mcpField
F
Finnhub

Finnhub

Remote
pipeworx.ioFinance0Updated 3d ago

About Finnhub

Finnhub connects the Finnhub Stock API’s market data to MCP for financial research and data-oriented agents. Its coverage includes stocks, crypto, forex, news, fundamentals, insider activity, sentiment, ETF holdings, and economic data.

Key features

  1. 01Retrieve real-time or delayed stock quotes
  2. 02Get company news and sentiment
  3. 03Check insider trades and ETF holdings
  4. 04Query earnings calendars and EPS estimates
  5. 05Access crypto, forex, and economic data

Use cases

  1. 01Combine quotes and news for company research
  2. 02Review earnings schedules and historical EPS surprises
  3. 03Research international stocks and market data
  4. 04Include insider activity in financial analysis

MCP support

It exposes Finnhub market-data tools to MCP clients. The Pipeworx gateway endpoint adds shared meta-tools, while the local stdio version provides only Finnhub tools; calls require the user’s own API key.

Configuration

Remote servers need no local install. Pick your client, then copy the config or install in one click.

Endpointhttps://gateway.pipeworx.io/finnhub/mcp
Copy install command
Claude CodeTerminal
Cursor.cursor/mcp.json
VS Code.vscode/mcp.json
Claude Desktopclaude_desktop_config.json
Windsurf~/.codeium/windsurf/mcp_config.json

Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.

Tools

Details

From the project README

Finnhub — Stock, Crypto, Forex Data

Finnhub's REST API. Real-time stock quotes, fundamentals, earnings, news, alternative data (insider transactions, sentiment, ETF holdings), economic data, crypto, and forex. Comprehensive market data with a strong free tier.

Part of Pipeworx — an MCP gateway connecting AI agents to 1879+ live data sources. This is an independent, unofficial integration — not affiliated with, endorsed by, or published by the upstream provider.

Why this matters for AI agents

Where Alpha Vantage is solid for fundamentals + technical, Finnhub is stronger for real-time quotes (1-minute resolution on free), insider transactions, news/sentiment, and international markets. For finance agents that need price data alongside news flow and insider activity, Finnhub is fit-for-purpose.

Common flows:

  • Quote. Real-time (or 15-min delayed on free tier) stock quote.
  • Company news. Recent news for a ticker with sentiment scoring.
  • Insider transactions. Form 4 filings with timing and dollar amounts (also see insider-trading pack which is SEC-EDGAR-direct).
  • Earnings. Upcoming earnings calendar, surprise history, EPS estimates.
  • International symbols. European, Asian, emerging-market tickers (where Alpha Vantage gets thinner).

Auth

Free key from https://finnhub.io/dashboard. Free tier: 60 API calls/minute. Pass via _apiKey.

Higher tiers ($50-200/mo) unlock real-time data, longer historical, and rate limits suitable for production. The free tier is sufficient for most agent research.

When to use Finnhub vs. Alpha Vantage

NeedBetter source
QuotesEither; Finnhub real-time-er on free tier
Income statement / balance sheetAlpha Vantage (cleaner schema)
Insider transactionsFinnhub or insider-trading pack
Company news with sentimentFinnhub
Earnings calendarFinnhub
International coverageFinnhub
Technical indicatorsAlpha Vantage
ETF holdingsFinnhub

For richest coverage, Pipeworx agents often mount both.

Earnings surprises vs. the earnings calendar

Two different tools answer two different earnings questions, and picking the wrong one silently misses data:

  • "What were last quarter's actual results vs. what analysts expected?" → get_earnings_surprises({ symbol }). Keyed by symbol, not by a date window, so it reaches back through the company's reporting history regardless of how long ago the quarter was. Returns EPS actual, EPS estimate, surprise amount and surprise percent per quarter (most recent first). EPS only — Finnhub's earnings-surprise history does not carry revenue, and the tool says so explicitly rather than returning a silent null.
  • "What's scheduled to report, or what happened in the last few weeks?" → get_earnings_calendar({ symbol, from, to }). Takes a date window and includes revenue estimate/actual alongside EPS — but measured live (2026-10-21), the calendar's historical reach on this plan is short: roughly the last several weeks from today, not the company's full reporting history. A quarter older than that returns found: false even though the company genuinely reported — that reads as "no earnings" if you don't know the window is truncated. Use get_earnings_surprises instead for anything older than a few weeks.

Common pitfalls

  • Symbol conventions. US tickers are simple (AAPL); international markets need exchange suffixes (NESN.SW for Nestlé on Swiss exchange). Finnhub has a symbol search.
  • Free tier real-time delay. "Real-time" on free is essentially current; on paid it's actually live. For agent research, free is fine. For trading, paid.
  • News sentiment is heuristic. Finnhub computes sentiment with NLP — it's directionally useful but noisy. Don't treat as ground truth; useful as a signal.
  • Insider transaction lag. SEC requires Form 4 filing within 2 business days of the transaction. Finnhub picks them up shortly after. Real-time alerting requires either polling or a separate real-time SEC feed.
  • Estimates vs. actuals. Finnhub provides analyst consensus estimates; mixing them with actuals in plots requires labeling carefully.
  • Free-tier coverage holes. Some smaller caps and OTC stocks have limited Finnhub coverage on free. Verify the data exists for your target tickers before designing flows around it.

Quick Start

Add to your MCP client (Claude Desktop, Cursor, Windsurf, etc.):

{
  "mcpServers": {
    "finnhub": {
      "url": "https://gateway.pipeworx.io/finnhub/mcp"
    }
  }
}
What this endpoint actually serves

tools/list at https://gateway.pipeworx.io/finnhub/mcp returns the tools in the table above plus the shared Pipeworx meta-tools — ask_pipeworx, discover_tools, search_within, remember/recall and the rest of the gateway-wide set. So the tool count you see is larger than this table: a single-pack endpoint currently lists roughly 30 shared tools alongside the pack's own. The connection's initialize response states its exact scope, and is the authoritative answer for a given day.

This is deliberate, not multiplexing by accident. The meta-tools are what let a scoped connection answer a question this pack does not cover — via ask_pipeworx, which routes across the whole catalog — without you adding a second MCP server. There is currently no way to mount a pack endpoint without them; if the extra schemas cost you more context than the routing is worth, connect to the full gateway once rather than to several pack endpoints.

Or connect to the full Pipeworx gateway to get every pack's tools listed directly, instead of just this one's:

{
  "mcpServers": {
    "pipeworx": {
      "url": "https://gateway.pipeworx.io/mcp"
    }
  }
}

Both URLs reach the same gateway and the same 1879+ data sources. The only difference is which pack's tools are listed directly; ask_pipeworx reaches all of them from either one.

No MCP client? Call it over HTTP

Our finnhub key is reserved for paid accounts, so an anonymous call to POST https://gateway.pipeworx.io/v1/tools/get_quote needs your own key passed as _apiKey alongside the arguments. Inspect any tool: GET https://gateway.pipeworx.io/v1/tools/get_quote. Find one: POST https://gateway.pipeworx.io/v1/tools/search_packs with {"query":"..."}.

Standalone (no gateway account)

This package also runs as a local stdio MCP server — no Pipeworx account, no gateway round-trip:

{
  "mcpServers": {
    "finnhub": {
      "command": "npx",
      "args": ["-y", "@pipeworx/mcp-finnhub"]
    }
  }
}

Or run it directly to confirm it starts:

npx -y @pipeworx/mcp-finnhub

It speaks MCP over stdin/stdout and answers initialize/tools/list/tools/call for only this pack's tools — none of the shared meta-tools the gateway connection above adds. Same source, same tools, no ask_pipeworx routing.

Using with ask_pipeworx

Instead of calling tools directly, you can ask questions in plain English — this works on the pack endpoint above as well as on the full gateway:

ask_pipeworx({ question: "your question about Finnhub data" })

The gateway picks the right tool and fills the arguments automatically.

More

License

MIT

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.