Jupyter MCP Server
Enables interaction with Jupyter notebooks through the Model Context Protocol, supporting code execution and markdown insertion within JupyterLab environments.
What is Jupyter MCP Server?
Jupyter MCP Server is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to enables interaction with jupyter notebooks through the model context protocol, supporting code execution and markdown insertion within jupyterlab environments.
Enables interaction with Jupyter notebooks through the Model Context Protocol, supporting code execution and markdown insertion within JupyterLab environments.
This server falls under the Developer Tools category on MCPgee, the world's largest MCP server directory with 33,000+ servers.
Features
- Enables interaction with Jupyter notebooks through the Model
Use Cases
Maintainer
Works with
Installation
Manual Installation
npx jupyter-mcp-serverConfiguration
Configuration Details
claude_desktop_config.json
Performance
Response Metrics
Resource Usage
How to Set Up and Use Jupyter MCP Server
Jupyter MCP Server bridges JupyterLab and MCP-compatible AI clients, enabling LLMs like Claude to read, write, and execute cells in live Jupyter notebooks through natural language. It exposes a comprehensive set of tools covering file listing, kernel management, cell insertion, cell execution, source editing, and notebook-level run-all operations, all backed by real-time collaboration via jupyter-collaboration. Data scientists and researchers use it to let AI assistants write and run analysis code directly inside their notebooks without copy-pasting between interfaces.
Prerequisites
- Python 3.8 or later with pip
- JupyterLab 4.4.1 and jupyter-collaboration 4.0.2 installed
- The uv package manager installed for uvx-based deployment
- A running JupyterLab instance with a known token
- An MCP-compatible client such as Claude Desktop, Cursor, or Windsurf
Install JupyterLab and collaboration dependencies
Install the specific versions of JupyterLab, the collaboration extension, and the MCP tools package required by the server.
pip install jupyterlab==4.4.1 jupyter-collaboration==4.0.2 jupyter-mcp-tools>=0.1.4 ipykernel pycrdtStart JupyterLab with a known token
Launch JupyterLab on a fixed port with a specific token so the MCP server can authenticate. Keep this terminal running while using the MCP server.
jupyter lab --port 8888 --IdentityProvider.token MY_TOKEN --ip 0.0.0.0Install and run the MCP server via uvx
In a separate terminal, install uv and launch the Jupyter MCP server using uvx, pointing it at your running JupyterLab instance.
pip install uv
uvx jupyter-mcp-server@latestConfigure your MCP client
Add the Jupyter MCP server to your Claude Desktop or compatible client configuration, passing the JupyterLab URL and token as environment variables.
{
"mcpServers": {
"jupyter": {
"command": "uvx",
"args": ["jupyter-mcp-server@latest"],
"env": {
"JUPYTER_URL": "http://localhost:8888",
"JUPYTER_TOKEN": "MY_TOKEN"
}
}
}
}Verify notebook access
In your MCP client, ask the AI to list available notebooks or kernels. This confirms the server has successfully authenticated with JupyterLab and can see your workspace.
Jupyter MCP Server Examples
Client configuration
Claude Desktop configuration for Jupyter MCP Server with authentication and image output enabled.
{
"mcpServers": {
"jupyter": {
"command": "uvx",
"args": ["jupyter-mcp-server@latest"],
"env": {
"JUPYTER_URL": "http://localhost:8888",
"JUPYTER_TOKEN": "MY_TOKEN",
"ALLOW_IMG_OUTPUT": "true",
"MCP_TOKEN": "your_mcp_token"
}
}
}
}Prompts to try
Example prompts that demonstrate cell execution, notebook editing, and data analysis workflows.
- "List all open notebooks and their active kernels"
- "Insert a Python cell at the end of my notebook that imports pandas and reads data.csv"
- "Execute the cell I just added and show me the output"
- "Read the source of cell 3 in my analysis notebook"
- "Run all cells in the notebook from top to bottom and summarize any errors"
- "Add a markdown cell explaining what the previous code block does"Troubleshooting Jupyter MCP Server
Connection refused or authentication errors when the MCP server tries to reach JupyterLab
Confirm JupyterLab is running on the port specified in JUPYTER_URL and that the JUPYTER_TOKEN exactly matches the --IdentityProvider.token value used when launching JupyterLab. If running in Docker, ensure the container network allows the MCP server to reach localhost:8888.
Cell execution hangs or times out
Check that the notebook has an active kernel ('list_kernels' tool or JupyterLab UI). If the kernel is idle or dead, restart it via the JupyterLab interface or use the 'restart_notebook' tool before retrying execution.
Image outputs from matplotlib or other libraries are not returned
Set the environment variable ALLOW_IMG_OUTPUT=true in your MCP client configuration. Without this flag the server filters out binary image outputs and only returns text. Restart the MCP server after adding the variable.
Frequently Asked Questions about Jupyter MCP Server
What is Jupyter MCP Server?
Jupyter MCP Server is a Model Context Protocol (MCP) server that enables interaction with jupyter notebooks through the model context protocol, supporting code execution and markdown insertion within jupyterlab environments. It connects AI assistants to external tools and data sources through a standardized interface.
How do I install Jupyter MCP Server?
Follow the installation instructions on the Jupyter MCP Server GitHub repository. Clone the repo, install dependencies, and add the server config to your AI client.
Which AI clients work with Jupyter MCP Server?
Jupyter MCP Server works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.
Is Jupyter MCP Server free to use?
Yes, Jupyter MCP Server is open source and available under the BSD 3-Clause license. You can use it freely in both personal and commercial projects.
Jupyter MCP Server Alternatives — Similar Developer Tools Servers
Looking for alternatives to Jupyter MCP Server? Here are other popular developer tools servers you can use with Claude, Cursor, and VS Code.
Ecc
★ 188.2kThe agent harness performance optimization system. Skills, instincts, memory, security, and research-first development for Claude Code, Codex, Opencode, Cursor and beyond.
Javaguide
★ 155.8kJava 面试 & 后端通用面试指南，覆盖计算机基础、数据库、分布式、高并发、系统设计与 AI 应用开发
Gemini CLI
★ 104.5kA secure MCP server that wraps the Google Gemini CLI, allowing clients to query Gemini models using local OAuth sessions without requiring an API key. It provides tools for model interaction and diagnostics with built-in protection against command in
Awesome MCP Servers
★ 87.3k⭐ Curated list of Model Context Protocol (MCP) servers - tools that extend Claude Desktop, Cursor, Windsurf, and other MCP clients with custom capabilities.
MCP Servers
★ 86.0kModel Context Protocol Servers
CC Switch
★ 77.5kA cross-platform desktop All-in-One assistant for Claude Code, Codex, OpenCode, OpenClaw, Gemini CLI & Hermes Agent. Only official website: ccswitch.io
Browse More Developer Tools MCP Servers
Explore all developer tools servers available in the MCPgee directory. Each server includes setup guides for Claude, Cursor, and VS Code.
Set Up Jupyter MCP Server in Your Editor
Choose your AI client for step-by-step setup instructions.
Quick Config Preview
Add this to your claude_desktop_config.json or .cursor/mcp.json
Ready to use Jupyter MCP Server?
Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.