Getting Started

Installing marimo gets you the marimo command-line interface (CLI), the entry point to all things marimo.

Installation

Before installing marimo, we recommend creating and activating a Python virtual environment.

Setting up a virtual environment

Python uses virtual environments to minimize conflicts among packages. Here’s a quickstart for pip users. If you use conda, please use a conda environment instead.

Run the following in the terminal:

  • create an environment with python -m venv marimo-env

  • activate the environment:

    • macOS/Unix: source marimo-env/bin/activate

    • Windows: marimo-env\Scripts\activate

Make sure the environment is activated before installing marimo and when using marimo. Install other packages you may need, such as numpy, pandas, matplotlib, and altair, in this environment. When you’re done, deactivate the environment with deactivate in the terimnal.

Learn more from the official Python tutorial.

Using uv?

uv is a next-generation Python package installer and manager that is 10-100x faster than pip, and also makes it easy to install Python and manage projects. With uv, creating a virtual environment is as easy as uv venv.

To install marimo, run the following in a terminal:

pip install marimo

uv pip install marimo

conda install -c conda-forge marimo

To check if the install worked, run

marimo tutorial intro

A tutorial notebook should open in your browser.

Installation issues?

Having installation issues? Reach out to us at GitHub or on Discord.

Tutorials

marimo tutorial intro opens the intro tutorial. List all tutorials with

marimo tutorial --help

Edit notebooks

Create and edit notebooks with marimo edit.

  • launch the notebook server to create new notebooks, and start or stop existing ones:

marimo edit

  • create or edit a single notebook with

marimo edit your_notebook.py

(If your_notebook.py doesn’t exist, marimo will create a blank notebook named your_notebook.py.)

Deploy as apps

Use marimo run to serve your notebook as an app, with Python code hidden and uneditable.

marimo run your_notebook.py

Convert from Jupyter to marimo

Automatically convert Jupyter notebooks to marimo notebooks with marimo convert:

marimo convert your_notebook.ipynb -o your_notebook.py

Then open the notebook with marimo edit your_notebook.py

Disable autorun on startup

marimo automatically runs notebooks when they are opened. If this is a problem for you (not all Jupyter notebooks are designed to be run on startup), you can disable autorun on startup via user configuration.

  1. Type marimo config show to get the location of your config file.

  2. If no config file exists, create it at ~/.marimo.toml or $XDG_CONFIG_HOME/marimo/marimo.toml.

  3. Update your config to include the following:

[runtime]
auto_instantiate = false

Enable more features with optional dependencies

Some features require additional dependencies, which are not installed by default. This includes:

To install the optional dependencies, run:

pip install "marimo[recommended]"

uv pip install "marimo[recommended]"

conda install -c conda-forge marimo duckdb altair polars openai ruff

This will install: duckdb, altair, polars, openai, and ruff.

GitHub Copilot and AI Assistant

The marimo editor natively supports GitHub Copilot, an AI pair programmer, similar to VS Code.

Get started with Copilot:

  1. Install Node.js.

  2. Enable Copilot via the settings menu in the marimo editor.

Note: Copilot is not yet available in our conda distribution; please install marimo from PyPI if you need Copilot.

marimo also comes with support for other copilots, and a built-in AI assistant that helps you write code.

VS Code extension

If you prefer VS Code over terminal, try our VS Code extension. Use this extension to edit and run notebooks directly from VS Code, and to list all marimo notebooks in your current directory.