pypy

JIT-compiled Python interpreter for speed

TLDR

Run Python script

$ pypy [script.py]

Start interactive mode

$ pypy

Run with arguments

$ pypy [script.py] [arg1] [arg2]

Execute code inline

$ pypy -c "[print('Hello')]"

Run a module as a script

$ pypy -m [module_name]

Install a package using pip through PyPy

$ pypy -m pip install [package]

Run a script with optimizations enabled (removes assert statements)

$ pypy -O [script.py]

Enter interactive mode after running a script (for debugging)

$ pypy -i [script.py]

SYNOPSIS

pypy [options] [script] [args]

PyPy is an alternative Python interpreter that uses a Just-In-Time (JIT) compiler to achieve significantly faster execution than CPython for many workloads. It is compatible with Python 2.7 and Python 3.x syntax and can run most pure-Python code without modification, offering speedups of 2-10x or more on long-running programs.The JIT compiler analyzes code at runtime and compiles frequently executed paths to machine code, so programs that loop heavily or perform repetitive computations benefit the most. PyPy also includes optimizations for memory usage through more compact object representations. Some C extension modules may not be compatible, though cffi-based extensions work well.

PARAMETERS

SCRIPT

Python script to run.

ARGS

Script arguments.

-c CODE

Execute code.

-m MODULE

Run module.

-i

Interactive after script.

-O

Optimize mode: removes assert statements and sets _debug_ to False.

-V, --version

Print the PyPy version and exit.

-u

Unbuffered binary stdout and stderr.

-W arg

Warning control (e.g., -Wall to show all warnings).

CAVEATS

Not all C extension packages are compatible; cffi-based extensions work well but some CPython C API extensions may not. JIT compilation has a warmup period, so short-lived scripts may not see speed improvements. Use pypy -m pip instead of system pip to install packages into the PyPy environment.

HISTORY

PyPy was created as a high-performance Python interpreter with JIT.