(Re-)introducing marimo Learn

marimo was born out of research, purpose-built to accelerate the path from exploration to insight. Education is core to that mission: reactive notebooks let educators transform static documents into living artifacts that learners explore, not just read.

Over the past couple of years, members of the marimo community have built lessons on topics ranging from basic probability to optimization and the Polars dataframe library. Today, we are relaunching their contributions, and some new additions, at https://marimo.io/learn. Over eighty notebooks are organized into eight courses: Altair, DuckDB, Optimization, Polars, Probability, Python, Queueing Theory, and SQL.

These notebooks are all freely licensed, and can all be downloaded for local use or run in molab without needing any installation. Unlike static tutorials, each notebook is fully reactive: learners can change inputs, tweak parameters, and see results update without writing a single line of code, or (if they’d rather) edit the Python and SQL embedded in these lessons to explore ideas of their own.

Along with the notebooks, we are also launching a new guide for educators that walks through the mechanics of using marimo notebooks in the classroom, and provides a set of pedagogical patterns for using them effectively. Some of these patterns are drawn from instructors’ experience with first-generation notebooks like Jupyter; others are based on our own experience, both in traditional classrooms and with AI-assisted learning.

We plan to add more material to our site soon. As the next post in this series discusses, we are also looking at building new notebook-native teaching tools. Feedback and contributions are very welcome: you can file issues in our GitHub repository, reach us by email at contact@marimo.io, or join our online community.