💬 Feature: Real-Time AI Chat via Socket.IO

[abhishek-nexgen-dev]

This feature adds real-time chat support using Socket.IO. It allows users to send messages to different AI models and receive complete AI-generated responses in real-time.

✅ Key Features

• ✅ No token-by-token streaming
• ✅ Easy to integrate with any frontend
• ✅ Works with local models (Ollama) and cloud models (Groq, OpenAI, etc.)

---

✅ Supported AI Providers

Provider	Type	Notes
Ollama	Local	Uses installed models (e.g., llama3)
Groq	Cloud	Requires user API key
OpenAI	Cloud	Supports GPT-3.5, GPT-4
Anthropic	Cloud	Supports Claude models
Other	Custom	Can be extended with custom logic

⚙️ How It Works

📥 userMessage Event (Client ➡️ Server)

The frontend sends a message to the AI model via socket:

{
  "model": {
    "provider": "ollama",         // or "groq", "openai", "custom"
    "name": "llama3",             // model name
    "isPaid": false               // optional
  },
  "messages": [
    { "role": "user", "content": "Tell me a fun fact about space." }
  ],
  "apiKey": "user-api-key-or-local-key"
}

📤 aiResponse Event (Server ➡️ Client)

The server responds with the full AI-generated reply:

{
  "model": "llama3",
  "response": "Space smells like seared steak, according to astronauts!",
  "timestamp": "2025-10-12T14:00:00Z"
}

🚫 error Event (Server ➡️ Client)

If something goes wrong:

{
  "error": true,
  "message": "Invalid API key or model not found."
}