What is TradingAgents?
TradingAgents is an open-source multi-agent LLM financial trading framework for research and development of algorithmic trading strategies.It orchestrates specialized LLM agents—fundamental analysts, sentiment experts, technical analysts, traders, and risk managers—to produce structured outputs for research managers, traders, and portfolio managers.
The framework provides multi-provider LLM support and a unified model catalog (GPT-5.x, Gemini 3.x, Claude 4.x, Grok 4.x, DeepSeek, Qwen, GLM, Azure OpenAI) for model selection and comparison.Key features include backtesting with date fidelity, LangGraph checkpoint and resume, persistent decision and memory logs, and a five-tier model rating system to support reproducible experiments.
Deployment and integration options include a CLI, Docker support, cross-platform compatibility, proxy support, and extensible provider hooks.Target users include quantitative researchers, algorithmic traders, data scientists, ML engineers, and fintech development teams building LLM-driven trading workflows.
TradingAgents user reviews
Based on 1 review, 100.0% of users recommend TradingAgents, rated highly for ease of use.
Liked for
Would you recommend TradingAgents?
TradingAgents's key features
-
Multi-agent orchestration of specialized LLMs (fundamental, sentiment, technical, trader, risk) producing structured outputs
-
Multi-provider LLM support with unified model catalog for model selection and comparison
-
Backtesting with date fidelity
-
LangGraph checkpoint/resume and persistent decision and memory logs
-
Deployment and integration via CLI, Docker, cross-platform compatibility, proxy support, and extensible provider hooks
TradingAgents use cases
-
Create reproducible, date-fidelity backtests for LLM-driven trading strategies using tradingagents' specialized analyst, trader and risk-manager agents, persisting decision logs and LangGraph checkpoints for auditability and research
-
Develop multi-model comparisons and hyperparameter optimization across the unified model catalog to identify best-performing LLM combinations, run experiments via CLI/Docker and rank strategies by risk-adjusted metrics
-
Deploy containerized, multi-agent trading workflows that automate signal generation, execution and risk control, maintain persistent logs for compliance, and iterate quickly on strategies using backtesting and reproducible checkpoints
Who is it for?
-
Data scientists
-
Research managers
-
Portfolio managers
-
Algorithmic traders
-
Quantitative researchers
