🤖 Qwen Code CLI Wrapper

Transform Qwen Code CLI into OpenAI-compatible endpoints using Cloudflare Workers. Access advanced AI capabilities through a standardized API interface, powered by OAuth2 authentication and seamless integration with the Qwen Code ecosystem.

✨ Features

🔐 OAuth2 Authentication - Uses your Qwen Code CLI credentials seamlessly
🎯 OpenAI-Compatible API - Drop-in replacement for OpenAI endpoints
📚 OpenAI SDK Support - Works with official OpenAI SDKs and libraries
🌐 Third-party Integration - Compatible with Open WebUI, Cline, and more
🛡️ API Key Security - Optional authentication layer for endpoint access
⚡ Cloudflare Workers - Global edge deployment with low latency
🔄 Smart Token Management - Automatic token refresh with KV storage
📡 Real-time Streaming - Server-sent events for live responses
🏗️ Clean Architecture - Well-structured, maintainable codebase
📊 Debug Logging - Comprehensive logging for troubleshooting

🚀 Quick Start

Prerequisites

Qwen Account with Code CLI access
Cloudflare Account with Workers enabled
Wrangler CLI installed (npm install -g wrangler)

Step 1: Get OAuth2 Credentials

You need OAuth2 credentials from the official Qwen Code CLI.

Using Qwen Code CLI

Install Qwen Code CLI:

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

Start Qwen Code and authenticate:
```
qwen
```
Select your preferred authentication method when prompted.

Locate the credentials file:

Windows:

C:\Users\USERNAME\.qwen\oauth_creds.json

macOS/Linux:

~/.qwen/oauth_creds.json

Copy the credentials: The file contains JSON in this format:

{
  "access_token": "your_access_token_here",
  "refresh_token": "your_refresh_token_here",
  "expiry_date": 1700000000000,
  "resource_url": "https://your-endpoint.com/v1",
  "token_type": "Bearer"
}

Step 2: Create KV Namespace

# Create a KV namespace for token caching
wrangler kv namespace create "QWEN_KV"

Note the namespace ID returned and update wrangler.toml:

kv_namespaces = [
  { binding = "QWEN_KV", id = "your-kv-namespace-id" }
]

Step 3: Environment Setup

Create a .dev.vars file:

# Required: Qwen Code CLI authentication JSON
QWEN_CLI_AUTH={"access_token":"your_access_token","refresh_token":"your_refresh_token","expiry_date":1700000000000,"resource_url":"https://your-endpoint.com/v1","token_type":"Bearer"}

# Optional: API key for client authentication (if set, users must provide it in Authorization header)
# OPENAI_API_KEY=sk-your-secret-key-here

# Optional: Default model override
# OPENAI_MODEL=qwen3-coder-plus

# Optional: Custom base URL (will use resource_url from OAuth if available)
# OPENAI_BASE_URL=https://api-inference.modelscope.cn/v1

For production, set the secrets:

wrangler secret put QWEN_CLI_AUTH
# Enter your OAuth credentials JSON

Step 4: Deploy

# Install dependencies
npm install

# Deploy to Cloudflare Workers
npm run deploy

# Or run locally for development
npm run dev

The service will be available at https://your-worker.your-subdomain.workers.dev

🔧 Configuration

Environment Variables

Core Configuration

Variable	Required	Description
`QWEN_CLI_AUTH`	✅	OAuth2 credentials JSON from Qwen Code CLI
`OPENAI_API_KEY`	❌	API key for client authentication
`OPENAI_MODEL`	❌	Default model override
`OPENAI_BASE_URL`	❌	Custom base URL (uses OAuth resource_url if available)

Authentication Security

When OPENAI_API_KEY is set, all /v1/* endpoints require authentication
Clients must include the header: Authorization: Bearer <your-api-key>
Recommended format: sk- followed by a random string
Without this variable, endpoints are publicly accessible (not recommended for production)

OAuth Token Management

Automatic Refresh: Tokens are automatically refreshed when expired
KV Persistence: Refreshed tokens are stored in Cloudflare KV
Fallback Logic: KV cache → environment → refresh → retry
Debug Logging: Comprehensive token source tracking

KV Namespaces

Binding	Purpose
`QWEN_KV`	OAuth token caching and refresh storage

🎯 API Endpoints

Base URL

https://your-worker.your-subdomain.workers.dev

OpenAI-Compatible Endpoints

Chat Completions

POST /v1/chat/completions
Authorization: Bearer sk-your-api-key-here (if OPENAI_API_KEY is set)
Content-Type: application/json

{
  "model": "qwen3-coder-plus",
  "messages": [
    {
      "role": "system",
      "content": "You are a helpful coding assistant."
    },
    {
      "role": "user",
      "content": "Write a Python function to calculate fibonacci numbers"
    }
  ],
  "stream": true,
  "temperature": 0.7,
  "max_tokens": 1000
}

List Models

GET /v1/models
Authorization: Bearer sk-your-api-key-here (if OPENAI_API_KEY is set)

Response:

{
  "object": "list",
  "data": [
    {
      "id": "qwen3-coder-plus",
      "object": "model",
      "created": 1700000000,
      "owned_by": "qwen"
    },
    {
      "id": "qwen3-coder-flash",
      "object": "model",
      "created": 1700000000,
      "owned_by": "qwen"
    }
  ]
}

Utility Endpoints

Health Check

GET /health

No authentication required

Response:

{
  "status": "ok",
  "uptime": 1700000000,
  "version": "qwen-worker-1.0.0"
}

💻 Usage Examples

OpenAI SDK (Python)

from openai import OpenAI

# Initialize with your worker endpoint
client = OpenAI(
    base_url="https://your-worker.workers.dev/v1",
    api_key="sk-your-secret-api-key-here"  # Only if OPENAI_API_KEY is set
)

# Chat completion
response = client.chat.completions.create(
    model="qwen3-coder-plus",
    messages=[
        {"role": "system", "content": "You are a helpful coding assistant."},
        {"role": "user", "content": "Write a binary search algorithm in Python"}
    ],
    temperature=0.2,
    max_tokens=500,
    stream=True
)

for chunk in response:
    if chunk.choices[0].delta.content:
        print(chunk.choices[0].delta.content, end="")

OpenAI SDK (JavaScript/TypeScript)

import OpenAI from 'openai';

const openai = new OpenAI({
  baseURL: 'https://your-worker.workers.dev/v1',
  apiKey: 'sk-your-secret-api-key-here', // Only if OPENAI_API_KEY is set
});

const stream = await openai.chat.completions.create({
  model: 'qwen3-coder-plus',
  messages: [
    { role: 'user', content: 'Explain async/await in JavaScript' }
  ],
  stream: true,
  temperature: 0.7,
});

for await (const chunk of stream) {
  const content = chunk.choices[0]?.delta?.content || '';
  process.stdout.write(content);
}

cURL Examples

# Chat completion (non-streaming)
curl -X POST https://your-worker.workers.dev/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer sk-your-secret-api-key-here" \
  -d '{
    "model": "qwen3-coder-plus",
    "messages": [
      {"role": "user", "content": "Hello! How are you?"}
    ],
    "temperature": 0.7
  }'

# Chat completion (streaming)
curl -N -X POST https://your-worker.workers.dev/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer sk-your-secret-api-key-here" \
  -d '{
    "model": "qwen3-coder-flash",
    "messages": [
      {"role": "user", "content": "Write a TypeScript hello world"}
    ],
    "stream": true
  }'

# List available models
curl https://your-worker.workers.dev/v1/models \
  -H "Authorization: Bearer sk-your-secret-api-key-here"

# Health check
curl https://your-worker.workers.dev/health

Open WebUI Integration

Add as OpenAI-compatible endpoint:
- Base URL: https://your-worker.workers.dev/v1
- API Key: sk-your-secret-api-key-here (only if OPENAI_API_KEY is set)
Auto-discovery: Open WebUI will automatically discover available models through the /v1/models endpoint.

🏗️ How It Works

graph TD
    A[Client Request] --> B[Cloudflare Worker]
    B --> C[API Key Validation]
    C --> D{Valid API Key?}
    D -->|No| E[401 Unauthorized]
    D -->|Yes| F{Token in KV Cache?}
    F -->|Yes| G[Use Cached Token]
    F -->|No| H[Check Environment Token]
    H --> I{Token Valid?}
    I -->|Yes| J[Cache & Use Token]
    I -->|No| K[Refresh Token via Qwen API]
    K --> L[Cache New Token]
    G --> M[Call Qwen API]
    J --> M
    L --> M
    M --> N{Streaming?}
    N -->|Yes| O[Forward SSE Stream]
    N -->|No| P[Return JSON Response]
    O --> Q[Client Receives Stream]
    P --> Q

The wrapper acts as a secure translation layer, managing OAuth2 authentication automatically while providing OpenAI-compatible responses.

🚨 Troubleshooting

Common Issues

401 Authentication Error

Verify your OPENAI_API_KEY is correctly set (if using API key auth)
Check if client is sending Authorization: Bearer <key> header
Ensure the API key format is valid

OAuth Token Issues

Check if your QWEN_CLI_AUTH credentials are valid
Ensure the refresh token hasn't expired
Verify the JSON format matches the expected structure

KV Storage Issues

Confirm KV namespace is correctly configured in wrangler.toml
Check KV namespace permissions in Cloudflare dashboard
Verify the binding name matches (QWEN_KV)

Streaming Problems

Check the worker logs for streaming-related errors
Ensure the upstream Qwen API supports streaming for the requested model
Verify the resource_url in your OAuth credentials

Debug Logging

The worker provides comprehensive debug logging:

# Run locally to see logs
npm run dev

Look for these log patterns:

=== New chat completion request ===
Environment loaded: { hasKv: true, hasCliAuth: true, ... }
Authentication passed
loadInitialCredentials called with json: present
Token validity check: { hasAccessToken: true, expiryDate: ..., tokenValid: true }
Base URL resolved: https://your-endpoint.com/v1
Making upstream request...
Captured usage in stream: {...}

Debug Endpoints

# Health check with detailed info
curl https://your-worker.workers.dev/health

# Check if KV credentials are loaded (logs will show)
curl https://your-worker.workers.dev/v1/models

🧠 Architecture

Project Structure

src/
├── types/                 # TypeScript type definitions
│   ├── bindings.ts       # Cloudflare bindings
│   ├── openai.ts         # OpenAI API types
│   ├── qwen.ts           # Qwen-specific types
│   └── common.ts         # Shared utilities
├── config/               # Configuration management
│   ├── constants.ts      # App constants
│   ├── validation.ts     # Request validation
│   └── index.ts          # Config exports
├── services/             # Business logic services
│   ├── qwenOAuthKvClient.ts    # OAuth client with KV storage
│   ├── qwenProxy.ts            # HTTP proxy to Qwen API
│   ├── openaiMapper.ts         # Request/response mapping
│   ├── auth.ts                 # API key authentication
│   └── credentials.ts          # Legacy (can be removed)
├── routes/               # HTTP route handlers
│   ├── chat.ts          # Chat completions endpoint
│   ├── health.ts        # Health check endpoint
│   └── models.ts        # Models listing endpoint
└── index.ts             # Hono bootstrap

Key Components

OAuth Client (services/qwenOAuthKvClient.ts)
- Manages Qwen OAuth tokens with KV persistence
- Automatic token refresh when expired
- Bootstrap from environment credentials
Request Mapping (services/openaiMapper.ts)
- Transforms OpenAI requests to Qwen-compatible format
- Maps sampling parameters (temperature, top_p, etc.)
- Validates request structure
HTTP Proxy (services/qwenProxy.ts)
- Calls Qwen API with proper authentication
- Resolves base URL from OAuth resource_url
- Handles both streaming and non-streaming responses
Authentication (services/auth.ts)
- Optional API key validation for endpoint access
- Bearer token format validation

🤝 Contributing

Fork the repository: https://github.com/gewoonjaap/qwen-code-cli-wrapper
Create a feature branch: git checkout -b feature-name
Make your changes and add tests
Run linting: npm run lint
Test thoroughly: npm test
Commit your changes: git commit -am 'Add feature'
Push to the branch: git push origin feature-name
Submit a pull request

Development Setup

git clone https://github.com/gewoonjaap/qwen-code-cli-wrapper.git
cd qwen-code-cli-wrapper
npm install
cp .dev.vars.example .dev.vars
# Edit .dev.vars with your Qwen OAuth credentials
npm run dev

Available Scripts

npm run dev          # Start development server
npm run deploy       # Deploy to Cloudflare Workers
npm run lint         # Run ESLint and TypeScript checks
npm run format       # Format code with Prettier
npm test             # Run test suite
npm run build        # Build the project

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

🙏 Acknowledgments

Built on Cloudflare Workers
Uses Hono web framework
Inspired by the official Qwen Code CLI
OAuth patterns adapted from Qwen Code CLI implementation

⚠️ Important: This project uses Qwen's API which may have usage limits and terms of service. Please ensure compliance with Qwen's policies when using this wrapper.

Name		Name	Last commit message	Last commit date
Latest commit History 4 Commits
.github		.github
docs		docs
src		src
test		test
.dev.vars.example		.dev.vars.example
.editorconfig		.editorconfig
.gitattributes		.gitattributes
.gitignore		.gitignore
.prettierrc		.prettierrc
Dockerfile		Dockerfile
README.md		README.md
docker-compose.yml		docker-compose.yml
eslint.config.mjs		eslint.config.mjs
package.json		package.json
start.sh		start.sh
tsconfig.json		tsconfig.json
vitest.config.mts		vitest.config.mts
worker-configuration.d.ts		worker-configuration.d.ts
wrangler.toml		wrangler.toml
yarn.lock		yarn.lock

Uh oh!

GewoonJaap/qwen-code-cli-wrapper

Folders and files

Latest commit

History

Repository files navigation

🤖 Qwen Code CLI Wrapper

✨ Features

🚀 Quick Start

Prerequisites

Step 1: Get OAuth2 Credentials

Using Qwen Code CLI

Step 2: Create KV Namespace

Step 3: Environment Setup

Step 4: Deploy

🔧 Configuration

Environment Variables

Core Configuration

Authentication Security

OAuth Token Management

KV Namespaces

🎯 API Endpoints

Base URL

OpenAI-Compatible Endpoints

Chat Completions

List Models

Utility Endpoints

Health Check

💻 Usage Examples

OpenAI SDK (Python)

OpenAI SDK (JavaScript/TypeScript)

cURL Examples

Open WebUI Integration

🏗️ How It Works

🚨 Troubleshooting

Common Issues

Debug Logging

Debug Endpoints

🧠 Architecture

Project Structure

Key Components

🤝 Contributing

Development Setup

Available Scripts

📄 License

🙏 Acknowledgments

About

Topics

Resources

Uh oh!

Stars

Watchers

Forks

Releases

Sponsor this project

Uh oh!

Packages 0

Uh oh!

Languages

Packages