Alibaba Cloud Model Studio:Qwen Code

Preparations

• Get an API key: Get an API key before you start.

• Check the Node.js version: Qwen Code requires Node.js version 20 or later. Run node -v to check the version. If the version is earlier than 20, you must reinstall it.

Install and configure Qwen Code

1. Install Qwen Code

   Run one of the following commands in your terminal to install Qwen Code.

   Install using npm (Recommended)

   npm install -g @qwen-code/qwen-code@latest

   Install from source

   git clone https://github.com/QwenLM/qwen-code.git
   cd qwen-code
   npm install
   npm install -g .

2. Start and configure Qwen Code

   1. Start Qwen Code

      Enter qwen in your terminal to start Qwen Code.

   2. Select an authentication method

      This example uses the OpenAI-compatible authentication provided by Model Studio. Select OpenAI and press Enter.

      

      Enter the following information on the page:

      Item	Description
      API Key	Your Model Studio API key.
      Base URL	Singapore: https://dashscope-intl.aliyuncs.com/compatible-mode/v1 Virginia: https://dashscope-us.aliyuncs.com/compatible-mode/v1 Beijing: https://dashscope.aliyuncs.com/compatible-mode/v1
      Model	The name of the model. Qwen Code is specifically optimized for the Qwen3-Coder series and is compatible with models that provide an OpenAI compatible API, such as Qwen-Max, Qwen-Plus, Qwen-Flash, and DeepSeek. We recommend models with strong coding capabilities, such as: qwen3-coder-plus qwen3-coder-480b-a35b-instruct qwen3-coder-flash qwen3-coder-30b-a3b-instruct

      To avoid entering this information every time you start Qwen Code, export environment variables as described below.

      Export environment variables

      Use global environment variables or a .env file. If both are configured, global environment variables take precedence over the .env file.

      .env file

      Create a .env file in the project root directory and add the following content. These environment variables apply only to the current project.

      # Replace YOUR_DASHSCOPE_API_KEY with your Model Studio API key
      OPENAI_API_KEY="YOUR_DASHSCOPE_API_KEY"
      # The following is the URL for the Singapore region. Different regions use separate base_url
      OPENAI_BASE_URL="https://dashscope-intl.aliyuncs.com/compatible-mode/v1"
      # This example uses the qwen3-coder-plus model. You can replace it with another model as needed.
      OPENAI_MODEL="qwen3-coder-plus"

      Global environment variables

      Set your Model Studio API key, base URL, and model as global environment variables. These variables apply to all projects and terminal sessions.

      macOS

      1. Run the following command in the terminal to check your default shell type.

         echo $SHELL

      2. Set the environment variables based on your shell type.

         Zsh

         # Replace YOUR_DASHSCOPE_API_KEY with your Model Studio API key
         # The following is the URL for the Singapore region. Different regions use separate base_url
         # This example uses the qwen3-coder-plus model. You can replace it with another model as needed.
         echo 'export OPENAI_API_KEY="YOUR_DASHSCOPE_API_KEY"' >> ~/.zshrc
         echo 'export OPENAI_BASE_URL="https://dashscope-intl.aliyuncs.com/compatible-mode/v1"' >> ~/.zshrc
         echo 'export OPENAI_MODEL="qwen3-coder-plus"' >> ~/.zshrc

         Bash

         # Replace YOUR_DASHSCOPE_API_KEY with your Model Studio API key
         # The following is the URL for the Singapore region. Different regions use separate base_url
         # This example uses the qwen3-coder-plus model. You can replace it with another model as needed.
         echo 'export OPENAI_API_KEY="YOUR_DASHSCOPE_API_KEY"' >> ~/.bash_profile
         echo 'export OPENAI_BASE_URL="https://dashscope-intl.aliyuncs.com/compatible-mode/v1"' >> ~/.bash_profile
         echo 'export OPENAI_MODEL="qwen3-coder-plus"' >> ~/.bash_profile

      3. Run the following command in the terminal to apply the environment variables.

         Zsh

         source ~/.zshrc

         Bash

         source ~/.bash_profile

      Windows

      CMD

      1. Run the following commands in CMD to set the environment variables.

         # Replace YOUR_DASHSCOPE_API_KEY with your Model Studio API key
         # The following is the URL for the Singapore region. Different regions use separate base_url
         # This example uses the qwen3-coder-plus model. You can replace it with another model as needed.
         setx OPENAI_API_KEY "YOUR_DASHSCOPE_API_KEY"
         setx OPENAI_BASE_URL "https://dashscope-intl.aliyuncs.com/compatible-mode/v1"
         setx OPENAI_MODEL "qwen3-coder-plus"

      2. Open a new CMD window and run the following commands to check if the environment variables are set correctly.

         echo %OPENAI_API_KEY%
         echo %OPENAI_BASE_URL%
         echo %OPENAI_MODEL%

      PowerShell

      1. Run the following commands in PowerShell to set the environment variables.

         # Replace YOUR_DASHSCOPE_API_KEY with your Model Studio API key
         # The following is the URL for the Singapore region. Different regions use separate base_url
         # This example uses the qwen3-coder-plus model. You can replace it with another model as needed.
         [Environment]::SetEnvironmentVariable("OPENAI_API_KEY", "YOUR_DASHSCOPE_API_KEY", [EnvironmentVariableTarget]::User)
         [Environment]::SetEnvironmentVariable("OPENAI_BASE_URL", "https://dashscope-intl.aliyuncs.com/compatible-mode/v1", [EnvironmentVariableTarget]::User)
         [Environment]::SetEnvironmentVariable("OPENAI_MODEL", "qwen3-coder-plus", [EnvironmentVariableTarget]::User)

      2. Open a new PowerShell window and run the following commands to check if the environment variables are set correctly.

         echo $env:OPENAI_API_KEY
         echo $env:OPENAI_BASE_URL
         echo $env:OPENAI_MODEL

Q: Why does Qwen Code consume tokens quickly?

A: Qwen Code may call the API multiple times, which can consume many tokens. Use the following methods to control token consumption:

• Streamline the working directory

  Start Qwen Code in a specific project directory. Having too many files in the startup directory, such as the root directory, can increase token consumption.

• Set a token limit

  Create a .qwen/settings.json file in the project root directory and restart Qwen Code. Use sessionTokenLimit to control the token usage for a single API call:

  {
    "sessionTokenLimit": 32000
  }

  Qwen Code may call the API multiple times, which can cause the token consumption for a single question to exceed the sessionTokenLimit value.

• Use compression/purge commands

  Enter the following commands to reduce token consumption:

  • /compress

    Compresses the conversation history to continue the dialogue within the token limit.

  • /clear

    Purges all conversation history and starts a new session.

Q: How to switch models?

A: The method depends on whether you have set environment variables:

• If you have not configured environment variables

  1. Enter /quit to exit Qwen Code.

  2. Enter qwen to start Qwen Code, select the OpenAI authentication method, and then enter the API key, base URL, and target model name.

• If you have configured environment variables

  1. Change the OPENAI_MODEL environment variable to the target model and apply the change. For more information, see Export environment variables.

  2. Enter /quit to exit Qwen Code, and then enter qwen to restart it.

Q: Why do I get a 401 Incorrect API key provided error?

A: This error may be caused by an incorrect API configuration. Compare the API key in the error message with your settings to ensure they match. Global environment variable configurations take precedence over environment file configurations. To use the configuration from the environment file, ensure that the relevant global environment variables are not set or clear them temporarily.

For more information, see Qwen Code troubleshooting guide.