pyston
A JIT for Python
Decision gist · record as of 2026-08-14
Yes, if you are running Python 3.7–3.10 on macOS or Linux and want a low-friction performance boost for web or batch workloads. Use the lite variant to avoid C extension recompilation. No, if you depend on active maintenance, need Python 3.11+ support, or require a clear license statement—the project is dormant and license terms are undocumented. Conditional yes for experimental evaluation; not recommended for production systems without thorough testing.AI-flagged interpretation of the facts on this page — verify before relying
Before you install
- Pyston-full requires recompilation of C extensions (not ABI-compatible with CPython); pyston-lite avoids this by using existing binary packages but is only available for Python 3.7–3.10.
- Medium install friction due to platform-specific wheels (cp37–cp310 across macOS and Linux).
- The project is dormant (last release September 2022, no updates in 1418 days) but the repository remains active with recent commits as of August 2024.
License · maintenance · safety
(unclear) — License treatment is unclear—no SPDX identifier or raw license text is available in the metadata. Verify the actual license terms on the GitHub repository before committing to production use.
last release 2022-09-26 (1418 days) · last repo commit 2024-08-12 · 2,494 stars
0 known vulnerabilities (OSV.dev, 2026-08-14) · 465,333 downloads/mo, #6,508 on PyPI
Alternatives
Verify before relying
# Install the JIT extension module
pip install pyston
# Then run Python normally; the JIT can be loaded at startup.
# To verify Pyston-full:
import sys
hasattr(sys, "pyston_version_info")
# To verify Pyston-lite:
"pyston_lite" in sys.modules- Current performance gains relative to modern CPython versions; the 30% and 10% figures cited lack a timestamp or baseline version reference.
- Whether the dormant maintenance status (no releases since 2022-09-26) affects compatibility with recent Python ecosystem changes.
- Exact license terms and any restrictions on commercial or proprietary use.
- Whether this package on PyPI is the same as the full Pyston distribution or refers to the lite variant.
What it is and what it does
Pyston is a JIT compiler that accelerates Python code by compiling hot paths to machine code at runtime. It comes in two flavors: pyston-full, a fork of CPython 3.8 with integrated optimizations, and pyston-lite, an extension module that works with Python 3.7–3.10 without requiring recompilation of C extensions. The full version uses low-overhead JIT compilation via DynASM, quickening, and aggressive attribute caching; the lite version trades some performance for ease of deployment and binary package compatibility.
Pyston is designed for workloads where Python startup and execution speed matter—particularly web servers and long-running processes. It requires no code changes to use (drop-in replacement) but pyston-full is not ABI-compatible with CPython, meaning C extensions must be recompiled. The project is no longer actively maintained (last release 2022-09-26), though the repository shows recent activity; adoption should weigh the performance upside against the risk of compatibility issues as the Python ecosystem evolves.
Use it for
- Accelerate web server workloads where request latency and throughput directly impact user experience.
- Speed up long-running batch jobs or data processing scripts without code refactoring.
- Reduce cloud compute costs by achieving higher throughput per instance with the same hardware.
- Evaluate JIT compilation techniques for Python without switching to a different language.
- Profile and optimize hot paths in existing Python applications with minimal integration effort.
Worth the install?
AI-flagged interpretation of the facts on this page. Verify before relying on it.
Yes, if you are running Python 3.7–3.10 on macOS or Linux and want a low-friction performance boost for web or batch workloads.
Use the lite variant to avoid C extension recompilation. No, if you depend on active maintenance, need Python 3.11+ support, or require a clear license statement—the project is dormant and license terms are undocumented. Conditional yes for experimental evaluation; not recommended for production systems without thorough testing.
Install
pyston on PyPI
Before you install
Medium install friction due to platform-specific wheels (cp37–cp310 across macOS and Linux). The project is dormant (last release September 2022, no updates in 1418 days) but the repository remains active with recent commits as of August 2024. No runtime dependencies simplifies deployment once installed.
Pyston-full requires recompilation of C extensions (not ABI-compatible with CPython); pyston-lite avoids this by using existing binary packages but is only available for Python 3.7–3.10.
License in practice
License treatment is unclear—no SPDX identifier or raw license text is available in the metadata. Verify the actual license terms on the GitHub repository before committing to production use.
Quickstart
# Install the JIT extension module
pip install pyston
# Then run Python normally; the JIT can be loaded at startup.
# To verify Pyston-full:
import sys
hasattr(sys, "pyston_version_info")
# To verify Pyston-lite:
"pyston_lite" in sys.modulesVerify before relying
- Current performance gains relative to modern CPython versions; the 30% and 10% figures cited lack a timestamp or baseline version reference.
- Whether the dormant maintenance status (no releases since 2022-09-26) affects compatibility with recent Python ecosystem changes.
- Exact license terms and any restrictions on commercial or proprietary use.
- Whether this package on PyPI is the same as the full Pyston distribution or refers to the lite variant.
Package facts
| License | Not declared unclear |
| Python support | Not specified |
| Install friction | Medium. Platform-specific wheel |
| Runtime dependencies | None |
| Maintenance | Dormant 1,418 days since the last release |
| Last repo commit | |
| First released | |
| Downloads | 465,333 / month, #6,508 on PyPI 30-day window, as of 2026-08-14 |
| Known vulnerabilities | None known OSV.dev, checked 2026-08-14 |
Evidence: pyston-2.3.5-cp310-cp310-macosx_10_16_x86_64.whl; pyston-2.3.5-cp310-cp310-macosx_11_0_arm64.whl; pyston-2.3.5-cp310-cp310-manylinux_2_17_aarch64.manylinux2014_aarch64.whl; pyston-2.3.5-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl; pyston-2.3.5-cp37-cp37m-macosx_10_16_x86_64.whl; pyston-2.3.5-cp37-cp37m-manylinux_2_17_aarch64.manylinux2014_aarch64.whl; pyston-2.3.5-cp37-cp37m-manylinux_2_17_x86_64.manylinux2014_x86_64.whl; pyston-2.3.5-cp38-cp38-macosx_10_16_x86_64.whl; pyston-2.3.5-cp38-cp38-macosx_11_0_arm64.whl; pyston-2.3.5-cp38-cp38-manylinux_2_17_aarch64.manylinux2014_aarch64.whl; pyston-2.3.5-cp38-cp38-manylinux_2_17_x86_64.manylinux2014_x86_64.whl; pyston-2.3.5-cp39-cp39-macosx_10_16_x86_64.whl; pyston-2.3.5-cp39-cp39-macosx_11_0_arm64.whl; pyston-2.3.5-cp39-cp39-manylinux_2_17_aarch64.manylinux2014_aarch64.whl; pyston-2.3.5-cp39-cp39-manylinux_2_17_x86_64.manylinux2014_x86_64.whl
Tags
Let your AI agent find packages like this
Example. Real query, live index.
You found this page by searching. An agent finds it by wishing: SkillFed indexes 14,416 PyPI packages by what they can do, searchable in plain language.
wish › “pyston jit”
- pystonPyston is a performance-optimizing JIT compiler for Python that runs…
- pyston-autoloadAutomatically loads and enables Pyston, a performance-optimizing JIT…
- nvidia-nvjitlinkProvides NVIDIA's JIT LTO compiler library for Python, enabling…
Give your agent the search over MCP, or paste the wish link into any chat.
More Software Development packages
Provides backported and experimental type hints for Python 3.9+, allowing use of newer typing features on older Python versions and enabling early experimentation with type system PEPs before they enter the standard library.
NumPy provides an N-dimensional array object and a comprehensive suite of mathematical, linear algebra, Fourier transform, and random number functions for scientific computing in Python.
FastAPI is a Python web framework for building REST APIs using type hints, with automatic request validation, serialization, and interactive API documentation.
Provides a way to document function parameters, class attributes, return types, and variables inline using Python's `Annotated` type hint syntax instead of traditional docstrings.
Typer builds command-line applications from Python functions using type hints, automatically generating help text, argument parsing, and shell completion.
Install it if you are building CLIs in Python.
Distlib provides low-level packaging utilities for building, distributing, and managing Python software—including metadata handling, version specifiers, wheel support, script installation, and dependency resolution.
See also pyston-autoload · nvidia-nvjitlink · tokenspeed-triton · nvidia-nvjitlink-cu12 · lupa · drjit · triton · charmcraftcache · cppyy-cling · numba