Moralis MCP Server
A TypeScript-based MCP server that implements a wrapper to the Moralis rest API.
## 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](https://github.com/modelcontextprotocol/spec), 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](https://admin.moralis.com) developer portal
2. Sign up and log in
3. Navigate to your [API Keys page](https://admin.moralis.com/api-keys) from the main menu
4. Copy your key and configure it in your config file (see next section), or set it in your environment:
\`\`\`bash
export MORALIS_API_KEY=
\`\`\`
> 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:
\`\`\`json
\{
"mcpServers": \{
"serverName": \{
"command": "npx @moralisweb3/api-mcp-server",
"args": [],
"env": \{
"MORALIS_API_KEY": ""
\}
\}
\}
\}
\`\`\`
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](https://smithery.ai/server/@MoralisWeb3/moralis-mcp-server):
\`\`\`bash
npx -y @smithery/cli install @MoralisWeb3/moralis-mcp-server --client claude
\`\`\`
## 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**:
\`\`\`bash
moralis-api-mcp --transport stdio
\`\`\`
2. **Using the \`web\` transport**:
\`\`\`bash
moralis-api-mcp --transport web
\`\`\`
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**:
\`\`\`bash
moralis-api-mcp --transport streamable-http
\`\`\`
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:
\`\`\`bash
npm install
\`\`\`
Build the server:
\`\`\`bash
npm run build
\`\`\`
For development with auto-rebuild:
\`\`\`bash
npm run watch
\`\`\`
### Debugging
Since MCP servers communicate over stdio, debugging can be challenging. We recommend using the [MCP Inspector](https://github.com/modelcontextprotocol/inspector), which is available as a package script:
\`\`\`bash
npm run inspector
\`\`\`
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:
\`\`\`
- What’s the current price of PEPE and Ethereum?
- What is the current trading sentiment for TOSHI on Base — bullish or bearish?
- Show me the NFTs owned by \`vitalik.eth\` on Base.
- What tokens does wallet \`0xab71...4321\` hold?
- When was wallet 0xabc...123 first and last seen active on Ethereum, Base, and Polygon?
- Show me the complete transaction history for 0xabc...123 across Ethereum, Base, and BNB Chain.
- What is the current net worth in USD of wallet 0xabc...123?
- Find wallet addresses that are likely associated with Coinbase.
- Analyze the current holder distribution of SPX6900 — include whales, small holders, and recent growth trends.
- Show me PEPE’s daily OHLC data for the past 30 days and provide a summary of the trend — is it bullish or bearish?
\`\`\`
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)](https://deep-index.moralis.io/api-docs-2.2/)**
This documentation covers endpoints for:
- Token pricing
- Wallet activity
- NFT metadata and ownership
- Transfers and transactions
- And more
