Best MCP Server for Stock Data in 2026

Many stock data providers now run an official MCP server: Alpha Vantage, Financial Modeling Prep (FMP), Financial Datasets, EODHD, Massive (formerly Polygon), Twelve Data, Finnhub, Fiscal.ai and Roic.ai. Brokers have followed: Interactive Brokers, Robinhood, Alpaca, tastytrade and Tradier offer MCP servers for account and trading data. Yahoo Finance has no official server; the community ones scrape its public endpoints.

The best MCP server for stock screening and fundamentals is one that gives Claude direct database access rather than forwarding each question to a rate-limited API. Shibui Finance connects Claude and ChatGPT to a pre-loaded database containing 64 years of US equity data: 31M+ daily price records, quarterly and yearly financial statements, 56 pre-calculated technical indicators, daily valuation metrics, earnings surprises, and 6.4 million SEC EDGAR filing records for 10,000+ companies. No API key, no per-query costs, fair-use limits only.

Most other servers forward each request to an external API, so a screening question that checks thousands of companies takes thousands of calls. Shibui answers it as a single query in seconds. The table below compares the main options.

The MCP options for financial data

The main MCP servers for stock market data each target a different slice of the problem. Alpha Vantage runs a hosted server (or a local one) with about 120 tools, one per API function. Its free key allows 25 requests per day, and real-time US data starts at $99.99/mo. Financial Modeling Prep (FMP) has a hosted server that groups its REST API into 27 tool modules, with deep fundamentals and SEC filings; the free plan allows 250 calls per day. Financial Datasets covers US fundamentals going back 30+ years and can extract sections of 10-K, 10-Q and 8-K filings. It has no free tier: $20 buys 1,000 requests. EODHD covers 60+ exchanges and 150,000 tickers, with a free tier of 20 calls per day. Massive (formerly Polygon) offers tick-level US stocks, options, forex and crypto; its free plan is end-of-day with 5 calls per minute. Yahoo Finance has no official server. Community servers wrap the unofficial yfinance library, and Yahoo's terms forbid automated data collection.

They don't all work the same way. Most expose one tool per API endpoint: when Claude needs data, it calls the provider's API, waits for the response, and works with whatever comes back. Some add a screener, but it filters on current values only. FMP's screener covers price, size and sector; EODHD's has 14 fields; Financial Datasets can filter on growth, but only for the latest period. Massive's self-hosted server goes a step further: it can store fetched results in an in-memory SQLite database and run SQL over them, up to 50 tables of 50,000 rows, but only over data the session has already fetched. Shibui runs SQL on the server, over the whole database.

The difference shows on a question like "which US companies grew revenue in each of the last 12 quarters?" On Shibui that is one query. On the other servers it means fetching quarterly statements one company at a time, thousands of calls, before any filtering can happen.

Here is how the main MCP servers for stock market data compare on the decisions that matter: how the data reaches Claude, whether you need an API key, what limits apply, what is covered, and what each one is best at.

Free tiers, rate limits, and tool counts change often; check each provider for current terms. The row that matters for your work is the last column: pick the server whose strength matches the question you ask most. For the free tiers and pricing of the data APIs behind these servers, see the financial data API comparison.

Other finance tools with an MCP server

Beyond the seven in the table, these finance tools offer official MCP servers:

• Market data: Twelve Data (hosted, an official Claude connector, free plan with 8 API credits per minute), Finnhub (hosted, about 70 tools), Intrinio (hosted, from $150/mo, options with Greeks), Fiscal.ai (an official Claude connector; the free trial covers 100 companies) and Roic.ai (hosted, 36 tools, free plan limited to 2 years of history).
• Options flow: Unusual Whales (options flow, dark pool and screeners; API plans from $150/mo).
• Brokerage: Interactive Brokers (hosted, an official Claude connector, no extra cost; Claude can draft trades but never submits orders), Robinhood (hosted, equities only, in beta), Alpaca (runs locally, paper trading by default), tastytrade (self-hosted) and Tradier (hosted).

Live API calls vs pre-loaded data

The core difference is where the data lives at query time. With a live-API server, each question triggers one or more external API calls. Ask for a company's financials and you get one call. Ask to compare 20 companies and you get 20 calls. Ask to screen the entire market and you hit rate limits before you get through 5% of the universe.

With Shibui, the data is already there. Claude writes a query that runs against 31M+ records in a single pass. There is no external call at query time, no per-request quota to hit, no API key to configure. This distinction matters most for cross-market screening, the kind of question where you need to check a condition across every company in Shibui's coverage.

This query checks 10,000+ companies in one pass. Claude writes a single query that examines quarterly financials across multiple periods, filters for the growth condition holding in each consecutive quarter, and pulls in cash flow data. With a live-API server, you would need thousands of individual API calls (one per company, often more), hitting rate limits and taking minutes or hours to complete.

What a pre-loaded database makes possible

Some questions are straightforward with a database and effectively impossible with individual API calls. These are queries that require combining millions of records across multiple data sources or checking conditions across long time periods.

The first query lines up daily technical indicators with valuation data to find co-occurring conditions, then measures forward price returns at a fixed offset. It touches millions of records across three data sources. The second checks a trend across five yearly periods for every technology company in Shibui's coverage. Both run in seconds against a pre-loaded database. With a live-API server, you would need to fetch data for each company individually, combine results client-side, and handle the rate limiting across thousands of calls.

When live API calls make more sense

A pre-loaded database is not the right tool for every job. If you need real-time or intraday price data, Alpha Vantage or Massive are better options; both need a paid plan for real-time data. Shibui updates daily after market close, so it cannot answer questions about what a stock is trading at right now.

If you need international markets, EODHD covers 60+ exchanges worldwide. Shibui covers US equities only (NYSE and NASDAQ). If you need options chains, Alpha Vantage and Massive have them, brokers like Interactive Brokers, tastytrade and Tradier expose them through their own MCP servers, and Unusual Whales covers options flow; Shibui has no options data. If you need the very latest quarterly filing on the day it is published, FMP's direct SEC feed may surface it faster than Shibui's daily import cycle.

Shibui's limitations

Shibui covers US equities only (NYSE and NASDAQ). Data is end-of-day, updated once a day after the close, not real-time or intraday. There is no options, forex, or cryptocurrency data. Coverage and accuracy are good for screening and analysis, but this is not institutional-grade data, and the price history is survivor-only: delisted companies are absent. Each query returns a maximum of 200 rows. GICS sector classifications have some gaps, particularly for ETFs and closed-end funds. These are real constraints, and for some workflows they are disqualifying.

Pairing Shibui with a qualitative layer

The servers above deliver data: prices, financial statements, filings. Some questions sit outside the data. Does the company have pricing power? How hard is it for customers to switch? How good is management? Baseline, from A5 Labs, is an MCP server built for that layer. It scores 3,600+ stocks from 1 to 5 on business quality, competitive dynamics, pricing power, management and risks, and each score comes with a written rationale. It has a free tier.

The two work in the same Claude session. Shibui narrows the market on the numbers, and Baseline explains the business behind each name on the shortlist.

For a full worked example, see low switching costs and high AI risk in US stocks, written with A5 Labs: Baseline flags 26 US stocks, and Shibui checks which of them the market has already marked down.

Frequently asked questions

What MCP servers are available for stock market data?

Many data providers now run official MCP servers, including Alpha Vantage, Financial Modeling Prep (FMP), Financial Datasets, EODHD, Massive (formerly Polygon), Twelve Data and Finnhub. Most forward each request to their data API and need an API key or an account with the provider. Yahoo Finance has no official server; community servers scrape its public endpoints. Shibui Finance takes a different approach: it connects Claude to a pre-loaded database of 64 years of US equity data, so queries run directly against the data with no external API calls.

What is the best MCP server for fundamental analysis and stock screening?

It depends on your needs. For cross-market screening with historical depth (checking conditions across 10,000+ companies and 64 years), Shibui Finance is built for that, with 31M+ daily records pre-loaded. For real-time or intraday data, Alpha Vantage or Massive are better suited. For international coverage, EODHD covers 60+ exchanges.

Do I need an API key to use stock data in Claude?

Not with Shibui Finance. It requires no API key, no account registration, and no configuration beyond a one-click connection in Claude. Other MCP servers for stock data (including Alpha Vantage, FMP, EODHD and Massive) need an API key or an account with the provider, each with its own signup and pricing.

Are there free stock data connectors for Claude?

Yes. Shibui Finance is free: it connects Claude to a pre-loaded database with 64 years of US equity data, with no API key and no per-request quotas. Fair-use limits apply to stop bulk scraping; a typical research session stays well under them. Several providers have free tiers on their own MCP servers: FMP allows 250 calls per day, Alpha Vantage 25 requests per day, EODHD 20 calls per day and Massive 5 calls per minute of end-of-day data. Community Yahoo Finance servers need no key, but rely on unofficial endpoints that may break.

Can I use Shibui as a Claude Desktop extension with Node.js?

Yes. Shibui Finance is available as a DXT extension for Claude Desktop. It runs on Node.js (version 18 or later). You can also connect via the MCP connector on claude.ai, which requires no local installation at all. Both methods give Claude access to the same database.

Is there an MCP server for brokerage or options data?

Yes. Interactive Brokers, Robinhood, Alpaca, tastytrade and Tradier offer official MCP servers for account and trading data, and Interactive Brokers is also an official Claude connector. For options data, Alpha Vantage, Massive and Intrinio include options through their servers, and Unusual Whales covers options flow. Shibui Finance has no options or brokerage data: it covers end-of-day US equities.

Detailed comparisons

Each of these pages goes deeper on the architectural trade-offs between Shibui and one specific competitor, with side-by-side tables, example prompts, and honest assessments of when the other tool is the better choice.

• Shibui vs Alpha Vantage MCP - pre-loaded database vs. rate-limited API passthrough (25 calls/day free)
• Shibui vs Yahoo Finance MCP - stable database vs. unofficial endpoint scraper (both free)
• Shibui vs Financial Datasets & FMP MCP - cross-market screening vs. deep fundamentals and SEC filing text
• Shibui vs Finviz - screening across time vs. preset dropdown filters
• Shibui vs TradingView - fundamental screening depth vs. real-time charting
• Shibui vs Koyfin - pattern detection vs. polished research dashboards