LAUNCHMongoDB 8.3 is built for the sub-100ms retrieval & zero downtime AI demands. Read blog >

AI DATAStop fighting your data layer. Get the memory & retrieval agents need to scale. Read blog >

MCP Server

TOOLS

MongoDB MCP Server

Bring database context into your agentic experiences with the official MongoDB MCP (Model Context Protocol) Server. Connect MongoDB deployments to your favorite AI tools and agents.

Read the DocsMongoDB MCP Server on GitHub

Illustration of a database producing charts.

Database context for better code

Your database holds the core data for your applications. Retrieving database configuration, collection schemas, and sample data enables coding agents to generate accurate, working code and optimized queries.

Illustration of hands typing on a laptop and a superimposed desktop window with some code on it.

Use with your favorite AI tools

The MongoDB MCP Server integrates with AI-powered IDEs and coding agents, providing the context developers and agents need to build with MongoDB.

Simplify development workflows

No more copy-pasting schemas or switching tools. Explore and manage MongoDB deployments without leaving your development environment.

Power your own agents

The MongoDB MCP Server lets you build agents that explore and query data autonomously. It acts as a plugin for any MCP-enabled agent.

What can you do with the MongoDB MCP Server?

View documentation

Bring valuable context from MongoDB databases into development environments for data exploration, database operations, and context-aware code generation.

View documentation

Connecting the MongoDB MCP Server to your AI development tools lets you use natural language to explore and manage databases while generating accurate application code.

Explore and query data

Manage databases

Generate code

Code Snippet

> Explain the schema of the "products" collection.

I'll help you explain the schema of the "products" collection and provide a summary of the data. Let me interact with your MongoDB database through the MCP server to get this information.
  Ran `collection-schema` - MongoDB (MCP Server)
Now, let me fetch some sample data from the collection to provide you with a summary:
  Ran `find` - MongoDB (MCP Server)
Let's also get a count of the total number of documents in the collection:
  Ran `count` - MongoDB (MCP Server)
Let me get some aggregated statistics about the data:
  Ran `aggregate` - MongoDB (MCP Server)
  Ran `aggregate` - MongoDB (MCP Server)
Schema and Data Summary for the "products" Collection
Schema Overview
The "products" collection has the following schema:
1. _id: ObjectId - MongoDB's default unique identifier for each document
2. name: String - The name of the product
3. price: Number - The price of the product in USD
4. category: String - The category the product belongs to
5. quantity: Number - The available quantity of the product
6. inStock: Boolean - Flag indicating whether the product is in stock or not

This collection appears to be a simple product catalog for a store with various categories of merchandise, including pricing and inventory information

The MCP Server connects to any MongoDB deployment, including Atlas, Community Edition, and Enterprise Advanced. A connection string grants database access, while Atlas service account credentials unlock Atlas-specific tools for cluster management.

Learn more about deployment options

Command line

Docker

Configuration file

Code Snippet

AI clients use different configuration files and syntax to set up MCP Servers. Use these templates to get started.

Learn more

VSCode (GitHub Copilot)

Cursor

Windsurf

Claude Code

OpenAI Codex CLI

Code Snippet

Explore and query data

Manage databases

Generate code

Code Snippet

> Explain the schema of the "products" collection.

This collection appears to be a simple product catalog for a store with various categories of merchandise, including pricing and inventory information

MongoDB MCP Server use cases

View all use cases

illustration of hands holding binoculars

DEVELOPMENT

Interact with your data in natural language

Explore and manage databases using natural language directly in your development environment to build faster and more easily.

DEVELOPMENT

Ship software faster

The MongoDB MCP Server enables coding agents to query and manage data, helping them iterate faster and build application features more reliably.

AGENTS

Power agentic systems

Integrate the MongoDB MCP Server with your agentic systems so they can explore and query MongoDB deployments effectively.

Learning hub

Quick and easy resources to start building with the MongoDB MCP Server.

Use the MongoDB MCP Server

Follow a step-by-step guide to get started with the MongoDB MCP Server.

Access the docs

Learn about best practices

Learn more about configuring and best practices when deploying the MongoDB MCP Server.

Access GitHub repository

MongoDB Developer Community

Ask questions, share ideas, get support, and connect with other developers through our global community of MongoDB users.

Join now

FAQ

An MCP server implements the Model Context Protocol (MCP), introduced by Anthropic in November 2024. MCP is a standard that defines how LLM-based applications connect with diverse data sources and systems, especially for agentic use cases. It enables interoperability and modularity in complex AI ecosystems, making deployment easier, more flexible, and scalable.

The MongoDB MCP Server connects any MongoDB deployment (Atlas, Community Edition, or Enterprise Advanced) to AI-powered developer tools like GitHub Copilot, Windsurf, Claude, or any MCP-enabled agentic system. It brings valuable context from MongoDB databases into developers’ environments, enabling data exploration, database operations, and context-aware code generation via natural language.

Additionally, the MongoDB MCP Server can serve as a building block for AI agents, giving them direct access to MongoDB databases for data retrieval and automated database operations.

Yes. The MongoDB MCP Server works with any MongoDB deployment, including Community Edition, Enterprise Advanced, and Atlas.

You can deploy the MongoDB MCP Server locally using stdio (for IDEs/agents like Windsurf or Cursor) or HTTP. It can also be self-hosted for broader access or packaged and deployed with Docker for easier containerized setups.

Learn more

Get started with the MongoDB MCP Server

Supercharge your coding agents with database context.

Read the DocsView on GitHub

WHY USE MCP SERVER?

Database context for better code
Works with your favorite AI tools

Simplified development workflows

MongoDB MCP Server

Database context for better code

Use with your favorite AI tools

Simplify development workflows

Power your own agents

What can you do with the MongoDB MCP Server?

MongoDB MCP Server use cases

Interact with your data in natural language

Ship software faster

Power agentic systems

Learning hub

Use the MongoDB MCP Server

Learn about best practices

MongoDB Developer Community

FAQ

What is an MCP Server?

How is the MongoDB MCP Server used?

Can I use the MongoDB MCP Server with MongoDB Community Edition?

How can I deploy the MongoDB MCP Server?

Get started with the MongoDB MCP Server