GPTeam
ActiveDescription
An open-source multi-agent simulation platform where multiple LLM-powered agents collaborate on complex tasks in shared environments, with customizable roles, memory systems, and environment interaction for studying multi-agent collaboration.
Key Features
- GPT-4 powered multi-agent simulation with autonomous collaboration on predefined goals
- Agent memory and reflection system inspired by Generative Agents research paper
- Spatial movement: agents move between locations and perform tasks in different environments
- Communication as a tool enabling agents to speak and collaborate in parallel
- Configurable world with customizable agents and locations via config.json
- Discord integration for observing agent interactions in real-time
Use Cases
💡 Studying emergent multi-agent collaboration behaviors and social dynamics
💡 Researching how LLM-powered agents communicate and coordinate tasks
💡 Simulating team workflows with autonomous agents in shared workspaces
💡 Exploring agent memory and reflection patterns in collaborative settings
💡 Building prototypes for multi-agent productivity tools and communication systems
Categories
Quick Start
```bash
git clone https://github.com/101dotxyz/GPTeam.git
cd gpteam
python setup.py
# Edit .env with your OpenAI API key
poetry run world
```
Use `poetry run world --turbo` for cheaper runs with gpt-3.5-turbo. Visit the `agents/` folder to see agent state summaries during the simulation.