uv

uv is an all-in-one Python package and project manager written in Rust by Astral — the team behind ruff. It replaces a sprawling ecosystem of Python tools (pip, pip-tools, pipx, poetry, pyenv, virtualenv) with a single, unified binary that is 10–100× faster than its Python-based equivalents.

Features

• Blazing fast — installs packages 10–100× faster than pip thanks to a parallel, Rust-native resolver and installer
• Drop-in pip replacement — uv pip install, uv pip freeze, uv pip sync are fully compatible with existing workflows
• Project management — manages pyproject.toml-based projects with lockfiles, similar to poetry or pdm
• Python version management — downloads and manages Python interpreters, replacing pyenv
• Tool runner — run CLI tools in isolated environments with uvx, replacing pipx
• Script runner — run single-file Python scripts with inline dependency declarations (uv run)
• Virtual environments — creates and manages .venv environments automatically
• Universal lockfile — uv.lock is cross-platform and human-readable
• Workspace support — manage monorepos with multiple Python packages from one root

Installation

# macOS / Linux
curl -LsSf https://astral.sh/uv/install.sh | sh

# Debian / Fedora
# Official install script (works on all Linux distributions):
# curl -LsSf https://astral.sh/uv/install.sh | sh

# via cargo
cargo install uv

# Homebrew (macOS)
brew install uv

# Windows (PowerShell)
powershell -c "irm https://astral.sh/uv/install.ps1 | iex"

Usage

As a pip replacement

# Install a package
uv pip install requests

# Install from a requirements file
uv pip install -r requirements.txt

# Sync an environment to exactly match requirements
uv pip sync requirements.txt

# Compile requirements.in to a pinned requirements.txt
uv pip compile requirements.in -o requirements.txt

Project management

# Create a new project
uv init my-project
cd my-project

# Add a dependency (updates pyproject.toml and uv.lock)
uv add requests

# Remove a dependency
uv remove requests

# Install all project dependencies
uv sync

# Run a command inside the project environment
uv run python main.py
uv run pytest

Python version management

# Install a specific Python version
uv python install 3.12

# List available Python versions
uv python list

# Pin the Python version for a project
uv python pin 3.11

Running tools with uvx

# Run a CLI tool without installing it globally
uvx ruff check .
uvx black --check .
uvx httpie GET httpbin.org/get

Single-file scripts

# Run a script, installing its declared dependencies automatically
uv run script.py

Declare dependencies inline in the script:

# /// script
# dependencies = ["requests", "rich"]
# ///
import requests
from rich import print
print(requests.get("https://httpbin.org/get").json())

Why uv over pip/poetry?

The speed difference is dramatic on cold installs — installing the scientific Python stack (numpy, pandas, matplotlib) takes seconds instead of minutes. Beyond speed, having a single binary that covers the entire Python toolchain (versions, environments, packages, scripts, tools) eliminates the confusion around which tool to use for which task.