CrewAI
CrewAI is an open-source Python framework for building production-ready multi-agent workflows. It provides high-level abstractions for autonomous agent collaboration through Crews and event-driven control through Flows, solving the problem of orchestrating multiple AI agents for complex tasks. It is designed for developers who need to build and deploy agentic automations with flexibility and control.
🎬 Getting Started with CrewAI Open Source · CrewAI
✨ Key features
- Crews for autonomous role-based agent collaboration
- Flows for event-driven workflow control with state
- Seamless integration of Crews and Flows
- Python-native customization of prompts, tools, and logic
- Agent-ready capabilities: tools, memory, knowledge, checkpointing
- Production-ready patterns: human input, structured outputs
🎯 Use cases
- Automating research and report generation with multiple specialized agents
- Building event-driven workflows that combine AI agents with deterministic logic
- Creating trip planners, stock analysis tools, and job description writers
- Scaffolding and deploying production-grade multi-agent systems
📦 Installation
🧰 Requirements: Python >=3.10 and <3.14, UV package manager, and API keys for your LLM provider (e.g., OpenAI) and optionally Serper.dev for web search.
1. Installation
CrewAI requires Python >=3.10 and <3.14. Check your version with:
python3 --version
CrewAI uses UV for dependency management. Install UV first:
macOS/Linux:
curl -LsSf https://astral.sh/uv/install.sh | sh
If your system doesn't have curl, use wget:
wget -qO- https://astral.sh/uv/install.sh | sh
Windows:
powershell -ExecutionPolicy ByPass -c "irm https://astral.sh/uv/install.ps1 | iex"
Then install the CrewAI CLI:
uv tool install crewai
If you encounter a PATH warning, run:
uv tool update-shell
Verify the install:
uv tool list
You should see something like:
crewai v0.102.0
- crewai
To upgrade the global CLI later:
uv tool install crewai --upgrade
🚀 Usage
2. Setting Up Your Crew
Create a new crew project:
crewai create crew my_project
cd my_project
Edit the generated files to define agents and tasks. For example, set agents/researcher.jsonc:
{
"role": "{topic} Senior Data Researcher",
"goal": "Uncover cutting-edge developments in {topic}",
"backstory": "You're a seasoned researcher who finds relevant information and presents it clearly.",
"llm": "openai/gpt-4o",
"tools": ["SerperDevTool"],
"settings": {
"verbose": true
}
}
Set your API keys in .env:
OPENAI_API_KEY=your_key_here
SERPER_API_KEY=your_key_here
Then install dependencies and run:
crewai install
crewai run
You should see output in the console, and output/report.md will be created.
❓ FAQ
What Python versions are supported?
CrewAI requires Python >=3.10 and <3.14.
How do I install CrewAI?
Install UV, then run uv tool install crewai.
What is the difference between Crews and Flows?
Crews are teams of autonomous agents for collaborative tasks, while Flows are event-driven workflows for precise control. They can be combined.
How do I create a new crew project?
Run crewai create crew <project_name> to scaffold a project with agents and tasks in JSONC files.
📊 Repository
🤖 Overview, features, install steps and FAQ were generated from the project's README on Sep 4, 2026. Always check the original source before running commands.