What are the best tools for creating packaged executables for Python?

Question

7

Options
Considered

10

User
Recs.

Aug 19, 2021

Last
Updated

Here’s the Deal

7 Options Considered

Best tools for creating packaged executables for Python	Price	Last Updated
-- Nuitka	-	Sep 7, 2020
-- cx_Freeze	-	May 18, 2021
-- py2app	-	Oct 29, 2019
-- pynsist	-	Oct 29, 2019
-- dh-virtualenv	-	Jul 25, 2015

See Full List

I Recommend...

Built By the Slant team

Find the best product instantly.

Lustre recommends the best products at their lowest prices – right on Amazon.

4.7 star rating

Try it now - it's free

score 2 · Answer 1

Nuitka

Top Pro

Excellent performance optimizations

It's the best tool for performance conscious users. It turns Python code into C/C++ code instead of freezing it, and, as an ahead-of-time compiler, performs several optimizations when compiling Python code into C/C++ code, improving performance significantly. Some of this optimizations are constant folding (which means that if a variable stores the result of 3+3, Nuitka already knows it is 6 at execution time), conditional-statement predictions, and quicker function calls. This benchmark can be useful for comparing Nuitka's performance against Python's performance. See More

Top Con

Needs to be used with caution

Even though it goes through multiple tests before release, things can still break. For example, it needs configuration tweaks to work with some third-party packages, ir could cause app crashes, or fail to include dependencies. See More

Top Pro

Compatible with most Python versions

Nuitka can replace the standard Python interpreter and compile to every construct that CPython 2.6, 2.7, 3.3, 3.4, 3.5, 3.6, and 3.7 has. See More

Top Con

Cross-compilation is not supported

Nuitka executables will only run on the platform/OS it was compiled in. This meaning that executables compiled on Windows won't run on Linux and vice-versa. See More

Top Pro

Easy to use

You can turn a Python script to an executable with all the modules embedded by running the following command python -m nuitka --follow-imports program.py. See More

Top Con

Stand-alone executables have larger file sizes

Stand-alone option's binaries come with CPython included, increasing the size of the executable. See More

Top Pro

Can produce standalone executables

The --standalone argument creates executables that can run on machines that don't have Python installed or other dependencies needed for the script to run all in one single file. See More

Top Con

Does not work with Pyenv on MacOS

Top Pro

Is optimized to avoid overhead

Avoids big file sizes by only including what is needed for your script to run. See More

Top Pro

Source code can be obfuscated

As it compiles to C/C++, no Python Bytecode is left. As a result, the code is obfuscated enough to prevent most cases of tampering. See More

Top Pro

Fully compatible with standard Python modules

Nuitka can be used with all Python standard libraries, modules, and extension modules. See More

Top Pro

Covers most use-cases

Can compile a Python App into an executable, an extension module or a program with modules embedded. See More

Top Pro

Supports all major OS platforms

Available on Linux, FreeBSD, NetBSD, macOS, and Windows 64/32 bits See More

See All

Built by Slant

Find the best product
instantly.

Stop wasting time searching endlessly. Lustre recommends the best products at their lowest prices – right on Amazon.

4.7 star rating

Add to Chrome - It's Free

Add to Edge - It's Free

Add to Firefox - It's Free

Add to Opera - It's Free

Add to Brave - It's Free

Add to Safari - It's Free

Try now - it's free

score 1 · Answer 2

cx_Freeze

Top Pro

It's cross platform

If you have a Python Script, you can use cx_Freeze to turn it into an executable on the platform you use it for. It will run on any platform where Python itself runs on. Note, however, that it can't cross-compile. See More

Top Con

Needs you to use a setup script

Won't detect that a file needs to be copied if it is not dynamically loaded by your program, you have to tell cx_Freeze about it in a setup script. See More

Top Pro

Egg support

Support for searching zip files (.egg files) for modules. See More

Top Con

Not a cross-compiler

While you can use cx_Freeze on any platform where python itself works on, you can't cross-compile (compile an app for another platform). See More

Top Con

Source code can be obtained

Since cx_Freeze does not provide solutions to reverse engineering or decompiling. Python bytecode can be obtained and the logic of your app can be exposed. See More

Top Con

Single file mode not supported

Can't compile all your script modules, dependencies and archives into a single executable stand alone file. See More

See All

Answer 3

py2app

Top Con

Mac OSX only

This tool is designed to "Create standalone Mac OS X applications" See More

Answer 4

pynsist

Top Con

Windows only

This tool is designed to build python applications for windows, and cannot build for multiple platforms. See More

Answer 5

dh-virtualenv

Add Video or Image

score 1 · Answer 6

Pyinstaller

Top Con

Not highly configurable

There is an option to dictate that the package should be completely run from one file, but other than a few other modification commands pyinstaller is not very configurable. See More

Top Pro

Easy to use

PyInstaller is one of the best tools for easily freezing Python Scripts and Modules into self-contained executables. It can be used via CLI with a single command and without many configurations. This tool pulls all the necessary files and libraries needed into a file to have your program run on any PC as an executable app. See More

KnowledgeablePhrike's Experience

This program is a waste of disk space. See More

Top Con

Fails to produce .exe

Produces nothing in the \dist directory, no .exe file, nothing except a \setup directory. A waste of disk space. I have a simple single python .py file, marginally above "hello", no user input, a simple parse of known binary file to a known text file. Copious unintelligible output to no effect. See More

Top Pro

Compatible with most 3rd-party packages out-of-the-box

Most packages will work without any configuration. The community maintains a list of tested packages here. The list is incomplete so any user can contribute to it by editing the table or writing to the mailing list. See More

standupberowdy's Experience

So going through all of the recommended libraries on here, pyinstaller was the most effortless along with nuitka (nuitka is used specifically for compiling to See More

Top Con

OS Specific

Any executable created can only run on the OS, bit size, and python version it was created on. If you want to create a package that runs on another machine you will need to know these exact parameters in order for it to function. See More

Top Pro

Executables can be compressed with UPX to reduce their size drastically

Executables can be compressed with UPX, a free utility. You can see how it works here. Note that currently, it has no effect on macOS 64-bit. See More

Top Pro

Easy install

Simply pip install pyinstaller and it is ready to go. See More

Top Pro

Can add binariess to the executable

The --add-binary flag allows you to bundle external executables along with your script. You can, for example, bundle chromium or selenium. See More

Top Pro

Can obfuscate Python byte code

Python bytecode can be obfuscated by specifying an encryption key with a CLI command. This should prevent most forms of tampering with the executable, such as reverse-engineering it to extract the underlying Python code. While the bytecode is encrypted using AES256, it should not be confused for security because the key can be extracted from the executable. See More

Top Pro

Compatible with Cython modules

Pyinstaller can resolve, use, and package Cython modules into the executable. See More

See All

Answer 7

py2exe

Add Video or Image

7 Options Considered

My Recommendation for Nuitka

My Recommendation for Nuitka

Excellent performance optimizations

Needs to be used with caution

Compatible with most Python versions

Cross-compilation is not supported

Easy to use

Stand-alone executables have larger file sizes

Can produce standalone executables

Does not work with Pyenv on MacOS

Is optimized to avoid overhead

Source code can be obfuscated

Fully compatible with standard Python modules

Covers most use-cases

Supports all major OS platforms

My Recommendation for cx_Freeze

My Recommendation for cx_Freeze

It's cross platform

Needs you to use a setup script

Egg support

Not a cross-compiler

Source code can be obtained

Single file mode not supported

My Recommendation for py2app

My Recommendation for py2app

Mac OSX only

My Recommendation for pynsist

My Recommendation for pynsist

Windows only

My Recommendation for dh-virtualenv

My Recommendation for dh-virtualenv

My Recommendation for Pyinstaller

My Recommendation for Pyinstaller

Not highly configurable

Easy to use

Fails to produce .exe

Compatible with most 3rd-party packages out-of-the-box

OS Specific

Executables can be compressed with UPX to reduce their size drastically

Easy install

Can add binariess to the executable

Can obfuscate Python byte code

Compatible with Cython modules

My Recommendation for py2exe

My Recommendation for py2exe

Don't see your favorite option? Add it.

One sec!