home / mcp / chainfetch mcp server
AI Powered Ethereum Intelligence API
Configuration
View docs{
"mcpServers": {
"chainfetch-chainfetch-mcp-server": {
"url": "https://your-ngrok-url.ngrok.app/streamable-http",
"headers": {
"CHAINFETCH_API_TOKEN": "your_api_token_here"
}
}
}
}You can run a ChainFETCH MCP Server to access AI-powered Ethereum blockchain intelligence via a local or remote MCP endpoint. This server connects to ChainFETCH APIs for semantic search, LLM-assisted queries, and multi‑modal data across addresses, transactions, blocks, tokens, and smart contracts, enabling fast exploratory analysis and detailed insights through your MCP client.
Connect to the MCP server from your client and choose your preferred access mode. Use the HTTP endpoint for a remote, streamable feed or run the server locally as a stdio process and interact via standard input/output.
Prerequisites: ensure you have Node.js 18.0.0 or higher installed on your system, and obtain a ChainFETCH API token from chainfetch.app.
# Option 1: Install the MCP server globally (recommended)
npm install -g chainfetch-mcp-serverOption 2: Install from source (for development or customization) shows how to clone, install, build, and link the package.
git clone https://github.com/chainfetch/chainfetch-mcp-server.git
cd chainfetch-mcp-server
npm install
npm run build
npm linkSet your ChainFETCH API token in the environment where the MCP server runs. You can export it in your shell before starting the server.
export CHAINFETCH_API_TOKEN="your_api_token_here"You can run the server either as a remote HTTP endpoint or as a local stdio process. The following configurations show explicit, usable examples.
{
"mcpServers": {
"chainfetch_http": {
"type": "http",
"name": "chainfetch_http",
"url": "https://your-ngrok-url.ngrok.app/streamable-http",
"args": []
},
"chainfetch_stdio": {
"type": "stdio",
"name": "chainfetch",
"command": "npx",
"args": ["chainfetch-mcp-server"],
"env": [
{"name": "CHAINFETCH_API_TOKEN", "value": "CHAINFETCH_API_TOKEN"}
]
}
}
}Run the MCP server in HTTP mode and expose it via ngrok to enable remote clients like Claude to connect.
MCP_HTTP_MODE=true chainfetch-mcp-serverngrok http 8000Semantic search examples and structured queries let you explore addresses, transactions, blocks, tokens, and contracts using AI-powered or JSON-filtered approaches.
AI-powered vector similarity search for addresses based on semantic similarity to a query.
JSON parameter-based address search with 150+ filters.
Natural language to smart parameter selection for address queries.
AI-generated summary for an address.
Detailed address information.
Semantic search over transactions using AI understanding.
JSON parameter-based transaction search with 254+ filters.
LLM-powered transaction queries.
AI-generated summary for a transaction.
Detailed transaction data.
AI-powered search over blocks.
JSON parameter-based block search with 120+ filters.
LLM-powered block queries.
AI-generated block summaries.
Detailed block information.
Semantic search for tokens.
JSON-based token filtering with extensive parameters.
LLM-powered token queries.
AI-generated token summaries.
Detailed token information.
NFT instance details.
AI-driven smart contract search.
JSON-based contract search with 50+ parameters.
LLM-powered smart contract queries.
AI-generated smart contract summaries.
Detailed contract information.