Messari MCP Server

The hosted MCP server provides the simplest way to connect AI assistants like Claude Desktop, Cursor, and Claude Code to Messari’s crypto intelligence platform. Unlike traditional MCP servers that require local installation, this server is fully hosted—just add a URL and authenticate. It provides tools for AI-powered document search, natural-language dataset queries, and direct timeseries data access.

Key Benefits

Zero Installation

No npm, no npx, no local dependencies. Just add a URL to your MCP client.

Secure OAuth

Browser-based authentication with your Messari API key. No keys in config files.

Purpose-Built Tools

Dedicated search and dataset agents, so the client picks the right one for the question.

Always Up-to-Date

Hosted infrastructure means you always have the latest capabilities.

Quick Setup

Add the Messari MCP server to your AI assistant in seconds: MCP Server URL:

https://mcp.messari.io/mcp

When you first connect, you’ll be redirected to authenticate with your Messari API key through a secure OAuth flow.

Configuring MCP Clients

Claude Connectors

To add Messari as a custom connector in Claude:

Navigate to claude.ai/settings/connectors
Click Add custom connector
Enter the MCP server URL: https://mcp.messari.io/mcp
Click Connect and enter your Messari API key when prompted

OpenAI ChatGPT App

To add Messari as a custom App in ChatGPT:

Navigate to chatgpt.com/admin/ca#available
Enable developer mode.
Click Create
Enter the MCP server URL: https://mcp.messari.io/mcp
Leave Authentication as OAuth.
Click Create and enter your Messari API key when prompted

Claude Code

Add to your Claude Code MCP settings:

{
  "mcpServers": {
    "messari": {
      "url": "https://mcp.messari.io/mcp"
    }
  }
}

Cursor

Add to your Cursor MCP configuration:

{
  "mcpServers": {
    "messari": {
      "url": "https://mcp.messari.io/mcp"
    }
  }
}

Other MCP Clients

Any MCP-compatible client that supports remote servers can connect using:

URL: https://mcp.messari.io/mcp
Authentication: OAuth 2.0 (handled automatically)

Authentication Flow

When you first connect:

Your MCP client will detect the server requires authentication
A browser window opens to the Messari OAuth page
Enter your Messari API key
You’re authenticated and ready to query

Your API key is securely stored and used to authenticate requests to Messari’s services. You can obtain an API key from your Messari account settings.

API Permissions

Access to the MCP server requires a Messari Enterprise subscription with access to the AI Toolkit. Access to each tool and its underlying data depends on your API key’s subscription tier. search_messari_documents and query_messari_datasets additionally require an authenticated Enterprise user whose team holds the ai_toolkit_permission API permission, shown as Messari AI API in team settings. A caller without it sees a FORBIDDEN error from the tool. For more information on API tiers, permissions, and rate limits, see our API Permissions documentation.

Rate Limits

search_messari_documents and query_messari_datasets each allow 30 requests per minute per user, counted in its own bucket — search traffic does not consume the dataset tool’s allowance. Exceeding a bucket returns a RATE_LIMITED error from that tool.

Credits

The search and dataset tools do not consume AI credits.

These two tools call the Agents API under the hood, and inherit its permissions, rate limits, and response shapes. Use that reference when you need the full endpoint detail.

Available Tools

The hosted MCP server provides the following tools:

Tool	Purpose
`search_messari_documents`	Search Messari research, news, and curated sources, and return the matching documents
`query_messari_datasets`	Answer quantitative questions by querying Messari’s datasets
`get_timeseries_catalog`	Discover available datasets, metrics, and granularities from the Messari Timeseries API
`get_timeseries`	Query timeseries data for crypto assets, protocols, and networks from the Messari Timeseries API

The `search_messari_documents` Tool

Searches Messari’s research, news, and curated sources for documents relevant to a question, and returns the matching documents themselves. It retrieves; it does not write the answer. Your assistant reads the documents and answers from them. sources holds the retrieved documents. Each carries a domain (research, news, external_news, diligence, governance, twitter, video_transcript and others), a title, a url, the document content, and published_at where the domain carries a date. error is set only when a run finds no usable document, and explains why.

The corpus spans years, and the Search agent reads the time window out of your wording. Name concrete dates whenever the question is time-sensitive — “in August 2026” or “since 2026-08-01”, not “latest” or “recently”. A query with no date returns the best matches from the whole corpus, most of which will be old.

Example Queries

“What is EigenLayer restaking?”
“Latest news about Ethereum”
“Explain Uniswap v4’s new features”
“What happened with FTX?”
“Recent DeFi funding rounds this month”

The `query_messari_datasets` Tool

Answers quantitative questions by translating them into SQL over Messari’s datasets and running it. It computes; it does not write the answer. Your assistant reads the rows and answers from them. results holds each query the Compute agent ran. dataset is the dataset it selected, query is the executed SQL, query_result holds the rows JSON-encoded as an array of arrays, and columns names each position in those rows, which are otherwise unlabelled. truncated reports that the row cap was reached, so more rows may match. query_error explains why an individual query returned nothing. error is set only when a run produces no usable result. A run can take several minutes.

The SQL and column names reflect Messari’s internal schema, which may change without notice. Read them, do not parse them.

Example Queries

“What is Bitcoin’s current price and 24h volume?”
“Show me the top 10 assets by market cap”
“What is Aave’s TVL?”
“When is the next ARB token unlock?”
“Solana daily active addresses this week”

The `get_timeseries_catalog` Tool

Discovers what timeseries data is available through Messari’s Timeseries API, including dataset slugs, supported time granularities, available metrics, and entity types. Your assistant uses this to determine what can be queried before fetching data. The datasets returned depend on your subscription tier.

Example Queries

“What are the available timeseries datasets for DeFi protocols?”
“Which metrics can I get for Ethereum, and at what granularity?”

The `get_timeseries` Tool

Queries timeseries data points for crypto assets, protocols, networks, and more through Messari’s Timeseries API. Covers 80+ datasets with configurable time ranges and granularities.

Example Queries

“What is Bitcoin’s daily price over the last 90 days?”
“Show me Ethereum’s weekly active addresses for the past year”
“Compare daily trading volumes for Bitcoin and Solana this quarter”
“Get Aave’s TVL trend over the last 6 months”

Data Coverage

Market Data

Real-time prices, volumes, and market caps for 40,000+ crypto assets

On-Chain Metrics

TVL, active addresses, fees, and transactions for 25+ networks and 200+ protocols

Fundraising

Funding rounds, M&A deals, investors, and company valuations

Token Unlocks

Vesting schedules, allocation breakdowns, and upcoming unlock events

News

Aggregated from 500+ sources with AI summarization

Research

Messari’s proprietary reports, diligence analyses, and quarterly reviews

Signals

Social mindshare and sentiment data from X/Twitter

Timeseries Data

Historical and real-time timeseries metrics across 80+ datasets via the Messari Timeseries API

Best Practices

For complex or multi-faceted questions, consider making multiple focused queries instead of one loaded question. Breaking down questions into specific, single-topic calls often yields better depth and accuracy.

Quantitative queries work best when you specify:

The exact metric (e.g., “TVL”, “market cap”, “24h trading volume”)
The asset or protocol name
Timeframe if relevant (e.g., “last 30 days”, “YTD”)
Comparison baseline if needed (e.g., “vs last month”, “vs Ethereum”)

Troubleshooting

Authentication Issues

OAuth window doesn’t open:

Ensure your MCP client supports OAuth 2.0 authentication
Check that popups aren’t blocked in your browser
Try restarting your MCP client

Invalid API key error:

Verify your API key at messari.io/account/api
Ensure your API key has the necessary permissions
Check that your API key hasn’t expired

Connection Issues

Server unreachable:

Verify your internet connection
Check if https://mcp.messari.io is accessible
Try again in a few minutes if there’s a temporary outage

Query Issues

Empty or unexpected responses:

Try rephrasing your question more specifically
Break complex questions into multiple simpler queries
Ensure you’re asking about supported data types

Results are older than expected:

Name concrete dates in the question. search_messari_documents reads the time window out of your wording, and a query with no date searches the whole corpus.

FORBIDDEN from the search or dataset tool:

Your team needs the ai_toolkit_permission API permission, shown as Messari AI API in team settings.

RATE_LIMITED from the search or dataset tool:

Each tool allows 30 requests per minute per user. Wait for the next minute before retrying. The two tools have separate buckets, so one being limited does not affect the other.

Support

Need help? Contact us:

Documentation: docs.messari.io
API Support: messari.io/contact
API Key Issues: messari.io/account/api

​Messari MCP Server

​Key Benefits

Zero Installation

Secure OAuth

Purpose-Built Tools

Always Up-to-Date

​Quick Setup

​Configuring MCP Clients

​Claude Connectors

​OpenAI ChatGPT App

​Claude Code

​Cursor

​Other MCP Clients

​Authentication Flow

​API Permissions

​Rate Limits

​Credits

​Available Tools

​The search_messari_documents Tool

​Example Queries

​The query_messari_datasets Tool

​Example Queries

​The get_timeseries_catalog Tool

​Example Queries

​The get_timeseries Tool

​Example Queries

​Data Coverage

Market Data

On-Chain Metrics

Fundraising

Token Unlocks

News

Research

Signals

Timeseries Data

​Best Practices

​Troubleshooting

​Authentication Issues

​Connection Issues

​Query Issues

​Support

Messari MCP Server

Key Benefits

Quick Setup

Configuring MCP Clients

Claude Connectors

OpenAI ChatGPT App

Claude Code

Cursor

Other MCP Clients

Authentication Flow

API Permissions

Rate Limits

Credits

Available Tools

The `search_messari_documents` Tool

Example Queries

The `query_messari_datasets` Tool

Example Queries

The `get_timeseries_catalog` Tool

Example Queries

The `get_timeseries` Tool

Example Queries

Data Coverage

Best Practices

Troubleshooting

Authentication Issues

Connection Issues

Query Issues

Support