What is CAMEL?

CAMEL is an open-source library for studying autonomous and communicative AI agents, designed to facilitate research into their behaviors, capabilities, and risks.

How does CAMEL achieve autonomous cooperation among agents?

CAMEL uses a novel role-playing framework with inception prompting to guide chat agents towards task completion while maintaining consistency with human intentions.

What is Inception Prompting?

Inception prompting guides chat agents towards task completion by providing initial instructions and context, allowing them to generate subsequent prompts and guide the conversation towards a desired outcome.

What is an MCP Server and how does CAMEL interact with it?

An MCP (Model Context Protocol) server acts as a bridge, allowing AI models to access and interact with external data sources and tools. CAMEL is designed to be compatible with MCP servers, enabling AI agents to leverage real-world data for complex tasks.

How can I install CAMEL?

You can install the base CAMEL library using pip: pip install camel-ai. Extra dependencies are needed for some features, such as Hugging Face agents or RAG, and can be installed using pip install camel-ai[huggingface-agent] or pip install camel-ai[tools] respectively.

Can I use open-source models with CAMEL?

Yes, CAMEL supports open-source models as backends, using an external server running LLM inference service (e.g., FastChat). The server must support OpenAI-compatible APIs, especially the method openai.ChatCompletion.create.

What types of models are supported by CAMEL?

CAMEL supports LLaMA2-based models (e.g., meta-llama/Llama-2-7b-chat-hf) and Vicuna-based models (e.g., lmsys/vicuna-7b-v1.5).

How does CAMEL integrate with UBOS.Tech?

UBOS.Tech is a full-stack AI Agent Development Platform that provides the infrastructure and tools necessary to build, train, and deploy CAMEL-based AI agents. UBOS accelerates development, enhances scalability, and simplifies management of CAMEL agents.

What are the licensing terms for CAMEL?

The source code is licensed under Apache 2.0. The datasets are licensed under CC BY NC 4.0, which permits only non-commercial usage. Models trained using the dataset should not be utilized for anything other than research purposes.

How can I contribute to CAMEL?

Contributions to CAMEL are welcome! Please refer to the contributing guidelines on the CAMEL GitHub repository for detailed instructions.