# CrewAI review

> CrewAI is an open-source Python framework for building multi-agent workflows, using role-based agent Crews and event-driven Flows.

Devtoolwatch Board, the week of Oct 25–31, 2026. Canonical page: https://index.devtoolwatch.com/tools/crewai/

## TL;DR

CrewAI is an open-source Python framework for orchestrating AI agents that work together on tasks. It is for Python developers building multi-agent automations. It offers two complementary approaches: Crews, for autonomous role-based collaboration, and Flows, for event-driven control with state and branching.

## CrewAI at a glance

- **Category:** [Agent frameworks](https://index.devtoolwatch.com/categories/agent-frameworks/)
- **Maintained by:** crewAI (organization, [@crewAIInc](https://github.com/crewAIInc) on GitHub)
- **License:** MIT
- **Language:** Python, JavaScript, CSS
- **Repository:** [github.com/crewAIInc/crewAI](https://github.com/crewAIInc/crewAI)
- **Website:** [crewai.com](https://crewai.com)
- **PyPI package:** [crewai](https://pypi.org/project/crewai/) 1.15.28
- **Latest release:** [1.15.28](https://github.com/crewAIInc/crewAI/releases/tag/1.15.28) on GitHub, Oct 14, 2026
- **First published:** Oct 27, 2023

## What CrewAI does

- CrewAI Crews model teams of specialized AI agents with roles, goals, backstories, tools, and tasks that collaborate and delegate work.
- CrewAI Flows build event-driven workflows with state, conditional branching, routing, single LLM calls, and native support for Crews.
- CrewAI runs crew tasks in a sequential process or a hierarchical process, where an automatically assigned manager delegates and validates work.
- CrewAI agents can use tools, memory, knowledge, checkpointing, async execution, and MCP/A2A support.
- CrewAI supports human input, structured outputs, and deterministic steps in workflows.
- CrewAI includes a crewai eval command, which can compare models with --models and report per-run cost and time.

### How it works

CrewAI is installed as a CLI through uv. The command crewai create crew scaffolds a project with agents in agents/*.jsonc and tasks in crew.jsonc. You add API keys to .env, then run crewai install and crewai run. Each agent names its LLM, such as openai/gpt-4o, and its tools. The optional commercial AMP Suite adds managed deployment, tracing, and governance.

## How to install CrewAI

**Install uv (macOS/Linux)**

```
curl -LsSf https://astral.sh/uv/install.sh | sh
```

**Install the CrewAI CLI with uv**

```
uv tool install crewai
```

### Requirements

- Python >=3.10 and <3.14
- The uv package manager for dependency management
- An API key for your model provider
- A Serper.dev API key if your agents use web search

### First steps

1. Create a JSON-first crew project with crewai create crew, or add --classic for the older Python/YAML scaffold.
2. Edit agents/*.jsonc to define each agent's role, goal, backstory, LLM, and tools, and define tasks in crew.jsonc.
3. Add your model provider API key to the .env file, plus SERPER_API_KEY if you use web search.
4. Run crewai install and then crewai run from the project directory to execute the crew.

## Who CrewAI is for

- Python developers building multi-agent workflows where role-based agents collaborate on tasks
- Teams that need event-driven automations with precise control over state and branching
- Enterprises that want managed deployment, tracing, and governance through the commercial AMP Suite

### Limitations

- CrewAI is a Python framework and supports only Python versions from 3.10 up to, but not including, 3.14.
- CrewAI's managed deployment, observability, governance, and enterprise support are part of the commercial AMP Suite, not the open-source framework.
- On Windows, CrewAI installs can fail with a chroma-hnswlib build error unless Visual Studio Build Tools with C++ are installed.

## Usage data (the week of Oct 25–31, 2026)

- Board rank: #307 overall, #31 in Agent frameworks, momentum score 30
- Downloads that week: 589.4k (−6% vs 4-week average)
- New GitHub stars per week (last 12 weeks): Aug 9: 337; Aug 16: 375; Aug 23: 354; Aug 30: 337; Sep 6: 328; Sep 13: 386; Sep 20: 290; Sep 27: 284; Oct 4: 263; Oct 11: 297; Oct 18: 300; Oct 25: 319

## CrewAI alternatives

| Rank | Tool | Stars | Downloads/wk |
|---|---|---|---|
| #1 | [Harness²](https://index.devtoolwatch.com/tools/google-research-harness2/) | 21.7k | – |
| #2 | [HyperFrames](https://index.devtoolwatch.com/tools/heygen-com-hyperframes/) | 76.6k | 1.5M |
| #3 | [AgentScope](https://index.devtoolwatch.com/tools/agentscope-ai-agentscope/) | 35k | 51.4k |
| #4 | [AG-UI](https://index.devtoolwatch.com/tools/ag-ui/) | 16.9k | 4.6M |
| #5 | [LangGraph](https://index.devtoolwatch.com/tools/langgraph/) | 44.2k | 18.5M |
| #6 | [Bifrost](https://index.devtoolwatch.com/tools/maximhq-bifrost/) | 9.3k | – |

## CrewAI FAQ

### Is CrewAI free and open source?

Yes. CrewAI's source code is on GitHub (crewAIInc/crewAI) under the MIT license, so you can use it for free. Its website also links to a pricing page (https://crewai.com/pricing) for any paid plans.

### How do I install CrewAI?

Run `curl -LsSf https://astral.sh/uv/install.sh | sh`. The crewai package needs Python 3.10 or newer, below 3.14. The install section above has the other options from the project's README.

### Is CrewAI still maintained?

Yes. The crewAIInc/crewAI repository was last updated on Oct 21, 2026, and the latest release, 1.15.28 on GitHub, came out on Oct 14, 2026.

### How popular is CrewAI?

CrewAI has 60.4k GitHub stars as of Nov 1, 2026. It gained 319 new stars in the week of Oct 25–31, 2026, against a 12-week average of 323 a week. Its main package had 589.4k downloads that week, −6% against its own 4-week average. That made it #31 among AI agent frameworks on the Devtoolwatch Board.

### CrewAI vs LangGraph: how do they compare?

CrewAI and LangGraph are both tracked by Devtoolwatch as AI agent frameworks. CrewAI has 60.4k GitHub stars and LangGraph has 44.2k. In the week of Oct 25–31, 2026, CrewAI's main package had 589.4k downloads and LangGraph's had 18.5M. LangGraph is a low-level orchestration framework and runtime for building, managing, and deploying long-running, stateful agents in Python or JavaScript. [LangGraph review](https://index.devtoolwatch.com/tools/langgraph/)

### CrewAI vs LangChain: how do they compare?

CrewAI and LangChain are both tracked by Devtoolwatch as AI agent frameworks. CrewAI has 60.4k GitHub stars and LangChain has 148.8k. In the week of Oct 25–31, 2026, CrewAI's main package had 589.4k downloads and LangChain's had 41.1M. [LangChain review](https://index.devtoolwatch.com/tools/langchain/)

### What are the best alternatives to CrewAI?

Other AI agent frameworks on the Devtoolwatch Board in the week of Oct 25–31, 2026, by momentum: Harness² (#1), HyperFrames (#2), AgentScope (#3), AG-UI (#4), LangGraph (#5). The alternatives table above has their numbers.

### What is the difference between Crews and Flows in CrewAI?

CrewAI Crews are teams of autonomous, role-based agents that delegate and collaborate on tasks. CrewAI Flows are event-driven workflows that give fine-grained control over execution paths, state, and conditional branching. You can combine them by running Crews inside Flows.

### Is CrewAI no-code?

The CrewAI open-source framework is code-first and Python-based. CrewAI's enterprise platform advertises a no-code visual editor that exports to Python, alongside a code-first API.

### How do I choose a model in CrewAI?

CrewAI sets the model per agent through the llm field in the agent's JSONC file, for example "openai/gpt-4o". You put your model provider API key in the project's .env file. Recent releases added native Gemini 3.8 Flash support and deepinfra as an OpenAI-compatible provider.

### Can coding agents help me write CrewAI code?

CrewAI publishes official CrewAI Skills that teach coding agents to scaffold Flows, configure Crews, and design agents and tasks. Claude Code installs them as a plugin. Cursor, Codex, Windsurf, and others use npx skills add crewaiinc/skills.

### Can I still use YAML to define CrewAI agents?

CrewAI supports YAML through the classic scaffold. Run crewai create crew --classic to get crew.py, config/agents.yaml, and config/tasks.yaml. The default scaffold is JSON-first.

## Sources

1. [crewAIInc/crewAI on GitHub](https://github.com/crewAIInc/crewAI)
2. [CrewAI README](https://github.com/crewAIInc/crewAI/blob/main/README.md)
3. [crewai on PyPI](https://pypi.org/project/crewai/)
4. [CrewAI website](https://crewai.com)
5. [Release 1.15.28 (Oct 14, 2026)](https://github.com/crewAIInc/crewAI/releases/tag/1.15.28)
6. [Release 1.15.27 (Oct 9, 2026)](https://github.com/crewAIInc/crewAI/releases/tag/1.15.27)
7. [Release 1.15.26 (Oct 8, 2026)](https://github.com/crewAIInc/crewAI/releases/tag/1.15.26)
8. [Devtoolwatch Board, the week of Oct 25–31, 2026](https://index.devtoolwatch.com/)