Finance MCP

v1.0.0Finance & Fintechstable

LLM-powered MCP server for building financial deep-research agents, integrating web search, Crawl4AI scraping, and entity extraction into composable analysis flows.

agentdeep-research-agentfinancefinance-mcpmcp
Share:
21
Stars
0
Downloads
0
Weekly
0/5

What is Finance MCP?

Finance MCP is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to llm-powered mcp server for building financial deep-research agents, integrating web search, crawl4ai scraping, and entity extraction into composable analysis flows.

LLM-powered MCP server for building financial deep-research agents, integrating web search, Crawl4AI scraping, and entity extraction into composable analysis flows.

This server falls under the Finance & Fintech and Search & Data Extraction categories on MCPgee, the world's largest MCP server directory with 33,000+ servers.

Features

  • LLM-powered MCP server for building financial deep-research

Use Cases

Build financial deep-research agents with web search and web scraping.
Extract and analyze financial data with entity recognition and composable flows.
FlowLLM-AI

Maintainer

LicenseApache-2.0
Languagepython
Versionv1.0.0
UpdatedMay 20, 2026
Statushealthy
Maintenanceactive

Works with

ClaudeOpenAIwindowsmacoslinux

Installation

Manual Installation

npx finance

Configuration

Configuration Details

Config File

claude_desktop_config.json

Performance

Response Metrics

Response Time< 200ms
ThroughputMedium

Resource Usage

Memory UsageLow
CPU UsageLow

How to Set Up and Use Finance MCP

Finance MCP is a Python-based MCP server and intelligent agent toolkit for financial deep-research scenarios, built on the FlowLLM framework. It integrates web search (via Tavily or DashScope), Crawl4AI web scraping, Tushare A-share market data, entity extraction, and Python/shell code execution into a composable set of 20+ pre-configured financial research flows. Analysts and developers can use it to build financial research agents that scrape earnings reports, search market news, perform price-volume analysis on A-share stocks, correlate financial entities, and execute custom analysis code — all through a single MCP-compatible server.

Prerequisites

  • Python 3.10 or higher installed
  • pip or uv package manager
  • At least one LLM API key for the FlowLLM backend (e.g., DashScope/Alibaba Cloud key set as FLOW_LLM_API_KEY)
  • Optional but recommended: Tushare API token (TUSHARE_API_TOKEN) for A-share market data, Tavily API key (TAVILY_API_KEY) for web search, DashScope API key (DASHSCOPE_API_KEY) for DashScope search
  • An MCP-compatible client such as Claude Desktop
1

Install Finance MCP

Install the finance-mcp package from PyPI using pip or uv.

pip install finance-mcp
# Or using uv:
uv pip install finance-mcp
2

Collect your API keys

Finance MCP integrates several services. Set the environment variables for the services you plan to use: FLOW_LLM_API_KEY for the LLM backend, DASHSCOPE_API_KEY for DashScope search, TUSHARE_API_TOKEN for A-share data, and TAVILY_API_KEY for Tavily web search.

export FLOW_LLM_API_KEY="your-flowllm-api-key"
export DASHSCOPE_API_KEY="your-dashscope-key"
export TUSHARE_API_TOKEN="your-tushare-token"
export TAVILY_API_KEY="your-tavily-key"
3

Add to your MCP client configuration in stdio mode

Configure Claude Desktop to run Finance MCP in stdio mode. Specify which config flows to load and which LLM model to use via command-line arguments.

4

Or start as an HTTP/SSE server for remote access

Run Finance MCP as a standalone SSE server accessible over the network. The service will be available at http://0.0.0.0:8001/sse.

finance-mcp \
  config=default,ths \
  mcp.transport=sse \
  mcp.host=0.0.0.0 \
  mcp.port=8001 \
  llm.default.model_name=qwen3-30b-a3b-thinking-2507 \
  disabled_flows='["tavily_search","mock_search","react_agent"]'
5

Test the available tools

Connect to the server and list available tools. Key tools include history_calculate for A-share price/volume analysis, crawl_url for scraping, dashscope_search and tavily_search for web search, and crawl_ths_* tools for TongHuaShun financial data.

Finance MCP Examples

Client configuration

Add this block to your claude_desktop_config.json to run Finance MCP in stdio mode with DashScope and Tushare integrations.

{
  "mcpServers": {
    "finance-mcp": {
      "command": "uvx",
      "args": [
        "finance-mcp",
        "config=default,ths",
        "mcp.transport=stdio",
        "llm.default.model_name=qwen3-30b-a3b-thinking-2507",
        "disabled_flows='[\"tavily_search\",\"mock_search\",\"react_agent\"]\']"
      ],
      "env": {
        "FLOW_LLM_API_KEY": "your-flowllm-api-key",
        "FLOW_LLM_BASE_URL": "https://dashscope.aliyuncs.com/compatible-mode/v1",
        "DASHSCOPE_API_KEY": "your-dashscope-key",
        "TUSHARE_API_TOKEN": "your-tushare-token",
        "TAVILY_API_KEY": "your-tavily-key"
      }
    }
  }
}

Prompts to try

Use these prompts with Claude to conduct financial research through the Finance MCP server.

- "Search for recent news about Zijin Mining and summarize the key developments"
- "Analyze the price-volume history for stock code 601899 over the past week"
- "Scrape and summarize the earnings report at https://example.com/earnings"
- "What is the shareholder structure and recent insider activity for stock 600519?"
- "Search for the latest analyst ratings and earnings forecasts for Kweichow Moutai"

Troubleshooting Finance MCP

Tools fail with missing API key errors

Each tool depends on specific environment variables. dashscope_search requires DASHSCOPE_API_KEY, history_calculate requires TUSHARE_API_TOKEN, and tavily_search requires TAVILY_API_KEY. Set only the keys for the tools you plan to use and disable the rest in the disabled_flows argument.

crawl_ths_* tools return empty results or IP block errors

The TongHuaShun crawler tools use Crawl4AI to scrape THS pages. High-frequency requests may result in IP throttling. Reduce request frequency and avoid running many crawl_ths_* calls in rapid succession.

Server fails to start with config parsing errors

Double-check the config=, mcp.transport=, and disabled_flows= arguments. The disabled_flows value must be a valid JSON array string. Use single quotes around the JSON on Linux/macOS to avoid shell escaping issues.

Frequently Asked Questions about Finance MCP

What is Finance MCP?

Finance MCP is a Model Context Protocol (MCP) server that llm-powered mcp server for building financial deep-research agents, integrating web search, crawl4ai scraping, and entity extraction into composable analysis flows. It connects AI assistants to external tools and data sources through a standardized interface.

How do I install Finance MCP?

Follow the installation instructions on the Finance MCP GitHub repository. Clone the repo, install dependencies, and add the server config to your AI client.

Which AI clients work with Finance MCP?

Finance MCP works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.

Is Finance MCP free to use?

Yes, Finance MCP is open source and available under the Apache-2.0 license. You can use it freely in both personal and commercial projects.

Browse More Finance & Fintech MCP Servers

Explore all finance & fintech servers available in the MCPgee directory. Each server includes setup guides for Claude, Cursor, and VS Code.

Quick Config Preview

{ "mcpServers": { "finance": { "command": "npx", "args": ["-y", "finance"] } } }

Add this to your claude_desktop_config.json or .cursor/mcp.json

Read the full setup guide →

Ready to use Finance MCP?

Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.

33,000+ ServersFree & Open SourceStep-by-Step Guides