Skip to the content
Know your tech before you buy it. Makers: get listed from $49/mo

uv

docs.astral.sh

  • Free plan
  • Windows
  • macOS
  • Linux
8.1 Club score Visit docs.astral.sh

What uv is

uv is a free Python package and project manager written in Rust for macOS, Linux, and Windows. It manages dependencies and environments, with lockfile and workspace support, and can install, list, pin, or uninstall Python versions. It can also run Python scripts with dependencies and run or install command-line tools from Python packages. Its pip-compatible interface can take on tasks handled by pip, pip-tools, pipx, poetry, pyenv, twine, and virtualenv, although its commands do not exactly match those tools. The documentation describes uv as 10–100x faster than pip. It supports source distributions, wheels, requirements.txt, pylock.toml, offline installation, and private registry authentication. By default, its first-index strategy restricts package candidates to the first index where a package is found. Credentials are not stored in uv.lock, but uv currently keeps credentials in a plaintext file; native system secret storage is available as an experimental preview. Installation options include standalone installers, package managers, Docker, GitHub Releases, and Cargo.

Who it suits

It suits Python developers managing dependencies, environments, and project workflows across supported desktop platforms. Teams using CI, containers, notebooks, or package registries can consult its integration guides.

Good to know, and what to watch

  • Manages dependencies, environments, lockfiles, and workspaces.
  • Can install and manage Python versions.
  • Supports offline installation and private registry authentication.
  • Available at no charge.
  • Credentials are currently stored in a plaintext file.
  • Native system secret storage is experimental.
  • Commands do not exactly match the tools it can replace.
  • Some platforms without a prebuilt wheel require a Rust toolchain.

uv

Free

Free plan

Getting started

  1. Open the uv documentation website
  2. Install uv with a standalone installer or a supported package manager
  3. Use uv on Linux, macOS, or Windows
  4. Create projects with dependencies, environments, lockfiles, or workspaces
  5. Install or select Python versions and run scripts or command-line tools

Limits to check

The free uv plan is the open-source package and project manager. Platforms without a prebuilt wheel may require a Rust toolchain to build from source, and native system secret storage is only an experimental preview.

The Club's review

Overview

uv is a free Python package and project manager for developers who want to manage dependencies, environments, Python versions, scripts, and command-line tools in one place. Its wide scope and stated speed make it a strong candidate for Python workflows, provided its command differences and credential handling fit the way you work.

Built by Astral, uv is designed to take on tasks commonly handled by pip, pip-tools, pipx, Poetry, pyenv, twine, and virtualenv. The documentation describes it as 10–100x faster than pip, but that is not a promise of the same gain in every workflow. Its pip-compatible interface also differs in command details and behavior from the tools it draws on, so teams should account for adjustment when replacing established processes.

Browse our Python Package Managers and Package Managers directories for other options.

Key features

Project dependencies and environments

uv resolves dependencies and manages project environments, lockfiles, and workspaces. It works with source distributions, wheels, requirements.txt, and pylock.toml, and supports both private registry authentication and offline installation. That combination suits projects that need reproducible dependency management or work with internal package sources, including environments with limited connectivity.

Its default first-index strategy limits package candidates to the first index where a package is found, helping reduce dependency-confusion risk. Credentials are not stored in uv.lock; uv can use environment variables, URLs, netrc, and keyring for credentials. However, uv currently stores credentials in a plaintext file. Native system secret storage is available only as an experimental preview, so teams handling sensitive registry credentials should weigh that limitation before adopting it.

Python versions, scripts, and tools

uv can install, list, pin, and uninstall Python versions. It can also run Python scripts with dependencies and run or install command-line tools distributed through Python packages. These capabilities broaden its appeal beyond package installation: developers can consolidate more of their Python setup and day-to-day tooling in one manager.

Integrations and installation

Official integration guides cover Docker, Jupyter, GitHub Actions, GitLab CI/CD, Pre-commit, PyTorch, FastAPI, Bazel, and package registries. uv uses rustls with aws-lc-rs for TLS and bundled Mozilla root certificates by default. It is distributed through a standalone installer, PyPI, Homebrew, MacPorts, WinGet, Scoop, Docker, GitHub Releases, and Cargo. On platforms without a prebuilt PyPI wheel, installing from PyPI requires a Rust toolchain to build from source.

Pricing

uv is free: its open-source Python package and project manager costs nothing. There is no paid tier or trial to weigh against the free offering. The free plan is the tool itself, rather than a reduced edition with stated feature or usage caps.

Platforms

uv supports Linux, macOS, and Windows, as well as CPython, PyPy, Pyodide, and GraalPy. Its package workflows cover source distributions, wheels, requirements.txt, and pylock.toml. The PyPI route has a practical caveat on platforms without a prebuilt wheel: building from source requires a Rust toolchain. Other installation routes include native package managers, Docker, GitHub Releases, and Cargo.

Who it's for

uv is a good fit for Python developers who want one manager for project dependencies, environments, Python versions, scripts, and command-line tools, especially when lockfiles, workspaces, offline installation, or private registries matter. Its supported platforms and integration guides also make it relevant to teams incorporating Python into varied development and automation workflows.

It is a less comfortable choice for workflows that depend on exact pip-like command behavior or require credentials to be stored in native system secret storage today. The experimental status of that storage and the current plaintext credential file deserve particular attention in organizations with strict secret-handling requirements.

Pros and cons

  • Broad Python tooling scope: It is designed to replace several separate package, environment, version, and command-line tool workflows.
  • Strong project management: Dependency resolution, lockfiles, workspaces, private registry authentication, and offline installation cover demanding project setups.
  • Wide compatibility: It supports Linux, macOS, and Windows, with CPython, PyPy, Pyodide, and GraalPy.
  • Command differences complicate migration: The pip-compatible interface does not exactly reproduce the commands and behavior of the tools uv draws on.
  • Credential storage is a material caveat: Credentials currently go to a plaintext file; native system storage remains experimental.
  • Some PyPI installs need build tooling: Platforms without a prebuilt wheel require a Rust toolchain to build from source.

Alternatives

Choose pip if you want the free Python package installer and do not need uv's broader project, environment, and Python-version management scope. For projects in another language or ecosystem, npm is the package manager to consider for public package authors, while Cargo is the free Rust package manager and build tool.

Yarn is another free package manager for Linux, macOS, and Windows. Gradle is a free open-source build system, and Apache Maven is a free option for Java builds that require JDK 8 or later for Maven 3.10. For PHP dependency management, consider Composer. Hex is an alternative package registry with an open-source plan for public packages.

Verdict

Choose uv if you want a free, broad Python manager that brings dependencies, environments, Python versions, scripts, and command-line tools together, with lockfiles and workspaces for project management. Look elsewhere if your workflow requires close command-for-command compatibility with existing tools or dependable native secret storage now: the interface differs, and system secret storage is still experimental.

The verdict

uv consolidates many Python packaging and environment tasks in one free tool. Review its command differences and credential-storage behavior before adopting it in workflows that handle private registry credentials.

Questions people ask

How much does uv cost?

The uv plan is free at no cost, and it is open source.

Which operating systems does uv support?

The documented platforms are Linux, macOS, and Windows. Supported Python runtimes include CPython, PyPy, Pyodide, and GraalPy.

Can uv work offline or with private registries?

Yes. uv supports offline installation, private registry authentication, and credentials through environment variables, URLs, netrc, and keyring.

Does uv replace pip?

uv includes a pip-compatible interface and is designed to consolidate pip and several related tools, but its commands are not exact matches.

How are credentials handled?

Credentials are not stored in uv.lock. uv currently stores credentials in a plaintext file, while native system secret storage is an experimental preview.

What the maker publishes

Consolidated tooling
uv is designed to replace tools including pip, pip-tools, pipx, poetry, pyenv, twine, and virtualenv. source
Credentials
Credentials are not stored in uv.lock; uv supports credentials through environment variables, URLs, netrc, and keyring. source
Credential storage
uv currently stores credentials in a plaintext file; native system secret storage is available as an experimental preview feature. source
Dependency confusion protection
By default, uv's first-index strategy limits package candidates to the first index where a package is found, to help prevent dependency confusion attacks. source
Dependency resolution
Yes source
Distribution
uv can be installed with a standalone installer, PyPI, Homebrew, MacPorts, WinGet, Scoop, Docker, GitHub Releases, or Cargo. source
Free plan
Yes source
Install methods
uv offers standalone installers and is also available through PyPI, Homebrew, MacPorts, WinGet, Scoop, Docker, GitHub Releases, and Cargo. source
Integrations
Official integration guides cover Docker, Jupyter, GitHub Actions, GitLab CI/CD, Pre-commit, PyTorch, FastAPI, Bazel, and package registries. source
Integrations
The documentation lists integrations and guides for Docker, Jupyter, marimo, GitHub Actions, GitLab CI/CD, Pre-commit, PyTorch, FastAPI, and several package registries. source
Intended users
Astral says it builds high-performance developer tools for the Python ecosystem to help developers ship software faster. source
Lockfile support
Yes source
Maker
Astral Software Inc. identifies Astral founder Charlie Marsh and describes the team as small and distributed. source
Maker
Astral says its mission is to make the Python ecosystem more productive by building high-performance developer tools, starting with Ruff. source
Offline installation
Yes source
Package formats
source distributions,wheels,requirements.txt,pylock.toml source
Performance
The documentation describes uv as 10–100x faster than pip. source
pip compatibility
uv includes a pip-compatible interface, though its commands do not exactly match the interfaces and behavior of the tools they are based on. source
pip compatibility
uv provides a pip-compatible interface for common pip, pip-tools, and virtualenv commands. source
Platform limits
The documentation lists macOS, Linux, and Windows support; the PyPI installation note says platforms without a prebuilt wheel require a Rust toolchain to build from source. source
Private registry auth
Yes source
Project management
uv manages project dependencies and environments and supports lockfiles and workspaces. source
Projects
uv manages project dependencies and environments and supports lockfiles and workspaces. source
Purpose
uv is an extremely fast Python package and project manager written in Rust. source

Checked against the maker's pages on 30 Sept 2026.