CrewAI home pagelight logodark logo
  • Start Cloud Trial
  • crewAIInc/crewAI
  • crewAIInc/crewAI
Documentation
Enterprise
API Reference
Examples
Changelog
  • Website
  • Forum
  • Blog
  • CrewGPT
  • Get Started
    • Introduction
    • Installation
    • Quickstart
    Guides
    • Strategy
    • Agents
    • Crews
    • Flows
    • Advanced
    Core Concepts
    • Agents
    • Tasks
    • Crews
    • Flows
    • Knowledge
    • LLMs
    • Processes
    • Collaboration
    • Training
    • Memory
    • Reasoning
    • Planning
    • Testing
    • CLI
    • Tools
    • Event Listeners
    MCP Integration
    • MCP Servers as Tools in CrewAI
    • Stdio Transport
    • SSE Transport
    • Streamable HTTP Transport
    • Connecting to Multiple MCP Servers
    • MCP Security Considerations
    Tools
    • Tools Overview
    • File & Document
    • Web Scraping & Browsing
    • Search & Research
    • Database & Data
    • AI & Machine Learning
    • Cloud & Storage
    • Integrations
    • Automation
    Observability
    • Overview
    • Arize Phoenix
    • LangDB Integration
    • Langfuse Integration
    • Langtrace Integration
    • Maxim Integration
    • MLflow Integration
    • Neatlogs Integration
    • OpenLIT Integration
    • Opik Integration
    • Patronus AI Evaluation
    • Portkey Integration
    • Weave Integration
    • TrueFoundry Integration
    Learn
    • Overview
    • Strategic LLM Selection Guide
    • Conditional Tasks
    • Coding Agents
    • Create Custom Tools
    • Custom LLM Implementation
    • Custom Manager Agent
    • Customize Agents
    • Image Generation with DALL-E
    • Force Tool Output as Result
    • Hierarchical Process
    • Human Input on Execution
    • Kickoff Crew Asynchronously
    • Kickoff Crew for Each
    • Connect to any LLM
    • Using Multimodal Agents
    • Replay Tasks from Latest Crew Kickoff
    • Sequential Processes
    • Using Annotations in crew.py
    Telemetry
    • Telemetry

    CrewAI Documentation

    Build collaborative AI agents, crews, and flows — fast.

    CrewAI

    Orchestrate multi‑agent systems and flows with guardrails, memory, knowledge, tools, and observability. Production‑ready by default.

    Get startedInstallationFlowsAgentsToolsAPI Reference

    Quickstart

    Create your first crew in minutes. Learn the core runtime, project layout, and dev loop.

    Installation

    Install via UV and set up environment keys for OpenAI and tools.

    CLI

    Create, run, train, test, deploy, and manage configs — all from the CLI.

    Agents

    Compose robust agents with memory, knowledge, and tools. Structure outputs with Pydantic.

    Flows

    Design complex workflows with start/listen/router, state, persistence, and resumability.

    Memory & Knowledge

    Short‑term, long‑term, and entity memory. Built‑in knowledge with vector storage.

    RAG & Vector Stores

    Provider‑neutral client. ChromaDB by default, Qdrant supported. Configure per use case.

    Observability

    Traces with batching, metrics, and integrations (Langfuse, Phoenix, and more).

    Enterprise

    Deploy, automate with triggers, manage orgs, and run at scale.
    CrewAI — multi‑agent orchestration

    ​
    Popular paths

    • Getting started: /en/quickstart
    • Install + requirements: /en/installation
    • Define crews via YAML: /en/guides/crews/first-crew
    • Build flows: /en/guides/flows/first-flow
    • API Reference: /en/api-reference/introduction
    Looking for examples? Check out the examples repo and cookbooks under /en/examples/cookbooks.

    Star us on GitHub

    If CrewAI helps you ship faster, give us a star and join the community.

    Join the community

    Ask questions, share feedback, and see what others are building.

    Was this page helpful?

    websitexgithublinkedinyoutubereddit
    Powered by Mintlify
    Assistant
    Responses are generated using AI and may contain mistakes.