Deploy marimo
Reactive Python notebook with password login and persistent storage
Just deployed
/data
Deploy and Host marimo on Railway
marimo is an open-source reactive Python notebook. When you run a cell or change a UI element, every dependent cell runs again, so code and outputs stay consistent. Notebooks are stored as plain Python files that work with git, run as scripts, and can be shared as interactive web apps.
About Hosting marimo
Hosting marimo means running its notebook server (marimo edit) on a public URL, protected by a login, with your notebooks kept on persistent storage. This template builds the official marimo 0.25.2 image (the -sql variant with pandas, Polars, Altair and DuckDB), pinned by digest, from a public GitHub repo. A generated MARIMO_PASSWORD protects the editor, because anyone who can open a notebook can run Python on the server. Notebooks, marimo's settings and packages you install from the UI live on a Railway volume at /data, so they survive redeploys. A /healthz check and a restart policy keep the service up.
Common Use Cases
- A private, always-on Python notebook you can open from any browser
- Data exploration with SQL cells over DuckDB, pandas and Polars
- Building small interactive data apps and dashboards to share with your team
Dependencies for marimo Hosting
- One marimo service built from the template repo (no database needed)
- A Railway volume at
/datafor notebooks, settings and installed packages
Deployment Dependencies
- marimo source and docs: https://github.com/marimo-team/marimo and https://docs.marimo.io
- marimo 0.25.2 release: https://github.com/marimo-team/marimo/releases/tag/0.25.2
- Template repo (Dockerfile, start script, changelog): https://github.com/aalfath/marimo-railway-template
Implementation Details
After deploy, open the service's Railway domain and log in with MARIMO_PASSWORD from the marimo service's Variables tab. The file browser opens on /data/notebooks, which starts with marimo's tutorials. Packages installed from marimo's package panel go to /data/.packages, and AI assistant keys you enter in settings are stored in /data/.config.
marimo uses about 150 MB of memory at rest and runs within a 0.5 GB service limit for light work. Each open notebook starts its own Python kernel, so give the service 1 GB or more for larger data sets. Thread pools for NumPy, Polars and Rayon are capped at 2; raise them as variables on bigger plans.
Updates: the service builds from the template repo. When we release a new marimo version, Railway shows you an "update available" notice and you choose when to apply it. Read the repo's CHANGELOG.md first; your data on the volume is kept.
Why Deploy marimo on Railway?
Railway is a singular platform to deploy your infrastructure stack. Railway will host your infrastructure so you don't have to deal with configuration, while allowing you to vertically and horizontally scale it.
By deploying marimo on Railway, you are one step closer to supporting a complete full-stack application with minimal burden. Host your servers, databases, AI agents, and more on Railway.
Template Content

