Documentation
Start typing to search…

Birdeye AI - MCP

🚧 Welcome to the Birdeye MCP Server

Birdeye MCP Server is currently in Beta. We're constantly improving, and your feedback is crucial. Please share any thoughts or suggestions.

📕 Overview

The official Birdeye MCP Server is now live, making Birdeye data readily available to your AI models and applications. With the Birdeye MCP, you can empower your agents to:

  • Access real-time market data: Get aggregated prices, market cap, and trading volume for over 15k+ coins on Birdeye.
  • Dive into onchain analytics: Query onchain DEX price and liquidity data for more than 8M tokens across 10+ networks.
  • Discover market trends: Instantly find trending coins, new token listings, top gainers/losers, and popular NFT collections.
  • Retrieve rich metadata: Pull essential details like project descriptions, logos, social links, contract addresses, security info, and more.
  • Analyze historical performance: Access historical price, market data, and OHLCV for any cryptocurrency.
  • Explore crypto categories: Effortlessly list coins within specific sectors like Meme, DeFi, Layer 1, AI agent, and more.

MCP Demo with Claude Desktop

🔥 Getting Started

Connecting your AI to Birdeye is simple. We offer several MCP server options to fit your needs, from keyless access for testing to authenticated connections for production applications.

Most MCP-compatible clients, like Claude Desktop, Gemini CLI, and Cursor, can be configured using a simple JSON file (e.g., claude_desktop_config.json)

📘 Prerequisites

  • Make sure your device has node installed. You can download it from nodejs.org/download
  • If you are using Claude Desktop, node version 18 or higher is required.
  • Get your API key (see "How to get Birdeye API Key" below)

How to get Birdeye API Key

  1. Login to bds.birdeye.so
Birdeye Website
  1. Go to Usages -> Security
Metrics Security
  1. Click Generate Authentication key -> Copy api key
API Key

Which MCP Server Should You Use

Here's a breakdown of the available options to help you choose the right one:

MCP Server Type

Best For

Setup Details

Remote Server
(Authenticated)

  • Scalable apps, AI agent integrations
    - Unlocks all tools available under your Demo/Pro plan
    - Higher, reliable rate limits. Get your API key here

https://mcp.birdeye.so/mcp

Remote Server (Public, Keyless)

The easiest way to get started. Just add the following to your client's mcp_config.json file.

📘 Client-Specific Config

The file name and location depend on your client. Find your config file here: modelcontextprotocol.io/quickstart

Add the following configuration to your birdeye_mcp_config.json:

{
    "mcpServers": {
        "birdeye-api-mcp": {
            "command": "npx",
            "args": [
               "-y",
                "[email protected]",
                "https://mcp.birdeye.so/mcp",
                "--header",
                "x-api-key:${API_KEY}"
            ],
            "env": {
                "API_KEY": <YOUR_API_KEY>
            }
        }
    }
}

Here's a quick 2-minute tutorial for setting up the public server with Claude Desktop:

How to Connect Birdeye MCP to Claude Desktop: Step-by-Step Guide - YouTube

How to Connect Birdeye MCP to Claude Desktop: Step-by-Step Guide

Remote Server (Authenticated)

Step 1: Add the configuration

Add the following configuration to your mcp_config.json:

JSON


{
    "mcpServers": {
        "birdeye-api-mcp": {
            "command": "npx",
            "args": [
                "-y",
                "[email protected]",
                "https://mcp.birdeye.so/mcp",
                "--header",
                "x-api-key:${API_KEY}"
            ],
            "env": {
                "API_KEY": <YOUR_API_KEY>
            }
        }
    }
}

Step 2: Authorize your MCP access

  • Simply paste in your Birdeye API key, and authorize to link your key to the MCP session.

✨ Don't have an API key yet? Upgrade to Pro today! Read more here.

  • You can also toggle between dynamic/static tools here. Learn more about

🚀 Connecting with Claude

Connecting Birdeye MCP to Claude is straightforward. The method varies slightly depending on your Claude plan.

For Claude Free Users (via Claude Desktop)

You must use the Claude Desktop app and modify the configuration file.

  1. Locate claude_desktop_config.json: Follow the instructions here to find the file on your system.
  2. Add a server config: Copy and paste one of the server configs above that matches your use case.
  3. Restart Claude Desktop: Close and reopen the app for the changes to take effect.

For Claude Pro Users

👍 Tips

You can also follow the same steps as the Free users by modifying the claude_desktop_config.json file.

  1. In Claude ( claude.ai or the Desktop app), click on 'Add connectors' in your chat.

  2. Click on 'Add custom connector'

  3. Click on 'Add', and you're ready to go!

🔌 Connecting with Other AI Tools

Birdeye MCP works with all major MCP-compatible AI tools. Below are setup guides for the most popular ones.

1. Claude Code (CLI)

Add the MCP server via the CLI command:

claude mcp add birdeye-mcp -- npx -y [email protected] https://mcp.birdeye.so/mcp --header "x-api-key:<YOUR_API_KEY>"

Or add it directly to your project's .mcp.json file:

{
    "mcpServers": {
        "birdeye-mcp": {
            "command": "npx",
            "args": [
                "-y",
                "[email protected]",
                "https://mcp.birdeye.so/mcp",
                "--header",
                "x-api-key:${API_KEY}"
            ],
            "env": {
                "API_KEY": "<YOUR_API_KEY>"
            }
        }
    }
}

2. Cursor

Add the following to .cursor/mcp.json (project-level) or ~/.cursor/mcp.json (global):

{
    "mcpServers": {
        "birdeye-mcp": {
            "command": "npx",
            "args": [
                "-y",
                "[email protected]",
                "https://mcp.birdeye.so/mcp",
                "--header",
                "x-api-key:${API_KEY}"
            ],
            "env": {
                "API_KEY": "<YOUR_API_KEY>"
            }
        }
    }
}

3. Windsurf

Add the following to ~/.codeium/windsurf/mcp_config.json:

{
    "mcpServers": {
        "birdeye-mcp": {
            "command": "npx",
            "args": [
                "-y",
                "[email protected]",
                "https://mcp.birdeye.so/mcp",
                "--header",
                "x-api-key:${API_KEY}"
            ],
            "env": {
                "API_KEY": "<YOUR_API_KEY>"
            }
        }
    }
}

4. GitHub Copilot (VS Code)

Add the following to .vscode/mcp.json in your project:

{
    "inputs": [
        {
            "type": "promptString",
            "id": "birdeye-api-key",
            "description": "Enter your Birdeye API Key",
            "password": true
        }
    ],
    "servers": {
        "birdeye-mcp": {
            "type": "stdio",
            "command": "npx",
            "args": [
                "-y",
                "[email protected]",
                "https://mcp.birdeye.so/mcp",
                "--header",
                "x-api-key:${input:birdeye-api-key}"
            ]
        }
    }
}

5. Gemini CLI

Add the following to your ~/.gemini/settings.json:

{
    "mcpServers": {
        "birdeye-mcp": {
            "command": "npx",
            "args": [
                "-y",
                "[email protected]",
                "https://mcp.birdeye.so/mcp",
                "--header",
                "x-api-key:${BIRDEYE_API_KEY}"
            ],
            "env": {
                "BIRDEYE_API_KEY": "<YOUR_API_KEY>"
            }
        }
    }
}

6. OpenAI Codex CLI

Add the following to ~/.codex/config.toml:

[mcp_servers.birdeye-mcp]
command = "npx"
args = ["-y", "[email protected]", "https://mcp.birdeye.so/mcp", "--header", "x-api-key:${BIRDEYE_API_KEY}"]

[mcp_servers.birdeye-mcp.env]
BIRDEYE_API_KEY = "<YOUR_API_KEY>"

Or use the CLI command:

codex mcp add birdeye-mcp -- npx -y [email protected] https://mcp.birdeye.so/mcp --header "x-api-key:<YOUR_API_KEY>"

7. Cline (VS Code Extension)

Open VS Code Settings, navigate to Cline > MCP Settings, and add:

{
    "mcpServers": {
        "birdeye-mcp": {
            "command": "npx",
            "args": [
                "-y",
                "[email protected]",
                "https://mcp.birdeye.so/mcp",
                "--header",
                "x-api-key:${API_KEY}"
            ],
            "env": {
                "API_KEY": "<YOUR_API_KEY>"
            }
        }
    }
}

8. Roo Code (VS Code Extension)

Add to your project's .roo/mcp.json (project-level) or configure via Roo Code settings in VS Code (global):

{
    "mcpServers": {
        "birdeye-mcp": {
            "command": "npx",
            "args": [
                "-y",
                "[email protected]",
                "https://mcp.birdeye.so/mcp",
                "--header",
                "x-api-key:${API_KEY}"
            ],
            "env": {
                "API_KEY": "<YOUR_API_KEY>"
            }
        }
    }
}

9. Amazon Q Developer

Add to ~/.aws/amazonq/mcp.json (global) or .amazonq/mcp.json (workspace):

{
    "mcpServers": {
        "birdeye-mcp": {
            "command": "npx",
            "args": [
                "-y",
                "[email protected]",
                "https://mcp.birdeye.so/mcp",
                "--header",
                "x-api-key:${API_KEY}"
            ],
            "env": {
                "API_KEY": "<YOUR_API_KEY>"
            }
        }
    }
}

10. JetBrains IDEs (IntelliJ, WebStorm, PyCharm, etc.)

Go to Settings > Tools > AI Assistant > Model Context Protocol (MCP) and add a new server with this config:

{
    "mcpServers": {
        "birdeye-mcp": {
            "command": "npx",
            "args": [
                "-y",
                "[email protected]",
                "https://mcp.birdeye.so/mcp",
                "--header",
                "x-api-key:${API_KEY}"
            ],
            "env": {
                "API_KEY": "<YOUR_API_KEY>"
            }
        }
    }
}

Note: Requires JetBrains IDE version 2025.1+ with the MCP Server plugin, or 2025.2+ which has built-in MCP support.

11. Continue (VS Code Extension)

Add the following to your Continue configuration (~/.continue/config.yaml or via the Continue settings UI):

{
    "mcpServers": {
        "birdeye-mcp": {
            "command": "npx",
            "args": [
                "-y",
                "[email protected]",
                "https://mcp.birdeye.so/mcp",
                "--header",
                "x-api-key:${API_KEY}"
            ],
            "env": {
                "API_KEY": "<YOUR_API_KEY>"
            }
        }
    }
}

12. Zed

Open Zed settings (Cmd+, on macOS) and add to the context_servers section:

{
    "context_servers": {
        "birdeye-mcp": {
            "command": {
                "path": "npx",
                "args": [
                    "-y",
                    "[email protected]",
                    "https://mcp.birdeye.so/mcp",
                    "--header",
                    "x-api-key:<YOUR_API_KEY>"
                ]
            }
        }
    }
}

💡 Example Prompts

Tap into the full potential of Birdeye data — use these prompts to kickstart your next AI build.

Simple Queries

  • What is the current price of Bitcoin in USD?
  • What is the current price of JUP in USD?
  • What is the top 10 BONK holders?
  • What are the top 3 trending coins on Birdeye right now?
  • What are the top AI coins on Birdeye now?

Advanced Queries

  • List tokens launched in the last 72 hours with liquidity above $500k. Include price, market cap, liquidity, and 24h volume. Display in an interactive table.
  • Generate a 14-day price chart for Solana (SOL) against USD. Include both price and trading volume.

Creative and Fun Ideas

  • Create a “Whale Moves” dashboard that highlights tokens with significant accumulation from top holders over the past 7 days. Include price change and liquidity. Try it here: claude.ai/public/artifacts

  • Build a “Market Mood” view that compares performance of major sectors like AI, Meme, and DeFi over the last 7 days. Display sector averages and total volume. Try it here: claude.ai/public/artifacts


Updated 9 days ago