Multi-Model Chat System

This repository contains two different implementations of a multi-model chat system that leverages various AI models for enhanced conversational capabilities.

Files

1. multi-chat.py

A chat system that allows users to switch between different AI models during conversation:

• Supports Gemini, OpenAI (O1), DeepSeek, and Qwen models
• Maintains conversation history across model switches
• Features command-line interface with colored output
• Commands:
  • --change: Switch to a different model
  • --clear: Clear chat history
  • --exit: Exit the chat
  • --help: Show available commands

2. frankenthought-chat.py

An advanced implementation that runs all reasoning models in parallel and synthesizes their responses:

• Parallel execution of:
  1. Google's Gemini (gemini-2.0-flash-thinking-exp-01-21)
  2. OpenAI's O1 model
  3. DeepSeek's R1 model (via OpenRouter)
  4. Qwen's 32B Preview model (via OpenRouter)
• Uses OpenAI's O1 model again for final synthesis of all responses
• Maintains a detailed chat log in JSON format
• Features:

  • Parallel processing of model responses

  • Response synthesis combining insights from all models

  • Detailed logging of all responses and final synthesis

  • Commands:

    • --clear: Clear chat history
    • --exit: Exit the chat
    • --help: Show available commands

🎥 Watch How It's Built!

Watch the complete build process on Patreon See exactly how this was created step by step, with detailed explanations and insights into the development process at my 1000x Cursor Course

❤️ Join my AI community & Get 400+ AI Projects

This is one of 400+ fascinating projects in my collection! Support me on Patreon to get:

• 🎯 Access to 400+ AI projects (and growing daily!)
  • Including advanced projects like 2 Agent Real-time voice template with turn taking
• 📥 Full source code & detailed explanations
• 📚 1000x Cursor Course
• 🎓 Live coding sessions & AMAs
• 💬 1-on-1 consultations (higher tiers)
• 🎁 Exclusive discounts on AI tools & platforms (up to $180 value)

Setup

1. Set up environment variables:

GEMINI_API_KEY=your_gemini_key
OPENAI_API_KEY=your_openai_key
OPENROUTER_API_KEY=your_openrouter_key

2. Install dependencies:

pip install -r requirements.txt

Usage

Running multi-chat.py

python multi-chat.py

• Select a model using numbers 1-4
• Chat normally
• Use --change to switch models
• Previous chat history is maintained when switching models

Running frankenthought-chat.py

python frankenthought-chat.py

• Type your message
• See responses from all models (anonymized as Model 1-4)
• Get a synthesized response combining insights from all models
• All interactions are logged in chat_responses.json

Features

Common Features

• Colored terminal output for better readability
• Error handling and graceful degradation
• UTF-8 encoding for all file operations
• Async API calls for better performance

multi-chat.py Specific

• Model switching while maintaining context
• Individual model conversations
• Chat history persistence across model changes

frankenthought-chat.py Specific

• Parallel model execution of 4 different models
• Two-stage process:
  1. Get parallel responses from all models
  2. Use O1 to synthesize a final response
• Comprehensive JSON logging
• Anonymous model responses (Model 1, Model 2, etc.)
• Detailed chat history with timestamps

Chat Log Format (frankenthought-chat.py)

{
  "timestamp": "ISO-8601 timestamp",
  "user_message": "User's input",
  "model_responses": {
    "Model 1": "Response from first model",
    "Model 2": "Response from second model",
    "Model 3": "Response from third model",
    "Model 4": "Response from fourth model"
  },
  "final_response": "Synthesized response"
}

Error Handling

• Graceful handling of API failures
• Informative error messages
• Continued operation if some models are unavailable
• Safe file operations with proper error catching

Dependencies

• google-generativeai
• openai
• termcolor
• httpx

Notes

• API keys must be set as environment variables
• Models may have different response times
• JSON log file grows with usage - consider periodic cleanup
• All models maintain conversation context
• O1 is used twice in frankenthought-chat.py:
  1. As one of the parallel responders
  2. As the final synthesizer of all responses