README.md

metadata

title: LLM AI Agent API
emoji: 🤖
colorFrom: blue
colorTo: purple
sdk: docker
app_port: 7860
pinned: false
license: mit

LLM AI Agent API

A secure, production-ready AI agent API powered by local LLM deployment on Hugging Face Spaces.

🚀 Quick Start

1. Set Environment Variables in your Space settings:

   • API_KEY_1: Your secure API key
   • API_KEY_2: Additional API key (optional)
   • MODEL_NAME: LLM model to use (default: microsoft/DialoGPT-medium)

2. Access your API at: https://your-username-your-space.hf.space

3. Test the API:

curl -X POST "https://your-space-url.hf.space/chat" \
  -H "Authorization: Bearer your-api-key" \
  -H "Content-Type: application/json" \
  -d '{"message": "Hello, how are you?"}'

📚 Documentation

• API Docs: /docs (Interactive Swagger UI)
• Health Check: /health
• Model Info: /models (requires authentication)

🔐 Security Features

• API key authentication
• Rate limiting (10 requests/minute per key)
• CORS support
• Request logging

🛠️ Integration

See client_examples.py and client_examples.js for integration examples in Python and JavaScript.

📝 API Reference

POST /chat

{
  "message": "Your input message",
  "max_length": 200,
  "temperature": 0.7,
  "system_prompt": "Optional system prompt"
}

Response:

{
  "response": "AI response",
  "model_used": "microsoft/DialoGPT-medium",
  "timestamp": "2024-01-01T12:00:00",
  "tokens_used": 45,
  "processing_time": 1.23
}

🔧 Configuration

Set these environment variables in your Space:

• API_KEY_1, API_KEY_2: Authentication keys
• MODEL_NAME: Hugging Face model ID
• MAX_LENGTH: Maximum response length
• TEMPERATURE: Response creativity (0.1-2.0)
• RATE_LIMIT: Requests per minute per key

📄 License

MIT License - see LICENSE file for details.