CrewAI

CrewAI is an open-source Python framework for building multi-agent systems, where autonomous agents are given roles, goals, and tools and then collaborate to complete a task. Its core abstraction is the Crew — a team of agents that delegate work to each other — layered on top of Flows, an event-driven mechanism for precise, deterministic control over how tasks are sequenced and when agents are invoked. Unlike some early agent libraries, CrewAI is built from scratch as a standalone framework with no LangChain dependency, which its maintainers cite as the reason for its lean footprint.

The framework is model- and tool-agnostic: it works with any LLM provider, ships built-in memory and checkpointing, and supports the Model Context Protocol (MCP) for connecting external tools and data sources. It has become one of the most widely adopted agent frameworks in the Python ecosystem, with broad enterprise use and a large community of certified developers running agentic automations in production.

CrewAI is distributed under the MIT license and is free to use commercially. Alongside the library, CrewAI, Inc. offers AMP, a managed platform that adds a visual Studio for composing crews, GitHub integration, one-click deployment, and execution monitoring, billed by workflow execution volume with a free entry tier and custom enterprise pricing. Teams typically prototype with the open-source library and adopt the platform when they need hosted deployment, observability, and collaboration around production agent workflows.