = "3.12"

[[tool.pyappdist.launchers]]
name = "myapp"              # produces myapp.exe (or a shell wrapper on Linux/macOS)
entry = "myapp:main"        # module:callable
# gui = true                # use pythonw.exe (no console window) on Windows
# icon = { windows = "assets/app.ico" }   # per-OS launcher icon table
# args = "--serve"          # fixed leading arguments

[[tool.pyappdist.targets]]
name = "windows"
platform = "windows-x86_64"
format = "msi"
manufacturer = "Example Inc."
# scope = "user"            # "user" (default, no admin) or "machine" (Program Files)

Then build the MSI package (on Windows):

uvx pyappdist build

The result lands under appdist/<target>/dist/.

How it works

pyappdist creates a dedicated Python runtime directory and installs your application and its dependencies into it with pip. That runtime directory itself becomes the setup package. The Python runtime comes from python-build-standalone, the same distribution Astral's uv uses to build its environments.

With this approach, binary files such as a package's DLLs and related files such as images are placed in the proper directories according to the Python language specification and the PyPA specifications. Because this environment is used for the setup package as-is, most applications can be expected to run unmodified, with no per-application adjustments. If your app runs under uv run, it almost certainly runs after pyappdist build.

Why pyappdist

Tools such as PyInstaller and Nuitka analyze your code, select only the necessary files from the Python interpreter and dependency packages, and build an executable or a directory from that minimal set of files.

The problem is that the selection is not always correct. Static analysis cannot reliably find dynamically imported modules, data files, or plugins, so these tools often need per-application adjustments — hidden-import declarations, data-file lists, and library-specific hooks — and adding a new dependency can break the build again.

Those tools trade complexity for smaller distributions. A typical Python runtime adds roughly 100–150 MB. That used to matter more than it does today.

pyappdist makes the opposite trade-off: it builds a complete environment according to the Python and PyPA specifications and creates the distribution package from it. What your application and its dependencies contain does not matter — there is nothing to hunt down and nothing to adjust per application. It also does not build a one-binary executable — it only provides small launcher executables where needed. Through this design, pyappdist creates a stable application environment that end users can rely on.

What it produces

Each [[tool.pyappdist.targets]] entry describes one output package, selected by its platform and format:

format	Platform	Output
msi	Windows	.msi installer
msix	Windows	.msix package (Microsoft Store / sideloading)
linux	Linux	self-extracting .run installer
macos	macOS	self-extracting .run installer
macapp	macOS	.app bundle
dmg	macOS	.dmg disk image
pkg	macOS	.pkg installer
image	any	plain .zip / .tar.gz archive of the install tree

Supported platform values: windows-x86_64, windows-arm64, linux-x86_64, linux-aarch64, macos-x86_64, macos-aarch64. The macOS formats (macapp / dmg / pkg) support code-signing and notarization.

Samples

Runnable example apps live under samples/, each with its own [tool.pyappdist] config:

• helloworld — smallest config, no dependencies; a good starting template
• pandascli — pandas + numpy (C extensions)
• datafiles — ships a bundled data file, located via sysconfig
• multiprocessingdemo — multiprocessing with spawn works unmodified
• pytorchdemo — CUDA PyTorch via a per-index pin (Windows/Linux)
• matplotlibdemo — TkAgg GUI using the bundled tkinter
• pygamedemo — pygame-ce GUI (C extensions)
• pyside6demo — PySide6 (large abi3 wheel, Qt plugins)
• niceguidemo — NiceGUI + pywebview, per-target extras for backend selection
• pyvisionmarker — Gradio + pywebview object detection; environment markers pick the backend, MSI license dialog

Status

Beta: Core packaging workflows are ready for real-world use, although configuration details may still change before 1.0.