Qwen Code | Qwen Cloud

Qwen Code is the official CLI tool for Qwen models, providing AI-powered coding assistance directly in your terminal. Connect it to Qwen Cloud's pay-as-you-go API for seamless access to Qwen3-Coder and other Qwen models.

Quick start

Get running in a few minutes:

# 1. Install
curl -fsSL https://qwen-code-assets.oss-cn-hangzhou.aliyuncs.com/installation/install-qwen.sh | bash
qwen --version

# 2. Configure (~/.qwen/settings.json)
{
  "env": {
    "QWENCLOUD_API_KEY": "sk-xxx"
  },
  "modelProviders": {
    "openai": [{
      "id": "qwen3.5-plus",
      "name": "[QwenCloud] qwen3.5-plus",
      "baseUrl": "https://dashscope-intl.aliyuncs.com/compatible-mode/v1",
      "envKey": "QWENCLOUD_API_KEY"
    }]
  }
}

# 3. Test (run qwen and ask)
qwen
Ask: "Write a Python function to reverse a string"

You should see: Qwen Code generates the string reversal function

Configuration

Basic setup

Configure Qwen Code to use Qwen Cloud:

API endpoint: https://dashscope-intl.aliyuncs.com/compatible-mode/v1
Authentication: API key required
Model selection: Configure multiple Qwen models

Free quota and billing:

First-time users get a free quota (valid for 90 days)
Enable Free quota only to prevent unexpected charges
Qwen OAuth provides 1,000 free calls/day (separate from Qwen Cloud quota)

For Coding Plan users: Use your exclusive base URL and API key instead. See Coding Plan for Qwen Code.

Installation

macOS/Linux
Windows

curl -fsSL https://qwen-code-assets.oss-cn-hangzhou.aliyuncs.com/installation/install-qwen.sh | bash
qwen --version

Open CMD as administrator:

curl -fsSL -o %TEMP%\install-qwen.bat https://qwen-code-assets.oss-cn-hangzhou.aliyuncs.com/installation/install-qwen.bat && %TEMP%\install-qwen.bat

Close CMD, reopen, and verify:

qwen --version

Authentication methods

API Key (Recommended)
Qwen OAuth (1000 free calls/day)

Edit ~/.qwen/settings.json (macOS/Linux) or C:\Users\username\.qwen\settings.json (Windows):

{
  "env": {
    "QWENCLOUD_API_KEY": "YOUR_API_KEY"
  },
  "modelProviders": {
    "openai": [
      {
        "id": "qwen3.5-plus",
        "name": "[QwenCloud] qwen3.5-plus",
        "baseUrl": "https://dashscope-intl.aliyuncs.com/compatible-mode/v1",
        "envKey": "QWENCLOUD_API_KEY",
        "generationConfig": {
          "extra_body": {
            "enable_thinking": true
          }
        }
      },
      {
        "id": "qwen3-coder-next",
        "name": "[QwenCloud] qwen3-coder-next",
        "baseUrl": "https://dashscope-intl.aliyuncs.com/compatible-mode/v1",
        "envKey": "QWENCLOUD_API_KEY"
      }
    ]
  }
}

Model configuration

Configure multiple models for different tasks:

Model	Best for	Configuration
`qwen3.5-plus`	General coding	Default, with thinking mode
`qwen3-coder-next`	Fast iterations	Lightweight, quick responses
`qwen3-coder-plus`	Complex logic	Deep code understanding
`qwen3-max`	Architecture	Advanced reasoning

Limitations

Token usage: Scans project files and conversation history
Context limit: Large projects may exceed token limits
File access: Cannot read files outside project directory

Examples

Basic usage

qwen
# Ask questions directly
"Explain this error: TypeError: 'NoneType' object is not iterable"

Code generation

"Create a FastAPI endpoint for user registration with email validation"

Code review

"Review this function for potential bugs and performance issues"

Refactoring

"Refactor this class to follow SOLID principles"

Testing

"Write comprehensive unit tests for the payment module"

Commands

Command	Description
`/auth`	Switch authentication method
`/model`	Change active model
`/compress`	Compress conversation history
`/clear`	Clear all history
`/settings`	View/edit settings
`/help`	Show available commands

Advanced features

Sub-agents

Qwen Code supports specialized sub-agents:

File Agent: Navigate and edit files
Web Agent: Search documentation
Terminal Agent: Execute commands

MCP support

Connect external tools via Model Context Protocol:

{
  "mcpServers": {
    "filesystem": {
      "command": "npx",
      "args": ["@modelcontextprotocol/server-filesystem"]
    }
  }
}

Custom skills

Create reusable workflows:

qwen /skill create "test-writer"
# Define the skill steps
qwen /skill use "test-writer"

Token optimization

Reduce consumption by:

Directory focus: Start Qwen Code in specific project folders
Settings optimization:

{
  "sessionTokenLimit": 32000
}

Regular compression: Use /compress to free tokens
Clear sessions: /clear between unrelated tasks
Model selection: Use lighter models for simple tasks

Troubleshooting

"401 Incorrect API key provided"

Solution:

Verify API key in settings.json

Check baseUrl matches your region

Ensure API key has quota

High token consumption

Solution:

Set sessionTokenLimit in settings

Work in focused directories

Use /compress regularly

Clear history with /clear

"Range of input length should be [1, xxx]"

Solution:

Reduce context with /compress

Lower sessionTokenLimit

Start new session with /clear

Model not responding

Solution:

Check /model to verify selection

Verify model in settings.json

Try switching to different model

FAQ

How to use 1,000 free calls/day?

Use /auth → Select "Qwen OAuth" → Authenticate via browser. This quota is separate from Qwen Cloud quota.

How to switch models?

Type /model to select from configured models, or add new models to settings.json.

How to reduce costs?

Use /compress and /clear regularly
Set sessionTokenLimit to limit context
Start Qwen Code in specific directories
Use appropriate models for tasks

Coding Plan: Setup with subscription →
Models & pricing: See pricing →
API docs: OpenAI-compatible reference →
Official docs: Qwen Code documentation →

​Quick start

​Configuration

​Basic setup

​Installation

​Authentication methods

​Model configuration

​Limitations

​Examples

​Basic usage

​Code generation

​Code review

​Refactoring

​Testing

​Commands

​Advanced features

​Sub-agents

​MCP support

​Custom skills

​Token optimization

​Troubleshooting

​FAQ

​How to use 1,000 free calls/day?

​How to switch models?

​How to reduce costs?

​Related resources

Quick start

Configuration

Basic setup

Installation

Authentication methods

Model configuration

Limitations

Examples

Basic usage

Code generation

Code review

Refactoring

Testing

Commands

Advanced features

Sub-agents

MCP support

Custom skills

Token optimization

Troubleshooting

FAQ

How to use 1,000 free calls/day?

How to switch models?

How to reduce costs?

Related resources