Massive Market Data MCP Server

The Massive.com MCP server gives LLMs access to the full Massive.com financial data API through an LLM-friendly interface. Rather than exposing one tool per endpoint, it ships three composable tools (search_endpoints, call_api, query_data) that cover the entire Massive.com API surface, including stock, options, forex, and crypto aggregates and bars, real-time and historical trades and quotes, market snapshots, ticker reference data, dividends, splits, IPOs, financial fundamentals, news, treasury data, and market status.

The server dynamically indexes all Massive.com endpoints at startup from the provider's llms.txt, so it automatically stays in sync with API changes. Results from call_api can be persisted into an in-memory SQLite database (default limits: 50 tables, 50,000 rows), which the LLM can then query with full SQL including CTEs and window functions. The server also bundles built-in financial functions for Black-Scholes Greeks (price, delta, gamma, theta, vega, rho), returns and risk-adjusted performance ratios (simple/log/cumulative return, Sharpe, Sortino), and technical indicators (SMA, EMA).

It is the official MCP server maintained by Massive.com under the massive-com GitHub org, written in Python and licensed MIT. Releases are active, with v0.10.0 published in May 2026.

Tools

Prerequisites

• Python 3.12+
• A Massive.com API key (set as MASSIVE_API_KEY)
• Astral UV v0.4.0+

Install

uv tool install "mcp_massive @ git+https://github.com/massive-com/mcp_massive@v0.10.0"

Claude Code

claude mcp add massive -e MASSIVE_API_KEY=your_api_key_here -- mcp_massive

Claude Desktop

Add to claude_desktop_config.json:

{
  "mcpServers": {
    "massive": {
      "command": "<path_to_mcp_massive>",
      "env": {
        "MASSIVE_API_KEY": "<your_api_key_here>",
        "HOME": "<your_home_directory>"
      }
    }
  }
}

Replace <path_to_mcp_massive> with the absolute path returned by which mcp_massive (or where mcp_massive on Windows) after the uv tool install step.

Built-in functions

Available via call_api post-processing or referenced from search_endpoints:

• Greeks: bs_price, bs_delta, bs_gamma, bs_theta, bs_vega, bs_rho
• Returns: simple_return, log_return, cumulative_return, sharpe_ratio, sortino_ratio
• Technical: sma, ema

• Pull historical OHLC bars for a basket of tickers, store them as SQLite tables, and run SQL to compute correlation matrices or rank by return.
• Compute Black-Scholes Greeks (delta, gamma, theta, vega) for an options chain pulled live from Massive.com.
• Build a daily market scan: fetch top gainers/losers, snapshots, and news, then summarize the macro tape.
• Backtest a moving-average crossover by calling the aggregates endpoint and applying the built-in sma/ema and sharpe_ratio functions.
• Query fundamentals, dividends, splits, and IPO calendar data to feed an equity research workflow.

• "Get the last 30 days of daily bars for AAPL, MSFT, and NVDA and compute their Sharpe ratios."
• "Find the call option on SPY expiring this Friday closest to ATM and return its Black-Scholes delta and gamma."
• "Show me today's top 10 stock market gainers and any related news headlines."
• "Pull the options chain for TSLA expiring next month and compute total open interest by strike using SQL."
• "Fetch QQQ minute aggregates from the last trading day and plot a 20-period EMA."

• Official server maintained by Massive.com under the massive-com GitHub org.
• Auto-syncs with the API via dynamic indexing of llms.txt, so new endpoints are picked up without code changes.
• Composable design (search + call + SQL) is more powerful than one-tool-per-endpoint wrappers for analytical workflows.
• Ships with finance-specific helpers (Black-Scholes Greeks, Sharpe/Sortino, SMA/EMA) so the LLM does not need to reimplement them.

• Requires a paid Massive.com API key for most non-trivial data access.
• Python 3.12+ and Astral UV are hard prerequisites; no prebuilt Docker image or hosted endpoint is documented.
• In-memory SQLite store has default caps (50 tables, 50,000 rows), so very large multi-ticker pulls need explicit pagination and table management.

• Alpha Vantage MCP server for free-tier equity, FX, and crypto data.
• Yahoo Finance MCP servers (community implementations) for no-key historical quotes.
• Finnhub MCP server for fundamentals, earnings, and news.