Installation

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

??? note "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](https://conda.io/projects/conda/en/latest/user-guide/tasks/manage-environments.html#creating-an-environment-with-commands)
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 terminal.

Learn more from the [official Python tutorial](https://docs.python.org/3/tutorial/venv.html#creating-virtual-environments).

/// admonition | Using uv? type: tip

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. Create a uv project with uv init; this creates and manages a virtual environment for you behind-the-scenes. For detailed information on using marimo with uv, see our uv guide. ///

/// admonition | Prefer VS Code/Cursor? type: tip

Try our extension, which works in VS Code, Cursor, and other VS Code forks. ///

Install with minimal dependencies

To install marimo, run the following in a terminal:

/// tab | install with pip

bash

pip install marimo

To check if the install worked, run

bash

marimo tutorial intro

///

/// tab | install with uv

bash

uv add marimo

To check if the install worked, run

bash

uv run marimo tutorial intro

///

/// tab | install with conda

bash

conda install -c conda-forge marimo

To check if the install worked, run

bash

marimo tutorial intro

///

A tutorial notebook should open in your browser.

/// admonition | Installation issues? type: note

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

Install with recommended dependencies

marimo is lightweight, with few dependencies, to maximize compatibility with your own environments.

To unlock additional features in the marimo editor, including SQL cells, AI completion, server-side plotting of dataframe columns, and more, we suggest installing marimo[recommended]:

/// tab | install with pip

bash

pip install "marimo[recommended]"

///

/// tab | install with uv

bash

uv add "marimo[recommended]"

///

/// tab | install with conda

bash

conda install -c conda-forge marimo "duckdb>=1.0.0" "altair>=5.4.0" pyarrow "polars>=1.9.0" "sqlglot[c]>=23.4" "openai>=1.55.3" "ruff" "nbformat>=5.7.0" "vegafusion>=2.0.0" "vl-convert-python>=1.0.0"

///

Installing marimo in this way installs the following additional dependencies and unlocks the following features:

Dependency	Feature
duckdb>=1.0.0	SQL cells
altair>=5.4.0	Plotting in datasource viewer
polars[pyarrow]>=1.9.0	SQL output back in Python
sqlglot[c]>=23.4	SQL cells parsing
openai>=1.55.3	AI features
ruff	Formatting
nbformat>=5.7.0	Export as IPYNB
vegafusion>=2.0.0	Performant charting
vl-convert-python>=1.0.0	Required by vegafusion