FinData MCP is a financial data server for AI agents that provides real-time and historical financial data to compatible clients through the Model Context Protocol (MCP).
Purpose It integrates multiple financial data sources into tools that AI can call directly, enabling AI to query stock quotes, company fundamentals, macroeconomic indicators, SEC filings, and cryptocurrency prices in conversational or programming environments without having to build its own data pipeline.
Main capabilities The server provides five categories of tools:
- Real-time stock prices: covering 50,000+ ticker symbols on NYSE and NASDAQ.
- Stock fundamentals: including price-to-earnings ratio (P/E), earnings, market capitalization, and more.
- Macroeconomic indicators: over 800,000 FRED economic series, such as GDP, CPI, inflation, interest rates, and Treasury yields. Commonly used series include GDP, CPIAUCSL (inflation), UNRATE (unemployment rate), DGS10 (10-year Treasury), and FEDFUNDS.
- SEC EDGAR filings: 10-K annual reports, 10-Q, 8-K, and more.
- Cryptocurrency prices: 8,000+ coins including Bitcoin and Ethereum. The data sources are Yahoo Finance, FRED, SEC EDGAR, and CoinGecko.
How to use This tool is compatible with Claude Desktop, Cursor, Windsurf, and any MCP client. Installation and configuration steps:
- Set the wallet private key (any EVM wallet holding USDC on Base), with the environment variable EVM_PRIVATE_KEY.
- Add to Claude Desktop: edit ~/Library/Application Support/Claude/claude_desktop_config.json.
- Add to Cursor: edit .cursor/mcp.json in the project root directory.
- Restart the client to use the five financial data tools. It can also be installed via pip (package name findata-mcp). This package is a lightweight MCP stdio server that proxies tool calls to a hosted backend and automatically handles x402 payment signing. It can also be run without installation using uvx.
Call examples include: getting stock quotes, querying company fundamentals, viewing economicindicators, and reading SEC filings. The tools accept ticker as a parameter and are also compatible with ticker_or_cik (legacy) or the symbol alias.
Pricing and payment The first 50 calls/day are free. After that, it is billed at $0.01/call, automatically settled in USDC on Base Mainnet through the x402 open micropayment protocol. No registration, no API key, no monthly fee. Requirements: an EVM wallet private key (EVM_PRIVATE_KEY), and a small USDC balance on Base Mainnet (about $1 can cover 100 calls). No account, no rate limits, no billing page.
Use cases Suitable for developers who need to integrate financial data query capabilities into AI agents, for example to quickly obtain quotes, financial reports, macroeconomic indicators, and crypto prices in AI assistants, programming tools, or automated workflows.
Limitations The materials do not provide information such as the latency of each data source, historical data depth, details of call frequency limits, data accuracy guarantees, or refund policy. Users need to provide their own EVM wallet and USDC, and payment occurs on-chain.
