OpenRouter Backend

A robust backend service that provides a unified API interface for interacting with multiple Large Language Model (LLM) providers, including OpenAI, Google Gemini, Anthropic, and DeepSeek. This project simplifies the integration of various AI models into your applications by offering a consistent chat interface, built-in API key management, automatic usage tracking, and pay-per-token billing.

🚀 Features

• Multi-Provider Support: Seamlessly switch between OpenAI, Google Gemini, Anthropic, and DeepSeek models through a single API.
• Unified Chat API: A single POST endpoint to interact with all supported LLMs — just specify the provider and model.
• API Key Management: Generate, list, and delete API keys securely, scoped per user.
• Automatic Usage Tracking: Every chat request automatically logs token consumption (input, output, total) per API key.
• Pay-Per-Token Billing: Price calculations based on a configurable pricing table (pricing table in Supabase) with a 5% commission layer.
• Usage & Billing Endpoint: Check aggregated usage stats and the final price for a user across all their API keys.
• Authentication: Two-layer auth middleware — user-only validation for key management, and user + API key validation for chat requests.
• Model Pricing Reference: Includes a comprehensive CSV (Token-Prices/Model_Pricings.csv) with per-token pricing for 79+ models across all providers.
• Scalable Architecture: Built with Node.js and Express 5, designed for easy provider extension via the factory pattern.

🛠️ Technology Stack

Category	Technology
Runtime	Node.js (v16+)
Framework	Express.js v5
Database / Auth	Supabase (PostgreSQL + Auth)
AI SDKs	openai (OpenAI, Gemini & DeepSeek via compatibility layer), @anthropic-ai/sdk
Utilities	dotenv (env config), nodemon (dev hot-reload)

📋 Prerequisites

Before you begin, ensure you have the following:

• Node.js (v16 or higher) & npm
• A Supabase project with the following tables configured:
  • Users — user records with User_name column
  • API_Keys — stores generated keys with API_key and User_name columns
  • pricing — model pricing with provider, model, and price_per_token columns
  • api_usage — usage logs with api_key, provider, model, input_tokens, output_tokens, total_tokens, and price columns
• API Keys for the providers you intend to use:
  • OpenAI API Key
  • Google Gemini API Key
  • Anthropic API Key
  • DeepSeek API Key

⚙️ Installation

1. Clone the repository

   git clone <repository_url>
   cd Backend

2. Install dependencies

   npm install

3. Configure Environment Variables Create a .env file in the root directory:

   PORT=3000
   
   # AI Provider Keys
   OPENAI_API_KEY=your_openai_api_key_here
   GEMINI_API_KEY=your_gemini_api_key_here
   ANTHROPIC_API_KEY=your_anthropic_api_key_here
   DEEPSEEK_API_KEY=your_deepseek_api_key_here
   
   # Supabase Configuration
   SUPABASE_URL=your_supabase_project_url
   SUPABASE_KEY=your_supabase_anon_key

🏃‍♂️ Running the Application

Development Mode

Use nodemon to automatically restart the server on file changes:

npm run dev

Production Mode

Start the server normally:

npm start

The server will start on http://localhost:3000 (or the port specified in your .env).

🔌 API Endpoints

All endpoints require authentication headers. See the Authentication section below for details.

---

Chat — /api/layers

POST /api/layers/chat

Send a message to any supported LLM provider.

Headers:

Header	Required	Description
user	✅	Your registered username
api-key	✅	A valid API key generated for your account

Request Body:

{
  "providerName": "openai",
  "model_name": "gpt-4o",
  "message": "Hello, how are you?"
}

Supported Providers: openai, gemini, anthropic, deepseek

Response:

{
  "content": "I'm doing great! How can I help you today?",
  "usage": {
    "inputTokens": 12,
    "outputTokens": 15,
    "totalTokens": 27
  }
}

Note: Usage is automatically logged to the api_usage table with price calculation after every successful request.

---

API Key Management — /api/keys

Manage API keys for accessing the service. These endpoints require a valid user header.

GET /api/keys/list

Retrieve all API keys associated with the authenticated user.

POST /api/keys/generate

Generate a new API key for the authenticated user.

DELETE /api/keys/delete

Revoke an existing API key.

Request Body:

{
  "key": "your_api_key_to_delete"
}

---

Usage & Billing — /api/usage

GET /api/usage/check

Get aggregated usage statistics and the calculated bill for the authenticated user.

Headers:

Header	Required	Description
user	✅	Your registered username
api-key	✅	A valid API key

Response:

{
  "priceToPay": {
    "username": "john_doe",
    "total_input_tokens": 15240,
    "total_output_tokens": 8320,
    "total_tokens": 23560,
    "base_price": 0.029450,
    "commission_5_percent": 0.001473,
    "final_price": 0.030923
  }
}

🔐 Authentication

The API uses two levels of middleware:

Middleware	Used By	Validates
auth.middleware.js	Chat, Usage	user header (against Users table) + api-key header (against API_Keys table)
authForKeys.middleware.js	Key management	user header only (against Users table)

💰 Pricing & Billing

The billing system works in three steps:

1. Per-Request Price Calculation (priceCal.logistics.js): After each chat request, the system looks up the per-token price from the pricing table for the given provider/model and logs the usage with the calculated cost.
2. Aggregated Billing (finalPrice.logistics.js): When a user checks their usage, all usage records across all their API keys are summed up, and a 5% commission is applied on top of the base price.
3. Model Pricing Reference: The Token-Prices/Model_Pricings.csv file contains per-token pricing for 79 models across all four providers, which can be used to seed the pricing table.

📂 Project Structure

├── Token-Prices/
│   └── Model_Pricings.csv      # Reference pricing data for 79+ models
├── src/
│   ├── controller/
│   │   ├── chat.controller.js   # Chat request handler
│   │   ├── keys.controller.js   # API key CRUD handler
│   │   └── usage.controller.js  # Usage & billing handler
│   ├── logistics/
│   │   ├── priceCal.logistics.js    # Per-request price calculation & usage logging
│   │   └── finalPrice.logistics.js  # Aggregated billing with commission
│   ├── middleware/
│   │   ├── auth.middleware.js        # User + API Key validation
│   │   ├── authForKeys.middleware.js # User-only validation
│   │   └── error.middleware.js       # Global error handler
│   ├── model/
│   │   ├── dbClient.js          # Database client setup
│   │   └── supabaseClient.js    # Supabase client initialization
│   ├── provider/
│   │   ├── openai.provider.js       # OpenAI integration
│   │   ├── gemini.provider.js       # Google Gemini integration (via OpenAI compat)
│   │   ├── anthropic.provider.js    # Anthropic Claude integration
│   │   ├── deepseek.provider.js     # DeepSeek integration (via OpenAI compat)
│   │   └── providerManager.provider.js  # Provider factory / registry
│   ├── routes/
│   │   ├── chat.routes.js       # /api/layers/* routes
│   │   ├── keys.routes.js       # /api/keys/* routes
│   │   └── usage.routes.js      # /api/usage/* routes
│   ├── services/
│   │   ├── chat.services.js         # Chat business logic + usage logging
│   │   └── generateKey.services.js  # API key generation logic
│   ├── app.js                   # Express app setup & route mounting
│   └── index.js                 # Server entry point
├── .env                         # Environment variables (not committed)
├── .gitignore
├── package.json
└── README.md

🤝 Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

1. Fork the project
2. Create your feature branch (git checkout -b feature/AmazingFeature)
3. Commit your changes (git commit -m 'Add some AmazingFeature')
4. Push to the branch (git push origin feature/AmazingFeature)
5. Open a Pull Request

📝 License

This project is licensed under the ISC License.