Overview

warning

🚨 BREAKING CHANGE Since version 0.6.0, the configuration has changed. Read more in the Release Notes.

Jupyter MCP Server is a Model Context Protocol (MCP) server implementation that enables real-time interaction with 📓 Jupyter Notebooks, allowing AI to edit, document and execute code for data analysis, visualization etc.

Compatible with any Jupyter deployment (local, JupyterHub, ...) and with Datalayer hosted Notebooks.

Key features include:

• ⚡ Real-time control: Instantly view notebook changes as they happen.
• 🔁 Smart execution: Automatically adjusts when a cell run fails thanks to cell output feedback.
• 🤝 MCP-Compatible: Works with any MCP client, such as Claude Desktop, Cursor, Cline, Windsurf and more.

To use Jupyter MCP Server:

1. Run a Jupyter Notebook: You can use the Datalayer hosted Notebooks or use the OSS JupyterLab or Notebook>6. Refer to Datalayer section or Jupyter section to setup your Notebook.

2. Configure your MCP client: Then you need to configure your MCP client:

   1. With stdio, your client will launch the MCP server for you, so you the default configuration is often enough. The server will be pulled from Docker Hub.
   2. With streamable-http, you need to run yourself a Jupyter MCP Server (on your machine or in the cloud), so you'd better look at the fine-grained configuration details of your server.

Looking for blog posts, videos or other resources related to Jupyter MCP Server?

👉 Check out the Resources section.

🧰 Dive into the Tools section to understand the tools powering the server.