--- title: "CrewAI Documentation" description: "Build collaborative AI agents, crews, and flows — fast." mode: "center" ---
CrewAI

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

Get started Installation Flows Agents Tools API Reference
Create your first crew in minutes. Learn the core runtime, project layout, and dev loop. Install via UV and set up environment keys for OpenAI and tools. Create, run, train, test, deploy, and manage configs — all from the CLI. Compose robust agents with memory, knowledge, and tools. Structure outputs with Pydantic. Design complex workflows with start/listen/router, state, persistence, and resumability. Short‑term, long‑term, and entity memory. Built‑in knowledge with vector storage. Provider‑neutral client. ChromaDB by default, Qdrant supported. Configure per use case. Traces with batching, metrics, and integrations (Langfuse, Phoenix, and more). Deploy, automate with triggers, manage orgs, and run at scale. CrewAI — multi‑agent orchestration
Compose agents, tasks, flows, and tools — production‑ready by default.
## Popular paths - Getting started: [/en/quickstart](/en/quickstart) - Install + requirements: [/en/installation](/en/installation) - Define crews via YAML: [/en/guides/crews/first-crew](/en/guides/crews/first-crew) - Build flows: [/en/guides/flows/first-flow](/en/guides/flows/first-flow) - API Reference: [/en/api-reference/introduction](/en/api-reference/introduction) Looking for examples? Check out the examples repo and cookbooks under /en/examples/cookbooks. If CrewAI helps you ship faster, give us a star and join the community. Ask questions, share feedback, and see what others are building.