Python agent framework
CrewAI
A Python framework for orchestrating role based autonomous agents in Crews and event driven Flows.
Quick decision
What it is
CrewAI is described in its repository as an open source Python framework with high level abstractions and low level APIs for building production ready multi agent workflows. It gives developers autonomous agent collaboration through Crews and precise event driven control through Flows.
Crews are teams of role based agents with defined goals and expertise that collaborate and delegate tasks. Flows provide event driven automations with fine grained control, state management, conditional branching and clean integration with production Python code, and Flows can run Crews natively.
The project also offers a commercial AMP suite and Crew Control Plane for managed deployment, observability, governance and enterprise support, while the open source framework remains MIT licensed and installable from PyPI. The repository states Python 3.10 to less than 3.14 is required.
What it does
- Orchestrates role based agents with goals, backstories, tools, memory and guardrails in Crews.
- Builds event driven Flows with precise workflow control, state management and conditional branching.
- Ships a large built in tools collection through crewai-tools for search, scraping, files, databases and automation.
- Supports structured task output and sequential or hierarchical crew processes.
- Offers an optional commercial control plane for tracing, observability and managed deployment.
What it does not do
- It is not a model provider; agents require a configured LLM provider and API key.
- It is not independent evidence of any third party tool service quality; tool API keys and accounts are still required.
- The open source framework alone does not include the commercial AMP control plane features described for enterprises.
Setup
Create and run a CrewAI project
The repository installation flow installs the CrewAI CLI with uv, scaffolds a crew project, then installs and runs it after keys are set in the environment file.
uv tool install crewai
crewai create crew <project_name>
crewai install
crewai runEnvironment variables (names only)
OPENAI_API_KEYSERPER_API_KEY
Source: https://github.com/crewAIInc/crewAI. Examples use placeholders only.
MCP support: CrewAI exposes MCP integration through MCPServerAdapter and MCPNativeTool, listed in the crewai-tools tool set in fetched package and repository material.
Health and maintenance
59,431
Checked 2026-10-07
8,660
Checked 2026-10-07
Stars and forks read from the GitHub repository page in this run. PyPI page shows release files for crewai 1.15.25.
Verified compatible tools
Via crewai_tools FirecrawlScrapeWebsiteTool (crewai[tools]). The CrewAI docs page documents FirecrawlScrapeWebsiteTool, requires FIRECRAWL_API_KEY and installs the Firecrawl SDK with the crewai[tools] package. Evidence. Setup detail
Via crewai_tools TavilySearchTool. The CrewAI docs page documents TavilySearchTool as an interface to the Tavily Search API for CrewAI agents and shows installation with crewai[tools] and tavily-python. Evidence. Setup detail
Via crewai_tools EXASearchTool. The crewai-tools PyPI page lists EXASearchTool under API Integrations in the CrewAI tools collection. Evidence. Setup detail
Via crewai_tools StagehandTool. The crewai-tools PyPI page lists StagehandTool among AI powered tools in the CrewAI tools collection. Evidence. Setup detail
Via crewai_tools SeleniumScrapingTool. The crewai-tools PyPI page lists SeleniumScrapingTool under Web Scraping in the CrewAI tools collection. Evidence. Setup detail
Via crewai_tools BraveSearchTool. A fetched CrewAI repository issue prints the crewai_tools import list and includes BraveSearchTool, BrowserbaseLoadTool, ApifyActorsTool and ComposioTool. Evidence. Setup detail
Via crewai_tools BrowserbaseLoadTool. The same fetched CrewAI repository issue import list includes BrowserbaseLoadTool in crewai_tools. Evidence. Setup detail
Via crewai_tools ComposioTool. The fetched CrewAI repository issue import list includes ComposioTool in crewai_tools. Evidence. Setup detail
Grouped view: tools for CrewAI.
Sources and freshness
- GitHub repository: https://github.com/crewAIInc/crewAI
- PyPI package: https://pypi.org/project/crewai/
- crewai-tools PyPI listing: https://pypi.org/project/crewai-tools/1.9.3/
Last checked 2026-10-07. Verification label: source verified. It does not mean AgentsUse ran the framework. Spotted an error? Send a correction. Back to the framework index.