CAMEL is an open-source community focused on discovering scaling laws for agents through large-scale studies. The framework supports various agents, tasks, prompts, models, and simulated environments to facilitate research. CAMEL's design principles emphasize evolvability, scalability, statefulness, and code-as-prompt for effective agent interaction. Researchers use CAMEL to simulate large-scale agent systems, enable dynamic communication, and equip agents with stateful memory. It supports multiple benchmarks, diverse agent types, and automates data generation, integrating seamlessly with various tools. CAMEL enables building data generation systems, task automation tools, and world simulations. Getting started is easy with pip install camel-ai, followed by setting up the OpenAI API key. The CAMEL tech stack includes modules for agents, agent societies, data generation, models, and memory. CAMEL encourages research contributions and provides synthetic datasets with visualizations of instructions and tasks. The platform offers practical cookbooks for building agents and multi-agent systems, including advanced features like RAG and knowledge graphs.