Framework
The CAMEL Multi-Agent Framework
A comprehensive framework designed for building scalable, evolvable, and stateful multi-agent systems with code-as-prompt architecture
Evolvability
Agents can dynamically evolve their capabilities through data generation, fine-tuning, and continuous learning processes.
Scalability
From single agents to massive workforces with thousands of agents working in parallel across distributed systems.
Statefulness
Built-in memory systems including vector databases, long-term memory, and contextual state management for persistent agent behavior.
Code-as-Prompt
Define agent behaviors, tools, and interactions using clean, type-safe code rather than brittle prompt templates.
Ready to Get Started?
Explore our comprehensive documentation and start building with CAMEL today.