yfin-mcp
An enhanced Yahoo Finance MCP server with pagination, caching, and LLM-optimized responses for stock data, financials, options, and analyst info.
README
yfin-mcp - Yahoo Finance MCP Server
Enhanced fork of yahoo-finance-mcp by Alex2Yang97
With intelligent pagination, caching, and LLM-optimized responses
<div align="right"> <a href="README.md">English</a> | <a href="README.zh.md">δΈζ</a> </div>
A high-performance Model Context Protocol (MCP) server that provides comprehensive financial data from Yahoo Finance with intelligent pagination, caching, and LLM-optimized responses.
π Attribution & Motivation
Original Work
This project is built upon the excellent foundation of yahoo-finance-mcp created by Alex2Yang97. The original implementation provides a comprehensive set of tools for accessing Yahoo Finance data through the Model Context Protocol.
Why These Enhancements?
While using the original implementation with LLM agents (Claude, ChatGPT, etc.), I encountered critical limitations:
Problem 1: Context Window Overflow
- Historical stock data with
period="max"could return thousands of rows - Option chains for popular stocks contained hundreds of contracts
- LLM context windows (typically 200K tokens) would overflow
- Responses would be truncated, losing critical data
Problem 2: No Data Persistence
- Large datasets couldn't be saved for offline analysis
- Repeated queries wasted API calls and time
- No way to export data for use in other tools
Problem 3: Poor LLM Readability
- JSON responses were hard for LLMs to parse when truncated
- No clear navigation guidance for paginated data
- Cache status was invisible to the LLM
Solution: Pagination, Caching & Export
This fork adds three key enhancements:
-
Token-Based Pagination (6,000 token limit per page)
- Prevents context window overflow
- Clear navigation guidance for LLMs
- Dynamic page sizing based on data complexity
-
Intelligent Caching (TTL: 5min-1hr)
- Reduces redundant API calls by 50-80%
- Sub-millisecond response times for cached data
- Automatic cache invalidation based on data volatility
-
JSON Export (File Download)
- Save full datasets for offline analysis
- Export to Excel, databases, or other tools
- Preserve complete data without pagination
Credit Where Credit is Due
Original Author: Alex2Yang97 deserves full credit for:
- β Complete Yahoo Finance API integration
- β All 9 MCP tools implementation
- β Robust error handling
- β Comprehensive documentation
This Fork Adds: Pagination, caching, and export features to make the server production-ready for LLM agents handling large financial datasets.
Demo

MCP Tools
The server exposes the following tools through the Model Context Protocol:
Stock Information
| Tool | Description |
|---|---|
get_historical_stock_prices |
Get historical OHLCV data for a stock with customizable period and interval |
get_stock_info |
Get comprehensive stock data including price, metrics, and company details |
get_yahoo_finance_news |
Get latest news articles for a stock |
get_stock_actions |
Get stock dividends and splits history |
Financial Statements
| Tool | Description |
|---|---|
get_financial_statement |
Get income statement, balance sheet, or cash flow statement (annual/quarterly) |
get_holder_info |
Get major holders, institutional holders, mutual funds, or insider transactions |
Options Data
| Tool | Description |
|---|---|
get_option_expiration_dates |
Get available options expiration dates |
get_option_chain |
Get options chain for a specific expiration date and type (calls/puts) |
Analyst Information
| Tool | Description |
|---|---|
get_recommendations |
Get analyst recommendations or upgrades/downgrades history |
Real-World Use Cases
With this MCP server, you can use Claude to:
Stock Analysis
- Price Analysis: "Show me the historical stock prices for AAPL over the last 6 months with daily intervals."
- Financial Health: "Get the quarterly balance sheet for Microsoft."
- Performance Metrics: "What are the key financial metrics for Tesla from the stock info?"
- Trend Analysis: "Compare the quarterly income statements of Amazon and Google."
- Cash Flow Analysis: "Show me the annual cash flow statement for NVIDIA."
Market Research
- News Analysis: "Get the latest news articles about Meta Platforms."
- Institutional Activity: "Show me the institutional holders of Apple stock."
- Insider Trading: "What are the recent insider transactions for Tesla?"
- Options Analysis: "Get the options chain for SPY with expiration date 2024-06-21 for calls."
- Analyst Coverage: "What are the analyst recommendations for Amazon over the last 3 months?"
Investment Research
- "Create a comprehensive analysis of Microsoft's financial health using their latest quarterly financial statements."
- "Compare the dividend history and stock splits of Coca-Cola and PepsiCo."
- "Analyze the institutional ownership changes in Tesla over the past year."
- "Generate a report on the options market activity for Apple stock with expiration in 30 days."
- "Summarize the latest analyst upgrades and downgrades in the tech sector over the last 6 months."
Requirements
- Python 3.11 or higher
- Dependencies as listed in
pyproject.toml, including:- mcp
- yfinance
- pandas
- pydantic
- and other packages for data processing
Installation
From PyPI (Recommended)
Install the package directly from PyPI:
pip install yfin-mcp
From Source
-
Clone this repository:
git clone https://github.com/fritzprix/yahoo-finance-mcp.git cd yahoo-finance-mcp -
Create and activate a virtual environment and install dependencies:
uv venv source .venv/bin/activate # On Windows: .venv\Scripts\activate uv pip install -e .
Usage
Integration with Claude for Desktop
After installing the package, you can integrate it with Claude for Desktop:
-
Install the package (if not already installed):
pip install yfin-mcp -
Configure Claude Desktop:
- MacOS: Edit
~/Library/Application Support/Claude/claude_desktop_config.json - Windows: Edit
%APPDATA%\Claude\claude_desktop_config.json
- MacOS: Edit
-
Add the server configuration:
Using uvx (Recommended - No installation needed):
{ "mcpServers": { "yfinance": { "command": "uvx", "args": ["yfin-mcp"] } } }Using Python directly (if installed via pip):
{ "mcpServers": { "yfinance": { "command": "python", "args": ["-m", "yfin_mcp"] } } }For development/source installation:
- macOS:
{ "mcpServers": { "yfinance": { "command": "uv", "args": [ "--directory", "/ABSOLUTE/PATH/TO/yahoo-finance-mcp", "run", "server.py" ] } } } - Windows:
{ "mcpServers": { "yfinance": { "command": "uv", "args": [ "--directory", "C:\\ABSOLUTE\\PATH\\TO\\yahoo-finance-mcp", "run", "server.py" ] } } }
- macOS:
-
Restart Claude for Desktop
Development Mode
For testing with MCP Inspector:
# From source
uv run yfin-mcp
# Or if installed via pip
python -m yfin_mcp
Publishing to PyPI
To build and publish the package, use the provided scripts. You can optionally provide an argument to bump the version:
Windows
# Just build and publish current version
publish_package.bat
# Bump version and then publish
publish_package.bat patch
publish_package.bat minor
publish_package.bat major
macOS/Linux
chmod +x publish_package.sh
# Just build and publish current version
./publish_package.sh
# Bump version and then publish
./publish_package.sh patch
./publish_package.sh minor
./publish_package.sh major
[!NOTE] The scripts will build the package into the
dist/directory and then usetwineto upload it. Ensure you have your PyPI credentials configured in~/.pypirc(or%HOME%\.pypircon Windows) or set theTWINE_PASSWORDenvironment variable.
Troubleshooting
[ERROR] os error 32: Process cannot access the file
If you see this error when running publish_package.bat or uv build, it means the MCP server is still running and locking the executable.
- Close Claude Desktop or any app using the yfinance MCP server.
- Stop any running MCP Inspector instances.
- If the error persists, manually kill the processes:
taskkill /F /IM yfin-mcp.exe /T taskkill /F /IM python.exe /T
License
MIT License
Original Work: Copyright (c) 2025 AlexYoung
Fork Enhancements: Copyright (c) 2026 SKTelecom
This project maintains the MIT License from the original yahoo-finance-mcp project. All enhancements (pagination, caching, export) are also released under MIT License.
See LICENSE file for full details.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.