Proposed Feature: Blind Testing and Ranking System for LLM Comparisons

1. Implement a voting system to select the better response between two LLM answers
2. Enable blind voting where LLM identities are hidden to prevent bias
3. Incorporate an ELO rating system to rank LLM performance
4. Display real-time LLM rankings in the right panel

This enhancement would allow users to:

• Compare responses objectively
• Contribute to a community-driven rating system
• Track LLM performance over time

I have implemented a quick prototype: