AI Multi-Agent Framework
16.7k 2026-04-13
camel-ai/camel
An open-source multi-agent framework dedicated to researching the scaling laws of AI agents and facilitating large-scale agent system development.
Core Features
Evolvability: Enables continuous evolution of multi-agent systems through data generation and environment interaction.
Scalability: Designed to support millions of agents for efficient coordination and resource management.
Statefulness: Agents maintain memory for multi-step interactions and complex task handling.
Code-as-Prompt: Code and comments serve as prompts, ensuring interpretability for both humans and agents.
Community-driven Research: A collective of researchers advancing frontier research in Multi-Agent Systems.
Detailed Introduction
CAMEL is an open-source community and framework focused on discovering the scaling laws of AI agents. It provides a robust platform for implementing and supporting various types of agents, tasks, prompts, models, and simulated environments. By enabling the study of agents at scale, CAMEL offers critical insights into their behaviors, capabilities, and potential risks. Its design principles emphasize evolvability, scalability, statefulness, and a "Code-as-Prompt" paradigm, making it an ideal tool for frontier research and development in multi-agent systems.