TradingAgents
Basic Information
TradingAgents is an open-source multi-agent framework for research into LLM-driven financial trading that models the structure and workflow of real trading firms. It composes specialized agents including fundamentals, sentiment, news, technical analysts, bullish and bearish researchers, a trader agent, and a risk management and portfolio manager to collaboratively analyze markets and propose trades. The project provides both a Python package interface and a command-line interface for selecting tickers, dates, models, and research depth. It is implemented with LangGraph for modularity and requires FinnHub for market data and OpenAI for LLMs. The framework supports configurable debate rounds, online tools or cached data from a curated Tauric TradingDB, simulated exchange execution, and explicit warnings that the code is for research and not financial advice.
Links
App Details
-
$$$$ -
$$$$