Moralis MCP Server

🧠 Overview

The Moralis MCP Server is a local or cloud-deployable engine that connects natural language prompts to real blockchain insights — allowing AI models to query wallet activity, token metrics, dapp usage, and more without custom code or SQL.

Built on top of the Model Context Protocol, this server makes it easy for LLMs to talk to Moralis APIs in a consistent, explainable, and extensible way.

• 🔗 Fully pluggable: swap LLMs, customize retrieval logic, or extend with your own tools
• 🧱 Works with OpenAI, Claude, and open-source models
• 🧠 Powers agents, devtools, bots, dashboards, and beyond

⚙️ Common Use Cases

• 🤖 AI agents & assistants: “What’s this wallet’s trading history?”
• 📈 Devtools: on-chain QA, testing, CLI integrations
• 📊 Dashboards: natural language to charts/data
• 📉 Monitoring: alerting & summarization for tokens/dapps
• 🧠 Trading bots: LLM-driven strategies with real blockchain grounding

🔐 Getting an API Key

To use this MCP server with Moralis APIs, you'll need an API key:

1. Go to Moralis developer portal
2. Sign up and log in
3. Navigate to your API Keys page from the main menu
4. Copy your key and configure it in your config file (see next section), or set it in your environment:

⚠️ Note: Some features and endpoints require a Moralis paid plan. For full access and production-grade performance, we recommend signing up for a paid tier.

🚀 Usage with a Client

To connect the MCP server to a compatible client (e.g. Claude Desktop, OpenAI-compatible agents, VS Code extensions, etc.), configure the client to launch the server as a subprocess.

Most clients support a simple config file - for example, you might create a file like mcp.json in the client’s configuration directory with the following:

This setup can be adapted for any client that supports MCP servers. Replace the example values with those specific to your use case.

Installing via Smithery

To install Moralis API Server for Claude Desktop automatically via Smithery:

🖥️ Using as a Server

The server accepts an optional --transport argument to specify the transport type. The available transport types are:

• stdio: Communicates over standard input/output (default).
• web: Starts a HTTP server for communication.
• streamable-http: Starts an HTTP server with streamable endpoints.

Examples

1. Using the default stdio transport:

2. Using the web transport:

This will start a HTTP server. You can send requests to the server using tools like curl or Postman.

3. Using the streamable-http transport:

This will start an HTTP server. You can send requests to the server using tools like curl or Postman.

Notes

• Ensure that the required environment variables (e.g., MORALIS_API_KEY) are set before starting the server.
• For custom configurations, you can pass additional arguments or environment variables as needed.
• Refer to the documentation for more details on each transport type.

🛠 Development

Install dependencies:

Build the server:

For development with auto-rebuild:

🐞 Debugging

Since MCP servers communicate over stdio, debugging can be challenging. We recommend using the MCP Inspector, which is available as a package script:

The Inspector will provide a URL to access debugging tools in your browser.

💬 Example Prompts

Here are some example prompts you can use with your AI agent through the MCP server:

These prompts are parsed and mapped to structured Moralis API calls using the MCP method registry.

💡 You can also build custom prompts based on any supported method.

📚 API Reference

The Moralis MCP Server wraps and translates prompts into Moralis REST API calls. You can explore the underlying API surface here:

🔗 Moralis Swagger Docs (v2.2)

This documentation covers endpoints for:

• Token pricing
• Wallet activity
• NFT metadata and ownership
• Transfers and transactions
• And more