Project Title

TradingAgents-CN — A Multi-Agent LLM-Based Chinese Financial Trading Framework

Overview

TradingAgents-CN is a Chinese financial trading decision framework based on multi-agent large language models, optimized for Chinese users and providing comprehensive analysis capabilities for A-shares, Hong Kong stocks, and US stocks. This project stands out for its full support for the Chinese market, integration of domestic large models, and a user-friendly interface that caters to the local audience.

Key Features

Native OpenAI support and Google AI integration
Custom endpoint configuration and intelligent model selection
Multi-LLM provider support and model selection persistence
Docker containerized deployment and professional report export
Complete A-share support and Chinese localization
FastAPI + Vue 3 architecture for enterprise-level performance
User permission management and configuration management center
Real-time notification system and batch analysis functionality
Smart stock screening and personalized stock management
Simulated trading system for strategy validation

Use Cases

Use case 1: Financial analysts and traders using the framework for in-depth analysis of Chinese and international stocks.
Use case 2: Researchers leveraging the framework's multi-agent LLM capabilities for financial market predictions and simulations.
Use case 3: Individual investors utilizing the framework for personalized stock analysis and portfolio management.

Advantages

Advantage 1: Tailored for the Chinese market with full A-share, Hong Kong stock, and US stock analysis capabilities.
Advantage 2: Integration of domestic large models and a user-friendly interface for local users.
Advantage 3: Enhanced performance and user experience with the new FastAPI + Vue 3 architecture.

Limitations / Considerations

Limitation 1: The project may require a steep learning curve for users unfamiliar with multi-agent LLM frameworks.
Limitation 2: The reliance on local data sources and models may limit its applicability in non-Chinese markets.

Project 1: TauricResearch/TradingAgents - The original multi-agent trading framework that TradingAgents-CN is based on, offering a broader range of features but without the Chinese market focus.
Project 2: QuantConnect/Lean - An open-source algorithmic trading engine that allows users to develop and backtest trading algorithms, differing in its focus on algorithmic trading rather than multi-agent LLMs.
Project 3: Zipline/Zipline - A Pythonic algorithmic trading simulator that is part of the Quantopian platform, offering a different approach to backtesting trading strategies.