TradingAgents is an open-source multi-agent framework that uses specialized LLM-powered agents to simulate real-world trading firms. The system includes analyst teams (fundamentals, sentiment, news, technical), researcher teams for debate-based decision making, trader agents, and risk management components. Built with LangGraph and designed for research purposes, it requires OpenAI and FinnHub APIs and offers both CLI and Python package interfaces for testing trading strategies.

6m read timeFrom github.com
Post cover image
Table of contents
TradingAgents FrameworkInstallation and CLITradingAgents PackageContributingCitation

Sort: