Connect Claude to live financial data via FMP MCP server
Kevin Meneses González walks through setting up Financial Modeling Prep's official MCP server to give Claude direct access to real-time stock prices, financial statements, and market data — eliminating manual copy-paste workflows.
Score breakdown
Developers building AI-powered financial tools can replace brittle scraping or manual data pipelines with a single MCP server config, giving Claude live access to institutional-grade financial data for portfolio monitoring, earnings analysis, and custom stock screening.
- 01Financial Modeling Prep (FMP) launched an official MCP server that connects Claude directly to its financial data catalog.
- 02Setup requires installing `@financialmodelingprep/mcp-server` via npm and editing `claude_desktop_config.json` with an API key.
- 03The MCP server covers real-time and historical stock prices, income statements, balance sheets, cash flow data, ratios, earnings calendars, analyst estimates, and crypto/forex/commodities data.
Kevin Meneses González argues that the bottleneck in AI-assisted investing isn't model quality — it's the data pipeline. Developers building Claude workflows hit a recurring wall: getting fresh, structured financial data into the model without maintaining a custom integration layer. Previous options (scraping Yahoo Finance, paying for a premium API and writing connector code from scratch, or processing CSV exports manually) were fragile and didn't scale. Financial Modeling Prep's official MCP server addresses this by exposing its full data catalog — real-time and historical stock prices, income statements, balance sheets, cash flow data, financial ratios, valuation metrics, earnings calendars, analyst estimates, sector performance, and crypto/forex/commodities data — directly to Claude with no middleware required.
For automated pipelines outside interactive sessions, the article also provides Python code snippets using FMP's REST API directly to pull income statements and TTM ratios.
Setup involves installing `@financialmodelingprep/mcp-server` globally via npm, then adding an `fmp` entry to `claude_desktop_config.json` that passes an `FMP_API_KEY` environment variable. After restarting Claude Desktop, the FMP tools appear automatically. The article walks through three concrete workflow examples: a morning portfolio brief that retrieves price, daily change, and P/E ratio for multiple tickers and flags positions down more than 2%; a pre-earnings deep-dive that pulls four quarters of revenue, gross margin, and EPS for a ticker and compares against analyst consensus; and a natural-language stock screener that finds S&P 500 companies meeting specific P/E, debt-to-equity, and revenue growth criteria sorted by free cash flow yield. For automated pipelines outside interactive sessions, the article also provides Python code snippets using FMP's REST API directly to pull income statements and TTM ratios.
Key facts
- 01Financial Modeling Prep (FMP) launched an official MCP server that connects Claude directly to its financial data catalog.
- 02Setup requires installing `@financialmodelingprep/mcp-server` via npm and editing `claude_desktop_config.json` with an API key.
- 03The MCP server covers real-time and historical stock prices, income statements, balance sheets, cash flow data, ratios, earnings calendars, analyst estimates, and crypto/forex/commodities data.
- 04A morning portfolio brief prompt across multiple tickers is described as taking 90 seconds versus 20 minutes of manual tab-hopping.
- 05Natural-language stock screening (e.g., P/E below 15, debt-to-equity under 0.5, revenue growth above 10% YoY) is possible without SQL or manual filter configuration.