I’ve been trying out the new uv from Astral for managing my Python project. It’s fast, simple, and honestly makes the usual pip + venv setup feel kinda clunky. So, I figured I’d write this down since I’m switching my Python project over to uv.

Why even switch?

Python’s tooling has always been a bit… scattered. You install stuff with pip, manage environments with venv, freeze those with pip-tools, or lock them using poetry. Each tool solves part of the problem, but juggling all of them gets tiresome fast.

For most of my Python projects, I just want to:

  • Create a virtual environment
  • Install the dependencies
  • Keep dependencies consistent
  • Move on

Python may be a slow language in terms of execution speed, but it’s fast to write and great for turning ideas into actionable projects. Still, you’ve got to go through those steps for each new one.

What is uv

It’s a Python package manager built by the team behind Ruff, that basically combines pip for installing packages, venv for managing environments, pip-tools for generating lock files, and pipx for installing and running Python tools. It’s pretty convenient for my use case, and it’s fast (well, it’s written in Rust).

Follow the official docs for installation instructions and more information.

How I use it

The basic flow I follow for a new project is:

# 1. Creating a new project
uv init new-project
cd new-project

# 2. Installing dependencies
uv add requests pandas

# 3. Running Python scripts
uv run main.py

If I’m cloning an existing project that already uses uv, the steps are just as simple:

# 1. Cloning a project
git clone https://github.com/odhyp/sipd-ri.git
cd sipd-ri

# 2. Installing dependencies
uv sync

# 3. Running Python scripts
uv run main.py

It’s that simple now. I don’t need to think about whether I’ve activated a virtual environment, I just start building (or using) the project.

Back to Writings