TradingAgents

What it is

TradingAgents is a multi-agent LLM financial trading framework designed to simulate the operations of real-world trading firms. It utilizes specialized agents for various roles such as fundamental analysis, sentiment analysis, and trading decisions.

How it works

The framework decomposes trading tasks into specialized roles:

• Analyst Team: Includes Fundamentals Analyst, Sentiment Analyst, News Analyst, and Technical Analyst.
• Researcher Team: Comprises bullish and bearish researchers who assess insights from the Analyst Team.
• Trader Agent: Compiles reports to make informed trading decisions.
• Risk Management and Portfolio Manager: Evaluates portfolio risks and approves transactions.

Getting started

Installation

Clone TradingAgents:

git clone https://github.com/TauricResearch/TradingAgents.git
cd TradingAgents

Create a virtual environment:

conda create -n tradingagents python=3.12
conda activate tradingagents

Install the package:

pip install .

Docker

Run with Docker:

docker compose run --rm tradingagents

For local models with Ollama:

docker compose --profile ollama run --rm tradingagents-ollama

Required APIs

Set the API key for your chosen provider:

export OPENAI_API_KEY=...

CLI Usage

Launch the interactive CLI:

tradingagents          # installed command
python -m cli.main     # alternative: run directly from source

Recent releases

• v0.3.1 (2026-07-05): Correctness and stability patch, including Alpha Vantage look-ahead filter.
• v0.3.0 (2026-06-22): Stabilization and extensibility release with expanded provider registry.
• v0.2.5 (2026-05-11): Introduced grounded Sentiment Analyst and Qwen/GLM/MiniMax dual-region support.
• v0.2.4 (2026-04-25): Added structured-output decision agents and Docker support.
• v0.2.3 (2026-03-29): Multi-language output and improved backtesting fidelity.

Traction

• Stars: 92,376
• Forks: 17,852

Caveats

• License: Apache-2.0
• Open issues: 292
• Created: 2024-12-28
• Last push: 2026-07-05