CrewAI

CrewAI is a Python framework for developers automating work with teams of AI agents. It combines agents that make decisions with explicit workflow logic, so a process can include both autonomous work and predictable steps.

• Specialized agents: Give agents distinct roles, goals, tools and tasks, then coordinate their work through Crews.
• Controlled workflows: Use Flows to manage state, branching and routing, with human input and structured results alongside agent decisions.
• Memory and tools: Add knowledge, memory and checkpoints, and connect agents to tools and other agents through MCP and A2A.
• Python customization: Control prompts, agent behavior and execution logic within your application code.

The open-source framework uses the MIT license and runs on your own hardware or servers. It supports local inference through Ollama, but defaults to the OpenAI API, which sends model requests to OpenAI.

A separate commercial platform adds a visual builder that exports Python, managed deployment, tracing, cost tracking, access controls and approval gates. Runtime hooks support personal-data redaction and policy checks. The commercial platform has on-premise and cloud deployment options.

The framework collects anonymous usage telemetry, which you can disable. Enabling the optional share_crew feature also shares task descriptions and agent attributes such as backstories and goals.