This directory contains interfaces for various LLM models that can power KaliGPT.

• GPT-4/GPT-3.5 (gpt.py) - OpenAI's models (requires API key)
• Mistral (mistral.py) - Mistral AI models (API or local)

• LLaMA (local_llama.py) - Meta's LLaMA models via Ollama
• Qwen (qwen.py) - Alibaba's Qwen models via Ollama

export OPENAI_API_KEY=your_api_key_here

# Install Ollama
curl -fsSL https://ollama.com/install.sh | sh

# Pull a model
ollama pull llama2
ollama pull mistral
ollama pull qwen

# Start Ollama server
ollama serve

Edit config/model_config.json:

{
  "model_type": "llama",
  "model": "llama2",
  "temperature": 0.7,
  "max_tokens": 2000
}

The model_selector.py automatically chooses the appropriate model interface based on your configuration.