# Qwen-Agent — Build AI Agents on Qwen Models

> Agent framework by Alibaba with function calling, code interpreter, RAG, and MCP support. Built for Qwen 3.0+. 15K+ stars.

## Install

Save as a script file and run:

# Qwen-Agent — Build AI Agents on Qwen Models

## Quick Use

```bash
pip install -U "qwen-agent[gui,rag,code_interpreter,mcp]"
```

```python
from qwen_agent.agents import Assistant

# Create an agent with tools
agent = Assistant(
    llm={"model": "qwen-max", "api_key": "YOUR_KEY"},
    function_list=["code_interpreter", "web_search"],
    system_message="You are a helpful data analyst."
)

# Run a task
messages = [{"role": "user", "content": "Analyze the trend of AI funding in 2025"}]
for response in agent.run(messages):
    print(response)
```

---

## Intro

Qwen-Agent is an open-source agent framework by the Alibaba Qwen team, designed to build LLM-powered applications on top of Qwen models (3.0+). With 15,800+ GitHub stars, it provides native function calling, a sandboxed code interpreter running in Docker, RAG pipelines for document understanding, MCP tool discovery, and a built-in Gradio GUI for rapid prototyping. It's the official way to build agents with Qwen — the leading open-weight LLM family from China.

Works with: Qwen models (qwen-max, qwen-plus, qwen-turbo), DashScope API, local Qwen deployments. Best for developers building AI agents on the Qwen ecosystem. Setup time: under 5 minutes.

---

## Qwen-Agent Architecture & Features

### Core Components

```
┌─────────────────────────────────────┐
│           Agent Layer               │
│  Assistant │ ReActChat │ Router     │
├─────────────────────────────────────┤
│           Tool Layer                │
│  Function Calling │ Code Interpreter│
│  Web Search │ MCP Tools │ Custom    │
├─────────────────────────────────────┤
│           LLM Layer                 │
│  Qwen-Max │ Qwen-Plus │ Local Qwen │
└─────────────────────────────────────┘
```

### Function Calling

Native function calling with automatic schema parsing:

```python
from qwen_agent.tools import BaseTool, register_tool

@register_tool("get_weather")
class WeatherTool(BaseTool):
    description = "Get weather for a city"
    parameters = [{"name": "city", "type": "string", "required": True}]

    def call(self, params, **kwargs):
        city = params["city"]
        return f"Weather in {city}: 25°C, sunny"
```

### Code Interpreter

Executes Python code in an isolated Docker sandbox:
- Matplotlib/Seaborn chart generation
- Pandas data analysis
- File I/O in sandboxed environment
- Automatic output capture and display

### RAG (Retrieval Augmented Generation)

Built-in document understanding pipeline:
- PDF, DOCX, HTML, Markdown parsing
- Chunking with configurable overlap
- Vector similarity search
- Citation tracking in responses

### MCP Support

Connect external tools via Model Context Protocol:

```python
agent = Assistant(
    llm={"model": "qwen-max"},
    mcp_servers=[{
        "url": "http://localhost:8080/mcp",
        "tools": ["search", "calculator"]
    }]
)
```

### Built-in GUI

Launch a Gradio chat interface instantly:

```python
from qwen_agent.gui import WebUI

WebUI(agent).run(server_port=7860)
```

### Multi-Agent Orchestration

Route tasks to specialized agents:

```python
from qwen_agent.agents import GroupChat, Router

agents = [coder_agent, researcher_agent, writer_agent]
router = Router(llm=llm, agents=agents)
```

---

## FAQ

**Q: What is Qwen-Agent?**
A: Qwen-Agent is an open-source Python framework by Alibaba for building AI agents on Qwen models, featuring function calling, code interpretation, RAG, and MCP support. 15,800+ GitHub stars.

**Q: Can I use Qwen-Agent with non-Qwen models?**
A: It's optimized for Qwen models but supports any OpenAI-compatible API endpoint. Best results come from Qwen-Max and Qwen-Plus which have native function calling.

**Q: Is Qwen-Agent free?**
A: Yes, Apache-2.0 licensed. The Qwen models themselves have varying licenses — Qwen 3.0 is Apache-2.0 for most sizes.

---

## Source & Thanks

> Created by [QwenLM](https://github.com/QwenLM) (Alibaba Qwen Team). Licensed under Apache-2.0.
>
> [Qwen-Agent](https://github.com/QwenLM/Qwen-Agent) — ⭐ 15,800+

Thanks to the Qwen team at Alibaba for open-sourcing both the models and the agent framework.

---

<!-- ZH -->

## 快速使用

```bash
pip install -U "qwen-agent[gui,rag,code_interpreter,mcp]"
```

```python
from qwen_agent.agents import Assistant

agent = Assistant(
    llm={"model": "qwen-max", "api_key": "YOUR_KEY"},
    function_list=["code_interpreter", "web_search"],
    system_message="你是一个数据分析师。"
)

messages = [{"role": "user", "content": "分析 2025 年 AI 融资趋势"}]
for response in agent.run(messages):
    print(response)
```

---

## 简介

Qwen-Agent 是阿里巴巴通义千问团队开源的 AI 代理框架，基于 Qwen 模型（3.0+）构建。拥有 15,800+ GitHub stars，提供原生函数调用、Docker 沙盒代码解释器、RAG 文档理解管线、MCP 工具发现和内置 Gradio 界面。这是在千问生态中构建 AI Agent 的官方方式。

适用于：Qwen 系列模型（qwen-max、qwen-plus、qwen-turbo）、DashScope API、本地 Qwen 部署。适合在千问生态中开发 AI 代理的开发者。

---

## 核心功能

### 函数调用
原生函数调用，自动解析工具 schema，支持自定义工具注册。

### 代码解释器
在隔离的 Docker 沙盒中执行 Python 代码，支持数据分析和图表生成。

### RAG 文档理解
内置文档解析、分块、向量检索和引用追踪。

### MCP 支持
通过模型上下文协议连接外部工具和数据源。

### 多 Agent 协作
支持 GroupChat 和 Router 模式，将任务路由到专门的代理。

---

## 来源与感谢

> Created by [QwenLM](https://github.com/QwenLM) (Alibaba Qwen Team). Licensed under Apache-2.0.
>
> [Qwen-Agent](https://github.com/QwenLM/Qwen-Agent) — ⭐ 15,800+


---
Source: https://tokrepo.com/en/workflows/d95df987-a5d5-4578-90d7-45e4c8b34f93
Author: TokRepo精选