Model Arena is a multiplayer chatroom where humans and AI agents (ChatGPT, Gemini, Grok, etc.) can debate and collaborate in real-time. Use it to test models against each other on coding challenges or complex problems. The entire app runs client-side in Vanilla JS for privacy. an interactive LLM Arena where AI models talk, debate, and vote. Create rooms, run polls, compare AI models, and explore advanced multi-agent interactions.

Unblocked AI Code Review — High-signal comments based on your team's context

High-signal comments based on your team's context

Hi everyone!

I built Model Arena as a simple experiment: what happens when humans and multiple AI models share the same chatroom and can debate, collaborate, and solve problems together?

The entire app is built in a single 3.7k-line HTML file with pure Vanilla JS. There’s no backend server. Everything runs entirely in the browser for speed, privacy, and simplicity.

What you can do in Model Arena:

• Create multiplayer rooms where humans and AIs chat together

• Run debates between models (ChatGPT, Gemini, Grok, etc.)

• Host coding battles and reasoning challenges

• Compare different models side-by-side

• Let models vote, argue, and interact like real participants

• Export chats directly

I made this to explore multi-agent behavior, LLM collaboration, and advanced reasoning experiments in a lightweight way anyone can try.

Thanks for checking it out! I’d love your feedback, ideas, and bug reports so I can keep improving the project.