Apex AI Proxy

Appearance

Guide ​

Welcome to the comprehensive Apex AI Proxy guide! This documentation will help you master all aspects of deploying, configuring, and using your personal AI gateway.

Overview ​

Apex AI Proxy is a powerful solution that aggregates multiple AI service providers behind a unified OpenAI-compatible API. It runs on Cloudflare Workers and helps you overcome rate limits, maximize free quotas, and simplify your AI integrations.

Getting Started ​

New to Apex AI Proxy? ​

If you're just getting started, we recommend following this learning path:

  1. Quick Start - Get up and running in 60 seconds
  2. Installation - Detailed installation instructions
  3. Configuration - Learn how to configure providers and models
  4. Basic Usage - Make your first API calls

Core Concepts ​

Understanding these concepts will help you get the most out of Apex AI Proxy:

🔄 Provider Aggregation ​

Apex AI Proxy acts as a smart router that distributes requests across multiple AI providers. This allows you to:

  • Overcome individual provider rate limits
  • Take advantage of multiple free tiers
  • Ensure high availability through automatic failover

🔑 Unified Authentication ​

Instead of managing multiple API keys in your applications, you configure them once in the proxy and use a single service key for all requests.

🤖 OpenAI Compatibility ​

All interactions use the familiar OpenAI API format, making integration seamless with existing codebases and tools.

Architecture ​

mermaid
graph TD
    A[Your Application] --> B[Apex AI Proxy]
    B --> C[Azure OpenAI]
    B --> D[DeepSeek]
    B --> E[Aliyun DashScope]
    B --> F[DeepInfra]
    B --> G[Custom Provider]
    
    subgraph "Cloudflare Workers"
        B
    end
    
    subgraph "AI Providers"
        C
        D
        E
        F
        G
    end

Key Features ​

🆓 Completely Free ​

  • Runs on Cloudflare Workers' free tier
  • No server maintenance required
  • No hidden costs or subscriptions

🔄 Smart Load Balancing ​

  • Automatic request distribution
  • Failover on provider errors
  • Round-robin and weighted strategies

💰 Cost Optimization ​

  • Maximize free quotas across providers
  • Intelligent provider selection
  • Usage tracking and optimization

🛡️ Robust Error Handling ​

  • Graceful failover between providers
  • Comprehensive error mapping
  • Retry logic with exponential backoff

🌐 Multi-Protocol Support ​

  • OpenAI Chat Completions API
  • OpenAI Embeddings API
  • Anthropic Messages API
  • Next-gen /v1/responses API

Quick Reference ​

Essential Commands ​

bash
# Install dependencies
pnpm install

# Start development server
pnpm run dev

# Deploy to production
pnpm run deploy

# View logs
wrangler tail

Basic Configuration ​

javascript
// wrangler-config.js
const providerConfig = {
  azure: {
    base_url: 'https://your-resource.openai.azure.com/openai/deployments/your-deployment',
    api_keys: ['your-azure-key'],
  },
  deepseek: {
    base_url: 'https://api.deepseek.com/v1',
    api_keys: ['your-deepseek-key'],
  },
};

const modelProviderConfig = {
  'gpt-4o-mini': {
    providers: [
      { provider: 'azure', model: 'gpt-4o-mini' },
      { provider: 'deepseek', model: 'deepseek-chat' },
    ],
  },
};

Basic Usage ​

python
from openai import OpenAI

client = OpenAI(
    base_url="https://your-proxy.workers.dev/v1",
    api_key="your-service-api-key"
)

response = client.chat.completions.create(
    model="gpt-4o-mini",
    messages=[{"role": "user", "content": "Hello!"}]
)

Documentation Sections ​

📚 Getting Started ​

Perfect for newcomers to understand the basics and get their first deployment running.

🚀 Advanced Usage ​

Deep dive into advanced features and optimization techniques.

🔌 Integrations ​

Learn how to integrate with popular tools and frameworks.

📊 API Reference ​

Complete API documentation with examples and best practices.

🔑 Provider Guides ​

Detailed setup instructions for each supported provider.

Best Practices ​

🎯 Configuration ​

  • Use multiple providers for the same model to ensure redundancy
  • Configure multiple API keys per provider to scale beyond rate limits
  • Set up proper fallback chains for critical applications

🔐 Security ​

  • Keep your configuration files secure and out of version control
  • Use environment variables for sensitive data in production
  • Regularly rotate your API keys

📈 Performance ​

  • Monitor your usage patterns to optimize provider selection
  • Use appropriate models for different use cases
  • Implement proper caching strategies in your applications

💰 Cost Management ​

  • Take advantage of free tiers from multiple providers
  • Monitor usage to avoid unexpected charges
  • Use cheaper models for development and testing

Community & Support ​

🤝 Getting Help ​

  • Documentation: You're reading it! Search for specific topics
  • GitHub Issues: Report bugs and request features
  • Discussions: Ask questions and share ideas with the community

🌟 Contributing ​

We welcome contributions! Check out our:

Next Steps ​

Choose your path based on your experience level:

🆕 New Users ​

  1. Complete the Quick Start tutorial
  2. Read about Configuration
  3. Explore Provider Setup

🔧 Developers ​

  1. Review the API Reference
  2. Check out Integration Guides
  3. Learn about Advanced Usage

🏢 Enterprise Users ​

  1. Understand Security Best Practices
  2. Set up Monitoring and Analytics
  3. Plan your Multi-Provider Strategy

Ready to dive in? Start with our Quick Start guide and have your proxy running in under 60 seconds!

Released under the MIT License.

Copyright © 2024-present Apex AI Proxy Team