Build a custom view for a Marimo notebook

Choose the cells and Python values an audience needs, then arrange them in a custom web page. The Marimo notebook keeps calculations, reactive controls, plots, tables, and widgets. Studio keeps each view in HTML and CSS beside the notebook.

uvx marimo-studio analysis.py

The command opens the Marimo editor and a live view preview in one workspace. Saved view changes refresh around the running Python session.

Keep Python and presentation together

Notebook	Studio view
Data loading and calculations	Page structure and navigation
Reactive dependencies	Audience-specific wording
Controls, plots, tables, and widgets	Layout, branding, and responsive design
Presentation values	Which results appear

Place a complete cell output with:

<marimo-cell name="forecast_chart"></marimo-cell>

Place a JSON-compatible Python value in page text with:

<time mo-value="report.updated_at"></time>

Both remain connected to the notebook kernel. While editing, a control change in the notebook or an attached preview updates the other open views and reruns the affected cells.

Build for each audience

One notebook can serve several views:

uvx marimo-studio view add executive analysis.py
uvx marimo-studio view add operations analysis.py

Views share notebook cells and aliases while keeping separate HTML, CSS, and static files.

Continue

• Create your first view from an existing notebook.
• Design a view with cells, values, loading space, and on-demand regions.
• Share a view through Marimo's server.
• Look up commands and settings in CLI and configuration.
• Use the Python API for notebook inspection or ASGI composition.