CryptoRank MCP Server

Overview

The CryptoRank MCP Server connects AI agents and LLMs directly to CryptoRank’s curated crypto and fundraising datasets via the Model Context Protocol (MCP). 

Instead of parsing REST responses, agents call high-level tools to access funding rounds, investor profiles, token unlocks, project teams, and market data in a format that models understand natively. 

Access to the CryptoRank MCP Server is available exclusively for users with a PRO (or higher) CryptoRank Public API plan.

Why use CryptoRank MCP?

Traditional APIs require you to write glue code, manage pagination, and normalize JSON before your AI agents can use the data. 

With MCP, CryptoRank exposes domain-specific tools like fundraising search, fund intelligence, and vesting schedules that agents can call autonomously from natural language prompts. 

This makes it easy to turn a general-purpose model into a crypto-native research, trading, or investment assistant powered by data that is difficult or impossible to scrape elsewhere.

Key datasets

CryptoRank MCP focuses on unique, research-grade datasets beyond basic prices and tickers.​

• Funding Rounds: ICO/IDO/IEO/private rounds, amounts raised, stages, participants, and dates.​

• Funds & Backers: VC profiles, tiers, strategies, portfolio breakdowns, recent deals, and check sizes.​

• Vesting & Token unlocks: detailed schedules, upcoming unlocks, percentages vs supply/market cap.​

• Project Teams: team composition, notable members, and links.

All of these are exposed as MCP tools so that agents can combine them in a single query, such as “Show AI projects that raised over $10M in 2025 and have major token unlocks next quarter”.

Server options

CryptoRank offers a single, authenticated remote MCP server. There is no free keyless tier and no local MCP server.​

All access goes through your existing CryptoRank Public API key, and MCP usage counts against your PRO (or higher) plan limits.

Requirements

To use the CryptoRank MCP Server you need:

• A CryptoRank Public API key on a PRO plan or higher, which unlocks the datasets exposed via MCP.

• An MCP-compatible client such as Claude Desktop, ChatGPT (Developer Mode with connectors), Cursor, Gemini CLI, or another MCP-enabled environment.

• Basic familiarity with editing your client’s MCP configuration file (for example, claude_desktop_config.json or mcp_config.json).​

There is no public, unauthenticated endpoint and no local/offline MCP server distribution.​

Authentication

The CryptoRank MCP Server uses the same API key mechanism as the CryptoRank Public API, so you manage keys and permissions in a single place. 

You pass your API key from the MCP client (for example as an environment variable) and the MCP server forwards it to the CryptoRank API for each tool call. Rate limits and quotas for MCP requests are derived from your underlying API plan (PRO, Business, etc.).​

Example environment configuration (conceptual):

export CRYPTORANK_API_KEY="YOUR_PRO_API_KEY"

Your MCP client configuration then injects this key into the env section for the CryptoRank MCP server process.​

Connecting from an MCP client

Most MCP clients support adding external servers via a JSON configuration file that defines the command to run and its arguments. Below is an example configuration for a remote CryptoRank MCP server using mcp-remote:

{

  "mcpServers": {

    "cryptorank_mcp": {

      "command": "npx",

      "args": [

        "mcp-remote",

        "https://api.cryptorank.io/mcp"

      ],

      "env": {

        "CRYPTORANK_API_KEY": "YOUR_PRO_API_KEY"

      }

    }

  }

}

After saving the configuration and restarting your client (for example Claude Desktop or Cursor), the CryptoRank tools will appear automatically and can be invoked by the model when relevant to a user’s query.

Sample tools

The exact tool names may evolve, but the CryptoRank MCP Server is designed around focused, high-signal tools similar to market endpoints.

Typical categories include:

• Project & asset discovery (search by name, symbol, sector, tags, or narratives).

• Fundraising & rounds (filter by date range, sector, minimum raise, or lead investor).​

• Fund & investor profiles (tier, thesis, portfolio composition, recent deals, geographies).​

• Token unlocks & vesting (upcoming unlocks, historical unlocks, impact metrics).​

Tools are designed to return structured, model-friendly objects so that AI agents can generate tables, run comparisons, and chain multiple calls without extra glue code.

Usage examples

Here are sample prompts that an AI agent can answer using CryptoRank MCP tools:

• “Show all AI projects that raised over 10M USD in 2025, with rounds and lead investors.”​

• “List major token unlocks for the next quarter sorted by unlock value and percentage of circulating supply.”​

• “Find all investments made by a16z in DeFi projects and summarize their portfolio size and recent exits.”​

In each case, the agent automatically selects and combines the appropriate tools (fundraising, funds, unlocks, teams, prices) through MCP, instead of you manually stitching multiple REST endpoints together.

Pricing and rate limits

There is no separate pricing for the CryptoRank MCP Server; access is bundled with the PRO and higher tiers of the CryptoRank Public API. 

MCP requests consume credits and rate limits from your existing plan, making it easy to upgrade or downgrade without managing a second billing surface. 

If you hit rate limits, you can either optimize your agent’s tool usage or move to a higher API tier, just as you would for direct REST usage.

Limitations

The CryptoRank MCP Server is a hosted remote service with no free keyless endpoint and no local/offline MCP distribution. 

Access to specific tools and datasets may vary slightly by plan and will follow the same entitlement model as the underlying CryptoRank Public API.