Model Comparison

Compares configured LLM targets with the same attack set.

Requirements

• Configure comparison.targets in config.json
• Provide needed API keys via target env vars

Run

uv run python main.py --module comparison --comparison

Output

• Per-target module summaries
• Combined success/failure totals
• Side-by-side vulnerability rates