Qwen Code

Qwen Code can integrate with CoAIWork through a custom OpenAI Provider.

Configuration Values

Item	Value
baseUrl	https://coaiwork.com/v1
Request path relative to baseUrl	/chat/completions
Default model	gpt-5.4
apiKey	CoAIWork API token
Protocol	OpenAI Compatible

Because baseUrl already includes /v1, Qwen Code requests /chat/completions. The full URL is https://coaiwork.com/v1/chat/completions.

Environment	Qwen Code config path
macOS	~/.qwen/settings.json
Linux	~/.qwen/settings.json
Windows	$HOME\.qwen\settings.json

Recommended ~/.qwen/settings.json

{
  "model": "gpt-5.4",
  "modelProvider": "coaiwork",
  "modelProviders": {
    "coaiwork": {
      "name": "CoAIWork",
      "baseUrl": "https://coaiwork.com/v1",
      "envKey": "OPENAI_API_KEY",
      "authType": "openai"
    }
  }
}

Warning

OPENAI_API_KEY must be set as an environment variable. Do not hard-code the real token in settings.json. envKey only tells Qwen Code which environment variable name to read.

macOS

npm install -g @qwen-code/qwen-code@latest
echo 'export OPENAI_API_KEY="<YOUR_TOKEN>"' >> ~/.zshrc
source ~/.zshrc
mkdir -p ~/.qwen

Save the configuration to ~/.qwen/settings.json, then run:

qwen

Linux

npm install -g @qwen-code/qwen-code@latest
echo 'export OPENAI_API_KEY="<YOUR_TOKEN>"' >> ~/.bashrc
source ~/.bashrc
mkdir -p ~/.qwen

Save the configuration to ~/.qwen/settings.json, then run:

qwen

Windows

npm install -g @qwen-code/qwen-code@latest
[Environment]::SetEnvironmentVariable("OPENAI_API_KEY", "<YOUR_TOKEN>", "User")
New-Item -ItemType Directory -Force $HOME\.qwen | Out-Null

Save the configuration to $HOME\.qwen\settings.json, then run:

qwen