CAMEL-AI
What is CAMEL-AI?
CAMEL-AI is an open-source agent framework for teams building single-agent and multi-agent workflows in code that supports role-based orchestration, data generation pipelines, and stateful runtime components. Its CAMEL Toolkit includes Workforce, Connect to RL, Evolvability, and MCP, plus the CAMEL Toolkit for agents, tools, memories, storage, and data loaders. The ecosystem is used by Amazon, Apple, DeepMind, and Bytedance, and it is self-hostable with docs at docs.camel-ai.org.
Last verifiedHow we evaluate
At a glance
- CAMEL-AI is best for teams building agent workflows that need scalable orchestration and stateful behavior.
What does CAMEL-AI do?
CAMEL-AI runs multi-agent workflows by combining role-based agent orchestration, data generation pipelines, and a toolkit for building agent behaviors in code. Its framework shows Code-as-Prompt, so teams define tools, interactions, and task logic in type-safe code instead of brittle prompt templates. The platform also supports single-agent and multi-agent patterns, plus memory, storage, and runtime components that keep agents stateful across longer jobs. At scale, CAMEL-AI is built around distributed workforces: the site describes thousands of agents working in parallel and positions the project around data generation, world simulation, and task automation. The community reports 42.5k stars, 200+ contributors, 30K community members, and 4K+ forks, while the ecosystem includes research work used by Amazon, Apple, Meta, Tesla, MIT, Stanford, and Oxford. It is open source and self-hostable, with docs available at docs.camel-ai.org for teams that want to build on their own infrastructure.
Why use CAMEL-AI?
- Stateful memory and storage let agents carry context across longer workflows instead of restarting from scratch.
- The framework is designed for distributed workforces, so teams can scale from one agent to thousands in parallel.
- A large research and contributor community provides visible momentum and a broad base of examples and extensions.
Who is CAMEL-AI for?
- ML engineers who want to orchestrate single-agent and multi-agent workflows in code.
- Research teams who need agent frameworks for data generation and world simulation.
- Applied AI developers who want memory, tools, and runtime components in one stack.
- Open-source contributors who prefer a self-hostable framework with active community support.
What are CAMEL-AI's key features?
Workforce
Build multi-agent teams for task automation, backed by 42.5k GitHub stars and 200+ contributors for a mature agent framework.
CAMEL Toolkit
Use the CAMEL Toolkit to assemble agents, tools, memories, storage, and data loaders in one framework for faster system design.
Connect to RL
Connect agents to reinforcement learning workflows, letting teams train and evaluate behaviors inside the same CAMEL environment.
Research Ecosystem
Tap a research ecosystem used by 100+ researchers and 30K community members, with docs at docs.camel-ai.org for shared patterns and examples.
Evolvability
Design agents that can change over time with code-as-prompt, verifier, and runtime components that support iterative updates and checks.
Scalability
Run from single-agent to multi-agent setups, with support for one million agents and integrations like OpenAI, Azure, and AWS Bedrock.
Human in the Loop
Insert human review into agent workflows so people can approve, correct, or steer actions before they continue through tools like Slack or Zapier.
MCP
Connect through MCP and related tools such as Playwright MCP and Minimax MCP to extend agents across external systems and browser tasks.
What does CAMEL-AI integrate with?
- Arxiv
- Bohrium
- Crawl4AI
- Dappier
- Dingtalk
- EdgeOne Pages
- Excel
- GitHub
- Gmail
- Google Calendar
- Google Drive
- Google Maps
- Google Scholar
- IMAP Mail
- Jina Reranker
- Klavis
- Lark
- MCP
- Meshy
- MinerU
- Minimax MCP
- Notion
- Open APIOpenAI Image
- OpenBB
- Origene
- Outlook Mail
- Playwright MCP
- PubMed
- Pulse MCP Search
What are CAMEL-AI's use cases?
ML engineers orchestrate agents
ML engineers who want to orchestrate single-agent and multi-agent workflows in code use CAMEL-AI to build task pipelines that can branch, coordinate, and hand off work. They lean on Workforce for structured agent execution and the CAMEL Toolkit to wire tools, memory, and runtime components into a reproducible stack.
Research teams generate data
Research teams who need agent frameworks for data generation and world simulation use CAMEL-AI to spin up controlled experiments and synthetic workflows. They combine Research Ecosystem with Data Generation and Environments to produce testable scenarios, then use Verifier to check outputs before analysis.
Applied AI apps with memory
Applied AI developers who want memory, tools, and runtime components in one stack use CAMEL-AI to ship assistants that remember context and call external systems. They rely on Memories, Tools, and Runtime to keep interactions stateful, while MCP helps connect the agent to existing services.
Self-hosted open-source agents
Open-source contributors who prefer a self-hostable framework with active community support use CAMEL-AI to extend agent behavior without locking into a closed platform. They use Evolvability and Scalability to adapt the framework over time, and Human in the Loop to keep critical actions reviewable.
How does CAMEL-AI work?
- Connect your first model or service through MCP, then choose whether to start with Single-agent or Multi-agent execution so the workflow matches your task complexity.
- Add Tools, Memories, and Storage from the CAMEL Toolkit to give agents context, external actions, and persistent state across turns and sessions.
- Define your workflow in Workforce, then use Runtime and Interpreters to run tasks, branch subtasks, and keep agent behavior reproducible in code.
- Test outputs with Verifier and Observe, watching traces in Overview so you can catch failures early and tune prompts, tools, or handoffs.
- Expand into Research Ecosystem or Connect to RL when you need data generation, simulation, or reinforcement learning loops, then keep improving with Human in the Loop.
Frequently asked questions
What is CAMEL-AI?
CAMEL-AI is an open-source agent framework for teams building single-agent and multi-agent workflows in code that supports role-based orchestration, data generation pipelines, and stateful runtime components. Its CAMEL Toolkit includes Workforce, Connect to RL, Evolvability, and MCP, and the ecosystem is used by Amazon, Apple, and DeepMind. It is self-hostable and backed by docs.camel-ai.org.
What is CAMEL-AI used for? Who is it for?
CAMEL-AI is used for Workforce, CAMEL Toolkit, and Connect to RL. It's built for ML engineers, Research teams, and Applied AI developers.
Does CAMEL-AI have an API and what does it integrate with?
CAMEL-AI doesn't publish a public API. It integrates with Arxiv, Bohrium, Crawl4AI, Dappier, Dingtalk, and 25 more.
Explore other Agent Frameworks
